US20150325528A1 - Method of manufacturing semiconductor device, and semiconductor device - Google Patents
Method of manufacturing semiconductor device, and semiconductor device Download PDFInfo
- Publication number
- US20150325528A1 US20150325528A1 US14/803,486 US201514803486A US2015325528A1 US 20150325528 A1 US20150325528 A1 US 20150325528A1 US 201514803486 A US201514803486 A US 201514803486A US 2015325528 A1 US2015325528 A1 US 2015325528A1
- Authority
- US
- United States
- Prior art keywords
- chip
- logic chip
- electrodes
- wiring substrate
- semiconductor
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Abandoned
Links
Images
Classifications
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
- H01L23/00—Details of semiconductor or other solid state devices
- H01L23/544—Marks applied to semiconductor devices or parts, e.g. registration marks, alignment structures, wafer maps
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
- H01L23/00—Details of semiconductor or other solid state devices
- H01L23/48—Arrangements for conducting electric current to or from the solid state body in operation, e.g. leads, terminal arrangements ; Selection of materials therefor
- H01L23/481—Internal lead connections, e.g. via connections, feedthrough structures
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
- H01L24/00—Arrangements for connecting or disconnecting semiconductor or solid-state bodies; Methods or apparatus related thereto
- H01L24/93—Batch processes
- H01L24/95—Batch processes at chip-level, i.e. with connecting carried out on a plurality of singulated devices, i.e. on diced chips
- H01L24/97—Batch processes at chip-level, i.e. with connecting carried out on a plurality of singulated devices, i.e. on diced chips the devices being connected to a common substrate, e.g. interposer, said common substrate being separable into individual assemblies after connecting
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
- H01L25/00—Assemblies consisting of a plurality of individual semiconductor or other solid state devices ; Multistep manufacturing processes thereof
- H01L25/03—Assemblies consisting of a plurality of individual semiconductor or other solid state devices ; Multistep manufacturing processes thereof all the devices being of a type provided for in the same subgroup of groups H01L27/00 - H01L33/00, or in a single subclass of H10K, H10N, e.g. assemblies of rectifier diodes
- H01L25/04—Assemblies consisting of a plurality of individual semiconductor or other solid state devices ; Multistep manufacturing processes thereof all the devices being of a type provided for in the same subgroup of groups H01L27/00 - H01L33/00, or in a single subclass of H10K, H10N, e.g. assemblies of rectifier diodes the devices not having separate containers
- H01L25/065—Assemblies consisting of a plurality of individual semiconductor or other solid state devices ; Multistep manufacturing processes thereof all the devices being of a type provided for in the same subgroup of groups H01L27/00 - H01L33/00, or in a single subclass of H10K, H10N, e.g. assemblies of rectifier diodes the devices not having separate containers the devices being of a type provided for in group H01L27/00
- H01L25/0657—Stacked arrangements of devices
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
- H01L25/00—Assemblies consisting of a plurality of individual semiconductor or other solid state devices ; Multistep manufacturing processes thereof
- H01L25/18—Assemblies consisting of a plurality of individual semiconductor or other solid state devices ; Multistep manufacturing processes thereof the devices being of types provided for in two or more different subgroups of the same main group of groups H01L27/00 - H01L33/00, or in a single subclass of H10K, H10N
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
- H01L25/00—Assemblies consisting of a plurality of individual semiconductor or other solid state devices ; Multistep manufacturing processes thereof
- H01L25/50—Multistep manufacturing processes of assemblies consisting of devices, each device being of a type provided for in group H01L27/00 or H01L29/00
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
- H01L2223/00—Details relating to semiconductor or other solid state devices covered by the group H01L23/00
- H01L2223/544—Marks applied to semiconductor devices or parts
- H01L2223/5442—Marks applied to semiconductor devices or parts comprising non digital, non alphanumeric information, e.g. symbols
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
- H01L2223/00—Details relating to semiconductor or other solid state devices covered by the group H01L23/00
- H01L2223/544—Marks applied to semiconductor devices or parts
- H01L2223/54426—Marks applied to semiconductor devices or parts for alignment
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
- H01L2223/00—Details relating to semiconductor or other solid state devices covered by the group H01L23/00
- H01L2223/544—Marks applied to semiconductor devices or parts
- H01L2223/54473—Marks applied to semiconductor devices or parts for use after dicing
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
- H01L2224/00—Indexing scheme for arrangements for connecting or disconnecting semiconductor or solid-state bodies and methods related thereto as covered by H01L24/00
- H01L2224/01—Means for bonding being attached to, or being formed on, the surface to be connected, e.g. chip-to-package, die-attach, "first-level" interconnects; Manufacturing methods related thereto
- H01L2224/10—Bump connectors; Manufacturing methods related thereto
- H01L2224/15—Structure, shape, material or disposition of the bump connectors after the connecting process
- H01L2224/16—Structure, shape, material or disposition of the bump connectors after the connecting process of an individual bump connector
- H01L2224/161—Disposition
- H01L2224/16151—Disposition the bump connector connecting between a semiconductor or solid-state body and an item not being a semiconductor or solid-state body, e.g. chip-to-substrate, chip-to-passive
- H01L2224/16221—Disposition the bump connector connecting between a semiconductor or solid-state body and an item not being a semiconductor or solid-state body, e.g. chip-to-substrate, chip-to-passive the body and the item being stacked
- H01L2224/16225—Disposition the bump connector connecting between a semiconductor or solid-state body and an item not being a semiconductor or solid-state body, e.g. chip-to-substrate, chip-to-passive the body and the item being stacked the item being non-metallic, e.g. insulating substrate with or without metallisation
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
- H01L2224/00—Indexing scheme for arrangements for connecting or disconnecting semiconductor or solid-state bodies and methods related thereto as covered by H01L24/00
- H01L2224/80—Methods for connecting semiconductor or other solid state bodies using means for bonding being attached to, or being formed on, the surface to be connected
- H01L2224/83—Methods for connecting semiconductor or other solid state bodies using means for bonding being attached to, or being formed on, the surface to be connected using a layer connector
- H01L2224/8319—Arrangement of the layer connectors prior to mounting
- H01L2224/83192—Arrangement of the layer connectors prior to mounting wherein the layer connectors are disposed only on another item or body to be connected to the semiconductor or solid-state body
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
- H01L2225/00—Details relating to assemblies covered by the group H01L25/00 but not provided for in its subgroups
- H01L2225/03—All the devices being of a type provided for in the same subgroup of groups H01L27/00 - H01L33/648 and H10K99/00
- H01L2225/04—All the devices being of a type provided for in the same subgroup of groups H01L27/00 - H01L33/648 and H10K99/00 the devices not having separate containers
- H01L2225/065—All the devices being of a type provided for in the same subgroup of groups H01L27/00 - H01L33/648 and H10K99/00 the devices not having separate containers the devices being of a type provided for in group H01L27/00
- H01L2225/06503—Stacked arrangements of devices
- H01L2225/06541—Conductive via connections through the device, e.g. vertical interconnects, through silicon via [TSV]
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
- H01L2225/00—Details relating to assemblies covered by the group H01L25/00 but not provided for in its subgroups
- H01L2225/03—All the devices being of a type provided for in the same subgroup of groups H01L27/00 - H01L33/648 and H10K99/00
- H01L2225/04—All the devices being of a type provided for in the same subgroup of groups H01L27/00 - H01L33/648 and H10K99/00 the devices not having separate containers
- H01L2225/065—All the devices being of a type provided for in the same subgroup of groups H01L27/00 - H01L33/648 and H10K99/00 the devices not having separate containers the devices being of a type provided for in group H01L27/00
- H01L2225/06503—Stacked arrangements of devices
- H01L2225/06548—Conductive via connections through the substrate, container, or encapsulation
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
- H01L2225/00—Details relating to assemblies covered by the group H01L25/00 but not provided for in its subgroups
- H01L2225/03—All the devices being of a type provided for in the same subgroup of groups H01L27/00 - H01L33/648 and H10K99/00
- H01L2225/04—All the devices being of a type provided for in the same subgroup of groups H01L27/00 - H01L33/648 and H10K99/00 the devices not having separate containers
- H01L2225/065—All the devices being of a type provided for in the same subgroup of groups H01L27/00 - H01L33/648 and H10K99/00 the devices not having separate containers the devices being of a type provided for in group H01L27/00
- H01L2225/06503—Stacked arrangements of devices
- H01L2225/06582—Housing for the assembly, e.g. chip scale package [CSP]
- H01L2225/06586—Housing with external bump or bump-like connectors
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
- H01L2924/00—Indexing scheme for arrangements or methods for connecting or disconnecting semiconductor or solid-state bodies as covered by H01L24/00
- H01L2924/06—Polymers
- H01L2924/078—Adhesive characteristics other than chemical
- H01L2924/07802—Adhesive characteristics other than chemical not being an ohmic electrical conductor
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
- H01L2924/00—Indexing scheme for arrangements or methods for connecting or disconnecting semiconductor or solid-state bodies as covered by H01L24/00
- H01L2924/15—Details of package parts other than the semiconductor or other solid state devices to be connected
- H01L2924/151—Die mounting substrate
- H01L2924/1517—Multilayer substrate
- H01L2924/15172—Fan-out arrangement of the internal vias
- H01L2924/15174—Fan-out arrangement of the internal vias in different layers of the multilayer substrate
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
- H01L2924/00—Indexing scheme for arrangements or methods for connecting or disconnecting semiconductor or solid-state bodies as covered by H01L24/00
- H01L2924/15—Details of package parts other than the semiconductor or other solid state devices to be connected
- H01L2924/151—Die mounting substrate
- H01L2924/153—Connection portion
- H01L2924/1531—Connection portion the connection portion being formed only on the surface of the substrate opposite to the die mounting surface
- H01L2924/15311—Connection portion the connection portion being formed only on the surface of the substrate opposite to the die mounting surface being a ball array, e.g. BGA
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
- H01L2924/00—Indexing scheme for arrangements or methods for connecting or disconnecting semiconductor or solid-state bodies as covered by H01L24/00
- H01L2924/15—Details of package parts other than the semiconductor or other solid state devices to be connected
- H01L2924/181—Encapsulation
Definitions
- the present invention relates to manufacturing techniques of semiconductor devices, and more specifically, to a technique effectively applied to a semiconductor device with a semiconductor chip mounted over another semiconductor chip.
- Patent Document 1 discloses a method for stacking semiconductor chips which involves respectively forming alignment marks on the surface of a parent chip (fixed device) and on the surface of a child chip (semiconductor chip), and stacking the child chip over the parent chip based on the result of detection of the alignment marks such that the respective alignment marks on both chips face each other.
- Patent Document 2 discloses a semiconductor device including a semiconductor chip with a penetrating electrode between a wiring substrate and a laminated body of a plurality of memory chips.
- Patent Document 1
- the inventors of the present application have studied about techniques for improving the performance of a semiconductor device including a laminated body of semiconductor chips stacked over a wiring substrate.
- the so-called system in package (SiP) semiconductor device has been studied, which forms a system only by one semiconductor device having a plurality of semiconductor chips (for example, a memory chip and a control chip for controlling the memory chip) mounted therein.
- Methods for stacking semiconductor chips includes a method which involves arranging and stacking an electrode of one (upper) semiconductor chip to face the other (lower) semiconductor chip, and electrically connecting the respective electrodes via a conductive member, such as solder material. This method connects the stacked semiconductor chips together without wirings, and thus can decrease a transmission distance between the semiconductor chips.
- a method for manufacturing a semiconductor device includes the step of (a) mounting a first semiconductor chip over a wiring substrate such that a first main surface of the first semiconductor chip faces the wiring substrate.
- the manufacturing method also includes the step of (b) mounting a second semiconductor chip over a wiring substrate such that a second main surface of the second semiconductor chip faces a first back surface of the first semiconductor chip.
- the first semiconductor chip is mounted.
- the second semiconductor chip is mounted.
- a plurality of first back surface side pads are formed on the first back surface, and a plurality of second main surface side pads are formed on the second main surface.
- the first back surface side pads are electrically connected with the second main surface, side pads via external terminals respectively formed on the second main surface side pads.
- the reliability of the semiconductor device can be improved.
- FIG. 1 is a perspective view of a semiconductor device according to one embodiment of the invention.
- FIG. 2 is a bottom view of the semiconductor device shown in FIG. 1 ;
- FIG. 3 is a perspective plan view showing an internal structure of the semiconductor device over a wiring substrate with a seal member shown in FIG. 1 removed therefrom;
- FIG. 4 is a cross-sectional view taken along the line A-A of FIG. 1 ;
- FIG. 5 is an exemplary explanatory diagram showing an example of a circuit structure of the semiconductor device shown in FIGS. 1 to 4 ;
- FIG. 6 is an enlarged cross-sectional view of a portion “A” shown in FIG. 4 ;
- FIG. 7 is a plan view showing a front surface side of a memory chip shown in FIG. 4 ;
- FIG. 8 is a plan view showing an example of a back surface side of a memory chip shown in FIG. 7 ;
- FIG. 9 is a plan view showing a front surface side of a logic chip shown in FIG. 4 ;
- FIG. 10 is a plan view showing an example of a back surface side of a logic chip shown in FIG. 9 ;
- FIG. 11 is an exemplary explanatory diagram showing alignment means (alignment marks) used in assembling the semiconductor device shown in FIG. 4 ;
- FIG. 12 is an explanatory diagram showing a studied example corresponding to FIG. 11 ;
- FIG. 13 is an explanatory diagram showing the outline of a manufacturing process of the semiconductor device explained with reference to FIGS. 1 to 11 ;
- FIG. 14 is a plan view showing the entire structure of a wiring substrate provided in a substrate provision step shown in FIG. 13 ;
- FIG. 15 is an enlarged plan view of one device region shown in FIG. 14 ;
- FIG. 16 is an enlarged cross-sectional view taken along the line A-A of FIG. 15 ;
- FIG. 17 is an enlarged plan view showing the surface opposite to that shown in FIG. 15 ;
- FIG. 18 is an enlarged cross-sectional view taken along the line B-B of FIG. 15 ;
- FIG. 19 is an enlarged plan view showing the state of an adhesive material arranged in a chip mounting region shown in FIG. 15 ;
- FIG. 20 is an enlarged cross-sectional view taken along the line A-A of FIG. 19 ;
- FIG. 21 is an exemplary explanatory diagram showing the outline of a manufacturing process of a semiconductor chip with a penetrating electrode shown in FIG. 6 ;
- FIG. 22 is an exemplary explanatory diagram showing the outline of another manufacturing process of the semiconductor chip, following the process shown in FIG. 21 ;
- FIG. 23 is an enlarged plan view showing the state of a logic chip LC mounted over the chip mounting region of the wiring substrate shown in FIG. 19 ;
- FIG. 24 is an enlarged cross-sectional view taken along the line A-A of FIG. 23 ;
- FIG. 25 is an exemplary explanatory diagram showing a main part of a first chip conveying step shown in FIG. 13 ;
- FIG. 26 is an exemplary explanatory diagram showing a main part of a first mark detection step shown in FIG. 13 ;
- FIG. 27 is an enlarged cross-sectional view showing a main part of a cross-sectional structure of an alignment mark of the logic chip shown in FIG. 26 ;
- FIG. 28 is an exemplary explanatory diagram showing a main part of a first alignment step shown in FIG. 13 ;
- FIG. 29 is an exemplary explanatory diagram showing the state of the logic chip moved to the wiring substrate after the first alignment step shown in FIG. 28 ;
- FIG. 30 is an explanatory diagram showing the state of a heating jig pressed against the back surface of the semiconductor chip after removal of a holding jig shown in FIG. 29 ;
- FIG. 31 is an enlarged plan view showing the state of the adhesive material arranged at the back surface of the semiconductor chip and its surroundings shown in FIG. 20 ;
- FIG. 32 is an enlarged cross-sectional view taken along the line A-A of FIG. 31 ;
- FIG. 33 is an exemplary explanatory diagram showing the outline of an assembly process of a laminated body of the memory chips shown in FIG. 4 ;
- FIG. 34 is an exemplary explanatory diagram showing the outline of another assembly process of the laminated body of the memory chips, following the process shown in FIG. 33 ;
- FIG. 35 is an enlarged plan view showing the state of the laminated body mounted over the back surface of the logic chip shown in FIG. 31 ;
- FIG. 36 is an enlarged cross-sectional view taken along the line A-A of FIG. 35 ;
- FIG. 37 is an exemplary explanatory diagram showing a main part of a second chip conveying step shown in FIG. 13 ;
- FIG. 38 is an exemplary explanatory diagram showing a main part of a second mark detection step shown in FIG. 13 ;
- FIG. 39 is an exemplary explanatory diagram showing a main part of a second alignment step shown in FIG. 13 ;
- FIG. 40 is an exemplary explanatory diagram showing the state of the logic chip moved to the wiring substrate after the second alignment step shown in FIG. 39 ;
- FIG. 41 is an explanatory diagram showing the state of heating jig pressed against the back surface of the semiconductor chip after removal of a holding jig shown in FIG. 40 ;
- FIG. 42 is an enlarged cross-sectional view showing the state of the stacked semiconductor chips sealed with a seal member formed over the wiring substrate shown in FIG. 36 ;
- FIG. 43 is a plan view showing the entire structure of the seal member shown in FIG. 42 ;
- FIG. 44 is an enlarged cross-sectional view showing the state of solder balls bonded to lands of the wiring substrate shown in FIG. 37 ;
- FIG. 45 is a cross-sectional view showing the singulated multi-piece wiring substrate shown in FIG. 44 ;
- FIG. 46 is an enlarged plan view showing the alignment mark shown in FIGS. 3 , 7 , 9 , and 10 ;
- FIG. 47 is an enlarged plan view showing a first modified example corresponding to FIG. 46 ;
- FIG. 48 is an enlarged plan view showing a second modified example corresponding to FIG. 46 ;
- FIG. 49 is an enlarged plan view showing a third modified example corresponding to FIG. 46 ;
- FIG. 50 is a plan view showing a modified example corresponding to FIG. 8 ;
- FIG. 51 is a perspective plan view showing a modified example corresponding to FIG. 3 .
- the following preferred embodiments may be described below by being divided into a plurality of sections or the like for convenience, if necessary, which are not independent from each other unless otherwise specified. Regardless of the order of the description of these sections, the sections indicate respective parts in a single example. Alternatively, one of the sections may be the details of a part of the other, or a modified example of a part or all of the other. In principle, parts having the same function will not be described repeatedly. Respective components of the preferred embodiments are not essential unless otherwise specified, except when limiting the number of the components in theory, and except when considered not to be definitely so from the context thereof.
- the term “X formed of A” or the like as to material, composition, and the like does not exclude elements other than the element “A”, unless otherwise specified and except when considered not to be definitely so from the context.
- the above term means “X containing A as a principal component”.
- the term “silicon member” is not limited to pure silicon, and may obviously include a SiGe (silicon-germanium) alloy, or a multi-component alloy containing silicon as a principal component, and another additive.
- the term “gold plating”, “Cu layer”, or “nickel plating” is not limited to pure one, but include a member containing gold, Cu, or nickel as a principal component unless otherwise specified.
- a cross-sectional view may omit hatching in some cases if the hatching possibly makes the sectional view complicated, or when a cavity or hole is easy to discriminate.
- the outline of a hole closed in a planar manner to and from the background may be omitted when clearly seen from the description or the like.
- a hatching or dot pattern is sometimes given even when the figure is not a cross-sectional view.
- SiP semiconductor device a semiconductor package in which a semiconductor chip (memory chip) with a memory circuit and another semiconductor chip (control chip) with a control circuit for controlling the operation of the memory circuit are mounted.
- FIG. 1 shows a perspective view of a semiconductor device of this embodiment.
- FIG. 2 is a bottom view of the semiconductor device shown in FIG. 1 .
- FIG. 3 shows a perspective plan view of an internal structure of the semiconductor device over a wiring substrate with a seal member shown in FIG. 1 removed therefrom.
- FIG. 4 is a cross-sectional view taken along the line A-A of FIG. 1 .
- the number of terminals is shown to be small in FIGS.
- FIG. 3 shows the contour of the logic chip LC by a dotted line for easy understanding of the positional relationship and a difference in planar size between the logic chip LC and the memory chip MC 4 in plan view.
- the wiring substrate 2 has an upper surface (surface, chip mounting surface) 2 a over which semiconductor chips 3 are mounted, a lower surface (surface, mounting surface) 2 b opposite to the upper surface 2 a , and side surfaces 2 c positioned between the upper surface 2 a and the lower surface 2 b .
- the wiring substrate 2 has a quadrilateral contour in the plan view shown in FIGS. 2 and 3 .
- the wiring substrate 2 is formed in a square shape having a planar size (size in plan view, sizes of the upper surface 2 a and the lower surface 2 b , contour size) of about 14 mm on a side.
- the wiring substrate 2 has a thickness (height), that is, a distance from the upper surface 2 a to the lower surface 2 b shown in FIG. 4 , for example, of about 0.3 to 0.5 mm.
- the wiring substrate 2 serves as an interposer for electrically connecting the semiconductor chips 3 mounted over the upper surface 2 a side with a mounting substrate (not shown), and has a plurality of wiring layers (four layers in an example shown in FIG. 4 ) electrically connecting the upper surface 2 a side with the lower surface 2 b side.
- the respective wiring layers have insulating layers 2 e for insulating the adjacent wirings 2 d as well as the adjacent wiring layers from each other.
- the wiring substrate 2 of this embodiment includes three insulating layers 2 e with the center insulating layer 2 e made of a core layer (core material).
- the wiring substrate 2 may be the so-called coreless substrate which does not have the core insulating layer 2 e .
- the wirings 2 d include wirings 2 d 1 formed on the upper or lower surface of the insulating layer 2 e , and via wirings 2 d 2 serving as an interlayer conductive path penetrating the insulating layer 2 e in the thickness direction.
- the upper surface 2 a of the wiring substrate 2 is provided with a plurality of bonding leads (terminals, chip mounting surface side terminals, electrodes) 2 f , which are terminals electrically connected with the semiconductor chips 3 .
- the lower surface 2 b of the wiring substrate 2 is provided with lands 2 g bonded to the solder balls 5 serving as a terminal for electrical connection to the mounting substrate (not shown), that is, as a terminal for external connection of the semiconductor device 1 .
- the bonding leads 2 f are electrically connected with the lands 2 g via the wirings 2 d .
- the wiring 2 d connected with the bonding lead 2 f or land 2 g is integrally formed with the bonding lead 2 f or land 2 g , whereby the bonding lead 2 f and land 2 g shown in FIG. 4 serve as a part of the wiring 2 d.
- the upper surface 2 a and lower surface 2 b of the wiring substrate 2 are covered with insulating films (solder resist films) 2 h and 2 k .
- the wiring 2 d formed on the upper surface 2 a of the wiring substrate 2 is covered with the insulating film 2 h .
- the insulating film 2 h is provided with openings, and at the openings, at least parts (bonding portions or bonding regions with the semiconductor chip 3 ) of the bonding leads 2 f are exposed from the insulating film 2 h .
- the wiring 2 d formed on the lower surface 2 b of the wiring substrate 2 is covered with the insulating film 2 k .
- the insulating film 2 k is provided with openings, and at the openings, at least parts (bonding portions with the solder balls 5 ) of the lands 2 g are exposed from the insulating film 2 k.
- the solder balls (external terminals, electrodes, external electrodes) 5 bonded to the lands 2 g at the lower surface 2 b of the wiring substrate 2 are arranged in columns and rows (array or matrix) as shown in FIG. 2 .
- the lands 2 g bonded to the solder balls 5 are also arranged in columns and rows (matrix).
- the semiconductor device including the external terminals (solder balls 5 , lands 2 g ) arranged on the mounting surface side of the wiring substrate 2 in the columns and rows are called an area array type semiconductor device.
- the area array type semiconductor device can effectively use the mounting surface (lower surface 2 b ) side of the wiring substrate 2 as an arrangement space for the external terminals.
- the area array type semiconductor device is preferable in that an increase in mounting area of the semiconductor device can be suppressed even when the number of external terminals is increased. That is, together with higher functionality and higher integration, the semiconductor device whose number of external terminals is increased can be mounted while saving space.
- the semiconductor device 1 includes the semiconductor chips 3 mounted over the wiring substrate 2 .
- the semiconductor chips 3 are stacked over the upper surface 2 a of the wiring substrate 2 .
- Each of the semiconductor chips 3 has a quadrilateral contour in plan view of FIG. 3 having a front surface (main surface, upper surface) 3 a , a back surface (main surface, lower surface) 3 b opposite to the front surface 3 a , and side surfaces 3 c positioned between the front surface 3 a and the back surface 3 b .
- the semiconductor chips are stacked, which can reduce amounting area of the semiconductor device 1 even when the semiconductor device 1 has higher functionality.
- the semiconductor chip 3 mounted as the lowermost level (in a position closest to the wiring substrate 2 ) is a logic chip (semiconductor chip) LC in which the arithmetic processing circuit PU (see FIG. 5 ) is formed.
- the semiconductor chips 3 mounted at the upper level of the logic chip LC include memory chips (semiconductor chips) MC 1 , MC 2 , MC 3 , and MC 4 provided with main storage circuits (storage circuits) MM (see FIG. 5 ) for storing therein data for communicating with the logic chip LC.
- the logic chip LC is provided with a control circuit for controlling the operation of the main storage circuit of each of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 , in addition to the above arithmetic processing circuit.
- the example of the circuit structure of the semiconductor device 1 will be described below.
- an adhesive material NCL (insulating adhesive material) is arranged between the adjacent semiconductor chips 3 .
- the adhesive material NCL is disposed to fill in a space between a front surface 3 a of the upper semiconductor chip 3 and a back surface 3 b of the lower semiconductor chip 3 (or, upper surface 2 a of the wiring substrate 2 ).
- the adhesive material NCL includes an adhesive material (insulating adhesive material) NCL 1 bonding and fixing the logic chip LC to the wiring substrate 2 , and an adhesive material (insulating adhesive material) NCL 2 bonding and fixing a laminated body MCS with the memory chips MC 1 , MC 2 , MC 3 , and MC 4 to over the logic chip.
- the adhesive materials NCL 1 and NCL 2 are made of insulating (non-conductive) material (for example, resin).
- the adhesive material NCL is disposed at a bonding portion between the logic chip LC and the wiring substrate 2 , and another bonding portion between the laminated body MCS and the logic chip LC, which can electrically insulate the electrodes provided in the respective bonding portions from each other.
- a seal member (seal member for the chip laminated body, resin for the chip laminated body) 6 other than the seal member 4 is arranged between the memory chips MC 1 , MC 2 , MC 3 , and MC 4 , so that the laminated body MCS of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 are sealed with the seal member 6 .
- the seal member 6 is embedded in intimate contact with the front surface 3 a and the back surface 3 b of each of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 .
- the laminated body MCS of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 is integral with each other by the bonding portion and the seal member 6 between the respective semiconductor chips 3 .
- the seal member 6 is made of insulating (non-conductive) material (for example, resin).
- the seal member 6 is arranged in the respective bonding portions between the memory chips MC 1 , MC 2 , MC 3 , and MC 4 , so that the electrodes provided in the bonding portions can be electrically insulated from each other. As shown in FIG.
- the surface 3 a of the memory chip MC 1 mounted at the lowermost level (in the position closest to the logic chip LC) in the laminated body MCS of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 is exposed from the seal member 6 .
- the back surface 3 b of the memory chip MC 4 mounted at the uppermost level in the laminated body MCS of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 is exposed from the seal member 6 .
- the semiconductor device 1 includes the seal member 4 for sealing the semiconductor chips 3 .
- the seal member 4 has a quadrilateral contour in plan view having an upper surface (surface, front surface) 4 a , a lower surface (surface, back surface, mounting surface) 4 b opposite to the upper surface 4 a (see FIG. 4 ), and side surfaces 4 c positioned between the upper surface 4 a and the lower surface 4 b .
- the planar size of the seal member 4 (size in plan view from the upper surface 4 a side, contour size of the upper surface 4 a ) is the same as that of the wiring substrate 2 .
- the side surface 4 c of the seal member 4 continues into the side surface 2 c of the wiring substrate 2 .
- the seal member 4 has a square shape having a planar size (size in plan view) of, for example, about 14 mm on a side.
- the seal member 4 is a resin member for protecting the semiconductor chips 3 .
- the seal member 4 is formed between the semiconductor chips 3 and in intimate contact with the semiconductor chip 3 and the wiring substrate 2 , which can suppress the damage on the thin semiconductor chips 3 .
- the seal member 4 is preferably formed, for example, of the following material from the viewpoint of improving the function as the protective member.
- the seal member 4 is required to be easily brought into intimate contact with the semiconductor chip 3 , the wiring substrate 2 , and the space between the semiconductor chips 3 , and to have some hardness after sealing. For this reason, the seal member 4 preferably contains thermosetting resin, such as epoxy resin.
- filler particles such as silica (silicon dioxide: SiO 2 ) particles
- Silica silicon dioxide: SiO 2
- the ratio of addition of the filler particles to the mixture is preferably adjusted to make a linear expansion coefficient of the semiconductor chip 3 equal to that of the seal member 4 for the purpose of suppressing the damage on the semiconductor chip 3 due to thermal deformation of the seal member 4 and chip 3 after formation of the seal member 4 .
- the logic chip LC is provided with the control circuit CU for controlling the operation of the main storage circuit MM of each of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 , in addition to the above arithmetic processing circuit PU.
- the logic chip LC is provided with an auxiliary storage circuit (storage circuit) SM having a smaller storage capacity than that of the main storage circuit MM, such as a cash memory for preliminary storing data.
- the arithmetic processing circuit PU, the control circuit CU, and the auxiliary storage circuit SM are collectively called a core circuit (main circuit) CR 1 by way of example.
- the circuits included in the core circuit CR 1 may include circuits other than the circuit described above.
- the logic chip LC is provided with an external interface circuit (external input/output circuit) GIF for inputting and outputting a signal to and from an external device (not shown).
- the external interface circuit GIF is connected with a signal line SG for transmitting a signal between the logic chip LC and the external device (not shown).
- the external interface circuit GIF is electrically connected with a core circuit CR 1 , and the core circuit CR 1 can transfer a signal to the external device via the external interface circuit GIF.
- the logic chip LC is provided with an internal interface circuit (internal input/output circuit) NIF for inputting and outputting a signal to and from internal devices (for example, memory chips MC 1 , MC 2 , MC 3 , and MC 4 ).
- the internal interface circuit NIF is connected with a data line (signal line) DS for transmitting a data signal, an address line (signal line) AS for transmitting an address signal, and a signal line OS for transmitting other signals.
- the date line DS, the address line AS, and the signal line OS are respectively connected with the internal interface circuit NIF of each of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 .
- circuits for inputting/outputting a signal to and from electronic parts other than the logic chip LC such as an external interface circuit GIF or an internal interface circuit NIF, are designated by an input/output circuit NS 1 .
- the logic chip LC includes a power supply circuit DR for supplying a potential for driving a core circuit CR 1 and the input/output circuit NS 1 .
- the power supply circuit DR includes a power supply circuit (power supply circuit for input/output) DR 1 for supplying a voltage for driving the input/output circuit NS 1 of the logic chip LC, and a power supply circuit (power supply circuit for a core) DR 2 for supplying a voltage for driving the core circuit CR 1 of the logic chip LC.
- first power supply potential and second power supply potential are supplied to the power supply circuit DR, and a difference in potential defines each of voltages to be applied to the core circuit CR 1 and the input/output circuit NS 1 .
- SoC system on a chip
- the formation of the main storage circuit MM shown in FIG. 5 in the logic chip LC can form the system using only one logic chip LC.
- a necessary capacity of the main storage circuit MM (see FIG. 5 ) varies depending on the device or system to be operated. For this reason, the formation of the main storage circuit MM in the semiconductor chip 3 other than the logic chip LC can improve the general versatility of the logic chip LC.
- the memory chips MC 1 , MC 2 , MC 3 , and MC 4 are connected according to a storage capacity required for the main storage circuit MM, which improves the flexibility in designing the capacity of the storage circuit included in the system.
- each of memory chips MC 1 , MC 2 MC 3 , and MC 4 is provided with the corresponding main storage circuit MM.
- the main storage circuit MM is designated as a core circuit (main circuit) CR 2 of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 .
- the circuits included in the core circuit CR 2 may include circuits other than the main storage circuit MM.
- Each of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 is provided with an internal interface circuit (internal input/output circuit) NIF for inputting and outputting a signal to and from internal devices (for example, logic chip LC).
- an internal interface circuit NIF for inputting and outputting the signal to and from electronic devices other than the memory chips MC 1 , MC 2 , MC 3 , and MC 4 is designated as the input/output circuit NS 2 .
- Each of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 includes a power supply circuit (drive circuit) DR for supplying a potential for driving the core circuit CR 2 and the input/output circuit NS 2 .
- the power supply circuit DR includes a power supply circuit (power supply circuit for input/output) DR 3 for supplying a voltage for driving the input/output circuit NS 2 of each of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 , and a power supply circuit (power supply circuit for a core) DR 4 for supplying a voltage for driving the core circuit CR 2 of each of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 .
- first power supply potential and second power supply potential are supplied to the power supply circuit DR, and a difference in potential defines each of voltages to be applied to the core circuit CR 2 and the input/output circuit NS 2 .
- the power circuit DR 1 of the logic chip LC and the power circuit DR 3 of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 are shared.
- the same voltage supplied from the power wire V 2 is applied to the input/output circuit NS 1 of the logic chip LC and the input/output circuit NS 2 of each of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 to drive the input/output circuits.
- a part or all of the power supply circuits DR can be shared to thereby decrease the number of power wires V 1 , V 2 , and V 3 for supplying a potential (driving voltage) to the power supply circuit.
- the decrease in number of the power wires V 1 , V 2 , and V 3 can decrease the number of electrodes formed in the logic chip LC.
- SiP system in Package
- memory chips MC 1 , MC 2 , MC 3 , and MC 4 are stacked over one logic chip LC in the example shown in FIG. 4
- the number of stacked semiconductor chips 3 can be modified as mentioned above.
- the invention can be applied to a modified example as the minimum structure in which one memory chip MC 1 is mounted over one logic chip LC.
- the planar size (size in plan view, size of the front surface 3 a and back surface 3 b , or contour size) of each of the logic chip LC and the memory chips MC 1 , MC 2 , MC 3 , and MC 4 is preferably minimized as long as the function of each semiconductor chip 3 can be achieved.
- the logic chip LC can have its planar size reduced by improving the integration degree of the circuit elements.
- Each of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 changes the capacity and transmission rate (for example, the amount of data transmitted depending on the width of a data bus) of the main storage circuit MM according to the planar size of the chip, and thus is limited in reduction of the planar size.
- the planar size of the memory chip MC 4 is larger than that of the logic chip LC.
- the memory chip MC 4 has a quadrilateral planar shape having a size of about 8 to 10 mm on a side
- the logic chip LC has a quadrilateral planar shape having a size of about 5 to 6 mm on a side.
- the planar size of each of the memory chips MC 1 , MC 2 , and MC 3 shown in FIG. 4 is the same as that of the memory chip MC 4 .
- the logic chip LC is provided with an external interface circuit GIF for inputting and outputting signal to and from an external device (not shown).
- the order of stacking the semiconductor chips 3 is preferably set such that the logic chip LC is positioned at the lowermost level, or closest to the wiring substrate 2 . That is, the semiconductor chips 3 (memory chips MC 1 , MC 2 , MC 3 , MC 4 ) having a larger planar size are preferably stacked over the semiconductor chip 3 (logic chip LC) having a smaller planar size, like the semiconductor device 1 .
- FIG. 6 shows an enlarged cross-sectional view of a portion “A” shown in FIG. 4 .
- FIG. 7 shows a plan view of the front surface side of the memory chip shown in FIG. 4 .
- FIG. 8 shows a plan view of one example of the back surface side of the memory chip shown in FIG. 7 .
- FIG. 9 shows a plan view of the front surface side of the logic chip shown in FIG. 4 .
- FIG. 10 shows a plan view of one example of the back surface side of the logic chip shown in FIG. 9 .
- FIG. 8 shows the back surface diagram of the memory chips MC 1 , MC 2 , and MC 3 .
- the structure of the back surface of the memory chip MC 4 (see FIG. 4 ) without the back surface electrode 3 bp is shown in FIG. 3 , and thus its illustration will be omitted in the figure.
- the inventors of the present application have studied about techniques for improving the performance of the SiP semiconductor device.
- One of the techniques is being studied to increase a signal transmission rate between the semiconductor chips mounted on SiP up to, for example, 12 Gbps (12 gigabit) or more.
- Methods for increasing a transmission rate between semiconductor chips mounted on the SiP includes a method which involves increasing the amount of data transmitted at one time by increasing the width of the data bus at the internal interface (hereinafter referred to as a “bus width increase”).
- Another method involves increasing the number of transmission per unit time (herein referred to as a “clock frequency increase”).
- a combination of the above bus width increase and the clock frequency increase can be applied.
- the semiconductor device 1 described with reference to FIGS. 1 to 5 uses the combination of the bus width increase and the clock frequency increase to thereby increase the transmission rate up to 12 Gbps or more at the internal interface.
- each of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 shown in FIG. 4 is the so-called wide I/O memory having a width of a data bus of 512 bit.
- each of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 has a width of the date bus of 128 bit, so that the total bus width of the four channels is 512 bit.
- the number of transmission per unit time of each channel is increased to be, for example, 3 Gbps or more.
- the logic chip LC is electrically connected with the memory chip MC 1 via a conductive member arranged between the logic chip LC and the memory chip MC 1 .
- the memory chips MC 1 , MC 2 , MC 3 , and MC 4 are electrically connected together via the conductive members arranged between the adjacent memory chips MC 1 , MC 2 , MC 3 , and MC 4 .
- the semiconductor device 1 does not include the wiring substrate 2 and a wire (not shown) (bonding wire) in a transmission path between the logic chip LC and the memory chip MC 1 .
- the semiconductor device 1 does not include the wiring substrate 2 and a wire (not shown) (bonding wire) in transmission paths among the memory chips MC 1 , MC 2 , MC 3 , and MC 4 .
- the direct connection between the semiconductor chips 3 employs the technique which involves forming penetrating electrodes penetrating the semiconductor chip 3 in the thickness direction, and connecting the semiconductor chips 3 together stacked via the penetrating electrodes.
- the logic chip LC includes front surface electrodes (electrodes, pads, front surface side pads) 3 ap formed on its front surface 3 a , and back surface electrodes (electrodes, pads, back surface side pads) 3 bp formed on its back surface 3 b .
- the logic chip LC further includes penetrating electrodes 3 tsv penetrating from one of the front surface 3 a and the back surface 3 b to the other and electrically connecting the front surface electrodes 3 ap with the back surface electrodes 3 bp.
- the semiconductor chip 3 includes a semiconductor substrate (not shown) made of, for example, silicon (Si), and has a plurality of semiconductor elements (not shown), such as a transistor, formed on its main surface (element formation surface) of the semiconductor substrate.
- a wiring layer (not shown) including an insulating film for insulating the wirings from each other is stacked over the main surface of the semiconductor substrate (front surface 3 a side).
- a plurality of wirings of the wiring layers are respectively electrically connected with the semiconductor elements to form circuits.
- the front surface electrodes 3 ap formed on the front surfaces 3 a (see FIG. 4 ) of the semiconductor chips 3 are electrically connected with the semiconductor elements via the wiring layers provided between the semiconductor substrate and the front surface 3 a to thereby form parts of the circuits.
- each penetrating electrode 3 tsv is formed to penetrate the semiconductor chip 3 in the thickness direction, and the front surface electrode 3 ap is electrically connected with the back surface electrode 3 bp via the penetrating electrode 3 tsv , so that the back surface electrode 3 bp can be electrically connected with the circuit of the semiconductor chip 3 formed on the front surface 3 a side. That is, as shown in FIG.
- the front surface electrode 3 ap of the memory chip MC 1 is electrically connected with the back surface electrode 3 bp of the logic chip LC via the conductive member, such as an external terminal (protrusion electrode, conductive member, bump electrode) 7 , so that the circuit of the memory chip MC 1 is electrically connected with the circuit of the logic chip LC via the penetrating electrodes 3 tsv.
- the conductive member such as an external terminal (protrusion electrode, conductive member, bump electrode) 7 , so that the circuit of the memory chip MC 1 is electrically connected with the circuit of the logic chip LC via the penetrating electrodes 3 tsv.
- the logic chip LC mounted in between the memory chip MC 1 and the wiring substrate 2 includes a plurality of the penetrating electrodes 3 tsv .
- the memory chip MC 1 is electrically connected with the logic chip LC via the penetrating electrodes 3 tsv , which can exclude the wiring substrate 2 and a wire (bonding wire) (not shown) from a transmission path between the logic chip LC and the memory chip MC 1 .
- an impedance component in the transmission path between the logic chip LC and the memory chip MC 1 can be reduced to suppress the influence of noise caused by increase in clock frequency. In other words, even upon increasing the signal transmission rate between the logic chip LC and the memory chip MC 1 , the transmission reliability of the semiconductor device can be improved.
- the memory chips MC 1 , MC 2 , MC 3 , and MC 4 are stacked over the logic chip LC.
- the signal transmission rate among the memory chips MC 1 , MC 2 , MC 3 , and MC 4 is preferably improved or increased.
- the memory chips MC 1 , MC 2 , and MC 3 under and above which the respective semiconductor chips 3 are arranged, among the memory chips MC 1 , MC 2 , MC 3 , and MC 4 include a plurality of penetrating electrodes 3 tsv , like the logic chip LC.
- each of the memory chips MC 1 , MC 2 , and MC 3 includes front surface electrodes (electrodes, pads) 3 ap formed on its front surface 3 a , and back surface electrodes (electrodes, pads) 3 bp formed on its back surface 3 b .
- the respective memory chips MC 1 , MC 2 , and MC 3 include the penetrating electrodes 3 tsv penetrating from one of the front surface 3 a and the back surface 3 b to the other and electrically connecting the front surface electrodes 3 ap to the back surface electrodes 3 bp.
- the front surface electrodes 3 ap of the upper semiconductor chip 3 among the memory chips MC 1 , MC 2 , MC 3 , and MC 4 are electrically connected with the back surface electrode 3 bp of the lower semiconductor chip 3 via the conductive members, including external terminals 7 , so that the circuits of the stacked semiconductor chips 3 are electrically connected with each other via the penetrating electrodes 3 tsv.
- the respective semiconductor chips 3 are connected with each other via the external terminals 7 (solder material 7 a ), which can exclude the wiring substrate 2 and the wires (bonding wires) (not shown) from the transmission paths among the memory chips MC 1 , MC 2 , MC 3 , and MC 4 .
- an impedance component in the transmission path between the stacked memory chips MC 1 , MC 2 , MC 3 , and MC 4 can be reduced to suppress the influence of noise caused by increase in clock or clock frequency.
- the transmission reliability of the semiconductor device can be improved.
- the memory chip MC 4 mounted at the upperrmost level is preferably connected with the memory chip MC 3 , whereby the surface electrodes 3 ap are formed, but the back surface electrodes 3 bp and the penetrating electrodes 3 tsv are not necessarily formed.
- the memory chip MC 4 mounted at the uppermost level takes the structure without having the back electrodes 3 bp and the penetrating electrodes 3 tsv , which can simplify the manufacturing process of the memory chip MC 4 .
- the memory chip MC 4 can include the back surface electrodes 3 bp and the penetrating electrodes 3 tsv , like the memory chips MC 1 , MC 2 , and MC 3 .
- the stacked memory chips MC 1 , MC 2 , MC 3 , and MC 4 have the same structure, which can improve the manufacturing efficiency.
- Each external terminal 7 is arranged between the adjacent stacked semiconductor chips 3 , and serves to electrically connect the front surface electrode 3 ap of the semiconductor chip 3 positioned at the upper level with the back surface electrode 3 bp of the semiconductor chip 3 positioned at the lower level.
- the external terminal 7 is made of the following material in the example shown in FIG. 6 . That is, the external terminal 7 electrically connecting the logic chip LC with the wiring substrate 2 includes a columnar (for example, cylindrical) member (protruding electrode 7 b ) mainly formed of copper (Cu), and a metal member formed of a nickel (Ni) film and a solder (for example, SnAg) film (solder material 7 a ) stacked on the tip of the columnar member. The electric connection of the external terminal 7 is established by bonding the solder film at its tip to the back surface electrode 3 bp.
- a columnar (for example, cylindrical) member protruding electrode 7 b ) mainly formed of copper (Cu)
- the semiconductor chips are connected together via the solder material 7 a without using the protruding electrode 7 b except for the bonding portion between the logic chip LC and the wiring substrate 2 .
- the solder material 7 a is bonded to the exposed surface of the front surface electrode 3 ap , and can serve as the external terminal (bump electrode called a microbump).
- the material for the external terminal 7 can be changed within the requirements of electrical characteristics, or bonding strength.
- the protruding electrodes 7 b are formed on the front surface electrode 3 ap of each of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 , so that the chips can be electrically connected together via the protruding electrode 7 b and the solder material 7 a .
- the solder material 7 a is directly bonded to the exposed surface of the front surface electrode 3 ap of the logic chip LC, so that the logic chip can be electrically connected with the wiring substrate 2 via the solder material 7 a.
- the semiconductor chip 3 including the penetrating electrodes 3 tsv is preferably thin, that is, the distance between the front surface 3 a and the back surface 3 b is preferably thin (small).
- the transmission distance of the penetrating electrode 3 tsv is shortened, which is preferable in reducing the impedance component.
- the processing accuracy of the opening for forming the penetrating electrodes 3 tsv can be improved.
- the diameters of the penetrating electrodes 3 tsv can be easily set to the same value (length or width of the semiconductor chip 3 in the direction perpendicular to the thickness direction), which facilitates the control of the impedance components of the transmission paths.
- the thickness of the logic chip LC is smaller than that of the laminated body MCS (see FIG. 4 ) formed of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 arranged over the logic chip LC.
- the thickness of the logic chip LC is also smaller than that of the memory chip MC 4 located at the uppermost level of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 , and not having the penetrating electrode 3 tsv .
- the thickness of the logic chip LC is 50 ⁇ m.
- the thickness of the memory chip MC 4 is in a range of about 80 to 100 ⁇ m.
- the thickness of the laminated body MCS (see FIG. 4 ) made of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 is about 260 ⁇ m.
- the seal member 4 is brought into intimate contact with the semiconductor chips 3 to seal the chips therewith.
- the seal member 4 can serve as a protection member for the semiconductor chips 3 to suppress the damage on the semiconductor chip 3 . That is, according to this embodiment, the semiconductor chips 3 can be sealed with resin to thereby improve the reliability (durability) of the semiconductor device 1 .
- the distance between the semiconductor chip 3 and the wiring substrate 2 is preferably narrow from the viewpoint of reducing the transmission distance.
- a distance between the front surface 3 a of the logic chip LC and an upper surface 2 a of the wiring substrate 2 is, for example, in a range of about 10 to 20 ⁇ m.
- the distance between the front surface 3 a of the memory chip MC 1 and the upper surface 2 a of the wiring substrate 2 is, for example, in a range of about 70 to 100 ⁇ m.
- the transmission distance is reduced by decreasing the distance between the semiconductor chips 3 and the thickness of each semiconductor chip 3 .
- the transmission distance between the logic chip LC and each of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 can be reduced.
- the front surface electrodes 3 ap included in the memory chips MC 1 , MC 2 , MC 3 , and MC 4 are collectively arranged in the center of the front surface 3 a .
- the back surface electrodes 3 bp included in the memory chips MC 1 , MC 2 , and MC 3 are collectively arranged in the center of each back surface 3 b .
- FIG. 8 shows that the front surface electrodes 3 ap included in the memory chips MC 1 , MC 2 , and MC 3 are collectively arranged in the center of each back surface 3 b .
- the front surface electrodes 3 ap of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 , and the back surface electrodes 3 bp of the memory chips MC 1 , MC 2 , and MC 3 are arranged to be superimposed over each other in the thickness direction.
- some parts (front surface electrodes 3 ap 1 ) of the front surface electrodes 3 ap included in the logic chip LC are collectively arranged in the center of the front surface 3 a .
- Other parts (front surface electrodes 3 ap 2 ) of the front surface electrodes 3 ap included in the logic chip LC are collectively arranged along each side (side surface 3 c ) of the front surface 3 a located at the peripheral edges of the front surface 3 a .
- the surface electrodes 3 ap 1 arranged in the center of the front surface 3 a among the front surface electrodes 3 ap shown in FIG. 9 are electrically connected with the back surface electrodes 3 bp via the penetrating electrodes 3 tsv shown in FIG.
- the front surface electrodes 3 ap 1 are electrodes for internal interface.
- the surface electrodes 3 ap 2 arranged at the peripheral edges of the front surface 3 a among the front surface electrodes 3 ap shown in FIG. 9 are electrically connected with the external device (not shown) via the wiring substrate 2 shown in FIG. 4 .
- the front surface electrode 3 ap 2 is electrically coupled to the bonding leads 2 f (see FIG. 4 ) via the external terminals 7 . That is, the front surface electrodes 3 ap 2 are electrodes for external interface.
- the front surface electrodes 3 ap and back surface electrodes 3 bp for the internal interface are superimposed to each other in the thickness direction, and connected together via the external terminals 7 .
- the planar size of the logic chip LC is smaller than that of each of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 .
- the center (central region) of the back surface 3 b of the logic chip LC is arranged to be superimposed over the center (central region) of the Memory chip MC 4 in plan view. That is, in plan view, four side surfaces 3 c of the memory chip MC 4 are arranged outside four side surfaces 3 c of the logic chip LC.
- the semiconductor chips 3 are stacked over the wiring substrate 2 such that the four side surfaces 3 c of the memory chip MC 4 are positioned between the four side surfaces 3 c of the logic chip LC and the four side surfaces 2 c of the wiring substrate 2 .
- the memory chips MC 1 , MC 2 , and MC 3 shown in FIG. 4 are arranged to be superimposed over (in the same position) as the memory chip MC 4 in plan view.
- peripheral edges (peripheral edges of the front surface 3 a and the back surface 3 b ) of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 are arranged to be superimposed over the outer periphery of the logic chip LC.
- the front surface electrodes Sap and the back surface electrodes 3 bp for internal interface of each semiconductor chip 3 shown in FIG. 6 are arranged such that the surface electrodes 3 ap and 3 bp are superimposed over each other in the thickness direction
- at least the front surface electrodes 3 ap and back surface electrodes 3 bp for the internal interface are preferably arranged to be superimposed over the logic chip LC in the thickness direction.
- the front surface electrodes 3 ap 2 for the external interface are arranged in the periphery of the logic chip LC.
- the front surface electrodes 3 ap 1 for the internal interface are preferably arranged collectively in the center of the front surface 3 a of the logic chip LC.
- a plurality of memory regions (storage circuit element arrangement regions) MR are formed on the front surface 3 a side (specifically, main surface of the semiconductor substrate) of each of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 .
- the four memory regions MR corresponding to the above four channels are formed.
- a plurality of memory cells (storage circuit elements) are arranged in an array.
- the front surface electrodes 3 ap are collectively arranged in the center of the front surface 3 a , the memory regions MR for the four channels can be arranged to enclose a region with a group of the front surface electrodes arranged.
- the distances from the respective memory regions MR to the front surface electrode 3 ap can be equalized. That is, preferably, the transmission distances of the channels can be made equal, which can decrease an error between the respective transmission rates for the channels.
- the front surface electrode 3 ap 1 When using the front surface electrodes 3 ap 1 collectively arranged in the center of the front surface 3 a of the logic chip LC shown in FIG. 9 as electrodes for the internal interface, the front surface electrode 3 ap 1 can act without being electrically connected with the wiring substrate 2 shown in FIG. 6 . As shown in FIG. 6 , a part of the front surface electrode 3 ap 1 is electrically connected with the bonding lead 2 f of the wiring board 2 , whereby the part of the front surface electrode 3 ap 1 can be preferably used as the electrode for the external interface.
- each of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 is provided with the power supply circuit DR for driving the main storage circuit MM shown in FIG. 5 .
- the part of the front surface electrode 3 ap 1 shown in FIG. 9 is used as a terminal for supplying the power supply potential (first reference potential) and the reference potential (second reference potential different from the first reference potential, for example, ground potential) to the power supply circuits DR.
- the surface electrodes 3 ap 1 arranged in the center, of the front surface 3 a of the logic chip LC include a first reference potential electrode to which the first reference potential (for example, power supply potential) is supplied, and a second reference potential electrode to which the second reference potential (for example, ground potential) different from the first reference potential is supplied.
- the front surface electrodes 3 ap 1 arranged in the center of the front surface 3 a of the logic chip LC include power wires V 2 and V 3 (see FIG. 5 ) for supplying a voltage for driving a circuit formed in the memory chip MC 1 .
- the transmission distance between a power supply source and a circuit consuming the power is preferably shortened.
- the parts of the front surface electrodes 3 ap 1 of the logic chip LC are electrically connected with the wiring board 2 to supply the first reference potential (for example, power supply potential) and the second reference potential (for example, ground potential), which can shorten the distance from each of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 with the circuit consuming the power to the corresponding driving circuit.
- the first reference potential (for example, power supply potential) is supplied to the first reference potential electrode.
- the second reference potential (for example, ground potential) different from the first reference potential is supplied to the second reference potential electrode.
- the first reference potential electrode and the second reference potential electrode are arranged such that the front surface electrode 3 ap is preferably superimposed over the back surface electrode 3 bp in the thickness direction as shown in FIG. 6 .
- the first reference potential electrode is preferably electrically connected with the second reference potential electrode via the penetrating electrodes 3 tsv.
- the inventors of the present application have studied and found that when the back surface electrode 3 bp of the lower semiconductor chip 3 faces and is electrically connected with the front surface electrode 3 ap of the upper semiconductor chip 3 via the external terminals 7 , the stacking of the semiconductor chips 3 requires the high alignment accuracy.
- the back surface electrodes 3 bp need to be superimposed over the front surface electrodes 3 ap in the thickness direction.
- the bonding area between the solder material 7 a and the back surface electrode 3 bp (or front surface electrode 3 ap ) becomes small, which might reduce the electrical characteristics.
- the number of the back surface electrodes 3 bp of the logic chip LC is large, the number of the objects of interest requiring the adequate alignment accuracy is increased, which needs the high alignment accuracy.
- a pitch between the back surface electrodes 3 bp of the logic chip LC becomes small, one electrode might come into contact with adjacent another electrode due to the misalignment. In this case, the alignment needs to be performed with higher accuracy.
- FIG. 11 is an exemplary explanatory diagram showing alignment means (alignment marks) used in assembling the semiconductor device shown in FIG. 4 .
- FIG. 12 is an explanatory diagram showing a studied example corresponding to FIG. 11 .
- an alignment mark 50 a formed in the wiring substrate 2 and an alignment mark 50 b formed in the front surface 3 a of the logic chip LC are detected (identified), and then aligned with each other.
- the logic chip LC is mounted over the wiring substrate 2 .
- the alignment mark 50 a formed in the wiring substrate 2 and an alignment mark 50 d formed in the front surface 3 a of the laminated body MCS are detected (identified), and then aligned with each other.
- the laminated body MCS is mounted over the logic chip LC.
- the alignment accuracy of the logic chip LC with respect to the wiring substrate 2 is defined mainly by a detection accuracy for detecting the positions of the alignment marks 50 a and 50 b , and a conveyance accuracy in mounting the logic chip LC.
- a detection accuracy for detecting the positions of the alignment marks 50 a and 50 b is 60
- the alignment accuracy of the logic chip LC with respect to the wiring substrate 2 is set within a tolerance of about ⁇ 5 to 10 ⁇ m, which can sufficiently ensure the electric connection reliability between the logic chip LC and the wiring substrate 2 .
- the alignment accuracy of the laminated body MCS with respect to the wiring substrate 2 is defined mainly by a detection accuracy for detecting the positions of the alignment marks 50 a and 50 d , and a conveyance accuracy in mounting the laminated body MCS.
- the use of the same detection device and mounting device as those in the step of mounting the above logic chip LC can set the alignment accuracy of the laminated body MCS with respect to the wiring substrate 2 within about a tolerance of ⁇ 5 to 10 ⁇ m.
- the alignment accuracy of the laminated body MCS with respect to the logic chip LC might be in a range of ⁇ 10 to 20 ⁇ m, taking into consideration the misalignment between the laminated body MCS and the logic chip LC.
- the alignment accuracy between the back surface electrode 3 bp of the logic chip LC and the front surface electrode 3 ap of the laminated body MCS needs to be determined by considering the formation positional accuracy of the respective electrodes.
- the alignment accuracy might be reduced. Even when two semiconductor chips 3 are stacked as shown in FIG. 12 , the reliability of electrical characteristics might be obviously reduced depending on a pitch between the electrodes.
- the structure shown in FIG. 11 includes alignment marks 50 c formed on the back surface 3 b of the logic chip LC, in addition to the structure shown in FIG. 12 .
- an alignment method shown in FIG. 11 first, an alignment mark 50 a formed in the wiring substrate 2 and an alignment mark 50 b formed in the front surface 3 a of the logic chip LC are detected (identified), and then aligned with each other. Thereafter, the logic chip LC is mounted over the wiring substrate 2 . This point is the same as that in the alignment method shown in FIG. 12 .
- the alignment mark 50 c formed in the logic chip LC and the alignment mark 50 d formed in the front surface 3 a of the laminated body MCS are detected (identified), and then aligned with each other. Thereafter, the laminated body MCS is mounted over the logic chip LC. That is, the alignment method shown in FIG. 11 differs from the alignment method shown in FIG. 12 in that the alignment is performed based on the alignment mark 50 c formed in the logic chip LC.
- the alignment mark 50 c formed on the back surface 3 b of the logic chip LC is used as the reference for alignment, so that the alignment accuracy of the laminated body MCS with respect to the logic chip LC does not need to take into consideration the misalignment of the logic chip LC.
- the use of the same detection device and mounting device as those in the above example can achieve the alignment accuracy within a tolerance of about ⁇ 5 to 10 ⁇ m.
- the alignment method shown in FIG. 11 can prevent the decrease in alignment accuracy even when the number of stacked semiconductor chips 3 is increased.
- the alignment accuracy of the upper semiconductor chip 3 with respect to the lower semiconductor chip 3 is not affected by the number of stacked semiconductor chips 3 . Even when the pitch between the electrodes is very small, the adequate alignment accuracy can be ensured according to the capability of a detector or mounting device, which can suppress the reduction in reliability of electrical characteristics between the semiconductor chips 3 .
- FIG. 3 and FIGS. 7 to 10 show an example of the layout of alignment marks 50 included in the semiconductor device 1 of the first embodiment.
- the detailed structure of the alignment mark and preferred embodiments will be described in more detail in describing the manufacturing method of the semiconductor device 1 .
- FIG. 13 shows an exemplary explanatory diagram of the outline of the manufacturing process of the semiconductor device described with reference to FIGS. 1 to 11 .
- the details of the respective steps will be described below with reference to FIGS. 14 to 45 .
- FIG. 14 shows a plan view of the entire structure of a wiring substrate provided in the substrate provision step shown in FIG. 13 .
- FIG. 15 shows an enlarged plan view of one device region shown in FIG. 14 .
- FIG. 16 shows an enlarged cross-sectional view taken along the line A-A of FIG. 15 .
- FIG. 17 shows an enlarged plan view of the surface opposite to that shown in FIG. 15 .
- FIG. 18 shows an enlarged cross-sectional view taken along the line B-B of FIG. 15 .
- the number of terminals is shown to be small in FIGS. 14 to 17 , the number of terminals (bonding lands 2 f and lands 2 g ) is not limited to those shown in FIGS. 14 to 17 .
- the wiring substrate 20 provided in this step includes device regions 20 a inside a frame (outer frame) 20 b .
- a plurality of device regions 20 a are arranged in columns and rows.
- Each of the device regions 20 a corresponds to the wiring substrate 2 shown in FIGS. 1 to 4 .
- the wiring substrate 20 is the so-called multi-piece substrate including a plurality of device regions 20 a , and dicing lines (dicing regions) 20 c positioned between the device regions 20 a . In this way, the use of the multi-piece substrate including the device regions 20 a can improve the manufacturing efficiency.
- each device region 20 a is provided with the components of the wiring substrate 2 described with reference to FIG. 4 .
- the wiring substrate 20 has an upper surface 2 a , a lower surface 2 b opposite to the upper surface 2 a , and a plurality of wiring layers (four layers in the example shown in FIG. 4 ) electrically connecting the upper surface 2 a with the lower surface 2 b .
- the respective wiring layers have insulating layers (core layers) 2 e for insulating the adjacent wirings 2 d as well as the adjacent wiring layers from each other.
- the wirings 2 d include wirings 2 d 1 formed on the upper or lower surface of the insulating layer 2 e , and via wirings 2 d 2 serving as an interlayer conductive path penetrating the insulating layer 2 e in the thickness direction.
- the upper surface 2 a of the wiring substrate 20 includes a chip mounting region (chip mounting portion) 2 p 1 where the logic chip LC shown in FIG. 9 is to be mounted in a first chip mounting step shown in FIG. 13 .
- the chip mounting region 2 p 1 exists in the center of the device region 20 a at the upper surface 2 a .
- the contours of the chip mounting region 2 p 1 , the device region 20 a , and the dicing lines 20 c are represented by respective two-dot chain lines.
- the chip mounting region 2 p 1 is a region where the logic chip LC is to be mounted as mentioned above, and thus does not necessarily include any boundary line actually viewable. Also, the device region 20 a and the dicing lines 20 c do not necessarily include any boundary lines actually viewable.
- the upper surface 2 a of the wiring substrate 20 is provided with a plurality of bonding leads (terminals, chip mounting surface side terminals, electrodes) 2 f .
- the bonding lead 2 f is a terminal electrically connected with the front surface electrodes 3 ap formed on the front surface 3 a of the logic chip LC shown in FIG. 9 in the first chip mounting step as shown in FIG. 13 .
- the logic chip LC is mounted with the front surface 3 a side of the chip LC opposite to the upper surface 2 a of the wiring substrate 20 by the so-called face down mounting method, whereby the bonding portions of the bonding leads 2 f are formed within the chip mounting region 2 p 1 .
- the upper surface 2 a of the wiring substrate 20 is covered with the insulating film (solder resist films) 2 h .
- the insulating film 2 h is provided with openings 2 hw , and in the openings 2 hw , at least parts (bonding portions or bonding regions with the semiconductor chip) of the bonding leads 2 f are exposed from the insulating film 2 h .
- an upper surface of the dicing line (dicing region) 20 c is also exposed from the insulting film 2 h .
- a plurality of lands 2 g are formed on the lower surface 2 b of the wiring substrate 20 .
- the lower surface 2 b of the wiring substrate 20 is covered with the insulating film (solder resist film) 2 k .
- the insulating film 2 k is provided with openings 2 kw , and at the openings 2 kw , at least parts (bonding portions with the solder balls 5 ) of the lands 2 g are exposed from the insulating film 2 k .
- a lower surface of the dicing line (dicing region) 20 c is also exposed from the insulting film 2 k .
- the insulating film 2 h can be prevented from being stuck in the dicing blade. That is, the cutting performance can be prevented from being reduced. If the cutting performance is not taken into consideration, the upper surface of the dicing line 20 c may be covered with the insulating film 2 k.
- the bonding leads 2 f are electrically connected with the lands 2 g via the wirings 2 d .
- These conductive patterns including the wirings 2 d , the bonding leads 2 f , and the lands 2 g , are formed of metal, for example, mainly containing copper (Cu)
- the wirings 2 d , the bonding leads 2 f , and the lands 2 g can be formed by the electrolytic plating.
- the wiring substrate 20 including four or more wiring layers (four layers in the example shown in FIG. 16 ) can be formed, for example, by the buildup technique.
- alignment marks 50 a are formed on the upper surface 2 a of the wiring substrate 20 .
- the alignment mark 50 a is made of a pattern of interest to be detected in the first mark detection step shown in FIG. 13 .
- the alignment mark 50 a is formed of the same metal as that of the wiring 2 d and bonding lead 2 f shown in FIG. 16 .
- an opening 2 hs is formed in the insulating film 2 h , and the alignment mark 50 a is exposed from the insulating film 2 h in the opening 2 hs.
- the position of the alignment mark 50 a is detected using an image sensor, such as a camera.
- the metal pattern exposed from the insulating film 2 h is used as the alignment mark 50 a , thereby improving the light reflection efficiency, which leads to the improvement of the position detection accuracy.
- the shape of the contour of the alignment mark 50 a may be identifiable.
- the alignment mark 50 a may be covered with the insulating film 2 h.
- the alignment mark 50 a can be formed at one time with the bonding leads 2 f and the wirings 2 d.
- the alignment mark 50 a is preferably formed within the device region 20 a .
- the alignment mark 50 a can be formed outside, the device region 20 a , that is, within the dicing line (dicing region) 20 c , or at a frame 20 b shown in FIG. 14 .
- the alignment mark 50 a is used for aligning the logic chip LC with the wiring substrate 2 shown in FIG. 11 .
- the alignment mark 50 a is preferably formed in a region near the chip mounting region 2 p 1 within the device region 20 a from the viewpoint of improving the positional accuracy in formation of the alignment mark 50 a .
- the alignment mark 50 a when the alignment mark 50 a is arranged within the dicing line (dicing region) 20 c , the alignment mark 50 a can hardly be covered with the adhesive material in the first adhesive material arrangement step to be described later.
- the alignment mark 50 a is preferably arranged within the dicing line 20 c.
- the alignment marks 50 a are preferably formed in positions within the device region 20 a . Even only one alignment mark 50 a can have its X and Y coordinates defined on a coordinate plane along the upper surface 2 a However, the formation of at least two or more alignment marks 50 a can specify the position of the device region 20 a in the direction having an angle ⁇ , thereby improving the position detection accuracy.
- the distance between two alignment marks 50 a is preferably increased.
- one alignment mark 50 a is arranged at one corner of the device region 20 a
- the other alignment mark 50 a is arranged at the other corner of the device region 20 a facing the above corner.
- two alignment marks 50 a are preferably arranged on a diagonal line of the device region 20 a.
- FIG. 19 shows an enlarged plan view of the state of an adhesive material arranged in the chip mounting region shown in FIG. 15 .
- FIG. 20 is an enlarged cross-sectional view taken along the line A-A of FIG. 19 .
- each of the chip mounting regions 2 p 1 and 2 p 2 is a region where the logic chip LC and the laminated body MCS are to be mounted as mentioned above, and thus does not necessarily include any boundary line actually viewable.
- the device region 20 a and the dicing lines 20 c do not necessarily include any boundary lines actually viewable.
- these regions and lines do not necessarily include any boundary line actually viewable.
- the semiconductor chip when mounting the semiconductor chip over the wiring substrate by the face down mounting method (flip-chip bonding), the semiconductor chip is electrically connected with the wiring substrate, and then a connection part is sealed with resin (by a post-charging method).
- the resin is supplied from a nozzle positioned near a clearance between the semiconductor chip and the wiring substrate, so that the resin fills in the clearance by capillary action.
- the adhesive material NCL 1 is disposed over the chip mounting region 2 p 1 , and the logic chip LC is pressed against and electrically connected with the wiring substrate 20 from above the adhesive material NCL 1 (a pre-coating method) to thereby mount the logic chip LC.
- the resin is embedded in the clearance by the capillary action, which takes a long processing time (time for charging the resin) to one device region 20 a .
- the adhesive material NCL 1 already fills in the clearance between the wiring substrate 20 and the logic chip LC.
- the pre-coating method is preferable in reducing the processing time for one device region 20 a and improving the manufacturing efficiency, as compared to the above post-charging method.
- the order of the first chip mounting step and the first adhesive material arrangement step shown in FIG. 13 is reversed, so that the post-charging method can be applied.
- the use of the post-charging method does not differ so much from the pre-coating method in processing time, and thus can suppress the reduction in manufacturing efficiency.
- the adhesive material NCL 1 used in the pre-coating method is made of insulating (non-conductive) material (for example, resin) as described above.
- the adhesive material NCL 1 is made of resin material that is cured (increases its hardness) by being subjected to energy.
- the adhesive material NCL 1 contains thermosetting resin.
- the adhesive material NCL 1 before being cured is softer than the external terminal 7 shown in FIG. 6 , and can be deformed by being pressed by the logic chip LC.
- the adhesive materials NCL 1 before being cured can be classified into two types based on the difference in handling.
- One of the adhesive materials is a paste-like resin (insulating paste) called a non-conductive paste (NCP).
- NCP non-conductive paste
- the resin discharged from the nozzle (not shown) is applied to the chip mounting region 2 p 1 .
- the other of the adhesive materials is a film-like resin (insulating film) previously molded in the form of film, and which is called a non-conductive film (NCF).
- the film-like resin is conveyed and bonded to the chip mounting region 2 p 1 while being in the state of film.
- the use of the insulating paste (NCP) does not need the step of bonding the insulating film (NCF), and thus can reduce stress on the semiconductor chips or the like as compared to the use of the insulating material film.
- the use of the insulating film (NCF) has the high shape retaining property as compared to the use of the insulating paste (NCP), which makes it easier to control the range and thickness of the adhesive material NCL 1 .
- the adhesive material NCL 1 made of the insulating film (NCF) is arranged at the chip mounting region 2 p 1 , and bonded in intimate contact with the upper surface 2 a of the wiring substrate 20 .
- the insulating paste (NCP) can also be used.
- the adhesive material NCL 1 is preferably arranged not to cover the alignment marks 50 a from the viewpoint of improving the light reflection efficiency of the alignment mark 50 a .
- the adhesive material NCL 1 is made of the material that transmits visible light
- the alignment marks 50 a may be covered with the adhesive material NCL 1 . In this case, the flexibility in selecting the material for the adhesive material NCL 1 is reduced.
- the alignment marks 50 a are preferably exposed from the adhesive material NCL 1 .
- the alignment marks 50 a are exposed from the adhesive material NCL 1 , the alignment marks 50 a are preferably positioned in the periphery of the device region 20 a.
- FIG. 21 shows an exemplary explanatory diagram of the outline of a manufacturing process of the semiconductor chip with the penetrating electrode shown in FIG. 6 .
- FIG. 22 shows an exemplary explanatory diagram of the outline of another manufacturing process of the semiconductor chip, following the process shown in FIG. 21 .
- a method for manufacturing the penetrating electrode 3 tsv and the back surface electrode 3 bp electrically connected with the penetrating electrode 3 tsv will be mainly described below. Illustration and description of formation steps of various circuits other than the penetrating electrode 3 tsv will be omitted.
- the manufacturing method of the semiconductor chip shown in FIGS. 21 and 22 can also be applied to the manufacturing method of the memory chips MC 1 , MC 2 , and MC 3 in addition to the logic chip LC shown in FIG. 4 .
- a wafer (semiconductor substrate) WH shown in FIG. 21 is provided.
- the wafer WH is a semiconductor substrate made of, for example, silicon (Si), and has a circular shape in plan view.
- the wafer WH has a front surface (main surface, upper surface) WHs as a semiconductor element formation surface, and a back surface (main surface, lower surface) WHb opposite to the front surface WHs.
- the thickness of the wafer WH is larger than that of each of the logic chip LC and the memory chips MC 1 , MC 2 , and MC 3 shown in FIG. 4 , and is, for example, about several hundreds of ⁇ m.
- holes (bores, openings) 3 tsh for forming the penetrating electrodes 3 tsv shown in FIG. 6 are formed.
- a mask 25 is arranged over the front surface WHs of the wafer WH, and subjected to etching to thereby form the holes 3 tsh .
- the semiconductor elements, including the logic chip LC, and the memory chips MC 1 , MC 2 , and MC 3 shown in FIG. 4 can be formed, for example, after the present step and before a wiring layer formation step.
- a wiring layer (chip wiring layer) 3 d is formed over the front surface WHs of the wafer WH.
- a plurality of front surface electrodes 3 ap shown in FIGS. 7 and 9 are formed, whereby the penetrating electrodes 3 tsv are respectively electrically connected with the front surface electrodes 3 ap .
- the front surface electrodes 3 ap and the uppermost wiring layer 3 d integrally formed with the front surface electrode 3 ap are formed of a metal film of, for example, aluminum (Al).
- the semiconductor elements including the logic chip LC and the memory chips MC 1 , MC 2 , and MC 3 shown in FIG. 4 , are electrically connected with the front surface electrodes 3 ap shown in FIGS. 7 and 9 via the wiring layer 3 d .
- the electric connection of the semiconductor elements including the logic chip LC and the memory chips MC 1 , MC 2 , and MC 3 , is established via the wiring layer 3 d.
- the alignment mark 50 b shown in FIG. 9 or the alignment mark 50 d shown in FIG. 7 is formed.
- the alignment marks 50 b and 50 d can be formed of the same material (for example, aluminum) as that of the front surface electrode 3 ap and the uppermost wiring layer 3 d .
- the marks, front surface electrodes, and uppermost wiring layer can be formed at one time. The details of the alignment mark 50 b shown in FIG. 9 and the alignment mark 50 d shown in FIG. 7 will be described later.
- external terminals 7 are formed over the front surface electrodes 3 ap (see FIGS. 7 and 9 ).
- the protruding electrodes 7 b are formed over the front surface electrodes 3 ap of the logic chip LC.
- the solder material 7 a is formed at the tip of each protruding electrode 7 b .
- the solder material 7 a is formed over each front surface electrode 3 ap of the memory chip MC 1 .
- the solder material 7 a serves as a bonding material in mounting the semiconductor chip 3 shown in FIG. 6 over the wiring substrate 2 or another lower semiconductor chip 3 .
- the back surface WHb (see FIG. 21 ) of the wafer WH is polished to decrease the thickness of the wafer WH.
- the back surface 3 b of the semiconductor chip 3 shown in FIG. 5 is exposed.
- the penetrating electrode 3 tsv penetrates the wafer WH in the thickness direction.
- the penetrating electrodes 3 tsv are exposed from the wafer WH at the back surface 3 b of the wafer WH.
- the back surface of the wafer WH is polished by a polishing jig 28 while the wafer WH is being supported by a support base 26 , such as a glass plate, and a protective layer 27 for protecting the external terminal 7 protecting the front surface WHs side.
- a back surface electrode formation step a plurality of back surface electrodes 3 bp are formed on the back surface 3 b to be electrically connected with the penetrating electrodes 3 tsv.
- the alignment marks 50 c shown in FIG. 10 are formed.
- the alignment marks 50 c can be formed of the same material (for example, copper) as that of the back surface electrode 3 bp .
- the alignment marks 50 c can be formed at one time in forming the back surface electrodes 3 bp . The details of the alignment mark 50 c shown in FIG. 10 will be described later.
- the wafer WH is divided along the dicing lines to produce a plurality of semiconductor chips 3 . Thereafter, each semiconductor chip is checked if necessary, whereby the semiconductor chips 3 (logic chip LC, and memory chips MC 1 , MC 2 , and MC 3 ) shown in FIG. 4 are obtained.
- a first chip mounting step shown in FIG. 13 as shown in FIGS. 23 and 24 , the logic chip LC is mounted over the wiring substrate 20 .
- the first chip conveying step, the first mark detection step, and the first alignment step shown in FIG. 13 can be regarded as sub-steps included in the step of mounting the logic chip LC over the wiring substrate 20 .
- the first chip conveying step, the first mark detection step, and the first alignment step are described below as the sub-steps included in the first chip mounting step.
- FIG. 23 shows an enlarged plan view of the state of a logic chip LC mounted over the chip mounting region of the wiring substrate shown in FIG. 19 .
- FIG. 24 shows an enlarged cross-sectional view taken along the line A-A of FIG. 23 .
- FIG. 25 shows an exemplary explanatory diagram of a main part of a first chip conveying step shown in FIG. 13 .
- FIG. 26 shows an exemplary explanatory diagram of a main part of a first mark detection step shown in FIG. 13 .
- FIG. 27 shows an enlarged cross-sectional view of a main part of a cross-sectional structure of alignment marks of the logic chip shown in FIG. 26 .
- FIG. 28 shows an exemplary explanatory diagram of a main part of a first alignment step shown in FIG. 13 .
- FIG. 13 shows an enlarged plan view of the state of a logic chip LC mounted over the chip mounting region of the wiring substrate shown in FIG. 19 .
- FIG. 24 shows an enlarged cross-sectional view taken along
- FIG. 29 shows an exemplary explanatory diagram of the state of the logic chip moved to the wiring substrate after the first alignment step shown in FIG. 28 .
- FIG. 30 shows an explanatory diagram of the state of a heating jig pressed against the back surface of the semiconductor chip after removal of a holding jig shown in FIG. 29 .
- the logic chip LC is mounted by the so-called face down mounting method (flip-chip bonding method) such that the front surface 3 a of the logic chip LC faces the upper surface 2 a of the wiring substrate 20 .
- the logic chip LC is electrically connected with the wiring substrate 20 .
- the front surface electrodes 3 ap formed on the front surface 3 a of the logic chip LC are electrically connected with the bonding leads 2 f formed on the upper surface 2 a of the wiring substrate 20 via the external terminals 7 (protruding electrode 7 b and solder material 7 a shown in FIG. 6 ).
- the first chip mounting step includes a first chip conveying step of arranging the logic chip LC (semiconductor chip 3 ) over the chip mounting region 2 p 1 of the wiring substrate 20 .
- the logic chip LC is conveyed to the chip mounting region 2 p 1 with the back surface 3 b being held by a holding jig (collet) 30 , and then arranged over the chip mounting region 2 p 1 (or adhesive material NCL 1 ) such that the front surface 3 a positioned on the element formation surface side faces the upper surface 2 a of the wiring substrate 20 .
- the holding jig 30 includes a holding surface 30 a for sucking and holding the back surface 3 b of the logic chip LC.
- the holding jig 30 conveys the logic chip LC while holding the logic chip LC by the holding surface 30 a.
- the protruding electrodes 7 b are formed on the front surface 3 a of the logic chip LC, and the solder material 7 a is formed at the tip of each protruding electrode 7 b .
- the solder material 7 a serving as a bonding material for electrically connecting with the protruding electrode 7 b is previously formed at a bonding portion of the bonding lead 2 f formed on the upper surface 2 a of the wiring substrate 20 .
- the logic chip LC Since in the first alignment step shown in FIG. 13 , the logic chip LC is aligned with the wiring substrate 20 with higher accuracy in this embodiment, the accuracy of the conveying position in the first chip conveying step may be low. In order to improve the positional accuracy in the first alignment step, a movement distance in the first alignment step may be preferably small.
- the logic chip LC is preferably arranged over the upper surface 2 a of the wiring substrate 20 with the accuracy that causes the logic chip LC to be arranged over the chip mounting region 2 p 1 .
- the first chip mounting step includes a first mark detection step of detecting (identifying) an alignment mark 50 a of the wiring substrate 20 and an alignment mark 50 b of the logic chip LC.
- a camera mark position detector, image sensor, imaging device 60 is arranged between the wiring substrate 20 and the logic chip LC, and then the alignment mark 50 a of the wiring substrate 20 and the alignment mark 50 b of the logic chip LC are detected (identified).
- the camera 60 includes a light receiver 60 a for receiving visible light, a photoelectric conversion circuit 60 b for converting the light received by the light receiver 60 a into an electric signal, and an output circuit 60 c for outputting the electric signal converted by the photoelectric conversion circuit to the outside.
- the camera 60 is electrically connected with a controller 61 .
- An electric signal output from the output circuit 60 c is transmitted to an image processing circuit (not shown) included in the controller 61 . After image processing is performed, positional data (coordinate data) of the alignment mark (alignment marks 50 a , 50 b ) are output.
- Positional data defining the positional relationship between the alignment mark 50 a and the bonding portion of the bonding lead 2 f , or another positional data defining the positional relationship between the alignment mark 50 b and the front surface electrode 3 ap are input to the controller 61 , which can calculate the position of the bonding portion of the bonding lead 2 f or the front surface electrode 3 ap by the controller 61 .
- a visible light reflector (mirror) 60 d is provided to reflect the visible light entering from the outside toward the light receiver 60 a .
- the camera 60 is connected with a driving device (not shown), and movably fixed over the wiring substrate 20 .
- a reflection angle of the visible light reflector 60 d is changed and the position of the camera 60 is moved, so that the positions of the alignment marks 50 a and 50 b can be detected (identified) by the same camera 60 .
- the alignment marks 50 a are formed over the wiring substrate 20
- the alignment marks 50 b are formed over the front surface 3 a side of the logic chip LC.
- the alignment marks 50 are respectively formed on the wiring substrate 20 and the front surface 3 a of the logic chip LC.
- the positions of the alignment marks 50 are respectively detected, which can provide coordinate data in the plane XY of FIG. 19 , and also data on the direction ⁇ indicative of the angle of inclination between the coordinate axes.
- the use of the coordinate date of the alignment mark 50 and the data on the direction ⁇ can accurately calculate the position of the bonding leads 2 f (see FIG. 15 ) or the position of the front surface electrodes 3 ap (see FIG. 9 ).
- the alignment marks 50 a are arranged at opposite corners such that the marks are positioned on one diagonal line of the device region 20 a forming the quadrilateral shape in plan view.
- the alignment marks 50 b are arranged at opposite corners such that the marks are positioned on one diagonal line of the front surface 3 a forming the quadrilateral in plan view.
- the alignment marks 50 are arranged at the opposite corners in this way, which can improve the accuracy of the data in the direction ⁇ . As a result, the accuracy of calculation of the positions of the bonding leads 2 f (see FIG. 15 ), or calculation of the positions of the front surface electrode 3 ap (see FIG. 9 ) can be improved.
- the alignment mark 50 b formed in the front surface 3 a of the logic chip LC is preferably exposed from the insulating film (protective film, passivation film) 3 p formed on the uppermost surface of the logic chip LC.
- openings 3 ps are formed in the insulating film (protective film, passivation film) 3 p formed on the uppermost surface of the logic chip LC.
- the alignment marks 50 b are exposed from the insulating film 3 p at the opening 3 ps .
- wiring layers are provided to electrically connect a semiconductor element formation surface (main surface) of the semiconductor substrate with the front surface electrodes Sap (see FIG. 9 ).
- FIG. 27 shows only the wafer WH which is a semiconductor substrate while omitting the illustration of the wiring layer.
- the metal pattern exposed from the insulating film 3 p is used as the alignment mark 50 b , thereby improving the light reflection efficiency, which lead to the improvement of the position detection accuracy.
- the contour shape of the alignment mark 50 b has only to be identified.
- the alignment mark 50 b may be covered with the insulating film 3 p.
- the alignment mark 50 a has just been described above with reference to FIG. 18 , and a redundant description thereof will be omitted.
- the first chip mounting step includes a first alignment step of aligning the wiring substrate 20 with the logic chip LC.
- the logic chip LC is relatively moved with respect to the wiring substrate 20 along the upper surface 2 a of the wiring substrate 20 , so that the bonding portions of the bonding leads 2 f face the external terminals 7 formed on the surface 3 a of the logic chip LC.
- this embodiment can calculate the positions of the bonding leads 2 f and the positions of the front surface electrodes 3 ap with high accuracy.
- the logic chip LC is relatively moved to the wiring substrate 20 , which can perform the alignment with high accuracy.
- a method for moving the relative position of the logic chip LC and the wiring substrate 20 involves moving the holding jig 30 holding the logic chip LC along the upper surface 2 a of the wiring substrate 20 as indicated by the arrow of FIG. 28 .
- the relative positional relationship between the logic chip LC and the wiring substrate 20 has only to be changed. Thus, one or both of the logic chip LC and the wiring substrate 20 can be moved.
- the logic chip LC is moved toward the wiring substrate 20 .
- the adhesive material NCL 1 is arranged over the wiring substrate 20 in its soft state before heating to cure.
- the front surface 3 a side of the logic chip LC is filled with the adhesive material NCL 1 .
- the alignment mark 50 b formed on the front surface 3 a side of the logic chip LC is covered with the adhesive material NCL 1 .
- the step of detecting the alignment mark 50 b (first mark detection step) is already completed, which does not cause any problem.
- the relative position of the logic chip LC and the wiring substrate 20 is aligned with each other with high accuracy.
- the bonding portions of the bonding leads 2 f can be kept opposite to the external terminals 7 formed on the surface 3 a of the logic chip LC.
- the logic chip LC is pressed against the wiring substrate 20 by pushing the heating jig 31 to the back surface 3 b of the logic chip LC.
- the adhesive material NCL 1 is in its soft state before curing, the logic chip LC is pushed into by the heating jig 31 , so that the logic chip LC approaches the wiring substrate 20 .
- the tip (specifically, solder material 7 a ) of each external terminal 7 formed on the front surface 3 a of the logic chip LC comes into contact with the corresponding bonding region of the bonding lead 2 f (specifically, solder material 7 a ).
- the thickness of the adhesive material NCL 1 is larger than at least the total of the height (height of a protruding part) of the external terminal 7 and the thickness of the bonding lead 2 f .
- the part of the front surface 3 a side of the logic chip LC is embedded in the adhesive material NCL 1 .
- at least a part of the front surface 3 a of the side surface of the logic chip LC is embedded in the adhesive material NCL 1 .
- the alignment marks 50 c and the back surface electrodes 3 bp are formed at the logic chip LC. It is necessary to prevent the alignment marks 50 c and back surface electrode 3 bp from being covered by the adhesive material NCL 1 coming around to the back surface 3 b side.
- a member (low elastic member), for example, a resin film (film) 32 which is softer than the heating jig 31 and the logic chip LC, preferably intervenes in between the heating jig 31 and the logic chip LC to cover the back surface 3 b of the logic chip LC with the resin film 32 .
- the resin film 32 comes into intimate contact with the back surface 3 b of the logic chip LC. Even though the adhesive material NCL 1 is thick, the NCL 1 can be prevented from coming around to the back surface 3 b of the logic chip LC.
- the resin film 32 of this embodiment is made of, for example, fluorine resin.
- the logic chip LC and the adhesive material NCL 1 are heated by the heating jig (heat source) 31 while pushing the heating jig 31 to the logic chip LC.
- the solder material 7 a on the bonding lead 2 f side and the solder material 7 a on the external terminal side are respectively melted and integral with each other into a bonding material electrically connecting the external terminal 7 with the bonding lead 2 f . That is, the logic chip LC is heated by the heating jig (heat source) 31 , so that the protruding electrodes 7 b are electrically connected with the bonding leads 2 f via the solder material 7 a.
- the adhesive material NCL 1 is heated to be cured. As a result, the cured adhesive material NCL 1 is obtained with a part of the logic chip LC embedded in the adhesive material.
- the alignment marks 50 c and the back surface electrodes 3 bp of the logic chip LC are covered by the resin film 32 , and exposed from the cured adhesive material NCL 1 .
- the adhesive material NCL 1 is not necessarily cured completely by the heat from the heating jig (heat source) 31 .
- the wiring substrate 20 is transferred to a heating furnace (not shown), and then the remainder of the thermosetting resin can be cured (fully cured). It takes some time to complete the full curing process for fully curing the entire thermosetting resin contained in the adhesive material NCL 1 . However, the full curing process can be performed in the heating furnace to improve the manufacturing efficiency.
- FIG. 31 shows an enlarged plan view of the state of the adhesive material arranged at the back surface of the semiconductor chip shown in FIG. 20 and its surroundings.
- FIG. 32 shows an enlarged cross-sectional view taken along the line A-A of FIG. 31 .
- the logic chip LC mounted at the lowermost level (for example, at the first level) and the memory chip MC 1 mounted at the second level from the bottom among the stacked semiconductor chips 3 are mounted by the face down mounting method (flip-chip bonding).
- the use of the above-mentioned pre-coating method is preferable in that the processing time for one device region 20 a (see FIGS. 31 and 32 ) can reduced to improve the manufacturing efficiency.
- the adhesive material NCL 2 to be used in the pre-coating method is made of insulating (non-conductive) material (for example, resin material) as described above.
- the adhesive material NCL 2 is made of resin that will be cured (increase its hardness) by being supplied with the energy.
- the adhesive material NCL 2 includes, for example, thermosetting resin.
- the adhesive material NCL 2 before being cured is softer than the protruding electrode 7 b shown in FIG. 6 , and is deformed by being pressed against the logic chip LC.
- the adhesive materials NCL 2 before being cured can be classified into two types based on the difference in handling, namely, a paste-like resin (insulating paste) called the NCP, and a film-like resin (insulating film) previously molded in the shape of film and called the NCF.
- the adhesive material NCL 2 used in this step can employ either the NCP or the NCF.
- the NCP is discharged from a nozzle 33 (see FIG. 32 ), so that the adhesive material NCL 2 is arranged over the back surface 3 b of the logic chip LC.
- the pre-coating method is common to the post-charging method explained in the description of the first adhesive material arrangement step, in that the paste-like adhesive material NCL 2 is discharged from the nozzle 33 .
- the adhesive material NCL 2 is mounted in advance before mounting the memory chip MC 1 as shown in FIG. 4 .
- the coating rate of the adhesive material NCL 2 can be greatly increased as compared to the post-charging method of charging the resin using the capillary action.
- the alignment marks 50 c formed on the back surface of the logic chip LC each are an object of interest to be detected in the second mark detection step shown in FIG. 13 , and the adhesive material NCL 2 is preferably arranged so as to expose the alignment marks 50 c .
- the adhesive material NCL 2 is formed of transparent or semi-transparent material to the visible light, the position of even the alignment mark 50 c covered by the adhesive material NCL 2 can be detected. From improving the flexibility in selecting material, as shown in FIG. 31 , the adhesive material NCL 2 is preferably arranged so as to expose the alignment marks 50 c .
- the insulating paste (NCP) can decrease the area of the adhesive material arranged as compared to the insulating film (NCF), thus making it easy to expose the alignment mark 50 c.
- the adhesive material NCL 2 has the function of bonding and fixing the memory chip MC 1 (see FIG. 4 ) to the logic chip LC (see FIG. 4 ) in the second chip mounting step shown in FIG. 13 .
- the adhesive material NCL 2 also has the function of sealing the bonding portion between the memory chip MC 1 and the logic chip LC.
- the sealing function includes a stress release function of diffusing and releasing the stress transferred to the bonding portion between the memory chip MC 1 and the logic chip LC to protect the bonding portion.
- the adhesive material NCL 2 is preferably arranged to surround the bonding portion between the memory chip MC 1 and the logic chip LC. For this reason, upon mounting at least the memory chip MC 1 , the external terminals 7 shown in FIG. 6 are preferably sealed with the adhesive material NCL 2 .
- the laminated body MCS with the memory chips MC 1 , MC 2 , MC 3 , and MC 4 shown in FIG. 4 is provided.
- the memory chips MC 1 , MC 2 , MC 3 , and MC 4 are stacked in turn over the logic chip LC.
- the memory chips MC 1 , MC 2 , MC 3 , and MC 4 are previously stacked to form the laminated body (memory chip laminated body, semiconductor chip laminated body) MCS shown in FIG. 34 .
- the formation of the laminated body MCS with the memory chips MC 1 , MC 2 , MC 3 , and MC 4 can be independently performed from the other steps in a place different from those for steps other than the second chip provision step shown in FIG. 13 .
- the laminated body MCS can be provided as a purchased part. This way can advantageously simplify the assembly process shown in FIG. 13 to improve the manufacturing efficiency as a whole.
- FIG. 33 shows an exemplary explanatory diagram of the outline of an assembly process of a laminated body of the memory chips shown in FIG. 4 .
- FIG. 34 shows an exemplary explanatory diagram of the outline of another assembly process of the laminated body of the memory chips, following the process shown in FIG. 33 .
- the manufacturing method of each of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 shown in FIGS. 33 and 34 can be performed by employing the manufacturing method of the semiconductor chip described with reference to FIGS. 21 and 22 , and thus a description thereof will be omitted below.
- a base (assembly base) 34 for assembling the laminated body MCS shown in FIG. 34 is prepared.
- the base 34 has an assembly surface 34 a over which the memory chips MC 1 , MC 2 , MC 3 , and MC 4 are stacked.
- the assembly surface 34 a is provided with an adhesive material layer 35 .
- the memory chips MC 1 , MC 2 , MC 3 , and MC 4 are stacked over the assembly surface 34 a of the base 34 .
- the memory chips MC 4 , MC 3 , MC 2 , and MC 1 are stacked in that order over the assembly surface 34 a of the base 34 such that the back surface 3 b of each semiconductor chip to be stacked faces the assembly surface 34 a of the base 34 .
- the back surface electrode 3 bp of the upper semiconductor chip 3 is connected with the front surface electrode 3 ap of the lower semiconductor chip 3 , for example, via the external terminals 7 (solder material 7 a ).
- the front surface electrodes 3 ap of the memory chip MC 1 disposed at the uppermost level are provided with the alignment marks 50 d as shown in FIG. 7 .
- a laminated body sealing step shown in FIG. 34 resin (underfill resin) is supplied to between the stacked semiconductor chips 3 to thereby form a seal member (seal member for a laminated body of chips, or resin for a laminated body of chips) 6 .
- the seal member 6 is formed by the post-charging method, which has been explained above in the description of the first adhesive material arrangement step.
- an underfill resin 6 a is supplied from the nozzle 36 , and fills in between the stacked semiconductor chips 3 .
- the underfill resin 6 a has a lower viscosity than that of resin for sealing used in the sealing step shown in FIG. 13 .
- the underfill resin 6 a can fill in between the semiconductor chips 3 by the capillary action.
- the underfill resin 6 a embedded in between the semiconductor chips 3 are cured to thereby obtain the seal member 6 .
- the method for forming the seal member 6 by the post-charging method is excellent in filling in a clearance as compared to the so-called transfer mold method, and can be effectively applied to the case where a clearance between the stacked semiconductor chips 3 is narrow.
- the underfill resin 6 a when the clearances to be filled with the underfill resin 6 a are stacked in layers, the underfill resin 6 a can collectively fill in all the clearances at one time. This can reduce the whole processing time.
- the base 34 and the adhesive material layer 35 are peeled off and removed from the back surface 3 b of the memory chip MC 4 .
- the method for removing the base 34 and the adhesive material layer 35 can be performed, for example, by using a method which involves curing a resin component (for example, ultraviolet curing resin) contained in the adhesive material layer 35 .
- a resin component for example, ultraviolet curing resin
- the laminated body MCS can be regarded as one memory chip having one front surface 3 a (corresponding to the front surface 3 a of the memory chip MC 1 ) with the front surface electrodes 3 ap formed thereat, and the other back surface 3 b (corresponding to the back surface 3 b of the memory chip MC 4 ) opposite to the front surface 3 a.
- the laminated body MCS is mounted over the back surface 3 b of the logic chip LC.
- the second chip conveying step, the second mark detection step, and the second alignment step shown in FIG. 13 can be regarded as the sub-steps included in the step of mounting the laminated body MCS over the logic chip LC.
- the second chip conveying step, the second mark detection step, and the second alignment step are described below as the sub-steps included in the second chip mounting step.
- FIG. 35 shows an enlarged plan view of the state of the laminated body mounted over the back surface of the logic chip shown in FIG. 31 .
- FIG. 36 shows an enlarged cross-sectional view taken along the line A-A of FIG. 35 .
- FIG. 37 shows an exemplary explanatory diagram of a main part of a second chip conveying step shown in FIG. 13 .
- FIG. 38 shows an exemplary explanatory diagram of a main part of a second mark detection step shown in FIG. 13 .
- FIG. 39 shows an exemplary explanatory diagram of a main part of a second alignment step shown in FIG. 13 .
- FIG. 40 shows an exemplary explanatory diagram of the state of the logic chip moved to the wiring substrate after the second alignment step shown in FIG. 39 .
- FIG. 41 shows an explanatory diagram of the state of a heating jig pressed against the back surface of the semiconductor chip after removal of a holding jig shown in FIG. 40 .
- the laminated body MCS is mounted by the so-called face down mounting method (flip-chip bonding) such that the front surface 3 a of the laminated body MCS (front surface 3 a of the memory chip MC 1 ) faces the back surface 3 b of the logic chip LC.
- the memory chips MC 1 , MC 2 , MC 3 , and MC 4 are electrically connected with the logic chip LC. Specifically, as shown in FIG.
- the front surface electrodes 3 ap formed on the front surface 3 a of the laminated body MCS (memory chip MC 1 ) are electrically connected with the back surface electrodes 3 bp formed on the back surface 3 b of the logic chip LC via the external terminals 7 (solder material 7 a shown in FIG. 6 ).
- the external terminals 7 solder material 7 a shown in FIG. 6 .
- the second chip mounting step includes a second chip conveying step of arranging the laminated body MCS (semiconductor chips 3 ) over the chip mounting region 2 p 2 of the wiring substrate 20 .
- the laminated body MCS is delivered to over the chip mounting region 2 p 2 while the back surface 3 b side is being held by a holding jig (collet) 30 , and arranged over the chip mounting region 2 p 2 (or adhesive material NCL 2 ) such that the front surface 3 a positioned on the element formation surface side faces the upper surface 2 a of the wiring substrate 20 .
- the holding jig 30 has a holding surface 30 a for sucking and holding the back surface 3 b of the laminated body MCS.
- the laminated body MCS is conveyed while being held by the holding surface 30 a .
- the chip mounting region 2 p 2 is a region where the laminated body MCS is to be mounted in this step, and does not necessarily includes the boundary line actually visible, like the chip mounting region 2 P 1 in the first chip mounting step described above.
- Each external terminal 7 is formed on the front surface 3 a side of the laminated body MCS.
- the solder material 7 a bonded to the front surface serves as the external terminal 7 .
- this embodiment can be applied to a modified example in which the protruding electrodes 7 b are formed to have its tip provided with the solder material 7 a.
- no bonding material such as solder material
- a solder material (not shown) can be formed over the back surface electrode 3 bp .
- the solder material covering the back surface electrode 3 bp is formed before the above first chip mounting step, the solder is melted by heating with the heating jig 31 shown in FIG. 30 , and at that time, the melted solder might be attached to the resin film 32 .
- a solder material is preferably applied after the first chip mounting step and before the first adhesive material arrangement step.
- the laminated body MCS is aligned with the wiring substrate 20 with higher accuracy in this embodiment, the accuracy of the conveying position in the second chip conveying step may be low. In order to improve the positional accuracy in the second alignment step, a movement distance in the second alignment step may be preferably small.
- the laminated body MCS is preferably arranged over the upper surface 2 a of the wiring substrate 20 with the accuracy that arranges the laminated body MCS over the chip mounting region 2 p 2 .
- the second chip mounting step includes a second mark detection step that detects (identifies) an alignment mark 50 c at the back surface 3 b of the logic chip LC, and another alignment mark 50 b of the laminated body MCS.
- the camera 60 is positioned between the wiring substrate 20 and the laminated body MCS, whereby detects (identifies) the alignment mark 50 c of the logic chip LC and the alignment mark 50 d of the laminated body MCS.
- the camera 60 , a controller 61 connected with the camera 60 , and a drive device (not shown) are substantially the same as those in the first mark detection step, and its redundant description will be omitted below.
- This step detects at least one or more alignment marks 50 c , and at least one or more alignment marks 50 d by use of the camera 60 .
- the alignment marks 50 a are disposed at the periphery of the device region 20 a , so that the alignment marks 50 a become viewable.
- the alignment mark 50 a of the wiring substrate 20 is used to perform alignment without forming any alignment mark 50 c at the back surface 3 b of the logic chip LC.
- the positions (coordinates in the XY plane shown in FIG. 23 ) or angles (angles in the direction ⁇ shown in FIG. 23 ) of the back surface electrodes 3 bp might become misaligned.
- the misalignment might be caused.
- an amount of misalignment between the external terminal 7 of the laminated body MCS and the back surface electrode 3 bp of the logic chip LC is the sum of an amount of misalignment between the wiring substrate 20 and the logic chip LC and an amount of misalignment between the wiring substrate 20 and the laminated body MCS. That is, the amount of misalignment between the external terminal 7 of the laminated body MCS and the back surface electrode 3 bp of the logic chip LC is larger than that between the respective semiconductor chips 3 .
- the alignment marks 50 c are formed on the back surface 3 b of the logic chip LC, and then used to align the laminated body MCS and logic chip LC with the substrate. That is, the amount of misalignment between the external terminal 7 of the laminated body MCS and the back surface electrode 3 bp of the logic LC is within a prescribed range of misalignment between the laminated body MCS and the logic chip LC.
- positional data for defining the positional relationship between the alignment mark 50 c and the back surface electrode 3 bp of the logic chip LC, or the positional relationship between the alignment mark 50 d and the front surface electrode 3 ap of the laminated body MCS is input to the controller 61 .
- the controller 61 can calculate the positions of the back surface electrodes 3 bp of the logic chip LC, and the front surface electrodes 3 ap of the laminated body MCS.
- the alignment marks 50 c are formed over the logic chip LC, and the alignment marks 50 d are formed over the front surface 3 a side of the laminated body MCS.
- the alignment marks 50 are formed on the back surface 3 b of the logic chip LC and at the front surface 3 a of the laminated body MCS, and then the positions of the alignment marks 50 are respectively detected, whereby data on the direction ⁇ indicative of an inclination of the coordinate axes can be obtained in addition to the coordinate data in the XY plane shown in FIG. 23 .
- the positions of the back surface electrodes 3 bp (see FIG. 10 ) of the logic chip LC, or the positions of the front surface electrode 3 ap (see FIG. 7 ) of the laminated body MCS can be accurately calculated by using the data on the coordinates and direction ⁇ of the alignment mark 50 .
- the alignment marks 50 c are arranged at opposite corners such that the marks are positioned on one diagonal line of the back surface 3 b forming the quadrilateral shape in plan view.
- the alignment marks 50 d are arranged at opposite corners such that the marks are positioned on one diagonal line of the front surface 3 a forming the quadrilateral shape in plan view.
- the alignment marks 50 are arranged at the opposite corners in this way, which can improve the accuracy of the data on the direction ⁇ . This can improve the accuracy of calculation of the positions of the back surface electrodes 3 bp (see FIG. 10 ), or the positions of the front surface electrodes 3 ap (see FIG. 7 ).
- the front surface electrodes 3 ap of each of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 are collectively arranged in the center of the front surface 3 a as shown in FIG. 7 .
- the alignment marks 50 d are preferably arranged outside (at the periphery of) the front surface electrodes 3 ap.
- the back surface electrodes 3 bp of the logic chip LC are arranged to face the front surface electrodes 3 ap of the memory chip MC 1 .
- the back surface electrodes 3 bp of the logic chip LC are collectively arranged in the center of the back surface 3 b of the logic chip LC.
- the alignment marks 50 c are preferably arranged outside (at the periphery of) the back surface electrodes 3 bp.
- each of the alignment marks 50 d is preferably arranged outside (at the periphery of) the memory regions MR at the front surface 3 a of the memory chip MC 1 .
- each of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 increases in proportion to the area of the memory region MR.
- the alignment marks 50 d are arranged outside (at the periphery of) the memory regions MR, so that each alignment mark 50 d is arranged at the periphery of the front surface 3 a .
- the distance between the alignment marks 50 d can be increased to thereby improve the alignment accuracy in the direction ⁇ as mentioned above.
- the alignment marks 50 d are preferably arranged at opposite corners in order to improve the alignment accuracy in the direction ⁇ as described above.
- the alignment mark 50 d formed in the front surface 3 a of the laminated body MCS is preferably exposed from the insulating film (protective film, passivation film) 3 p formed on the uppermost surface of the laminated body MCS.
- the opening 3 ps is formed in the insulating film (protective film, passivation film) 3 p formed on the uppermost surface of the laminated body MCS.
- Each alignment mark 50 d is exposed from the insulating film 3 p at the opening 3 ps.
- the alignment marks 50 c are formed on the back surface 3 b of the logic chip LC as shown in FIG. 38 without forming the insulating film or protective film covering the back surface 3 b .
- the adhesive material NCL 2 can be arranged to avoid the alignment marks 50 c to easily expose the alignment marks 50 c.
- the second chip mounting step includes a second alignment step that aligns the logic chip LC with the laminated body MCS.
- the relative position between the laminated body MCS and the logic chip LC (in other words, wiring substrate 20 ) are moved along the upper surface 2 a of the wiring substrate 20 , so that the back surface electrodes 3 b of the logic chi LC face the external terminals 7 formed on the front surface 3 a of the laminated body MCS.
- the positions of the back surface electrodes 3 bp of the logic chip LC and the positions of the front surface electrodes 3 ap of the laminated body MCS can be calculated with high accuracy. Based on the calculation data, the relative position between the laminated body MCS and the logic chip LC can be moved to perform alignment with high accuracy.
- a method for moving the relative position between the logic chip LC and the laminated body MCS involves moving the holding jig 30 holding the laminated body MCS along the upper surface 2 a of the wiring substrate 20 as indicated by the arrow of FIG. 39 .
- the relative relationship between the laminated body MCS and the logic chip LC can be moved by moving one or both of the laminated body MCS and the logic chip LC.
- the laminated body MCS is moved toward the logic chip LC.
- the relative position between the laminated body MCS and the logic chip LC is aligned with high accuracy.
- the back surface electrodes 3 bp of the logic chip LC and the external terminals 7 formed on the front surface 3 a of the laminated body MCS can be kept facing and opposite to each other by linearly moving the laminated body MCS toward the logic chip LC.
- the laminated body MCS is pressed against the logic chip LC by pushing the heating jig 31 to the back surface 3 b of the laminated body MCS.
- the adhesive material NCL 2 is soft before being cured.
- the laminated body MCS approaches the logic chip LC.
- the tip (specifically, solder material 7 a ) of each of the external terminals 7 formed on the front surface 3 a of the laminated body MCS comes into contact with the bonding region (specifically, solder material 7 a ) of the back surface electrode 3 bp.
- the adhesive material NCL 2 applied between the laminated body MCS and the logic chip LC is spread along the back surface 3 b of the logic chip LC.
- the periphery of the adhesive material NCL 2 does not reach the periphery of the back surface 3 b of the logic chip LC.
- the adhesive material NCL 2 may be spread over the periphery of the logic chip LC to cover the side surface of the logic chip LC or the side surface of the adhesive material NCL 1 .
- the alignment marks 50 c of the back surface 3 b of the logic chip LC are covered with the adhesive material NCL 2 .
- the position of the alignment mark 50 c is already detected, which does not raise any problem.
- a contact area between the adhesive material NCL 2 and the semiconductor chip 3 can be expanded to improve the strength of bonding between the adhesive material NCL 2 and the semiconductor chip 3 (laminated body MCS and logic chip LC).
- the alignment marks 50 and back surface electrodes 3 bp are not formed on the back surface 3 b of the laminated body MCS.
- the resin film 32 shown in FIG. 41 can be omitted.
- the provision of the resin film 32 can disperse the force received by the laminated body MCS in pressing the film with the heating jig 31 .
- the provision of the resin film 32 can perform a second chip mounting step in the same mechanism as that of the first chi mounting step, which can simplify the manufacturing process. In the second chip mounting step, the resin film 32 preferably intervenes in between the laminated body MCS and the heating jig 31 .
- the laminated body MCS and the adhesive material NSL 2 are heated by the heating jig (heat source) 31 while the laminated body MCS being pressed against the heating jig 31 .
- the solder material 7 a on the external terminal side is melt at a bonding portion between the laminated body MCS and the logic chip LC to be bonded with the back surface electrode 3 bp of the logic chip LC. That is, the laminated body MCS is heated by the heating jig (heat source) 31 , so that the front surface electrodes Sap of the laminated body MCS are electrically connected with the back surface electrodes 3 bp of the logic chip LC via the solder materials 7 a.
- the adhesive material NCL 2 is heated to be cured.
- the adhesive material NCL 1 is not necessarily cured fully by the heat from the heating jig (heat source) 31 , and a part of the thermosetting resin contained in the adhesive material NCL 1 is cured (semi-cured) to such an extent that can fix the laminated body MCS.
- the log chip LC is transferred to a heating furnace (not shown), and the remainder of the thermosetting resin is cured (fully cured). It takes time to complete the full curing process of fully curing the thermosetting resin components contained in the adhesive material NCL 1 , but the full curing process is performed in the heating furnace, which can improve the manufacturing efficiency.
- FIG. 42 shows an enlarged cross-sectional view of the state of the stacked semiconductor chips sealed with the seal member formed over the wiring substrate shown in FIG. 36 .
- FIG. 43 shows a plan view of the entire structure of the seal member shown in FIG. 42 .
- the seal member 4 is formed to seal the device regions 20 a at one time.
- the formation of the seal member 4 is called the “block molding” method.
- a semiconductor package manufactured by the block molding method is called the “multi array package (MAP) semiconductor device”.
- MAP multi array package
- the distance between the adjacent device regions 20 a can be decreased, so that an effective area of one piece of the wiring substrate 20 is increased. In other words, the number of products that can be obtained from one wiring substrate 20 is increased. In this way, the effective area of the one wiring substrate 20 is increased, which can make the manufacturing process more efficient.
- the seal member 4 is formed by the so-called transfer molding method which involves press-fitting resin heated and softened in a molding die (not shown), molding the resin therein, and thermally curing the resin.
- the seal member 4 formed by the transfer mold method has a high durability, for example, like the seal member 6 sealing the laminated body MCS shown in FIG. 42 , as compared to a member produced by curing a liquid resin.
- the seal member 4 is appropriate as a protective member. Filler particles, such as silica (silicone dioxide; SiO 2 ) particles, are mixed into the thermosetting resin, which can improve the function of the seal member 4 (for example, resistance to warpage).
- bonding portions (electric connection portions) between the stacked semiconductor chips 3 are sealed by the adhesive materials NCL 1 , NCL 2 , and seal member 6 .
- the seal member 4 is not formed. In this case, the full sealing step can be omitted.
- FIG. 44 shows an enlarged cross-sectional view of the state of the solder balls bonded to lands of the wiring substrate shown in FIG. 37 .
- the solder balls 5 are arranged over the respective lands 2 g exposed at the lower surface 2 b of the wiring substrate 20 , and then heated, so that the solder balls 5 are bonded to the lands 2 g .
- the solder balls 5 are electrically coupled to the semiconductor chips 3 (logic chip LC, and memory chips MC 1 , MC 2 , MC 3 , and MC 4 ) via the wiring substrate 20 .
- the technique described in this embodiment is not limited to the application to the so-called ball grid array (BGA) semiconductor device with solder balls 5 bonded thereto in an array.
- BGA ball grid array
- a modified example of this embodiment can be applied to the so-called land grid array (LGA) semiconductor device which is shipped with the lands 2 g exposed without forming the solder balls 5 , or with a solder paste thinner than the solder ball 5 attached to the lands 2 g .
- the LGA semiconductor device can omit the ball mount step.
- FIG. 45 shows a cross-sectional view of the singulated multi-piece wiring substrate shown in FIG. 44 .
- the wiring substrate 20 and the seal member 4 are cut along the dicing lines (dicing regions) 20 c to obtain the singulated semiconductor devices 1 (see FIG. 4 ).
- cutting methods are not specifically limited, but in the example shown in FIG. 45 , the wiring substrate 20 and seal member 4 bonded and fixed to a tape material (dicing tape) 41 are cut using the dicing blade (rotary blade) 40 from the lower surface 2 b of the wiring substrate 20 .
- the technique described in this embodiment is applied not only to the case of using the wiring substrate 20 as a multi-piece substrate with the device regions 20 a , but also to, for example, a semiconductor device having semiconductor chips 3 stacked over the wiring substrate 2 (see FIG. 4 ) corresponding to only one semiconductor device. In this case, the singulating step can be omitted.
- the semiconductor device 1 described with reference to FIGS. 1 to 11 can be obtained. Thereafter, necessary checking and testing, such as an appearance check or an electric test, are performed on each semiconductor device. Then, the semiconductor device is shipped, or mounted on a mounting substrate (not shown).
- FIG. 46 shows an enlarged plan view of the alignment mark shown in FIGS. 3 , 7 , 9 , and 10 .
- FIGS. 47 to 49 are enlarged plan views of modified examples corresponding to FIG. 46 .
- An alignment mark 50 shown in FIG. 46 forms a polygonal shape (a L-like shape shown in FIG. 46 ) in plan view.
- the polygonal shape has an asymmetric shape with respect to the center of a circumcircle 51 of the polygonal shape.
- the planar shape of the alignment mark 50 is set to the asymmetric shape, so that not only the position of the alignment mark 50 , but also an inclination of the alignment mark 50 in plan view can be detected.
- the X and Y coordinates of the alignment mark 50 and the position of the direction ⁇ can be specified.
- the alignment marks 50 d shown in FIG. 7 , and the alignment marks 50 b shown in FIG. 9 are set in the asymmetric shape shown in FIG. 46 , which can prevent or suppress the improper mounting, including the displacement of the mounting direction of the semiconductor chip 3 by 180 degrees, in the first chip mounting step and the second chip mounting step.
- the alignment mark having a circular shape in plan view can improve the processing accuracy as the circular shape can be easily processed as compared to the polygonal shape shown in FIG. 46 . Even a part of the circular alignment mark 52 can hardly break, which can improve the detection accuracy.
- At least one of the alignment marks 50 a , 50 b , 50 c , and 50 d employs the circular alignment mark 52 , which can improve the position detection accuracy of the position in which the alignment mark 52 is arranged.
- the alignment marks 50 a , 50 b , 50 c , and 50 d can be formed in a circular shape, like the alignment mark 52 .
- the back surface electrodes 3 bp of the logic chip LC have only to surely face the front surface electrodes 3 ap of the memory chip MC 1 .
- various modified examples can be applied within a range that can ensure the required alignment accuracy.
- an alignment mark 53 having a quadrilateral shape in plan view can be applied.
- an alignment mark 54 having an X-like shape in plan view can be applied.
- a combination of the alignment marks 50 , 52 , 53 , and 54 can be applied.
- the technique for stacking the semiconductor chips 3 using the result of detection of the alignment mark 50 is applied in the first chip mounting step and the second chip mounting step.
- the technique can applied to the step of assembling the laminated body MCS, that is, the second chip provision step.
- FIG. 50 shows a plan view of a modified example corresponding to FIG. 8 .
- the alignment marks 50 d are formed over the respective surfaces 3 a of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 forming the laminated body MCS (see FIG. 4 ).
- alignment marks 50 e are formed over the back surface 3 b of each of the memory chips MC 1 , MC 2 , and MC 3 .
- the alignment accuracy of the memory chips can be improved to thereby improve the electric connection reliability of the memory chips MC 1 , MC 2 , MC 3 , and MC 4 .
- FIG. 51 shows a perspective plane view of a modified example corresponding to FIG. 3 .
- the above embodiment has described a SiP semiconductor device 1 composed of the logic chip LC and the memory chips MC 1 , MC 2 , MC 3 , and MC 4 as an example of the semiconductor chips 3 to be stacked (used).
- the invention can also be applied to any combination other than a combination of the memory chip and the logic chip as long as the stacked chips 3 are electrically connected with each other.
- the back surface electrodes 3 bp are arranged on the side opposite to the front surface electrodes 3 ap of the logic chip LC.
- the back surface electrodes 3 bp are electrically connected with the front surface electrodes 3 ap via the penetrating electrodes 3 tsv .
- the invention can also be applied to a semiconductor device including back surface electrodes 3 bp formed at the lower semiconductor chip 3 , and front surface electrodes 3 ap formed at the upper semiconductor chip 3 electrically connected with the back surface electrodes 3 bp via the external terminals 7 .
Abstract
Provided is a semiconductor device with improved reliability. A logic chip (first semiconductor chip) and a laminated body (second semiconductor chip) are stacked in that order over a wiring substrate. An alignment mark formed over the wiring substrate is aligned with an alignment mark formed on a front surface of the logic chip, whereby the logic chip is mounted over the wiring substrate. An alignment mark formed on a back surface of the logic chip is aligned with an alignment mark formed on a front surface of the laminated body, whereby the laminated body is mounted over the back surface of the logic chip LG.
Description
- The disclosure of Japanese Patent Application No. 2013-061087 filed on Mar. 22, 2013 including the specification, drawings and abstract is incorporated herein by reference in its entirety.
- The present invention relates to manufacturing techniques of semiconductor devices, and more specifically, to a technique effectively applied to a semiconductor device with a semiconductor chip mounted over another semiconductor chip.
- Japanese Unexamined Patent Publication No. 2001-217387 (Patent Document 1) discloses a method for stacking semiconductor chips which involves respectively forming alignment marks on the surface of a parent chip (fixed device) and on the surface of a child chip (semiconductor chip), and stacking the child chip over the parent chip based on the result of detection of the alignment marks such that the respective alignment marks on both chips face each other.
- Japanese Unexamined Patent Publication No. 2011-187574 (Patent Document 2) discloses a semiconductor device including a semiconductor chip with a penetrating electrode between a wiring substrate and a laminated body of a plurality of memory chips.
- Japanese Unexamined Patent Publication No. 2001-217387
- Japanese Unexamined Patent Publication No. 2011-187574
- The inventors of the present application have studied about techniques for improving the performance of a semiconductor device including a laminated body of semiconductor chips stacked over a wiring substrate. As one of the techniques, the so-called system in package (SiP) semiconductor device has been studied, which forms a system only by one semiconductor device having a plurality of semiconductor chips (for example, a memory chip and a control chip for controlling the memory chip) mounted therein.
- Methods for stacking semiconductor chips includes a method which involves arranging and stacking an electrode of one (upper) semiconductor chip to face the other (lower) semiconductor chip, and electrically connecting the respective electrodes via a conductive member, such as solder material. This method connects the stacked semiconductor chips together without wirings, and thus can decrease a transmission distance between the semiconductor chips.
- In order to electrically connect the electrodes from the viewpoint of improving the reliability of the semiconductor device, it is necessary to improve the alignment accuracy of the upper semiconductor chip with respect to the lower semiconductor chip.
- Other problems and new features of the present invention will be better understood after a reading of the following detailed description in connection with the accompanying drawings.
- A method for manufacturing a semiconductor device according to one embodiment of the invention includes the step of (a) mounting a first semiconductor chip over a wiring substrate such that a first main surface of the first semiconductor chip faces the wiring substrate. The manufacturing method also includes the step of (b) mounting a second semiconductor chip over a wiring substrate such that a second main surface of the second semiconductor chip faces a first back surface of the first semiconductor chip. In the above step (a), after detection and alignment of a first alignment mark formed on the wiring substrate and a second alignment mark formed on the first main surface side of the first semiconductor chip, the first semiconductor chip is mounted. In the above (b) step, after detection and alignment of a third alignment mark formed on the first back surface side of the first semiconductor chip and a fourth alignment mark formed at the second main surface side of the second semiconductor chip, the second semiconductor chip is mounted. A plurality of first back surface side pads are formed on the first back surface, and a plurality of second main surface side pads are formed on the second main surface. In the above step (b), the first back surface side pads are electrically connected with the second main surface, side pads via external terminals respectively formed on the second main surface side pads.
- According to the one embodiment of the present invention, the reliability of the semiconductor device can be improved.
-
FIG. 1 is a perspective view of a semiconductor device according to one embodiment of the invention; -
FIG. 2 is a bottom view of the semiconductor device shown inFIG. 1 ; -
FIG. 3 is a perspective plan view showing an internal structure of the semiconductor device over a wiring substrate with a seal member shown inFIG. 1 removed therefrom; -
FIG. 4 is a cross-sectional view taken along the line A-A ofFIG. 1 ; -
FIG. 5 is an exemplary explanatory diagram showing an example of a circuit structure of the semiconductor device shown inFIGS. 1 to 4 ; -
FIG. 6 is an enlarged cross-sectional view of a portion “A” shown inFIG. 4 ; -
FIG. 7 is a plan view showing a front surface side of a memory chip shown inFIG. 4 ; -
FIG. 8 is a plan view showing an example of a back surface side of a memory chip shown inFIG. 7 ; -
FIG. 9 is a plan view showing a front surface side of a logic chip shown inFIG. 4 ; -
FIG. 10 is a plan view showing an example of a back surface side of a logic chip shown inFIG. 9 ; -
FIG. 11 is an exemplary explanatory diagram showing alignment means (alignment marks) used in assembling the semiconductor device shown inFIG. 4 ; -
FIG. 12 is an explanatory diagram showing a studied example corresponding toFIG. 11 ; -
FIG. 13 is an explanatory diagram showing the outline of a manufacturing process of the semiconductor device explained with reference toFIGS. 1 to 11 ; -
FIG. 14 is a plan view showing the entire structure of a wiring substrate provided in a substrate provision step shown inFIG. 13 ; -
FIG. 15 is an enlarged plan view of one device region shown inFIG. 14 ; -
FIG. 16 is an enlarged cross-sectional view taken along the line A-A ofFIG. 15 ; -
FIG. 17 is an enlarged plan view showing the surface opposite to that shown inFIG. 15 ; -
FIG. 18 is an enlarged cross-sectional view taken along the line B-B ofFIG. 15 ; -
FIG. 19 is an enlarged plan view showing the state of an adhesive material arranged in a chip mounting region shown inFIG. 15 ; -
FIG. 20 is an enlarged cross-sectional view taken along the line A-A ofFIG. 19 ; -
FIG. 21 is an exemplary explanatory diagram showing the outline of a manufacturing process of a semiconductor chip with a penetrating electrode shown inFIG. 6 ; -
FIG. 22 is an exemplary explanatory diagram showing the outline of another manufacturing process of the semiconductor chip, following the process shown inFIG. 21 ; -
FIG. 23 is an enlarged plan view showing the state of a logic chip LC mounted over the chip mounting region of the wiring substrate shown inFIG. 19 ; -
FIG. 24 is an enlarged cross-sectional view taken along the line A-A ofFIG. 23 ; -
FIG. 25 is an exemplary explanatory diagram showing a main part of a first chip conveying step shown inFIG. 13 ; -
FIG. 26 is an exemplary explanatory diagram showing a main part of a first mark detection step shown inFIG. 13 ; -
FIG. 27 is an enlarged cross-sectional view showing a main part of a cross-sectional structure of an alignment mark of the logic chip shown inFIG. 26 ; -
FIG. 28 is an exemplary explanatory diagram showing a main part of a first alignment step shown inFIG. 13 ; -
FIG. 29 is an exemplary explanatory diagram showing the state of the logic chip moved to the wiring substrate after the first alignment step shown inFIG. 28 ; -
FIG. 30 is an explanatory diagram showing the state of a heating jig pressed against the back surface of the semiconductor chip after removal of a holding jig shown inFIG. 29 ; -
FIG. 31 is an enlarged plan view showing the state of the adhesive material arranged at the back surface of the semiconductor chip and its surroundings shown inFIG. 20 ; -
FIG. 32 is an enlarged cross-sectional view taken along the line A-A ofFIG. 31 ; -
FIG. 33 is an exemplary explanatory diagram showing the outline of an assembly process of a laminated body of the memory chips shown inFIG. 4 ; -
FIG. 34 is an exemplary explanatory diagram showing the outline of another assembly process of the laminated body of the memory chips, following the process shown inFIG. 33 ; -
FIG. 35 is an enlarged plan view showing the state of the laminated body mounted over the back surface of the logic chip shown inFIG. 31 ; -
FIG. 36 is an enlarged cross-sectional view taken along the line A-A ofFIG. 35 ; -
FIG. 37 is an exemplary explanatory diagram showing a main part of a second chip conveying step shown inFIG. 13 ; -
FIG. 38 is an exemplary explanatory diagram showing a main part of a second mark detection step shown inFIG. 13 ; -
FIG. 39 is an exemplary explanatory diagram showing a main part of a second alignment step shown inFIG. 13 ; -
FIG. 40 is an exemplary explanatory diagram showing the state of the logic chip moved to the wiring substrate after the second alignment step shown inFIG. 39 ; -
FIG. 41 is an explanatory diagram showing the state of heating jig pressed against the back surface of the semiconductor chip after removal of a holding jig shown inFIG. 40 ; -
FIG. 42 is an enlarged cross-sectional view showing the state of the stacked semiconductor chips sealed with a seal member formed over the wiring substrate shown inFIG. 36 ; -
FIG. 43 is a plan view showing the entire structure of the seal member shown inFIG. 42 ; -
FIG. 44 is an enlarged cross-sectional view showing the state of solder balls bonded to lands of the wiring substrate shown inFIG. 37 ; -
FIG. 45 is a cross-sectional view showing the singulated multi-piece wiring substrate shown inFIG. 44 ; -
FIG. 46 is an enlarged plan view showing the alignment mark shown inFIGS. 3 , 7, 9, and 10; -
FIG. 47 is an enlarged plan view showing a first modified example corresponding toFIG. 46 ; -
FIG. 48 is an enlarged plan view showing a second modified example corresponding toFIG. 46 ; -
FIG. 49 is an enlarged plan view showing a third modified example corresponding toFIG. 46 ; -
FIG. 50 is a plan view showing a modified example corresponding toFIG. 8 ; and -
FIG. 51 is a perspective plan view showing a modified example corresponding toFIG. 3 . - In the present application, the following preferred embodiments may be described below by being divided into a plurality of sections or the like for convenience, if necessary, which are not independent from each other unless otherwise specified. Regardless of the order of the description of these sections, the sections indicate respective parts in a single example. Alternatively, one of the sections may be the details of a part of the other, or a modified example of a part or all of the other. In principle, parts having the same function will not be described repeatedly. Respective components of the preferred embodiments are not essential unless otherwise specified, except when limiting the number of the components in theory, and except when considered not to be definitely so from the context thereof.
- Similarly, in the description of the embodiments, the term “X formed of A” or the like as to material, composition, and the like does not exclude elements other than the element “A”, unless otherwise specified and except when considered not to be definitely so from the context. For example, as to the component, the above term means “X containing A as a principal component”. For example, the term “silicon member” is not limited to pure silicon, and may obviously include a SiGe (silicon-germanium) alloy, or a multi-component alloy containing silicon as a principal component, and another additive. The term “gold plating”, “Cu layer”, or “nickel plating” is not limited to pure one, but include a member containing gold, Cu, or nickel as a principal component unless otherwise specified.
- Even when referring to a specific numeral value or amount, the number of elements or the like may be greater than, or less than the specific numeral number, unless otherwise specified, except when limited to the specific number in theory, and except when considered not to be definitely so from the context.
- In each drawing of the embodiments, the same or like parts are indicated by the same or similar reference character or number, and its description will not be repeated in principle.
- In the accompanying drawings, even a cross-sectional view may omit hatching in some cases if the hatching possibly makes the sectional view complicated, or when a cavity or hole is easy to discriminate. In this context, the outline of a hole closed in a planar manner to and from the background may be omitted when clearly seen from the description or the like. Further, in order to represent a part which is not a cavity or hole, or in order to clearly represent a boundary between regions, a hatching or dot pattern is sometimes given even when the figure is not a cross-sectional view.
- The following embodiments will describe as one example of a SiP semiconductor device, a semiconductor package in which a semiconductor chip (memory chip) with a memory circuit and another semiconductor chip (control chip) with a control circuit for controlling the operation of the memory circuit are mounted.
- As an example of the semiconductor device with a laminated body of a plurality of semiconductor chips, this embodiment is directed to a semiconductor device in which semiconductor chips having memory circuits formed therein are stacked over another semiconductor chip with an arithmetic processing circuit formed therein.
FIG. 1 shows a perspective view of a semiconductor device of this embodiment.FIG. 2 is a bottom view of the semiconductor device shown inFIG. 1 .FIG. 3 shows a perspective plan view of an internal structure of the semiconductor device over a wiring substrate with a seal member shown inFIG. 1 removed therefrom.FIG. 4 is a cross-sectional view taken along the line A-A ofFIG. 1 . For easy understanding, although the number of terminals is shown to be small inFIGS. 1 to 4 , the number of terminals (bonding leads 2 f, lands 2 g, and solder balls 5) are not limited to those shown inFIGS. 1 to 4 .FIG. 3 shows the contour of the logic chip LC by a dotted line for easy understanding of the positional relationship and a difference in planar size between the logic chip LC and the memory chip MC4 in plan view. - As shown in
FIG. 4 , thewiring substrate 2 has an upper surface (surface, chip mounting surface) 2 a over whichsemiconductor chips 3 are mounted, a lower surface (surface, mounting surface) 2 b opposite to theupper surface 2 a, andside surfaces 2 c positioned between theupper surface 2 a and thelower surface 2 b. Thewiring substrate 2 has a quadrilateral contour in the plan view shown inFIGS. 2 and 3 . In the example shown inFIGS. 2 and 3 , thewiring substrate 2 is formed in a square shape having a planar size (size in plan view, sizes of theupper surface 2 a and thelower surface 2 b, contour size) of about 14 mm on a side. Thewiring substrate 2 has a thickness (height), that is, a distance from theupper surface 2 a to thelower surface 2 b shown inFIG. 4 , for example, of about 0.3 to 0.5 mm. - The
wiring substrate 2 serves as an interposer for electrically connecting thesemiconductor chips 3 mounted over theupper surface 2 a side with a mounting substrate (not shown), and has a plurality of wiring layers (four layers in an example shown inFIG. 4 ) electrically connecting theupper surface 2 a side with thelower surface 2 b side. The respective wiring layers have insulatinglayers 2 e for insulating theadjacent wirings 2 d as well as the adjacent wiring layers from each other. Thewiring substrate 2 of this embodiment includes three insulatinglayers 2 e with thecenter insulating layer 2 e made of a core layer (core material). Alternatively, thewiring substrate 2 may be the so-called coreless substrate which does not have the core insulatinglayer 2 e. Thewirings 2 d includewirings 2d 1 formed on the upper or lower surface of the insulatinglayer 2 e, and viawirings 2d 2 serving as an interlayer conductive path penetrating the insulatinglayer 2 e in the thickness direction. - The
upper surface 2 a of thewiring substrate 2 is provided with a plurality of bonding leads (terminals, chip mounting surface side terminals, electrodes) 2 f, which are terminals electrically connected with thesemiconductor chips 3. On the other hand, thelower surface 2 b of thewiring substrate 2 is provided withlands 2 g bonded to thesolder balls 5 serving as a terminal for electrical connection to the mounting substrate (not shown), that is, as a terminal for external connection of thesemiconductor device 1. The bonding leads 2 f are electrically connected with thelands 2 g via thewirings 2 d. Thewiring 2 d connected with thebonding lead 2 f orland 2 g is integrally formed with thebonding lead 2 f orland 2 g, whereby thebonding lead 2 f andland 2 g shown inFIG. 4 serve as a part of thewiring 2 d. - The
upper surface 2 a andlower surface 2 b of thewiring substrate 2 are covered with insulating films (solder resist films) 2 h and 2 k. Thewiring 2 d formed on theupper surface 2 a of thewiring substrate 2 is covered with the insulatingfilm 2 h. The insulatingfilm 2 h is provided with openings, and at the openings, at least parts (bonding portions or bonding regions with the semiconductor chip 3) of the bonding leads 2 f are exposed from the insulatingfilm 2 h. Thewiring 2 d formed on thelower surface 2 b of thewiring substrate 2 is covered with the insulatingfilm 2 k. The insulatingfilm 2 k is provided with openings, and at the openings, at least parts (bonding portions with the solder balls 5) of thelands 2 g are exposed from the insulatingfilm 2 k. - As shown in
FIG. 4 , the solder balls (external terminals, electrodes, external electrodes) 5 bonded to thelands 2 g at thelower surface 2 b of thewiring substrate 2 are arranged in columns and rows (array or matrix) as shown inFIG. 2 . Although not shown inFIG. 2 , thelands 2 g bonded to the solder balls 5 (seeFIG. 4 ) are also arranged in columns and rows (matrix). The semiconductor device including the external terminals (solder balls 5, lands 2 g) arranged on the mounting surface side of thewiring substrate 2 in the columns and rows are called an area array type semiconductor device. The area array type semiconductor device can effectively use the mounting surface (lower surface 2 b) side of thewiring substrate 2 as an arrangement space for the external terminals. Thus, the area array type semiconductor device is preferable in that an increase in mounting area of the semiconductor device can be suppressed even when the number of external terminals is increased. That is, together with higher functionality and higher integration, the semiconductor device whose number of external terminals is increased can be mounted while saving space. - The
semiconductor device 1 includes thesemiconductor chips 3 mounted over thewiring substrate 2. The semiconductor chips 3 are stacked over theupper surface 2 a of thewiring substrate 2. Each of thesemiconductor chips 3 has a quadrilateral contour in plan view ofFIG. 3 having a front surface (main surface, upper surface) 3 a, a back surface (main surface, lower surface) 3 b opposite to thefront surface 3 a, andside surfaces 3 c positioned between thefront surface 3 a and theback surface 3 b. The semiconductor chips are stacked, which can reduce amounting area of thesemiconductor device 1 even when thesemiconductor device 1 has higher functionality. - In the example shown in
FIGS. 3 and 4 , thesemiconductor chip 3 mounted as the lowermost level (in a position closest to the wiring substrate 2) is a logic chip (semiconductor chip) LC in which the arithmetic processing circuit PU (seeFIG. 5 ) is formed. The semiconductor chips 3 mounted at the upper level of the logic chip LC include memory chips (semiconductor chips) MC1, MC2, MC3, and MC4 provided with main storage circuits (storage circuits) MM (seeFIG. 5 ) for storing therein data for communicating with the logic chip LC. The logic chip LC is provided with a control circuit for controlling the operation of the main storage circuit of each of the memory chips MC1, MC2, MC3, and MC4, in addition to the above arithmetic processing circuit. The example of the circuit structure of thesemiconductor device 1 will be described below. - As shown in
FIG. 4 , an adhesive material NCL (insulating adhesive material) is arranged between theadjacent semiconductor chips 3. The adhesive material NCL is disposed to fill in a space between afront surface 3 a of theupper semiconductor chip 3 and aback surface 3 b of the lower semiconductor chip 3 (or,upper surface 2 a of the wiring substrate 2). Specifically, the adhesive material NCL includes an adhesive material (insulating adhesive material) NCL1 bonding and fixing the logic chip LC to thewiring substrate 2, and an adhesive material (insulating adhesive material) NCL2 bonding and fixing a laminated body MCS with the memory chips MC1, MC2, MC3, and MC4 to over the logic chip. The adhesive materials NCL1 and NCL2 are made of insulating (non-conductive) material (for example, resin). The adhesive material NCL is disposed at a bonding portion between the logic chip LC and thewiring substrate 2, and another bonding portion between the laminated body MCS and the logic chip LC, which can electrically insulate the electrodes provided in the respective bonding portions from each other. - In the example shown in
FIG. 4 , a seal member (seal member for the chip laminated body, resin for the chip laminated body) 6 other than theseal member 4 is arranged between the memory chips MC1, MC2, MC3, and MC4, so that the laminated body MCS of the memory chips MC1, MC2, MC3, and MC4 are sealed with theseal member 6. Theseal member 6 is embedded in intimate contact with thefront surface 3 a and theback surface 3 b of each of the memory chips MC1, MC2, MC3, and MC4. The laminated body MCS of the memory chips MC1, MC2, MC3, and MC4 is integral with each other by the bonding portion and theseal member 6 between therespective semiconductor chips 3. Theseal member 6 is made of insulating (non-conductive) material (for example, resin). Theseal member 6 is arranged in the respective bonding portions between the memory chips MC1, MC2, MC3, and MC4, so that the electrodes provided in the bonding portions can be electrically insulated from each other. As shown inFIG. 4 , thesurface 3 a of the memory chip MC1 mounted at the lowermost level (in the position closest to the logic chip LC) in the laminated body MCS of the memory chips MC1, MC2, MC3, and MC4 is exposed from theseal member 6. As shown inFIGS. 3 and 4 , theback surface 3 b of the memory chip MC4 mounted at the uppermost level in the laminated body MCS of the memory chips MC1, MC2, MC3, and MC4 is exposed from theseal member 6. - The
semiconductor device 1 includes theseal member 4 for sealing thesemiconductor chips 3. Theseal member 4 has a quadrilateral contour in plan view having an upper surface (surface, front surface) 4 a, a lower surface (surface, back surface, mounting surface) 4 b opposite to theupper surface 4 a (seeFIG. 4 ), andside surfaces 4 c positioned between theupper surface 4 a and thelower surface 4 b. In the example shown inFIG. 1 , the planar size of the seal member 4 (size in plan view from theupper surface 4 a side, contour size of theupper surface 4 a) is the same as that of thewiring substrate 2. Theside surface 4 c of theseal member 4 continues into theside surface 2 c of thewiring substrate 2. In the example shown inFIG. 1 , theseal member 4 has a square shape having a planar size (size in plan view) of, for example, about 14 mm on a side. - The
seal member 4 is a resin member for protecting thesemiconductor chips 3. Theseal member 4 is formed between thesemiconductor chips 3 and in intimate contact with thesemiconductor chip 3 and thewiring substrate 2, which can suppress the damage on thethin semiconductor chips 3. Theseal member 4 is preferably formed, for example, of the following material from the viewpoint of improving the function as the protective member. Theseal member 4 is required to be easily brought into intimate contact with thesemiconductor chip 3, thewiring substrate 2, and the space between thesemiconductor chips 3, and to have some hardness after sealing. For this reason, theseal member 4 preferably contains thermosetting resin, such as epoxy resin. In order to improve the function of theseal member 4 after being cured, for example, filler particles, such as silica (silicon dioxide: SiO2) particles, are preferably mixed in the resin material. For example, the ratio of addition of the filler particles to the mixture is preferably adjusted to make a linear expansion coefficient of thesemiconductor chip 3 equal to that of theseal member 4 for the purpose of suppressing the damage on thesemiconductor chip 3 due to thermal deformation of theseal member 4 andchip 3 after formation of theseal member 4. - Next, an example of the circuit structure of the
semiconductor device 1 will be described. As shown inFIG. 5 , the logic chip LC is provided with the control circuit CU for controlling the operation of the main storage circuit MM of each of the memory chips MC1, MC2, MC3, and MC4, in addition to the above arithmetic processing circuit PU. The logic chip LC is provided with an auxiliary storage circuit (storage circuit) SM having a smaller storage capacity than that of the main storage circuit MM, such as a cash memory for preliminary storing data. Referring toFIG. 5 , the arithmetic processing circuit PU, the control circuit CU, and the auxiliary storage circuit SM are collectively called a core circuit (main circuit) CR1 by way of example. The circuits included in the core circuit CR1 may include circuits other than the circuit described above. - The logic chip LC is provided with an external interface circuit (external input/output circuit) GIF for inputting and outputting a signal to and from an external device (not shown). The external interface circuit GIF is connected with a signal line SG for transmitting a signal between the logic chip LC and the external device (not shown). The external interface circuit GIF is electrically connected with a core circuit CR1, and the core circuit CR1 can transfer a signal to the external device via the external interface circuit GIF.
- The logic chip LC is provided with an internal interface circuit (internal input/output circuit) NIF for inputting and outputting a signal to and from internal devices (for example, memory chips MC1, MC2, MC3, and MC4). The internal interface circuit NIF is connected with a data line (signal line) DS for transmitting a data signal, an address line (signal line) AS for transmitting an address signal, and a signal line OS for transmitting other signals. The date line DS, the address line AS, and the signal line OS are respectively connected with the internal interface circuit NIF of each of the memory chips MC1, MC2, MC3, and MC4. Referring to
FIG. 5 , circuits for inputting/outputting a signal to and from electronic parts other than the logic chip LC, such as an external interface circuit GIF or an internal interface circuit NIF, are designated by an input/output circuit NS1. - The logic chip LC includes a power supply circuit DR for supplying a potential for driving a core circuit CR1 and the input/output circuit NS1. The power supply circuit DR includes a power supply circuit (power supply circuit for input/output) DR1 for supplying a voltage for driving the input/output circuit NS1 of the logic chip LC, and a power supply circuit (power supply circuit for a core) DR2 for supplying a voltage for driving the core circuit CR1 of the logic chip LC. For example, a plurality of different potentials (first power supply potential and second power supply potential) are supplied to the power supply circuit DR, and a difference in potential defines each of voltages to be applied to the core circuit CR1 and the input/output circuit NS1.
- One
semiconductor chip 3 into which circuits required for the operation of a device or system are integrated, such as a logic chip LC, is called a SoC (system on a chip). The formation of the main storage circuit MM shown inFIG. 5 in the logic chip LC can form the system using only one logic chip LC. However, a necessary capacity of the main storage circuit MM (seeFIG. 5 ) varies depending on the device or system to be operated. For this reason, the formation of the main storage circuit MM in thesemiconductor chip 3 other than the logic chip LC can improve the general versatility of the logic chip LC. - The memory chips MC1, MC2, MC3, and MC4 are connected according to a storage capacity required for the main storage circuit MM, which improves the flexibility in designing the capacity of the storage circuit included in the system. In the example shown in
FIG. 5 , each of memory chips MC1, MC2 MC3, and MC4 is provided with the corresponding main storage circuit MM. Referring toFIG. 5 , the main storage circuit MM is designated as a core circuit (main circuit) CR2 of the memory chips MC1, MC2, MC3, and MC4. The circuits included in the core circuit CR2 may include circuits other than the main storage circuit MM. - Each of the memory chips MC1, MC2, MC3, and MC4 is provided with an internal interface circuit (internal input/output circuit) NIF for inputting and outputting a signal to and from internal devices (for example, logic chip LC). Referring to FIG. 5, the internal interface circuit NIF for inputting and outputting the signal to and from electronic devices other than the memory chips MC1, MC2, MC3, and MC4 is designated as the input/output circuit NS2.
- Each of the memory chips MC1, MC2, MC3, and MC4 includes a power supply circuit (drive circuit) DR for supplying a potential for driving the core circuit CR2 and the input/output circuit NS2. The power supply circuit DR includes a power supply circuit (power supply circuit for input/output) DR3 for supplying a voltage for driving the input/output circuit NS2 of each of the memory chips MC1, MC2, MC3, and MC4, and a power supply circuit (power supply circuit for a core) DR4 for supplying a voltage for driving the core circuit CR2 of each of the memory chips MC1, MC2, MC3, and MC4. For example, a plurality of different potentials (first power supply potential and second power supply potential) are supplied to the power supply circuit DR, and a difference in potential defines each of voltages to be applied to the core circuit CR2 and the input/output circuit NS2.
- In the example shown in
FIG. 5 , the power circuit DR1 of the logic chip LC and the power circuit DR3 of the memory chips MC1, MC2, MC3, and MC4 are shared. In other words, the same voltage supplied from the power wire V2 is applied to the input/output circuit NS1 of the logic chip LC and the input/output circuit NS2 of each of the memory chips MC1, MC2, MC3, and MC4 to drive the input/output circuits. In this way, a part or all of the power supply circuits DR can be shared to thereby decrease the number of power wires V1, V2, and V3 for supplying a potential (driving voltage) to the power supply circuit. The decrease in number of the power wires V1, V2, and V3 can decrease the number of electrodes formed in the logic chip LC. - One
semiconductor device 1 into which circuits required for the operation of a device or system are integrated, such as thesemiconductor device 1, is called a SiP (system in Package). Although four memory chips MC1, MC2, MC3, and MC4 are stacked over one logic chip LC in the example shown inFIG. 4 , in various modified examples, the number of stackedsemiconductor chips 3 can be modified as mentioned above. Although not shown in the figure, for example, the invention can be applied to a modified example as the minimum structure in which one memory chip MC1 is mounted over one logic chip LC. - From the viewpoint of improving the general versatility of the logic chip LC and the memory chips MC1, MC2, MC3, and MC4, the planar size (size in plan view, size of the
front surface 3 a andback surface 3 b, or contour size) of each of the logic chip LC and the memory chips MC1, MC2, MC3, and MC4 is preferably minimized as long as the function of eachsemiconductor chip 3 can be achieved. The logic chip LC can have its planar size reduced by improving the integration degree of the circuit elements. Each of the memory chips MC1, MC2, MC3, and MC4 changes the capacity and transmission rate (for example, the amount of data transmitted depending on the width of a data bus) of the main storage circuit MM according to the planar size of the chip, and thus is limited in reduction of the planar size. - In the example shown in
FIG. 4 , the planar size of the memory chip MC4 is larger than that of the logic chip LC. For example, thememory chip MC 4 has a quadrilateral planar shape having a size of about 8 to 10 mm on a side, while the logic chip LC has a quadrilateral planar shape having a size of about 5 to 6 mm on a side. Although not shown, the planar size of each of the memory chips MC1, MC2, and MC3 shown inFIG. 4 is the same as that of the memory chip MC4. - As mentioned above, the logic chip LC is provided with an external interface circuit GIF for inputting and outputting signal to and from an external device (not shown). From the viewpoint of reducing a transmission distance to the external device, the order of stacking the
semiconductor chips 3 is preferably set such that the logic chip LC is positioned at the lowermost level, or closest to thewiring substrate 2. That is, the semiconductor chips 3 (memory chips MC1, MC2, MC3, MC4) having a larger planar size are preferably stacked over the semiconductor chip 3 (logic chip LC) having a smaller planar size, like thesemiconductor device 1. - The following will refer to the details of the logic chip LC and the memory chips MC1, MC2, MC3, and MC4 shown in
FIG. 4 , and the electric connection of thesemiconductor chips 3.FIG. 6 shows an enlarged cross-sectional view of a portion “A” shown inFIG. 4 .FIG. 7 shows a plan view of the front surface side of the memory chip shown inFIG. 4 .FIG. 8 shows a plan view of one example of the back surface side of the memory chip shown inFIG. 7 .FIG. 9 shows a plan view of the front surface side of the logic chip shown inFIG. 4 .FIG. 10 shows a plan view of one example of the back surface side of the logic chip shown inFIG. 9 . For easy understanding, although the number of electrodes is shown to be small inFIGS. 6 to 10 , the number of electrodes (front surface electrodes 3 ap, backsurface electrodes 3 bp, penetratingelectrodes 3 tsv) are not limited to those shown inFIGS. 6 to 10 .FIG. 8 shows the back surface diagram of the memory chips MC1, MC2, and MC3. The structure of the back surface of the memory chip MC4 (seeFIG. 4 ) without theback surface electrode 3 bp is shown inFIG. 3 , and thus its illustration will be omitted in the figure. - The inventors of the present application have studied about techniques for improving the performance of the SiP semiconductor device. One of the techniques is being studied to increase a signal transmission rate between the semiconductor chips mounted on SiP up to, for example, 12 Gbps (12 gigabit) or more. Methods for increasing a transmission rate between semiconductor chips mounted on the SiP includes a method which involves increasing the amount of data transmitted at one time by increasing the width of the data bus at the internal interface (hereinafter referred to as a “bus width increase”). Another method involves increasing the number of transmission per unit time (herein referred to as a “clock frequency increase”). Alternatively, a combination of the above bus width increase and the clock frequency increase can be applied. The
semiconductor device 1 described with reference toFIGS. 1 to 5 uses the combination of the bus width increase and the clock frequency increase to thereby increase the transmission rate up to 12 Gbps or more at the internal interface. - For example, each of the memory chips MC1, MC2, MC3, and MC4 shown in
FIG. 4 is the so-called wide I/O memory having a width of a data bus of 512 bit. Specifically, each of the memory chips MC1, MC2, MC3, and MC4 has a width of the date bus of 128 bit, so that the total bus width of the four channels is 512 bit. The number of transmission per unit time of each channel is increased to be, for example, 3 Gbps or more. - In using the combination of the clock frequency increase and the bus width increase, it is necessary to operate a number of data lines at high speeds. In order to reduce the influence of noise, the transmission distance for data needs to be shortened. As shown in
FIG. 4 , the logic chip LC is electrically connected with the memory chip MC1 via a conductive member arranged between the logic chip LC and the memory chip MC1. The memory chips MC1, MC2, MC3, and MC4 are electrically connected together via the conductive members arranged between the adjacent memory chips MC1, MC2, MC3, and MC4. In other words, thesemiconductor device 1 does not include thewiring substrate 2 and a wire (not shown) (bonding wire) in a transmission path between the logic chip LC and the memory chip MC1. Thesemiconductor device 1 does not include thewiring substrate 2 and a wire (not shown) (bonding wire) in transmission paths among the memory chips MC1, MC2, MC3, and MC4. - In this embodiment, the direct connection between the
semiconductor chips 3 employs the technique which involves forming penetrating electrodes penetrating thesemiconductor chip 3 in the thickness direction, and connecting thesemiconductor chips 3 together stacked via the penetrating electrodes. Specifically, as shown inFIG. 6 , the logic chip LC includes front surface electrodes (electrodes, pads, front surface side pads) 3 ap formed on itsfront surface 3 a, and back surface electrodes (electrodes, pads, back surface side pads) 3 bp formed on itsback surface 3 b. The logic chip LC further includes penetratingelectrodes 3 tsv penetrating from one of thefront surface 3 a and theback surface 3 b to the other and electrically connecting thefront surface electrodes 3 ap with theback surface electrodes 3 bp. - Various circuits (semiconductor element and wiring connected therewith) included in the
semiconductor chip 3 are formed on thefront surface side 3 a of thesemiconductor chip 3. Specifically, thesemiconductor chip 3 includes a semiconductor substrate (not shown) made of, for example, silicon (Si), and has a plurality of semiconductor elements (not shown), such as a transistor, formed on its main surface (element formation surface) of the semiconductor substrate. A wiring layer (not shown) including an insulating film for insulating the wirings from each other is stacked over the main surface of the semiconductor substrate (front surface 3 a side). A plurality of wirings of the wiring layers are respectively electrically connected with the semiconductor elements to form circuits. Thefront surface electrodes 3 ap formed on thefront surfaces 3 a (seeFIG. 4 ) of thesemiconductor chips 3 are electrically connected with the semiconductor elements via the wiring layers provided between the semiconductor substrate and thefront surface 3 a to thereby form parts of the circuits. - As shown in
FIG. 6 , each penetratingelectrode 3 tsv is formed to penetrate thesemiconductor chip 3 in the thickness direction, and thefront surface electrode 3 ap is electrically connected with theback surface electrode 3 bp via the penetratingelectrode 3 tsv, so that theback surface electrode 3 bp can be electrically connected with the circuit of thesemiconductor chip 3 formed on thefront surface 3 a side. That is, as shown inFIG. 6 , thefront surface electrode 3 ap of the memory chip MC1 is electrically connected with theback surface electrode 3 bp of the logic chip LC via the conductive member, such as an external terminal (protrusion electrode, conductive member, bump electrode) 7, so that the circuit of the memory chip MC1 is electrically connected with the circuit of the logic chip LC via the penetratingelectrodes 3 tsv. - In this embodiment, the logic chip LC mounted in between the memory chip MC1 and the
wiring substrate 2 includes a plurality of the penetratingelectrodes 3 tsv. The memory chip MC1 is electrically connected with the logic chip LC via the penetratingelectrodes 3 tsv, which can exclude thewiring substrate 2 and a wire (bonding wire) (not shown) from a transmission path between the logic chip LC and the memory chip MC1. As a result, an impedance component in the transmission path between the logic chip LC and the memory chip MC1 can be reduced to suppress the influence of noise caused by increase in clock frequency. In other words, even upon increasing the signal transmission rate between the logic chip LC and the memory chip MC1, the transmission reliability of the semiconductor device can be improved. - In the example shown in
FIG. 6 , the memory chips MC1, MC2, MC3, and MC4 are stacked over the logic chip LC. Thus, the signal transmission rate among the memory chips MC1, MC2, MC3, and MC4 is preferably improved or increased. The memory chips MC1, MC2, and MC3 under and above which therespective semiconductor chips 3 are arranged, among the memory chips MC1, MC2, MC3, and MC4, include a plurality of penetratingelectrodes 3 tsv, like the logic chip LC. Specifically, each of the memory chips MC1, MC2, and MC3 includes front surface electrodes (electrodes, pads) 3 ap formed on itsfront surface 3 a, and back surface electrodes (electrodes, pads) 3 bp formed on itsback surface 3 b. The respective memory chips MC1, MC2, and MC3 include the penetratingelectrodes 3 tsv penetrating from one of thefront surface 3 a and theback surface 3 b to the other and electrically connecting thefront surface electrodes 3 ap to theback surface electrodes 3 bp. - Like the above logic chip LC, the
front surface electrodes 3 ap of theupper semiconductor chip 3 among the memory chips MC1, MC2, MC3, and MC4 are electrically connected with theback surface electrode 3 bp of thelower semiconductor chip 3 via the conductive members, includingexternal terminals 7, so that the circuits of the stackedsemiconductor chips 3 are electrically connected with each other via the penetratingelectrodes 3 tsv. - The
respective semiconductor chips 3 are connected with each other via the external terminals 7 (solder material 7 a), which can exclude thewiring substrate 2 and the wires (bonding wires) (not shown) from the transmission paths among the memory chips MC1, MC2, MC3, and MC4. As a result, an impedance component in the transmission path between the stacked memory chips MC1, MC2, MC3, and MC4 can be reduced to suppress the influence of noise caused by increase in clock or clock frequency. In other words, even upon increasing the signal transmission rate among the memory chips MC1, MC2, MC3, and MC4, the transmission reliability of the semiconductor device can be improved. - In the example shown in
FIG. 6 , the memory chip MC4 mounted at the upperrmost level is preferably connected with the memory chip MC3, whereby thesurface electrodes 3 ap are formed, but theback surface electrodes 3 bp and the penetratingelectrodes 3 tsv are not necessarily formed. The memory chip MC4 mounted at the uppermost level takes the structure without having theback electrodes 3 bp and the penetratingelectrodes 3 tsv, which can simplify the manufacturing process of the memory chip MC4. Although not shown, in a modified example, the memory chip MC4 can include theback surface electrodes 3 bp and the penetratingelectrodes 3 tsv, like the memory chips MC1, MC2, and MC3. In this case, the stacked memory chips MC1, MC2, MC3, and MC4 have the same structure, which can improve the manufacturing efficiency. - Each
external terminal 7 is arranged between the adjacent stackedsemiconductor chips 3, and serves to electrically connect thefront surface electrode 3 ap of thesemiconductor chip 3 positioned at the upper level with theback surface electrode 3 bp of thesemiconductor chip 3 positioned at the lower level. Theexternal terminal 7 is made of the following material in the example shown inFIG. 6 . That is, theexternal terminal 7 electrically connecting the logic chip LC with thewiring substrate 2 includes a columnar (for example, cylindrical) member (protrudingelectrode 7 b) mainly formed of copper (Cu), and a metal member formed of a nickel (Ni) film and a solder (for example, SnAg) film (solder material 7 a) stacked on the tip of the columnar member. The electric connection of theexternal terminal 7 is established by bonding the solder film at its tip to theback surface electrode 3 bp. - In the example shown, the semiconductor chips are connected together via the
solder material 7 a without using the protrudingelectrode 7 b except for the bonding portion between the logic chip LC and thewiring substrate 2. In this case, thesolder material 7 a is bonded to the exposed surface of thefront surface electrode 3 ap, and can serve as the external terminal (bump electrode called a microbump). - In various modified examples, the material for the
external terminal 7 can be changed within the requirements of electrical characteristics, or bonding strength. For example, the protrudingelectrodes 7 b are formed on thefront surface electrode 3 ap of each of the memory chips MC1, MC2, MC3, and MC4, so that the chips can be electrically connected together via the protrudingelectrode 7 b and thesolder material 7 a. Alternatively, thesolder material 7 a is directly bonded to the exposed surface of thefront surface electrode 3 ap of the logic chip LC, so that the logic chip can be electrically connected with thewiring substrate 2 via thesolder material 7 a. - Like the logic chip LC and the memory chips MC1, MC2, and MC3 shown in
FIG. 6 , thesemiconductor chip 3 including the penetratingelectrodes 3 tsv is preferably thin, that is, the distance between thefront surface 3 a and theback surface 3 b is preferably thin (small). As the thickness of thesemiconductor chip 3 is decreased, the transmission distance of the penetratingelectrode 3 tsv is shortened, which is preferable in reducing the impedance component. When openings (including a through hole and a non-penetrating hole) are formed in the semiconductor substrate in the thickness direction, the processing accuracy is reduced with increasing depth of the opening. In other words, as thesemiconductor chip 3 becomes thinner, the processing accuracy of the opening for forming the penetratingelectrodes 3 tsv can be improved. As a result, the diameters of the penetratingelectrodes 3 tsv can be easily set to the same value (length or width of thesemiconductor chip 3 in the direction perpendicular to the thickness direction), which facilitates the control of the impedance components of the transmission paths. - In the example shown in
FIG. 6 , the thickness of the logic chip LC is smaller than that of the laminated body MCS (seeFIG. 4 ) formed of the memory chips MC1, MC2, MC3, and MC4 arranged over the logic chip LC. The thickness of the logic chip LC is also smaller than that of the memory chip MC4 located at the uppermost level of the memory chips MC1, MC2, MC3, and MC4, and not having the penetratingelectrode 3 tsv. For example, the thickness of the logic chip LC is 50 μm. In contrast, the thickness of the memory chip MC4 is in a range of about 80 to 100 μm. The thickness of the laminated body MCS (seeFIG. 4 ) made of the memory chips MC1, MC2, MC3, and MC4 is about 260 μm. - As mentioned above, in producing the
thinner semiconductor chip 3, the state of exposing thesemiconductor chip 3 might lead to damage on thesemiconductor chip 3. In this embodiment, as shown inFIG. 4 , theseal member 4 is brought into intimate contact with thesemiconductor chips 3 to seal the chips therewith. Thus, theseal member 4 can serve as a protection member for thesemiconductor chips 3 to suppress the damage on thesemiconductor chip 3. That is, according to this embodiment, thesemiconductor chips 3 can be sealed with resin to thereby improve the reliability (durability) of thesemiconductor device 1. - In the
semiconductor device 1 including the laminated body of thesemiconductor chips 3 with the penetratingelectrodes 3 tsv, the distance between thesemiconductor chip 3 and thewiring substrate 2 is preferably narrow from the viewpoint of reducing the transmission distance. For example, in the example shown inFIG. 6 , a distance between thefront surface 3 a of the logic chip LC and anupper surface 2 a of thewiring substrate 2 is, for example, in a range of about 10 to 20 μm. The distance between thefront surface 3 a of the memory chip MC1 and theupper surface 2 a of thewiring substrate 2 is, for example, in a range of about 70 to 100 μm. In thesemiconductor device 1 including the laminated body of thesemiconductor chips 3 with the penetratingelectrodes 3 tsv, it is preferable that the transmission distance is reduced by decreasing the distance between thesemiconductor chips 3 and the thickness of eachsemiconductor chip 3. - In the layout of the
front surface electrodes 3 ap and theback surface electrodes 3 bp in plan view of this embodiment, the transmission distance between the logic chip LC and each of the memory chips MC1, MC2, MC3, and MC4 can be reduced. - As shown in
FIG. 7 , thefront surface electrodes 3 ap included in the memory chips MC1, MC2, MC3, and MC4 are collectively arranged in the center of thefront surface 3 a. As shown inFIG. 8 , theback surface electrodes 3 bp included in the memory chips MC1, MC2, and MC3 are collectively arranged in the center of eachback surface 3 b. As shown inFIG. 6 , thefront surface electrodes 3 ap of the memory chips MC1, MC2, MC3, and MC4, and theback surface electrodes 3 bp of the memory chips MC1, MC2, and MC3 are arranged to be superimposed over each other in the thickness direction. - As shown in
FIG. 9 , some parts (front surface electrodes 3 ap 1) of thefront surface electrodes 3 ap included in the logic chip LC are collectively arranged in the center of thefront surface 3 a. Other parts (front surface electrodes 3 ap 2) of thefront surface electrodes 3 ap included in the logic chip LC are collectively arranged along each side (side surface 3 c) of thefront surface 3 a located at the peripheral edges of thefront surface 3 a. Thesurface electrodes 3ap 1 arranged in the center of thefront surface 3 a among thefront surface electrodes 3 ap shown inFIG. 9 are electrically connected with theback surface electrodes 3 bp via the penetratingelectrodes 3 tsv shown inFIG. 6 . That is, thefront surface electrodes 3ap 1 are electrodes for internal interface. Thesurface electrodes 3ap 2 arranged at the peripheral edges of thefront surface 3 a among thefront surface electrodes 3 ap shown inFIG. 9 are electrically connected with the external device (not shown) via thewiring substrate 2 shown inFIG. 4 . Specifically, thefront surface electrode 3ap 2 is electrically coupled to the bonding leads 2 f (seeFIG. 4 ) via theexternal terminals 7. That is, thefront surface electrodes 3ap 2 are electrodes for external interface. - From the viewpoint of reducing the transmission distance between the
respective semiconductor chips 3, preferably, as shown inFIG. 6 , thefront surface electrodes 3 ap and backsurface electrodes 3 bp for the internal interface are superimposed to each other in the thickness direction, and connected together via theexternal terminals 7. - As mentioned above, the planar size of the logic chip LC is smaller than that of each of the memory chips MC1, MC2, MC3, and MC4. In the
semiconductor device 1 shown inFIG. 3 , the center (central region) of theback surface 3 b of the logic chip LC is arranged to be superimposed over the center (central region) of the Memory chip MC4 in plan view. That is, in plan view, fourside surfaces 3 c of the memory chip MC4 are arranged outside fourside surfaces 3 c of the logic chip LC. In other words, thesemiconductor chips 3 are stacked over thewiring substrate 2 such that the fourside surfaces 3 c of the memory chip MC4 are positioned between the fourside surfaces 3 c of the logic chip LC and the fourside surfaces 2 c of thewiring substrate 2. The memory chips MC1, MC2, and MC3 shown inFIG. 4 are arranged to be superimposed over (in the same position) as the memory chip MC4 in plan view. - Thus, in plan view, the peripheral edges (peripheral edges of the
front surface 3 a and theback surface 3 b) of the memory chips MC1, MC2, MC3, and MC4 are arranged to be superimposed over the outer periphery of the logic chip LC. In other words, there is no logic chip LC existing between the periphery of each of the memory chips MC1, MC2, MC3, and MC4 and the wiring substrate 2 (see, for example,FIG. 4 ). - In order to arrange the front surface electrodes Sap and the
back surface electrodes 3 bp for internal interface of eachsemiconductor chip 3 shown inFIG. 6 such that thesurface electrodes 3 ap and 3 bp are superimposed over each other in the thickness direction, at least thefront surface electrodes 3 ap and backsurface electrodes 3 bp for the internal interface are preferably arranged to be superimposed over the logic chip LC in the thickness direction. As shown inFIG. 9 , thefront surface electrodes 3ap 2 for the external interface are arranged in the periphery of the logic chip LC. Thefront surface electrodes 3ap 1 for the internal interface are preferably arranged collectively in the center of thefront surface 3 a of the logic chip LC. - As shown in
FIG. 7 , a plurality of memory regions (storage circuit element arrangement regions) MR are formed on thefront surface 3 a side (specifically, main surface of the semiconductor substrate) of each of the memory chips MC1, MC2, MC3, and MC4. In the example shown inFIG. 7 , the four memory regions MR corresponding to the above four channels are formed. In each memory region MR, a plurality of memory cells (storage circuit elements) are arranged in an array. When as shown inFIG. 7 , thefront surface electrodes 3 ap are collectively arranged in the center of thefront surface 3 a, the memory regions MR for the four channels can be arranged to enclose a region with a group of the front surface electrodes arranged. As a result, the distances from the respective memory regions MR to thefront surface electrode 3 ap can be equalized. That is, preferably, the transmission distances of the channels can be made equal, which can decrease an error between the respective transmission rates for the channels. - When using the
front surface electrodes 3ap 1 collectively arranged in the center of thefront surface 3 a of the logic chip LC shown inFIG. 9 as electrodes for the internal interface, thefront surface electrode 3ap 1 can act without being electrically connected with thewiring substrate 2 shown inFIG. 6 . As shown inFIG. 6 , a part of thefront surface electrode 3ap 1 is electrically connected with thebonding lead 2 f of thewiring board 2, whereby the part of thefront surface electrode 3ap 1 can be preferably used as the electrode for the external interface. - For example, each of the memory chips MC1, MC2, MC3, and MC4 is provided with the power supply circuit DR for driving the main storage circuit MM shown in
FIG. 5 . The part of thefront surface electrode 3ap 1 shown inFIG. 9 is used as a terminal for supplying the power supply potential (first reference potential) and the reference potential (second reference potential different from the first reference potential, for example, ground potential) to the power supply circuits DR. In other words, in the example shown inFIG. 9 , thesurface electrodes 3ap 1 arranged in the center, of thefront surface 3 a of the logic chip LC include a first reference potential electrode to which the first reference potential (for example, power supply potential) is supplied, and a second reference potential electrode to which the second reference potential (for example, ground potential) different from the first reference potential is supplied. In other words, in the example shown inFIG. 9 , thefront surface electrodes 3ap 1 arranged in the center of thefront surface 3 a of the logic chip LC include power wires V2 and V3 (seeFIG. 5 ) for supplying a voltage for driving a circuit formed in the memory chip MC1. - In order to suppress the instability of the operation due to an instant voltage drop in increasing the signal transmission rate, the transmission distance between a power supply source and a circuit consuming the power is preferably shortened. Preferably, the parts of the
front surface electrodes 3ap 1 of the logic chip LC are electrically connected with thewiring board 2 to supply the first reference potential (for example, power supply potential) and the second reference potential (for example, ground potential), which can shorten the distance from each of the memory chips MC1, MC2, MC3, and MC4 with the circuit consuming the power to the corresponding driving circuit. The first reference potential (for example, power supply potential) is supplied to the first reference potential electrode. The second reference potential (for example, ground potential) different from the first reference potential is supplied to the second reference potential electrode. The first reference potential electrode and the second reference potential electrode are arranged such that thefront surface electrode 3 ap is preferably superimposed over theback surface electrode 3 bp in the thickness direction as shown inFIG. 6 . The first reference potential electrode is preferably electrically connected with the second reference potential electrode via the penetratingelectrodes 3 tsv. - The inventors of the present application have studied and found that when the
back surface electrode 3 bp of thelower semiconductor chip 3 faces and is electrically connected with thefront surface electrode 3 ap of theupper semiconductor chip 3 via theexternal terminals 7, the stacking of thesemiconductor chips 3 requires the high alignment accuracy. - As shown in
FIG. 6 , in order to electrically connect theback surface electrodes 3 bp of thelower semiconductor chip 3 to thefront surface electrodes 3 ap of theupper semiconductor chip 3, theback surface electrodes 3 bp need to be superimposed over thefront surface electrodes 3 ap in the thickness direction. When an area where theback surface electrode 3 bp is superimposed over thefront surface electrode 3 ap is small, the bonding area between thesolder material 7 a and theback surface electrode 3 bp (orfront surface electrode 3 ap) becomes small, which might reduce the electrical characteristics. Thus, it is necessary to improve the alignment accuracy between theback surface electrode 3 bp of thelower semiconductor chip 3 and thefront surface electrode 3 ap of theupper semiconductor chip 3. - Referring to
FIG. 10 , when the number of theback surface electrodes 3 bp of the logic chip LC is large, the number of the objects of interest requiring the adequate alignment accuracy is increased, which needs the high alignment accuracy. As a pitch between theback surface electrodes 3 bp of the logic chip LC (distance between the adjacent electrodes) becomes small, one electrode might come into contact with adjacent another electrode due to the misalignment. In this case, the alignment needs to be performed with higher accuracy. - For this reason, the inventors of the present application have studied techniques for improving the alignment accuracy in stacking the
semiconductor chips 3 over thewiring substrate 2.FIG. 11 is an exemplary explanatory diagram showing alignment means (alignment marks) used in assembling the semiconductor device shown inFIG. 4 .FIG. 12 is an explanatory diagram showing a studied example corresponding toFIG. 11 . - As shown in
FIG. 12 , the inventors of the present application have studied methods for forming alignment marks 50 in thewiring substrate 2 and thefront surfaces 3 a of thesemiconductor chips 3. In an alignment method shown inFIG. 12 , first, analignment mark 50 a formed in thewiring substrate 2 and analignment mark 50 b formed in thefront surface 3 a of the logic chip LC are detected (identified), and then aligned with each other. Thereafter, the logic chip LC is mounted over thewiring substrate 2. Then, thealignment mark 50 a formed in thewiring substrate 2 and analignment mark 50 d formed in thefront surface 3 a of the laminated body MCS are detected (identified), and then aligned with each other. Thereafter, the laminated body MCS is mounted over the logic chip LC. - In the method shown in
FIG. 12 , the alignment accuracy of the logic chip LC with respect to thewiring substrate 2 is defined mainly by a detection accuracy for detecting the positions of the alignment marks 50 a and 50 b, and a conveyance accuracy in mounting the logic chip LC. For example, when a pitch between the electrodes (distance between adjacent electrodes) is 60, the alignment accuracy of the logic chip LC with respect to thewiring substrate 2 is set within a tolerance of about ±5 to 10 μm, which can sufficiently ensure the electric connection reliability between the logic chip LC and thewiring substrate 2. - The alignment accuracy of the laminated body MCS with respect to the
wiring substrate 2 is defined mainly by a detection accuracy for detecting the positions of the alignment marks 50 a and 50 d, and a conveyance accuracy in mounting the laminated body MCS. Thus, the use of the same detection device and mounting device as those in the step of mounting the above logic chip LC can set the alignment accuracy of the laminated body MCS with respect to thewiring substrate 2 within about a tolerance of ±5 to 10 μm. - In the method shown in
FIG. 12 , the alignment accuracy of the laminated body MCS with respect to the logic chip LC might be in a range of ±10 to 20 μm, taking into consideration the misalignment between the laminated body MCS and the logic chip LC. The alignment accuracy between theback surface electrode 3 bp of the logic chip LC and thefront surface electrode 3 ap of the laminated body MCS needs to be determined by considering the formation positional accuracy of the respective electrodes. In the method shown inFIG. 12 , as the number of the stackedsemiconductor chips 3 is increased, the alignment accuracy might be reduced. Even when twosemiconductor chips 3 are stacked as shown inFIG. 12 , the reliability of electrical characteristics might be obviously reduced depending on a pitch between the electrodes. - The inventors of the present application have further studied and found the structure shown in
FIG. 11 . That is, the structure shown inFIG. 11 includes alignment marks 50 c formed on theback surface 3 b of the logic chip LC, in addition to the structure shown inFIG. 12 . In an alignment method shown inFIG. 11 , first, analignment mark 50 a formed in thewiring substrate 2 and analignment mark 50 b formed in thefront surface 3 a of the logic chip LC are detected (identified), and then aligned with each other. Thereafter, the logic chip LC is mounted over thewiring substrate 2. This point is the same as that in the alignment method shown inFIG. 12 . - Then, the
alignment mark 50 c formed in the logic chip LC and thealignment mark 50 d formed in thefront surface 3 a of the laminated body MCS are detected (identified), and then aligned with each other. Thereafter, the laminated body MCS is mounted over the logic chip LC. That is, the alignment method shown inFIG. 11 differs from the alignment method shown inFIG. 12 in that the alignment is performed based on thealignment mark 50 c formed in the logic chip LC. - In the alignment method shown in
FIG. 11 , thealignment mark 50 c formed on theback surface 3 b of the logic chip LC is used as the reference for alignment, so that the alignment accuracy of the laminated body MCS with respect to the logic chip LC does not need to take into consideration the misalignment of the logic chip LC. Thus, the use of the same detection device and mounting device as those in the above example can achieve the alignment accuracy within a tolerance of about ±5 to 10 μm. The alignment method shown inFIG. 11 can prevent the decrease in alignment accuracy even when the number of stackedsemiconductor chips 3 is increased. - That is, the alignment accuracy of the
upper semiconductor chip 3 with respect to thelower semiconductor chip 3 is not affected by the number of stackedsemiconductor chips 3. Even when the pitch between the electrodes is very small, the adequate alignment accuracy can be ensured according to the capability of a detector or mounting device, which can suppress the reduction in reliability of electrical characteristics between thesemiconductor chips 3. -
FIG. 3 andFIGS. 7 to 10 show an example of the layout of alignment marks 50 included in thesemiconductor device 1 of the first embodiment. The detailed structure of the alignment mark and preferred embodiments will be described in more detail in describing the manufacturing method of thesemiconductor device 1. - Next, a manufacturing process of the
semiconductor device 1 described with reference toFIGS. 1 to 11 will be described below. Thesemiconductor device 1 is manufactured along a flowchart ofFIG. 13 .FIG. 13 shows an exemplary explanatory diagram of the outline of the manufacturing process of the semiconductor device described with reference toFIGS. 1 to 11 . The details of the respective steps will be described below with reference toFIGS. 14 to 45 . - First, in a substrate provision step shown in
FIG. 13 , awiring substrate 20 is provided as shown inFIGS. 14 to 18 . FIG. 14 shows a plan view of the entire structure of a wiring substrate provided in the substrate provision step shown inFIG. 13 .FIG. 15 shows an enlarged plan view of one device region shown inFIG. 14 .FIG. 16 shows an enlarged cross-sectional view taken along the line A-A ofFIG. 15 .FIG. 17 shows an enlarged plan view of the surface opposite to that shown inFIG. 15 .FIG. 18 shows an enlarged cross-sectional view taken along the line B-B ofFIG. 15 . For easy understanding, although the number of terminals is shown to be small inFIGS. 14 to 17 , the number of terminals (bonding lands 2 f and lands 2 g) is not limited to those shown inFIGS. 14 to 17 . - As shown in
FIG. 14 , thewiring substrate 20 provided in this step includesdevice regions 20 a inside a frame (outer frame) 20 b. Specifically, a plurality ofdevice regions 20 a (27 regions inFIG. 14 ) are arranged in columns and rows. Each of thedevice regions 20 a corresponds to thewiring substrate 2 shown inFIGS. 1 to 4 . Thewiring substrate 20 is the so-called multi-piece substrate including a plurality ofdevice regions 20 a, and dicing lines (dicing regions) 20 c positioned between thedevice regions 20 a. In this way, the use of the multi-piece substrate including thedevice regions 20 a can improve the manufacturing efficiency. - As shown in
FIGS. 15 and 16 , eachdevice region 20 a is provided with the components of thewiring substrate 2 described with reference toFIG. 4 . Thewiring substrate 20 has anupper surface 2 a, alower surface 2 b opposite to theupper surface 2 a, and a plurality of wiring layers (four layers in the example shown inFIG. 4 ) electrically connecting theupper surface 2 a with thelower surface 2 b. The respective wiring layers have insulating layers (core layers) 2 e for insulating theadjacent wirings 2 d as well as the adjacent wiring layers from each other. Thewirings 2 d includewirings 2d 1 formed on the upper or lower surface of the insulatinglayer 2 e, and viawirings 2d 2 serving as an interlayer conductive path penetrating the insulatinglayer 2 e in the thickness direction. - As shown in
FIG. 15 , theupper surface 2 a of thewiring substrate 20 includes a chip mounting region (chip mounting portion) 2p 1 where the logic chip LC shown inFIG. 9 is to be mounted in a first chip mounting step shown inFIG. 13 . The chip mounting region 2p 1 exists in the center of thedevice region 20 a at theupper surface 2 a. In order to indicate the positions of the chip mounting region 2p 1, thedevice region 20 a, and the dicing lines 20 c inFIG. 15 , the contours of the chip mounting region 2p 1, thedevice region 20 a, and the dicing lines 20 c are represented by respective two-dot chain lines. The chip mounting region 2p 1 is a region where the logic chip LC is to be mounted as mentioned above, and thus does not necessarily include any boundary line actually viewable. Also, thedevice region 20 a and the dicing lines 20 c do not necessarily include any boundary lines actually viewable. - The
upper surface 2 a of thewiring substrate 20 is provided with a plurality of bonding leads (terminals, chip mounting surface side terminals, electrodes) 2 f. Thebonding lead 2 f is a terminal electrically connected with thefront surface electrodes 3 ap formed on thefront surface 3 a of the logic chip LC shown inFIG. 9 in the first chip mounting step as shown inFIG. 13 . In this embodiment, the logic chip LC is mounted with thefront surface 3 a side of the chip LC opposite to theupper surface 2 a of thewiring substrate 20 by the so-called face down mounting method, whereby the bonding portions of the bonding leads 2 f are formed within the chip mounting region 2p 1. - The
upper surface 2 a of thewiring substrate 20 is covered with the insulating film (solder resist films) 2 h. The insulatingfilm 2 h is provided withopenings 2 hw, and in theopenings 2 hw, at least parts (bonding portions or bonding regions with the semiconductor chip) of the bonding leads 2 f are exposed from the insulatingfilm 2 h. Although not shown, in this embodiment, an upper surface of the dicing line (dicing region) 20 c is also exposed from theinsulting film 2 h. With this arrangement, in use of a dicing blade (rotary blade) in the following singulating step (seeFIG. 13 ), the insulatingfilm 2 h can be prevented from being stuck in the dicing blade. That is, the cutting performance can be prevented from being reduced. If the cutting performance is not taken into consideration, the upper surface of the dicingline 20 c may be covered with the insulatingfilm 2 h. - As shown in
FIG. 17 , a plurality oflands 2 g are formed on thelower surface 2 b of thewiring substrate 20. Thelower surface 2 b of thewiring substrate 20 is covered with the insulating film (solder resist film) 2 k. The insulatingfilm 2 k is provided withopenings 2 kw, and at theopenings 2 kw, at least parts (bonding portions with the solder balls 5) of thelands 2 g are exposed from the insulatingfilm 2 k. Although not shown, in this embodiment, a lower surface of the dicing line (dicing region) 20 c is also exposed from theinsulting film 2 k. With this arrangement, in use of the dicing blade (rotary blade) in the following singulating step (seeFIG. 13 ), the insulatingfilm 2 h can be prevented from being stuck in the dicing blade. That is, the cutting performance can be prevented from being reduced. If the cutting performance is not taken into consideration, the upper surface of the dicingline 20 c may be covered with the insulatingfilm 2 k. - As shown in
FIG. 16 , the bonding leads 2 f are electrically connected with thelands 2 g via thewirings 2 d. These conductive patterns, including thewirings 2 d, the bonding leads 2 f, and thelands 2 g, are formed of metal, for example, mainly containing copper (Cu) Thewirings 2 d, the bonding leads 2 f, and thelands 2 g can be formed by the electrolytic plating. As shown inFIG. 16 , thewiring substrate 20 including four or more wiring layers (four layers in the example shown inFIG. 16 ) can be formed, for example, by the buildup technique. - As shown in
FIGS. 15 and 18 , alignment marks 50 a are formed on theupper surface 2 a of thewiring substrate 20. Thealignment mark 50 a is made of a pattern of interest to be detected in the first mark detection step shown inFIG. 13 . Thealignment mark 50 a is formed of the same metal as that of thewiring 2 d andbonding lead 2 f shown inFIG. 16 . In the example shown inFIG. 18 , anopening 2 hs is formed in the insulatingfilm 2 h, and thealignment mark 50 a is exposed from the insulatingfilm 2 h in theopening 2 hs. - As will be described later in details, in the first mark detection step shown in
FIG. 13 , the position of thealignment mark 50 a is detected using an image sensor, such as a camera. The metal pattern exposed from the insulatingfilm 2 h is used as thealignment mark 50 a, thereby improving the light reflection efficiency, which leads to the improvement of the position detection accuracy. The shape of the contour of thealignment mark 50 a may be identifiable. When the insulatingfilm 2 h is formed of material that can transmit visible light, thealignment mark 50 a may be covered with the insulatingfilm 2 h. - For example, when forming the bonding leads 2 f and the
wirings 2 d, thealignment mark 50 a can be formed at one time with the bonding leads 2 f and thewirings 2 d. - As shown in
FIG. 15 , thealignment mark 50 a is preferably formed within thedevice region 20 a. In modified examples of the embodiment, thealignment mark 50 a can be formed outside, thedevice region 20 a, that is, within the dicing line (dicing region) 20 c, or at aframe 20 b shown inFIG. 14 . As mentioned above, thealignment mark 50 a is used for aligning the logic chip LC with thewiring substrate 2 shown inFIG. 11 . Thus, thealignment mark 50 a is preferably formed in a region near the chip mounting region 2p 1 within thedevice region 20 a from the viewpoint of improving the positional accuracy in formation of thealignment mark 50 a. In the modified example, when thealignment mark 50 a is arranged within the dicing line (dicing region) 20 c, thealignment mark 50 a can hardly be covered with the adhesive material in the first adhesive material arrangement step to be described later. Thus, in order to surely expose thealignment mark 50 a, thealignment mark 50 a is preferably arranged within the dicingline 20 c. - As shown in
FIG. 15 , the alignment marks 50 a are preferably formed in positions within thedevice region 20 a. Even only onealignment mark 50 a can have its X and Y coordinates defined on a coordinate plane along theupper surface 2 a However, the formation of at least two or more alignment marks 50 a can specify the position of thedevice region 20 a in the direction having an angle θ, thereby improving the position detection accuracy. - In order to improve the accuracy for specifying the position in the direction θ, the distance between two alignment marks 50 a is preferably increased. Thus, more preferably, as shown in
FIG. 15 , onealignment mark 50 a is arranged at one corner of thedevice region 20 a, while theother alignment mark 50 a is arranged at the other corner of thedevice region 20 a facing the above corner. In other words, two alignment marks 50 a are preferably arranged on a diagonal line of thedevice region 20 a. - Then, in the first adhesive material arrangement step shown in
FIG. 13 , as shown inFIGS. 19 and 20 , an adhesive material NCL1 is arranged over the chip mounting region 2p 1 at theupper surface 2 a of thewiring substrate 20.FIG. 19 shows an enlarged plan view of the state of an adhesive material arranged in the chip mounting region shown inFIG. 15 .FIG. 20 is an enlarged cross-sectional view taken along the line A-A ofFIG. 19 . In order to indicate the positions of the chip mounting regions 2p 1 and 2p 2, thedevice region 20 a, and the dicing lines 20 c, the contours of the chip mounting regions 2p 1 and 2p 2, thedevice region 20 a, and the dicing lines 20 c are represented by respective two-dot chain lines. Each of the chip mounting regions 2p 1 and 2p 2 is a region where the logic chip LC and the laminated body MCS are to be mounted as mentioned above, and thus does not necessarily include any boundary line actually viewable. Also, thedevice region 20 a and the dicing lines 20 c do not necessarily include any boundary lines actually viewable. When illustrating the chip mounting regions 2p 1 and 2p 2, thedevice region 20 a, and the dicing lines 20 c in the plane view, these regions and lines do not necessarily include any boundary line actually viewable. - In general, when mounting the semiconductor chip over the wiring substrate by the face down mounting method (flip-chip bonding), the semiconductor chip is electrically connected with the wiring substrate, and then a connection part is sealed with resin (by a post-charging method). In this case, the resin is supplied from a nozzle positioned near a clearance between the semiconductor chip and the wiring substrate, so that the resin fills in the clearance by capillary action.
- In the example described in this embodiment, before mounting the logic chip LC (see
FIG. 9 ) over thewiring substrate 20 in the first chip mounting step to be described later, the adhesive material NCL1 is disposed over the chip mounting region 2p 1, and the logic chip LC is pressed against and electrically connected with thewiring substrate 20 from above the adhesive material NCL1 (a pre-coating method) to thereby mount the logic chip LC. - In the above post-charging method, the resin is embedded in the clearance by the capillary action, which takes a long processing time (time for charging the resin) to one
device region 20 a. In contrast, in the above pre-coating method, at the time when the tip of the logic chip LC (solder material 7 a formed at the tip of the protrudingelectrode 7 b shown inFIG. 6 ) comes into contact with the bonding portion of thebonding lead 2 f, the adhesive material NCL1 already fills in the clearance between thewiring substrate 20 and the logic chip LC. The pre-coating method is preferable in reducing the processing time for onedevice region 20 a and improving the manufacturing efficiency, as compared to the above post-charging method. - In a modified example of this embodiment, the order of the first chip mounting step and the first adhesive material arrangement step shown in
FIG. 13 is reversed, so that the post-charging method can be applied. For example, when there is a small area of product formation regions formed at one time, the use of the post-charging method does not differ so much from the pre-coating method in processing time, and thus can suppress the reduction in manufacturing efficiency. - The adhesive material NCL1 used in the pre-coating method is made of insulating (non-conductive) material (for example, resin) as described above.
- The adhesive material NCL1 is made of resin material that is cured (increases its hardness) by being subjected to energy. In this embodiment, the adhesive material NCL1 contains thermosetting resin. The adhesive material NCL1 before being cured is softer than the
external terminal 7 shown inFIG. 6 , and can be deformed by being pressed by the logic chip LC. - The adhesive materials NCL1 before being cured can be classified into two types based on the difference in handling. One of the adhesive materials is a paste-like resin (insulating paste) called a non-conductive paste (NCP). The resin discharged from the nozzle (not shown) is applied to the chip mounting region 2
p 1. The other of the adhesive materials is a film-like resin (insulating film) previously molded in the form of film, and which is called a non-conductive film (NCF). The film-like resin is conveyed and bonded to the chip mounting region 2p 1 while being in the state of film. The use of the insulating paste (NCP) does not need the step of bonding the insulating film (NCF), and thus can reduce stress on the semiconductor chips or the like as compared to the use of the insulating material film. The use of the insulating film (NCF) has the high shape retaining property as compared to the use of the insulating paste (NCP), which makes it easier to control the range and thickness of the adhesive material NCL1. - In the example shown in
FIGS. 19 and 20 , the adhesive material NCL1 made of the insulating film (NCF) is arranged at the chip mounting region 2p 1, and bonded in intimate contact with theupper surface 2 a of thewiring substrate 20. Although not shown, in another modified example, the insulating paste (NCP) can also be used. - In this step, as shown in
FIG. 19 , the adhesive material NCL1 is preferably arranged not to cover the alignment marks 50 a from the viewpoint of improving the light reflection efficiency of thealignment mark 50 a. When the adhesive material NCL1 is made of the material that transmits visible light, the alignment marks 50 a may be covered with the adhesive material NCL1. In this case, the flexibility in selecting the material for the adhesive material NCL1 is reduced. Thus, the alignment marks 50 a are preferably exposed from the adhesive material NCL1. - Since the alignment marks 50 a are exposed from the adhesive material NCL1, the alignment marks 50 a are preferably positioned in the periphery of the
device region 20 a. - In the first chip provision step shown in
FIG. 13 , the logic chip LC shown inFIGS. 9 and 10 is provided.FIG. 21 shows an exemplary explanatory diagram of the outline of a manufacturing process of the semiconductor chip with the penetrating electrode shown inFIG. 6 .FIG. 22 shows an exemplary explanatory diagram of the outline of another manufacturing process of the semiconductor chip, following the process shown inFIG. 21 . Referring toFIGS. 21 and 22 , a method for manufacturing the penetratingelectrode 3 tsv and theback surface electrode 3 bp electrically connected with the penetratingelectrode 3 tsv will be mainly described below. Illustration and description of formation steps of various circuits other than the penetratingelectrode 3 tsv will be omitted. The manufacturing method of the semiconductor chip shown inFIGS. 21 and 22 can also be applied to the manufacturing method of the memory chips MC1, MC2, and MC3 in addition to the logic chip LC shown inFIG. 4 . - First, in a wafer provision step, a wafer (semiconductor substrate) WH shown in
FIG. 21 is provided. The wafer WH is a semiconductor substrate made of, for example, silicon (Si), and has a circular shape in plan view. The wafer WH has a front surface (main surface, upper surface) WHs as a semiconductor element formation surface, and a back surface (main surface, lower surface) WHb opposite to the front surface WHs. The thickness of the wafer WH is larger than that of each of the logic chip LC and the memory chips MC1, MC2, and MC3 shown inFIG. 4 , and is, for example, about several hundreds of μm. - Then, in a hole formation step, holes (bores, openings) 3 tsh for forming the penetrating
electrodes 3 tsv shown inFIG. 6 are formed. In the example shown inFIG. 21 , amask 25 is arranged over the front surface WHs of the wafer WH, and subjected to etching to thereby form theholes 3 tsh. The semiconductor elements, including the logic chip LC, and the memory chips MC1, MC2, and MC3 shown inFIG. 4 can be formed, for example, after the present step and before a wiring layer formation step. - Then, metal material, such as copper (Cu), is embedded in the
holes 3 tsh to form the penetratingelectrodes 3 tsv. Next, in the wiring layer formation step, a wiring layer (chip wiring layer) 3 d is formed over the front surface WHs of the wafer WH. In this step, a plurality offront surface electrodes 3 ap shown inFIGS. 7 and 9 are formed, whereby the penetratingelectrodes 3 tsv are respectively electrically connected with thefront surface electrodes 3 ap. Thefront surface electrodes 3 ap and theuppermost wiring layer 3 d integrally formed with thefront surface electrode 3 ap are formed of a metal film of, for example, aluminum (Al). - In this step, the semiconductor elements, including the logic chip LC and the memory chips MC1, MC2, and MC3 shown in
FIG. 4 , are electrically connected with thefront surface electrodes 3 ap shown inFIGS. 7 and 9 via thewiring layer 3 d. In this way, the electric connection of the semiconductor elements, including the logic chip LC and the memory chips MC1, MC2, and MC3, is established via thewiring layer 3 d. - In this step, the
alignment mark 50 b shown inFIG. 9 or thealignment mark 50 d shown inFIG. 7 is formed. The alignment marks 50 b and 50 d can be formed of the same material (for example, aluminum) as that of thefront surface electrode 3 ap and theuppermost wiring layer 3 d. Thus, in forming thefront surface electrodes 3 ap, the marks, front surface electrodes, and uppermost wiring layer can be formed at one time. The details of thealignment mark 50 b shown inFIG. 9 and thealignment mark 50 d shown inFIG. 7 will be described later. - Then, in an external terminal formation step,
external terminals 7 are formed over thefront surface electrodes 3 ap (seeFIGS. 7 and 9 ). In this step, as shown inFIG. 6 , the protrudingelectrodes 7 b are formed over thefront surface electrodes 3 ap of the logic chip LC. Thesolder material 7 a is formed at the tip of each protrudingelectrode 7 b. Alternatively, thesolder material 7 a is formed over eachfront surface electrode 3 ap of the memory chip MC1. Thesolder material 7 a serves as a bonding material in mounting thesemiconductor chip 3 shown inFIG. 6 over thewiring substrate 2 or anotherlower semiconductor chip 3. - Then, in a back surface polishing step shown in
FIG. 22 , the back surface WHb (seeFIG. 21 ) of the wafer WH is polished to decrease the thickness of the wafer WH. Thus, theback surface 3 b of thesemiconductor chip 3 shown inFIG. 5 is exposed. In other words, the penetratingelectrode 3 tsv penetrates the wafer WH in the thickness direction. The penetratingelectrodes 3 tsv are exposed from the wafer WH at theback surface 3 b of the wafer WH. In the example shown inFIG. 22 , in the back polishing step, the back surface of the wafer WH is polished by a polishingjig 28 while the wafer WH is being supported by asupport base 26, such as a glass plate, and aprotective layer 27 for protecting theexternal terminal 7 protecting the front surface WHs side. - Then, in a back surface electrode formation step, a plurality of
back surface electrodes 3 bp are formed on theback surface 3 b to be electrically connected with the penetratingelectrodes 3 tsv. - In this step, the alignment marks 50 c shown in
FIG. 10 are formed. Thus, the alignment marks 50 c can be formed of the same material (for example, copper) as that of theback surface electrode 3 bp. The alignment marks 50 c can be formed at one time in forming theback surface electrodes 3 bp. The details of thealignment mark 50 c shown inFIG. 10 will be described later. - Then, in a singulating step, the wafer WH is divided along the dicing lines to produce a plurality of
semiconductor chips 3. Thereafter, each semiconductor chip is checked if necessary, whereby the semiconductor chips 3 (logic chip LC, and memory chips MC1, MC2, and MC3) shown inFIG. 4 are obtained. - Next, in a first chip mounting step shown in
FIG. 13 , as shown inFIGS. 23 and 24 , the logic chip LC is mounted over thewiring substrate 20. The first chip conveying step, the first mark detection step, and the first alignment step shown inFIG. 13 can be regarded as sub-steps included in the step of mounting the logic chip LC over thewiring substrate 20. Thus, in this embodiment, the first chip conveying step, the first mark detection step, and the first alignment step are described below as the sub-steps included in the first chip mounting step. -
FIG. 23 shows an enlarged plan view of the state of a logic chip LC mounted over the chip mounting region of the wiring substrate shown inFIG. 19 .FIG. 24 shows an enlarged cross-sectional view taken along the line A-A ofFIG. 23 .FIG. 25 shows an exemplary explanatory diagram of a main part of a first chip conveying step shown inFIG. 13 .FIG. 26 shows an exemplary explanatory diagram of a main part of a first mark detection step shown inFIG. 13 .FIG. 27 shows an enlarged cross-sectional view of a main part of a cross-sectional structure of alignment marks of the logic chip shown inFIG. 26 .FIG. 28 shows an exemplary explanatory diagram of a main part of a first alignment step shown inFIG. 13 .FIG. 29 shows an exemplary explanatory diagram of the state of the logic chip moved to the wiring substrate after the first alignment step shown inFIG. 28 .FIG. 30 shows an explanatory diagram of the state of a heating jig pressed against the back surface of the semiconductor chip after removal of a holding jig shown inFIG. 29 . - In this step, as shown in
FIG. 24 , the logic chip LC is mounted by the so-called face down mounting method (flip-chip bonding method) such that thefront surface 3 a of the logic chip LC faces theupper surface 2 a of thewiring substrate 20. In this step, the logic chip LC is electrically connected with thewiring substrate 20. Specifically, thefront surface electrodes 3 ap formed on thefront surface 3 a of the logic chip LC are electrically connected with the bonding leads 2 f formed on theupper surface 2 a of thewiring substrate 20 via the external terminals 7 (protrudingelectrode 7 b andsolder material 7 a shown inFIG. 6 ). The detailed flow of the respective steps will be described below with reference toFIGS. 25 to 30 . - As shown in
FIG. 25 , the first chip mounting step includes a first chip conveying step of arranging the logic chip LC (semiconductor chip 3) over the chip mounting region 2p 1 of thewiring substrate 20. The logic chip LC is conveyed to the chip mounting region 2p 1 with theback surface 3 b being held by a holding jig (collet) 30, and then arranged over the chip mounting region 2 p 1 (or adhesive material NCL1) such that thefront surface 3 a positioned on the element formation surface side faces theupper surface 2 a of thewiring substrate 20. The holdingjig 30 includes a holdingsurface 30 a for sucking and holding theback surface 3 b of the logic chip LC. The holdingjig 30 conveys the logic chip LC while holding the logic chip LC by the holdingsurface 30 a. - The protruding
electrodes 7 b are formed on thefront surface 3 a of the logic chip LC, and thesolder material 7 a is formed at the tip of each protrudingelectrode 7 b. Thesolder material 7 a serving as a bonding material for electrically connecting with the protrudingelectrode 7 b is previously formed at a bonding portion of thebonding lead 2 f formed on theupper surface 2 a of thewiring substrate 20. - Since in the first alignment step shown in
FIG. 13 , the logic chip LC is aligned with thewiring substrate 20 with higher accuracy in this embodiment, the accuracy of the conveying position in the first chip conveying step may be low. In order to improve the positional accuracy in the first alignment step, a movement distance in the first alignment step may be preferably small. For example, in the first chip conveying step, the logic chip LC is preferably arranged over theupper surface 2 a of thewiring substrate 20 with the accuracy that causes the logic chip LC to be arranged over the chip mounting region 2p 1. - As shown in
FIG. 26 , the first chip mounting step includes a first mark detection step of detecting (identifying) analignment mark 50 a of thewiring substrate 20 and analignment mark 50 b of the logic chip LC. As shown inFIG. 26 , in the first mark detection step, while the logic chip LC being arranged over thewiring substrate 20, a camera (mark position detector, image sensor, imaging device) 60 is arranged between thewiring substrate 20 and the logic chip LC, and then thealignment mark 50 a of thewiring substrate 20 and thealignment mark 50 b of the logic chip LC are detected (identified). - The
camera 60 includes alight receiver 60 a for receiving visible light, aphotoelectric conversion circuit 60 b for converting the light received by thelight receiver 60 a into an electric signal, and anoutput circuit 60 c for outputting the electric signal converted by the photoelectric conversion circuit to the outside. Thecamera 60 is electrically connected with acontroller 61. An electric signal output from theoutput circuit 60 c is transmitted to an image processing circuit (not shown) included in thecontroller 61. After image processing is performed, positional data (coordinate data) of the alignment mark (alignment marks 50 a, 50 b) are output. Positional data defining the positional relationship between thealignment mark 50 a and the bonding portion of thebonding lead 2 f, or another positional data defining the positional relationship between thealignment mark 50 b and thefront surface electrode 3 ap are input to thecontroller 61, which can calculate the position of the bonding portion of thebonding lead 2 f or thefront surface electrode 3 ap by thecontroller 61. - In the example shown in
FIG. 26 , a visible light reflector (mirror) 60 d is provided to reflect the visible light entering from the outside toward thelight receiver 60 a. Thecamera 60 is connected with a driving device (not shown), and movably fixed over thewiring substrate 20. Thus, a reflection angle of thevisible light reflector 60 d is changed and the position of thecamera 60 is moved, so that the positions of the alignment marks 50 a and 50 b can be detected (identified) by thesame camera 60. - In the example shown in
FIG. 26 , the alignment marks 50 a are formed over thewiring substrate 20, and the alignment marks 50 b are formed over thefront surface 3 a side of the logic chip LC. In this way, the alignment marks 50 are respectively formed on thewiring substrate 20 and thefront surface 3 a of the logic chip LC. The positions of the alignment marks 50 are respectively detected, which can provide coordinate data in the plane XY ofFIG. 19 , and also data on the direction θ indicative of the angle of inclination between the coordinate axes. The use of the coordinate date of thealignment mark 50 and the data on the direction θ can accurately calculate the position of the bonding leads 2 f (seeFIG. 15 ) or the position of thefront surface electrodes 3 ap (seeFIG. 9 ). - As shown in
FIG. 19 , the alignment marks 50 a are arranged at opposite corners such that the marks are positioned on one diagonal line of thedevice region 20 a forming the quadrilateral shape in plan view. As shown inFIG. 9 , the alignment marks 50 b are arranged at opposite corners such that the marks are positioned on one diagonal line of thefront surface 3 a forming the quadrilateral in plan view. The alignment marks 50 are arranged at the opposite corners in this way, which can improve the accuracy of the data in the direction θ. As a result, the accuracy of calculation of the positions of the bonding leads 2 f (seeFIG. 15 ), or calculation of the positions of thefront surface electrode 3 ap (seeFIG. 9 ) can be improved. - As shown in
FIG. 27 , thealignment mark 50 b formed in thefront surface 3 a of the logic chip LC is preferably exposed from the insulating film (protective film, passivation film) 3 p formed on the uppermost surface of the logic chip LC. In the example shown inFIG. 27 ,openings 3 ps are formed in the insulating film (protective film, passivation film) 3 p formed on the uppermost surface of the logic chip LC. The alignment marks 50 b are exposed from the insulatingfilm 3 p at theopening 3 ps. Referring toFIG. 27 , strictly, under the alignment marks 50 b, wiring layers are provided to electrically connect a semiconductor element formation surface (main surface) of the semiconductor substrate with the front surface electrodes Sap (seeFIG. 9 ).FIG. 27 shows only the wafer WH which is a semiconductor substrate while omitting the illustration of the wiring layer. - Like the
alignment mark 50 a described with reference toFIG. 18 , the metal pattern exposed from the insulatingfilm 3 p is used as thealignment mark 50 b, thereby improving the light reflection efficiency, which lead to the improvement of the position detection accuracy. The contour shape of thealignment mark 50 b has only to be identified. When the insulatingfilm 3 p is formed of material that can transmit visible light, thealignment mark 50 b may be covered with the insulatingfilm 3 p. - The
alignment mark 50 a has just been described above with reference toFIG. 18 , and a redundant description thereof will be omitted. - As shown in
FIG. 28 , the first chip mounting step includes a first alignment step of aligning thewiring substrate 20 with the logic chip LC. As shown inFIG. 28 , in the first alignment step, the logic chip LC is relatively moved with respect to thewiring substrate 20 along theupper surface 2 a of thewiring substrate 20, so that the bonding portions of the bonding leads 2 f face theexternal terminals 7 formed on thesurface 3 a of the logic chip LC. - As mentioned above, this embodiment can calculate the positions of the bonding leads 2 f and the positions of the
front surface electrodes 3 ap with high accuracy. Thus, based on the calculated data, the logic chip LC is relatively moved to thewiring substrate 20, which can perform the alignment with high accuracy. - In the example shown in
FIG. 28 , a method for moving the relative position of the logic chip LC and thewiring substrate 20 involves moving the holdingjig 30 holding the logic chip LC along theupper surface 2 a of thewiring substrate 20 as indicated by the arrow ofFIG. 28 . The relative positional relationship between the logic chip LC and thewiring substrate 20 has only to be changed. Thus, one or both of the logic chip LC and thewiring substrate 20 can be moved. - Then, as shown in
FIG. 29 , the logic chip LC is moved toward thewiring substrate 20. At this time, the adhesive material NCL1 is arranged over thewiring substrate 20 in its soft state before heating to cure. Thus, thefront surface 3 a side of the logic chip LC is filled with the adhesive material NCL1. In the example shown inFIG. 29 , thealignment mark 50 b formed on thefront surface 3 a side of the logic chip LC is covered with the adhesive material NCL1. The step of detecting thealignment mark 50 b (first mark detection step) is already completed, which does not cause any problem. - In the above first alignment step, the relative position of the logic chip LC and the
wiring substrate 20 is aligned with each other with high accuracy. When the logic chip LC is linearly moved along thewiring substrate 20, the bonding portions of the bonding leads 2 f can be kept opposite to theexternal terminals 7 formed on thesurface 3 a of the logic chip LC. - Then, as shown in
FIG. 30 , the logic chip LC is pressed against thewiring substrate 20 by pushing theheating jig 31 to theback surface 3 b of the logic chip LC. As mentioned above, since the adhesive material NCL1 is in its soft state before curing, the logic chip LC is pushed into by theheating jig 31, so that the logic chip LC approaches thewiring substrate 20. When the logic chip LC approaches thewiring substrate 20, the tip (specifically,solder material 7 a) of eachexternal terminal 7 formed on thefront surface 3 a of the logic chip LC comes into contact with the corresponding bonding region of thebonding lead 2 f (specifically,solder material 7 a). - The thickness of the adhesive material NCL1 is larger than at least the total of the height (height of a protruding part) of the
external terminal 7 and the thickness of thebonding lead 2 f. When pushed by theheating jig 31, the part of thefront surface 3 a side of the logic chip LC is embedded in the adhesive material NCL1. In other words, at least a part of thefront surface 3 a of the side surface of the logic chip LC is embedded in the adhesive material NCL1. - The alignment marks 50 c and the
back surface electrodes 3 bp are formed at the logic chip LC. It is necessary to prevent the alignment marks 50 c and backsurface electrode 3 bp from being covered by the adhesive material NCL1 coming around to theback surface 3 b side. As shown inFIG. 30 , a member (low elastic member), for example, a resin film (film) 32, which is softer than theheating jig 31 and the logic chip LC, preferably intervenes in between theheating jig 31 and the logic chip LC to cover theback surface 3 b of the logic chip LC with theresin film 32. By pressing the logic chip LC via theresin film 32, theresin film 32 comes into intimate contact with theback surface 3 b of the logic chip LC. Even though the adhesive material NCL1 is thick, the NCL1 can be prevented from coming around to theback surface 3 b of the logic chip LC. Theresin film 32 of this embodiment is made of, for example, fluorine resin. - Then, as shown in
FIG. 30 , the logic chip LC and the adhesive material NCL1 are heated by the heating jig (heat source) 31 while pushing theheating jig 31 to the logic chip LC. In the bonding portion between the logic chip LC and thewiring substrate 20, thesolder material 7 a on thebonding lead 2 f side and thesolder material 7 a on the external terminal side are respectively melted and integral with each other into a bonding material electrically connecting theexternal terminal 7 with thebonding lead 2 f. That is, the logic chip LC is heated by the heating jig (heat source) 31, so that the protrudingelectrodes 7 b are electrically connected with the bonding leads 2 f via thesolder material 7 a. - The adhesive material NCL1 is heated to be cured. As a result, the cured adhesive material NCL1 is obtained with a part of the logic chip LC embedded in the adhesive material. The alignment marks 50 c and the
back surface electrodes 3 bp of the logic chip LC are covered by theresin film 32, and exposed from the cured adhesive material NCL1. The adhesive material NCL1 is not necessarily cured completely by the heat from the heating jig (heat source) 31. In another embodiment, after curing (temporarily curing) a part of the thermosetting resin contained in the adhesive material NCL1 to such an extent that can fix the logic chip LC, thewiring substrate 20 is transferred to a heating furnace (not shown), and then the remainder of the thermosetting resin can be cured (fully cured). It takes some time to complete the full curing process for fully curing the entire thermosetting resin contained in the adhesive material NCL1. However, the full curing process can be performed in the heating furnace to improve the manufacturing efficiency. - Next, in a second adhesive material arrangement step shown in
FIG. 13 , as shown inFIG. 31 , the adhesive material NCL2 is arranged at theback surface 3 b of the logic chip LC (semiconductor chip 3).FIG. 31 shows an enlarged plan view of the state of the adhesive material arranged at the back surface of the semiconductor chip shown inFIG. 20 and its surroundings.FIG. 32 shows an enlarged cross-sectional view taken along the line A-A ofFIG. 31 . - As shown in
FIG. 6 , in thesemiconductor device 1 of this embodiment, the logic chip LC mounted at the lowermost level (for example, at the first level) and the memory chip MC1 mounted at the second level from the bottom among the stackedsemiconductor chips 3 are mounted by the face down mounting method (flip-chip bonding). Thus, as described in the above first adhesive material arrangement step, the use of the above-mentioned pre-coating method is preferable in that the processing time for onedevice region 20 a (seeFIGS. 31 and 32 ) can reduced to improve the manufacturing efficiency. - The adhesive material NCL2 to be used in the pre-coating method is made of insulating (non-conductive) material (for example, resin material) as described above. The adhesive material NCL2 is made of resin that will be cured (increase its hardness) by being supplied with the energy. In this embodiment, the adhesive material NCL2 includes, for example, thermosetting resin. The adhesive material NCL2 before being cured is softer than the protruding
electrode 7 b shown inFIG. 6 , and is deformed by being pressed against the logic chip LC. - The adhesive materials NCL2 before being cured can be classified into two types based on the difference in handling, namely, a paste-like resin (insulating paste) called the NCP, and a film-like resin (insulating film) previously molded in the shape of film and called the NCF. The adhesive material NCL2 used in this step can employ either the NCP or the NCF. In the example shown in
FIGS. 31 and 32 , the NCP is discharged from a nozzle 33 (seeFIG. 32 ), so that the adhesive material NCL2 is arranged over theback surface 3 b of the logic chip LC. - The pre-coating method is common to the post-charging method explained in the description of the first adhesive material arrangement step, in that the paste-like adhesive material NCL2 is discharged from the
nozzle 33. In this embodiment, before mounting the memory chip MC1 as shown inFIG. 4 , the adhesive material NCL2 is mounted in advance. The coating rate of the adhesive material NCL2 can be greatly increased as compared to the post-charging method of charging the resin using the capillary action. - As shown in
FIG. 31 , the alignment marks 50 c formed on the back surface of the logic chip LC each are an object of interest to be detected in the second mark detection step shown inFIG. 13 , and the adhesive material NCL2 is preferably arranged so as to expose the alignment marks 50 c. When the adhesive material NCL2 is formed of transparent or semi-transparent material to the visible light, the position of even thealignment mark 50 c covered by the adhesive material NCL2 can be detected. From improving the flexibility in selecting material, as shown inFIG. 31 , the adhesive material NCL2 is preferably arranged so as to expose the alignment marks 50 c. The insulating paste (NCP) can decrease the area of the adhesive material arranged as compared to the insulating film (NCF), thus making it easy to expose thealignment mark 50 c. - The adhesive material NCL2 has the function of bonding and fixing the memory chip MC1 (see
FIG. 4 ) to the logic chip LC (seeFIG. 4 ) in the second chip mounting step shown inFIG. 13 . The adhesive material NCL2 also has the function of sealing the bonding portion between the memory chip MC1 and the logic chip LC. The sealing function includes a stress release function of diffusing and releasing the stress transferred to the bonding portion between the memory chip MC1 and the logic chip LC to protect the bonding portion. - In order to satisfy the above sealing function, the adhesive material NCL2 is preferably arranged to surround the bonding portion between the memory chip MC1 and the logic chip LC. For this reason, upon mounting at least the memory chip MC1, the
external terminals 7 shown inFIG. 6 are preferably sealed with the adhesive material NCL2. - In the second chip provision step shown in
FIG. 13 , the laminated body MCS with the memory chips MC1, MC2, MC3, and MC4 shown inFIG. 4 is provided. In a modified example of this embodiment, the memory chips MC1, MC2, MC3, and MC4 are stacked in turn over the logic chip LC. In this embodiment to be describe later, the memory chips MC1, MC2, MC3, and MC4 are previously stacked to form the laminated body (memory chip laminated body, semiconductor chip laminated body) MCS shown inFIG. 34 . As will be described later, the formation of the laminated body MCS with the memory chips MC1, MC2, MC3, and MC4 can be independently performed from the other steps in a place different from those for steps other than the second chip provision step shown inFIG. 13 . For example, the laminated body MCS can be provided as a purchased part. This way can advantageously simplify the assembly process shown inFIG. 13 to improve the manufacturing efficiency as a whole. -
FIG. 33 shows an exemplary explanatory diagram of the outline of an assembly process of a laminated body of the memory chips shown inFIG. 4 .FIG. 34 shows an exemplary explanatory diagram of the outline of another assembly process of the laminated body of the memory chips, following the process shown inFIG. 33 . The manufacturing method of each of the memory chips MC1, MC2, MC3, and MC4 shown inFIGS. 33 and 34 can be performed by employing the manufacturing method of the semiconductor chip described with reference toFIGS. 21 and 22 , and thus a description thereof will be omitted below. - In an assembly base provision step, a base (assembly base) 34 for assembling the laminated body MCS shown in
FIG. 34 is prepared. Thebase 34 has anassembly surface 34 a over which the memory chips MC1, MC2, MC3, and MC4 are stacked. The assembly surface 34 a is provided with anadhesive material layer 35. - Then, in the chip stacking step, the memory chips MC1, MC2, MC3, and MC4 are stacked over the
assembly surface 34 a of thebase 34. In the example shown inFIG. 33 , the memory chips MC4, MC3, MC2, and MC1 are stacked in that order over theassembly surface 34 a of the base 34 such that theback surface 3 b of each semiconductor chip to be stacked faces theassembly surface 34 a of thebase 34. Theback surface electrode 3 bp of theupper semiconductor chip 3 is connected with thefront surface electrode 3 ap of thelower semiconductor chip 3, for example, via the external terminals 7 (solder material 7 a). Thefront surface electrodes 3 ap of the memory chip MC1 disposed at the uppermost level are provided with the alignment marks 50 d as shown inFIG. 7 . - Then, in a laminated body sealing step shown in
FIG. 34 , resin (underfill resin) is supplied to between thestacked semiconductor chips 3 to thereby form a seal member (seal member for a laminated body of chips, or resin for a laminated body of chips) 6. Theseal member 6 is formed by the post-charging method, which has been explained above in the description of the first adhesive material arrangement step. Specifically, after previously stacking thesemiconductor chips 3, anunderfill resin 6 a is supplied from thenozzle 36, and fills in between thestacked semiconductor chips 3. Theunderfill resin 6 a has a lower viscosity than that of resin for sealing used in the sealing step shown inFIG. 13 . Theunderfill resin 6 a can fill in between thesemiconductor chips 3 by the capillary action. Thereafter, theunderfill resin 6 a embedded in between thesemiconductor chips 3 are cured to thereby obtain theseal member 6. - The method for forming the
seal member 6 by the post-charging method is excellent in filling in a clearance as compared to the so-called transfer mold method, and can be effectively applied to the case where a clearance between thestacked semiconductor chips 3 is narrow. As shown inFIG. 34 , when the clearances to be filled with theunderfill resin 6 a are stacked in layers, theunderfill resin 6 a can collectively fill in all the clearances at one time. This can reduce the whole processing time. - Then, in an assembly base removal step, the
base 34 and theadhesive material layer 35 are peeled off and removed from theback surface 3 b of the memory chip MC4. The method for removing thebase 34 and theadhesive material layer 35 can be performed, for example, by using a method which involves curing a resin component (for example, ultraviolet curing resin) contained in theadhesive material layer 35. Through the above steps, the memory chips MC1, MC2, MC3, and MC4 are stacked with the connection portions between the memory chips MC1, MC2, MC3, and MC4 sealed with theseal member 6, which produces the laminated body MCS. The laminated body MCS can be regarded as one memory chip having onefront surface 3 a (corresponding to thefront surface 3 a of the memory chip MC1) with thefront surface electrodes 3 ap formed thereat, and theother back surface 3 b (corresponding to theback surface 3 b of the memory chip MC4) opposite to thefront surface 3 a. - Next, in a second chip mounting step shown in
FIG. 13 , as shown inFIGS. 35 and 36 , the laminated body MCS is mounted over theback surface 3 b of the logic chip LC. The second chip conveying step, the second mark detection step, and the second alignment step shown inFIG. 13 can be regarded as the sub-steps included in the step of mounting the laminated body MCS over the logic chip LC. In this embodiment, the second chip conveying step, the second mark detection step, and the second alignment step are described below as the sub-steps included in the second chip mounting step. -
FIG. 35 shows an enlarged plan view of the state of the laminated body mounted over the back surface of the logic chip shown inFIG. 31 .FIG. 36 shows an enlarged cross-sectional view taken along the line A-A ofFIG. 35 .FIG. 37 shows an exemplary explanatory diagram of a main part of a second chip conveying step shown inFIG. 13 .FIG. 38 shows an exemplary explanatory diagram of a main part of a second mark detection step shown inFIG. 13 .FIG. 39 shows an exemplary explanatory diagram of a main part of a second alignment step shown inFIG. 13 .FIG. 40 shows an exemplary explanatory diagram of the state of the logic chip moved to the wiring substrate after the second alignment step shown inFIG. 39 .FIG. 41 shows an explanatory diagram of the state of a heating jig pressed against the back surface of the semiconductor chip after removal of a holding jig shown inFIG. 40 . - In this step, as shown in
FIG. 36 , the laminated body MCS is mounted by the so-called face down mounting method (flip-chip bonding) such that thefront surface 3 a of the laminated body MCS (front surface 3 a of the memory chip MC1) faces theback surface 3 b of the logic chip LC. In this step, the memory chips MC1, MC2, MC3, and MC4 are electrically connected with the logic chip LC. Specifically, as shown inFIG. 6 , thefront surface electrodes 3 ap formed on thefront surface 3 a of the laminated body MCS (memory chip MC1) are electrically connected with theback surface electrodes 3 bp formed on theback surface 3 b of the logic chip LC via the external terminals 7 (solder material 7 a shown inFIG. 6 ). The detailed flow of this step will be described below with reference toFIGS. 37 to 41 . - As shown in
FIG. 37 , the second chip mounting step includes a second chip conveying step of arranging the laminated body MCS (semiconductor chips 3) over the chip mounting region 2p 2 of thewiring substrate 20. The laminated body MCS is delivered to over the chip mounting region 2p 2 while theback surface 3 b side is being held by a holding jig (collet) 30, and arranged over the chip mounting region 2 p 2 (or adhesive material NCL2) such that thefront surface 3 a positioned on the element formation surface side faces theupper surface 2 a of thewiring substrate 20. The holdingjig 30 has a holdingsurface 30 a for sucking and holding theback surface 3 b of the laminated body MCS. The laminated body MCS is conveyed while being held by the holdingsurface 30 a. The chip mounting region 2p 2 is a region where the laminated body MCS is to be mounted in this step, and does not necessarily includes the boundary line actually visible, like the chip mounting region 2P1 in the first chip mounting step described above. - Each
external terminal 7 is formed on thefront surface 3 a side of the laminated body MCS. In an example shown inFIG. 37 , thesolder material 7 a bonded to the front surface serves as theexternal terminal 7. Like the logic chip LC shown inFIG. 25 , this embodiment can be applied to a modified example in which the protrudingelectrodes 7 b are formed to have its tip provided with thesolder material 7 a. - In the example shown in
FIG. 37 , no bonding material, such as solder material, is formed over the upper surface of theback surface electrode 3 bp formed over theback surface 3 b of the logic chip LC. In a modified example of the embodiment corresponding toFIG. 37 , a solder material (not shown) can be formed over theback surface electrode 3 bp. When the solder material covering theback surface electrode 3 bp is formed before the above first chip mounting step, the solder is melted by heating with theheating jig 31 shown inFIG. 30 , and at that time, the melted solder might be attached to theresin film 32. When the solder material is formed on theback surface electrodes 3 bp of the logic chip LC, a solder material is preferably applied after the first chip mounting step and before the first adhesive material arrangement step. - Since in the second alignment step shown in
FIG. 13 , the laminated body MCS is aligned with thewiring substrate 20 with higher accuracy in this embodiment, the accuracy of the conveying position in the second chip conveying step may be low. In order to improve the positional accuracy in the second alignment step, a movement distance in the second alignment step may be preferably small. Thus, for example, in the second chip conveying step, the laminated body MCS is preferably arranged over theupper surface 2 a of thewiring substrate 20 with the accuracy that arranges the laminated body MCS over the chip mounting region 2p 2. - Then, as shown in
FIG. 38 , the second chip mounting step includes a second mark detection step that detects (identifies) analignment mark 50 c at theback surface 3 b of the logic chip LC, and anotheralignment mark 50 b of the laminated body MCS. - As shown in
FIG. 38 , in the second mark detection step, while the laminated body MCS being arranged over thewiring substrate 20, thecamera 60 is positioned between thewiring substrate 20 and the laminated body MCS, whereby detects (identifies) thealignment mark 50 c of the logic chip LC and thealignment mark 50 d of the laminated body MCS. Thecamera 60, acontroller 61 connected with thecamera 60, and a drive device (not shown) are substantially the same as those in the first mark detection step, and its redundant description will be omitted below. This step detects at least one or more alignment marks 50 c, and at least one or more alignment marks 50 d by use of thecamera 60. - As shown in
FIG. 35 , the alignment marks 50 a are disposed at the periphery of thedevice region 20 a, so that the alignment marks 50 a become viewable. Thealignment mark 50 a of thewiring substrate 20 is used to perform alignment without forming anyalignment mark 50 c at theback surface 3 b of the logic chip LC. - In the first chip mounting step, even when the alignment is performed with high accuracy, the positions (coordinates in the XY plane shown in
FIG. 23 ) or angles (angles in the direction θ shown inFIG. 23 ) of theback surface electrodes 3 bp might become misaligned. For example, after the first alignment step, in a step of moving the logic chip LC toward thewiring substrate 20, or upon switching from the holdingjig 30 shown inFIG. 29 to theheating jig 31 shown inFIG. 30 , the misalignment might be caused. - Thus, when the alignment marks 50 a of the
wiring substrate 20 are used to align the laminated body MCS in this step, an amount of misalignment between theexternal terminal 7 of the laminated body MCS and theback surface electrode 3 bp of the logic chip LC is the sum of an amount of misalignment between thewiring substrate 20 and the logic chip LC and an amount of misalignment between thewiring substrate 20 and the laminated body MCS. That is, the amount of misalignment between theexternal terminal 7 of the laminated body MCS and theback surface electrode 3 bp of the logic chip LC is larger than that between therespective semiconductor chips 3. - In this embodiment, as mentioned above, the alignment marks 50 c are formed on the
back surface 3 b of the logic chip LC, and then used to align the laminated body MCS and logic chip LC with the substrate. That is, the amount of misalignment between theexternal terminal 7 of the laminated body MCS and theback surface electrode 3 bp of the logic LC is within a prescribed range of misalignment between the laminated body MCS and the logic chip LC. - As explained in the description of the above first mark detection process, positional data for defining the positional relationship between the
alignment mark 50 c and theback surface electrode 3 bp of the logic chip LC, or the positional relationship between thealignment mark 50 d and thefront surface electrode 3 ap of the laminated body MCS is input to thecontroller 61. Thus, thecontroller 61 can calculate the positions of theback surface electrodes 3 bp of the logic chip LC, and thefront surface electrodes 3 ap of the laminated body MCS. - In the example shown in
FIG. 38 , the alignment marks 50 c are formed over the logic chip LC, and the alignment marks 50 d are formed over thefront surface 3 a side of the laminated body MCS. The alignment marks 50 are formed on theback surface 3 b of the logic chip LC and at thefront surface 3 a of the laminated body MCS, and then the positions of the alignment marks 50 are respectively detected, whereby data on the direction θ indicative of an inclination of the coordinate axes can be obtained in addition to the coordinate data in the XY plane shown inFIG. 23 . The positions of theback surface electrodes 3 bp (seeFIG. 10 ) of the logic chip LC, or the positions of thefront surface electrode 3 ap (seeFIG. 7 ) of the laminated body MCS can be accurately calculated by using the data on the coordinates and direction θ of thealignment mark 50. - As shown in
FIG. 10 , the alignment marks 50 c are arranged at opposite corners such that the marks are positioned on one diagonal line of theback surface 3 b forming the quadrilateral shape in plan view. As shown inFIG. 7 , the alignment marks 50 d are arranged at opposite corners such that the marks are positioned on one diagonal line of thefront surface 3 a forming the quadrilateral shape in plan view. The alignment marks 50 are arranged at the opposite corners in this way, which can improve the accuracy of the data on the direction θ. This can improve the accuracy of calculation of the positions of theback surface electrodes 3 bp (seeFIG. 10 ), or the positions of thefront surface electrodes 3 ap (seeFIG. 7 ). - In this embodiment, the
front surface electrodes 3 ap of each of the memory chips MC1, MC2, MC3, and MC4 are collectively arranged in the center of thefront surface 3 a as shown inFIG. 7 . From the viewpoint of ensuring the arrangement space for thefront surface electrodes 3 ap, as shown inFIG. 7 , the alignment marks 50 d are preferably arranged outside (at the periphery of) thefront surface electrodes 3 ap. - As shown in
FIG. 6 , theback surface electrodes 3 bp of the logic chip LC are arranged to face thefront surface electrodes 3 ap of the memory chip MC1. As shown inFIG. 10 , theback surface electrodes 3 bp of the logic chip LC are collectively arranged in the center of theback surface 3 b of the logic chip LC. In order to ensure the arrangement space for theback surface electrodes 3 bp of the logic chip LC, the alignment marks 50 c are preferably arranged outside (at the periphery of) theback surface electrodes 3 bp. - As shown in
FIG. 7 described above, in each of the memory chips MC1, MC2, MC3, and MC4, the memory regions MR for four channels are arranged to enclose the region (central part) where the group of the front surface electrodes are arranged. This arrangement can equalize the distances from the respective memory regions MR to thefront surface electrodes 3 ap. In this way, from the viewpoint of equalizing the distances from the respective memory regions MR to thefront surface electrodes 3 ap, the alignment marks 50 d should not be arranged between the memory region MR and eachfront surface electrode 3 ap. Thus, as shown inFIG. 7 , each of the alignment marks 50 d is preferably arranged outside (at the periphery of) the memory regions MR at thefront surface 3 a of the memory chip MC1. - The storage capacity of each of the memory chips MC1, MC2, MC3, and MC4 increases in proportion to the area of the memory region MR. As mentioned above, the alignment marks 50 d are arranged outside (at the periphery of) the memory regions MR, so that each
alignment mark 50 d is arranged at the periphery of thefront surface 3 a. As a result, the distance between the alignment marks 50 d can be increased to thereby improve the alignment accuracy in the direction θ as mentioned above. - The alignment marks 50 d are preferably arranged at opposite corners in order to improve the alignment accuracy in the direction θ as described above.
- As shown in
FIG. 27 , thealignment mark 50 d formed in thefront surface 3 a of the laminated body MCS is preferably exposed from the insulating film (protective film, passivation film) 3 p formed on the uppermost surface of the laminated body MCS. In the example shown inFIG. 27 , theopening 3 ps is formed in the insulating film (protective film, passivation film) 3 p formed on the uppermost surface of the laminated body MCS. Eachalignment mark 50 d is exposed from the insulatingfilm 3 p at theopening 3 ps. - The alignment marks 50 c are formed on the
back surface 3 b of the logic chip LC as shown inFIG. 38 without forming the insulating film or protective film covering theback surface 3 b. Thus, in the second adhesive material arrangement step, the adhesive material NCL2 can be arranged to avoid the alignment marks 50 c to easily expose the alignment marks 50 c. - Then, as shown in
FIG. 39 , the second chip mounting step includes a second alignment step that aligns the logic chip LC with the laminated body MCS. As shown inFIG. 39 , in the second alignment step, the relative position between the laminated body MCS and the logic chip LC (in other words, wiring substrate 20) are moved along theupper surface 2 a of thewiring substrate 20, so that theback surface electrodes 3 b of the logic chi LC face theexternal terminals 7 formed on thefront surface 3 a of the laminated body MCS. - As mentioned above, in this embodiment, the positions of the
back surface electrodes 3 bp of the logic chip LC and the positions of thefront surface electrodes 3 ap of the laminated body MCS can be calculated with high accuracy. Based on the calculation data, the relative position between the laminated body MCS and the logic chip LC can be moved to perform alignment with high accuracy. - In the example shown in
FIG. 39 , a method for moving the relative position between the logic chip LC and the laminated body MCS involves moving the holdingjig 30 holding the laminated body MCS along theupper surface 2 a of thewiring substrate 20 as indicated by the arrow ofFIG. 39 . The relative relationship between the laminated body MCS and the logic chip LC can be moved by moving one or both of the laminated body MCS and the logic chip LC. - Then, as shown in
FIG. 40 , the laminated body MCS is moved toward the logic chip LC. In the above second alignment step, the relative position between the laminated body MCS and the logic chip LC is aligned with high accuracy. Thus, theback surface electrodes 3 bp of the logic chip LC and theexternal terminals 7 formed on thefront surface 3 a of the laminated body MCS can be kept facing and opposite to each other by linearly moving the laminated body MCS toward the logic chip LC. - Then, as shown in
FIG. 41 , the laminated body MCS is pressed against the logic chip LC by pushing theheating jig 31 to theback surface 3 b of the laminated body MCS. At this time, the adhesive material NCL2 is soft before being cured. Upon pushing the laminated body MCS by theheating jig 31, the laminated body MCS approaches the logic chip LC. When the laminated body MCS approaches the logic chip LC, the tip (specifically,solder material 7 a) of each of theexternal terminals 7 formed on thefront surface 3 a of the laminated body MCS comes into contact with the bonding region (specifically,solder material 7 a) of theback surface electrode 3 bp. - The adhesive material NCL2 applied between the laminated body MCS and the logic chip LC is spread along the
back surface 3 b of the logic chip LC. In the example shown inFIG. 41 , the periphery of the adhesive material NCL2 does not reach the periphery of theback surface 3 b of the logic chip LC. In a modified example, the adhesive material NCL2 may be spread over the periphery of the logic chip LC to cover the side surface of the logic chip LC or the side surface of the adhesive material NCL1. In this case, the alignment marks 50 c of theback surface 3 b of the logic chip LC are covered with the adhesive material NCL2. The position of thealignment mark 50 c, however, is already detected, which does not raise any problem. A contact area between the adhesive material NCL2 and thesemiconductor chip 3 can be expanded to improve the strength of bonding between the adhesive material NCL2 and the semiconductor chip 3 (laminated body MCS and logic chip LC). - In this embodiment, the alignment marks 50 and back
surface electrodes 3 bp are not formed on theback surface 3 b of the laminated body MCS. Thus, theresin film 32 shown inFIG. 41 can be omitted. The provision of theresin film 32 can disperse the force received by the laminated body MCS in pressing the film with theheating jig 31. Also, the provision of theresin film 32 can perform a second chip mounting step in the same mechanism as that of the first chi mounting step, which can simplify the manufacturing process. In the second chip mounting step, theresin film 32 preferably intervenes in between the laminated body MCS and theheating jig 31. - Then, as shown in
FIG. 41 , the laminated body MCS and the adhesive material NSL2 are heated by the heating jig (heat source) 31 while the laminated body MCS being pressed against theheating jig 31. Thesolder material 7 a on the external terminal side is melt at a bonding portion between the laminated body MCS and the logic chip LC to be bonded with theback surface electrode 3 bp of the logic chip LC. That is, the laminated body MCS is heated by the heating jig (heat source) 31, so that the front surface electrodes Sap of the laminated body MCS are electrically connected with theback surface electrodes 3 bp of the logic chip LC via thesolder materials 7 a. - The adhesive material NCL2 is heated to be cured. Like the first chip mounting step described, in another embodiment, the adhesive material NCL1 is not necessarily cured fully by the heat from the heating jig (heat source) 31, and a part of the thermosetting resin contained in the adhesive material NCL1 is cured (semi-cured) to such an extent that can fix the laminated body MCS. Then, the log chip LC is transferred to a heating furnace (not shown), and the remainder of the thermosetting resin is cured (fully cured). It takes time to complete the full curing process of fully curing the thermosetting resin components contained in the adhesive material NCL1, but the full curing process is performed in the heating furnace, which can improve the manufacturing efficiency.
- Then, in a sealing step shown in
FIG. 13 , as shown inFIG. 42 , theupper surface 2 a of thewiring substrate 20, the logic chip LC, and the laminated body MCS with the memory chips MC1, MC2, MC3, and MC4 are sealed with resin to thereby form theseal member 4.FIG. 42 shows an enlarged cross-sectional view of the state of the stacked semiconductor chips sealed with the seal member formed over the wiring substrate shown inFIG. 36 .FIG. 43 shows a plan view of the entire structure of the seal member shown inFIG. 42 . - In this embodiment, as shown in
FIG. 43 , theseal member 4 is formed to seal thedevice regions 20 a at one time. The formation of theseal member 4 is called the “block molding” method. A semiconductor package manufactured by the block molding method is called the “multi array package (MAP) semiconductor device”. In the block molding method, the distance between theadjacent device regions 20 a can be decreased, so that an effective area of one piece of thewiring substrate 20 is increased. In other words, the number of products that can be obtained from onewiring substrate 20 is increased. In this way, the effective area of the onewiring substrate 20 is increased, which can make the manufacturing process more efficient. - In this embodiment, the
seal member 4 is formed by the so-called transfer molding method which involves press-fitting resin heated and softened in a molding die (not shown), molding the resin therein, and thermally curing the resin. Theseal member 4 formed by the transfer mold method has a high durability, for example, like theseal member 6 sealing the laminated body MCS shown inFIG. 42 , as compared to a member produced by curing a liquid resin. Theseal member 4 is appropriate as a protective member. Filler particles, such as silica (silicone dioxide; SiO2) particles, are mixed into the thermosetting resin, which can improve the function of the seal member 4 (for example, resistance to warpage). - In this embodiment, bonding portions (electric connection portions) between the
stacked semiconductor chips 3 are sealed by the adhesive materials NCL1, NCL2, and sealmember 6. In a modified example to which this embodiment can be applied, theseal member 4 is not formed. In this case, the full sealing step can be omitted. - Then, in a ball mounting step shown in
FIG. 13 , as shown inFIG. 44 , thesolder balls 5 serving as the external terminal are bonded to thelands 2 g formed on thelower surface 2 b of thewiring substrate 20.FIG. 44 shows an enlarged cross-sectional view of the state of the solder balls bonded to lands of the wiring substrate shown inFIG. 37 . - In this step, as shown in
FIG. 44 , after turning thewiring substrate 20 upside down, thesolder balls 5 are arranged over therespective lands 2 g exposed at thelower surface 2 b of thewiring substrate 20, and then heated, so that thesolder balls 5 are bonded to thelands 2 g. In this step, thesolder balls 5 are electrically coupled to the semiconductor chips 3 (logic chip LC, and memory chips MC1, MC2, MC3, and MC4) via thewiring substrate 20. The technique described in this embodiment is not limited to the application to the so-called ball grid array (BGA) semiconductor device withsolder balls 5 bonded thereto in an array. For example, a modified example of this embodiment can be applied to the so-called land grid array (LGA) semiconductor device which is shipped with thelands 2 g exposed without forming thesolder balls 5, or with a solder paste thinner than thesolder ball 5 attached to thelands 2 g. The LGA semiconductor device can omit the ball mount step. - Then, in a singulating step shown in
FIG. 13 , as shown inFIG. 45 , thewiring substrate 20 is divided everydevice region 20 a.FIG. 45 shows a cross-sectional view of the singulated multi-piece wiring substrate shown inFIG. 44 . - In this step, as shown in
FIG. 45 , thewiring substrate 20 and theseal member 4 are cut along the dicing lines (dicing regions) 20 c to obtain the singulated semiconductor devices 1 (seeFIG. 4 ). In the embodiment, cutting methods are not specifically limited, but in the example shown inFIG. 45 , thewiring substrate 20 andseal member 4 bonded and fixed to a tape material (dicing tape) 41 are cut using the dicing blade (rotary blade) 40 from thelower surface 2 b of thewiring substrate 20. The technique described in this embodiment is applied not only to the case of using thewiring substrate 20 as a multi-piece substrate with thedevice regions 20 a, but also to, for example, a semiconductor device havingsemiconductor chips 3 stacked over the wiring substrate 2 (seeFIG. 4 ) corresponding to only one semiconductor device. In this case, the singulating step can be omitted. - Throughout the above steps, the
semiconductor device 1 described with reference toFIGS. 1 to 11 can be obtained. Thereafter, necessary checking and testing, such as an appearance check or an electric test, are performed on each semiconductor device. Then, the semiconductor device is shipped, or mounted on a mounting substrate (not shown). - Then, preferable embodiments of the shape of the alignment mark in the plan view will be described later.
FIG. 46 shows an enlarged plan view of the alignment mark shown inFIGS. 3 , 7, 9, and 10.FIGS. 47 to 49 are enlarged plan views of modified examples corresponding toFIG. 46 . - An
alignment mark 50 shown inFIG. 46 forms a polygonal shape (a L-like shape shown inFIG. 46 ) in plan view. The polygonal shape has an asymmetric shape with respect to the center of acircumcircle 51 of the polygonal shape. Thus, the planar shape of thealignment mark 50 is set to the asymmetric shape, so that not only the position of thealignment mark 50, but also an inclination of thealignment mark 50 in plan view can be detected. - Thus, for example, even when the alignment marks 50 a, 50 b, 50 c, and 50 d are individually formed, the X and Y coordinates of the
alignment mark 50 and the position of the direction θ can be specified. - When two or more asymmetric alignment marks 50 are arranged in the same plane, it can be detected that the position of the chip in the direction shown in
FIG. 46 is displaced by 180 degrees. For example, the alignment marks 50 d shown inFIG. 7 , and the alignment marks 50 b shown inFIG. 9 are set in the asymmetric shape shown inFIG. 46 , which can prevent or suppress the improper mounting, including the displacement of the mounting direction of thesemiconductor chip 3 by 180 degrees, in the first chip mounting step and the second chip mounting step. - Like an
alignment mark 52 shown inFIG. 47 , the alignment mark having a circular shape in plan view can improve the processing accuracy as the circular shape can be easily processed as compared to the polygonal shape shown inFIG. 46 . Even a part of thecircular alignment mark 52 can hardly break, which can improve the detection accuracy. - Thus, for example, at least one of the alignment marks 50 a, 50 b, 50 c, and 50 d employs the
circular alignment mark 52, which can improve the position detection accuracy of the position in which thealignment mark 52 is arranged. The alignment marks 50 a, 50 b, 50 c, and 50 d can be formed in a circular shape, like thealignment mark 52. - In this embodiment, as shown in
FIG. 6 , theback surface electrodes 3 bp of the logic chip LC have only to surely face thefront surface electrodes 3 ap of the memory chip MC1. Thus, various modified examples can be applied within a range that can ensure the required alignment accuracy. For example, as shown inFIG. 48 , analignment mark 53 having a quadrilateral shape in plan view can be applied. For example, as shown inFIG. 49 , analignment mark 54 having an X-like shape in plan view can be applied. As mentioned above, a combination of the alignment marks 50, 52, 53, and 54 can be applied. - The invention made by the inventors has been specifically described based on the embodiments. However, the invention is not limited to the above embodiments, and various modifications and changes can be made without departing from the scope of the invention.
- For example, in the above embodiment, the technique for stacking the
semiconductor chips 3 using the result of detection of thealignment mark 50 is applied in the first chip mounting step and the second chip mounting step. The technique can applied to the step of assembling the laminated body MCS, that is, the second chip provision step.FIG. 50 shows a plan view of a modified example corresponding toFIG. 8 . - When applying the technique for stacking the
semiconductor chips 3 using the result of detection of thealignment mark 50 in the second chip preparation step, as shown inFIG. 7 , the alignment marks 50 d are formed over therespective surfaces 3 a of the memory chips MC1, MC2, MC3, and MC4 forming the laminated body MCS (seeFIG. 4 ). As shown inFIG. 50 , alignment marks 50 e are formed over theback surface 3 b of each of the memory chips MC1, MC2, and MC3. - In the present modified example, in stacking the memory chips MC1, MC2, MC3, and MC4, the alignment accuracy of the memory chips can be improved to thereby improve the electric connection reliability of the memory chips MC1, MC2, MC3, and MC4.
- The above embodiment has described the planar size of the laminated body MCS mounted at the upper level is larger than the planar size of the logic chip LC mounted at the lower level. However, like the
semiconductor device 1 a shown inFIG. 51 , as a second modified example, the invention can be applied to the case where the planar size of the laminated body MCS is smaller than that of the logic chip LC mounted at the lower level.FIG. 51 shows a perspective plane view of a modified example corresponding toFIG. 3 . - The above embodiment has described a
SiP semiconductor device 1 composed of the logic chip LC and the memory chips MC1, MC2, MC3, and MC4 as an example of thesemiconductor chips 3 to be stacked (used). However, as a third example, the invention can also be applied to any combination other than a combination of the memory chip and the logic chip as long as thestacked chips 3 are electrically connected with each other. - In this embodiment, as shown in
FIG. 6 , theback surface electrodes 3 bp are arranged on the side opposite to thefront surface electrodes 3 ap of the logic chip LC. Theback surface electrodes 3 bp are electrically connected with thefront surface electrodes 3 ap via the penetratingelectrodes 3 tsv. However, as a fourth modified example, the invention can also be applied to a semiconductor device including backsurface electrodes 3 bp formed at thelower semiconductor chip 3, andfront surface electrodes 3 ap formed at theupper semiconductor chip 3 electrically connected with theback surface electrodes 3 bp via theexternal terminals 7. - A combination of the modified examples can be applied without departing from the scope of the technical idea described in the embodiments.
Claims (9)
1-20. (canceled)
21. A semiconductor device, comprising:
a first semiconductor chip including a first surface, a plurality of first electrodes formed on the first surface, and a first alignment mark formed on the first surface; and
a second semiconductor chip including a second surface, a plurality of second electrodes formed on the second surface, and a second alignment mark formed on the second surface;
wherein the second semiconductor chip is mounted over the first semiconductor chip such that the second surface faces the first surface, and
wherein the plurality of second electrodes are electrically connected with the plurality of first electrodes via a plurality of first terminals, respectively.
22. The semiconductor device according to claim 21 ,
wherein, in plan view, the plurality of first electrodes are formed at a central area of the first surface, and
wherein, in plan view, the plurality of second electrodes are formed at a central area of the second surface.
23. The semiconductor device according to claim 22 ,
wherein, in plan view, the first alignment mark is formed at a peripheral area outside the central area of the first surface, and
wherein, in plan view, the second alignment mark is formed at a peripheral area outside the central area of the second surface.
24. The semiconductor device according to claim 23 ,
wherein, in plan view, a size of the second semiconductor chip is greater than a size of the first semiconductor chip.
25. The semiconductor device according to claim 23 ,
wherein the whole of the first surface of the first semiconductor chip is covered by the second semiconductor chip.
26. The semiconductor device according to claim 21 ,
wherein the first semiconductor chip further includes a third surface opposed to the first surface, and a plurality of third electrodes formed on the third surface,
wherein the plurality of third electrodes are electrically connected with the plurality of first electrodes via a plurality of fourth electrodes, respectively, and
wherein each of the plurality of fourth electrodes penetrates the first semiconductor chip from one of the first surface and the third surface to the other.
27. The semiconductor device according to claim 26 ,
wherein the first semiconductor chip is mounted over a wiring substrate such that the third surface of the first semiconductor chip faces a fourth surface of the wiring substrate, and
wherein the plurality of third electrodes are electrically connected with a plurality of fifth electrodes formed on the fourth surface of the wiring substrate via a plurality of second terminals, respectively.
28. The semiconductor device according to claim 21 ,
wherein the second semiconductor chip is a memory chip, and
wherein the first semiconductor chip is a logic chip controlling the memory chip.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US14/803,486 US20150325528A1 (en) | 2013-03-22 | 2015-07-20 | Method of manufacturing semiconductor device, and semiconductor device |
Applications Claiming Priority (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2013-061087 | 2013-03-22 | ||
JP2013061087A JP6207190B2 (en) | 2013-03-22 | 2013-03-22 | Manufacturing method of semiconductor device |
US14/194,890 US9117826B2 (en) | 2013-03-22 | 2014-03-03 | Method of manufacturing semiconductor device, and semiconductor device |
US14/803,486 US20150325528A1 (en) | 2013-03-22 | 2015-07-20 | Method of manufacturing semiconductor device, and semiconductor device |
Related Parent Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US14/194,890 Continuation US9117826B2 (en) | 2013-03-22 | 2014-03-03 | Method of manufacturing semiconductor device, and semiconductor device |
Publications (1)
Publication Number | Publication Date |
---|---|
US20150325528A1 true US20150325528A1 (en) | 2015-11-12 |
Family
ID=51552137
Family Applications (2)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US14/194,890 Active US9117826B2 (en) | 2013-03-22 | 2014-03-03 | Method of manufacturing semiconductor device, and semiconductor device |
US14/803,486 Abandoned US20150325528A1 (en) | 2013-03-22 | 2015-07-20 | Method of manufacturing semiconductor device, and semiconductor device |
Family Applications Before (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US14/194,890 Active US9117826B2 (en) | 2013-03-22 | 2014-03-03 | Method of manufacturing semiconductor device, and semiconductor device |
Country Status (6)
Country | Link |
---|---|
US (2) | US9117826B2 (en) |
JP (1) | JP6207190B2 (en) |
KR (1) | KR20140117285A (en) |
CN (1) | CN104064479B (en) |
HK (1) | HK1198562A1 (en) |
TW (1) | TWI596714B (en) |
Families Citing this family (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP5876000B2 (en) * | 2012-06-11 | 2016-03-02 | 株式会社新川 | Bonding apparatus and bonding method |
KR102149150B1 (en) * | 2013-10-21 | 2020-08-28 | 삼성전자주식회사 | Electronic Device |
JP6363854B2 (en) * | 2014-03-11 | 2018-07-25 | キヤノン株式会社 | Forming method and article manufacturing method |
TWI566305B (en) * | 2014-10-29 | 2017-01-11 | 巨擘科技股份有限公司 | Method for manufacturing three-dimensional integrated circuit |
KR102012788B1 (en) | 2015-09-23 | 2019-08-21 | 주식회사 엘지화학 | Adhesive film, preparation method of semiconductor device, and semiconductor device |
US9953963B2 (en) * | 2015-11-06 | 2018-04-24 | Taiwan Semiconductor Manufacturing Company, Ltd. | Integrated circuit process having alignment marks for underfill |
KR102022267B1 (en) * | 2017-12-28 | 2019-09-18 | 삼성전자주식회사 | Fan-out semiconductor package |
US10998247B2 (en) | 2018-08-16 | 2021-05-04 | Samsung Electronics Co., Ltd. | Board with embedded passive component |
KR102164793B1 (en) * | 2018-08-16 | 2020-10-14 | 삼성전자주식회사 | Printed circuit board with embedded passive component |
DE102018133319A1 (en) * | 2018-12-21 | 2020-06-25 | Rittal Gmbh & Co. Kg | Method for robot-assisted wiring of electrical components of an electrical switchgear arranged on a mounting plate |
US11430909B2 (en) * | 2019-07-31 | 2022-08-30 | Taiwan Semiconductor Manufacturing Company, Ltd. | BSI chip with backside alignment mark |
KR20210030774A (en) * | 2019-09-10 | 2021-03-18 | 삼성전자주식회사 | Package on package(POP) type semiconductor package |
KR20210035546A (en) * | 2019-09-24 | 2021-04-01 | 삼성전자주식회사 | Semiconductor package |
KR20210143568A (en) * | 2020-05-20 | 2021-11-29 | 에스케이하이닉스 주식회사 | Stack package including core die stacked over controlling die |
Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5434745A (en) * | 1994-07-26 | 1995-07-18 | White Microelectronics Div. Of Bowmar Instrument Corp. | Stacked silicon die carrier assembly |
US6972487B2 (en) * | 2001-03-30 | 2005-12-06 | Fujitsu Limited | Multi chip package structure having a plurality of semiconductor chips mounted in the same package |
US20080258287A1 (en) * | 2006-12-12 | 2008-10-23 | Shinko Electric Industries Co., Ltd. | Semiconductor device and method of manufacturing the same |
US8097490B1 (en) * | 2010-08-27 | 2012-01-17 | Stats Chippac, Ltd. | Semiconductor device and method of forming stepped interconnect layer for stacked semiconductor die |
US8519537B2 (en) * | 2010-02-26 | 2013-08-27 | Taiwan Semiconductor Manufacturing Company, Ltd. | 3D semiconductor package interposer with die cavity |
US8629543B2 (en) * | 2007-06-11 | 2014-01-14 | Invensas Corporation | Electrically interconnected stacked die assemblies |
US8780600B2 (en) * | 2011-12-07 | 2014-07-15 | Apple Inc. | Systems and methods for stacked semiconductor memory devices |
US9040348B2 (en) * | 2011-09-16 | 2015-05-26 | Altera Corporation | Electronic assembly apparatus and associated methods |
US9052187B2 (en) * | 2012-03-21 | 2015-06-09 | Kabushiki Kaisha Toshiba | Inspection apparatus and inspection method for semiconductor device |
US9153520B2 (en) * | 2011-11-14 | 2015-10-06 | Micron Technology, Inc. | Stacked semiconductor die assemblies with multiple thermal paths and associated systems and methods |
Family Cites Families (21)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP3565319B2 (en) * | 1999-04-14 | 2004-09-15 | シャープ株式会社 | Semiconductor device and manufacturing method thereof |
JP2001217387A (en) * | 2000-02-03 | 2001-08-10 | Rohm Co Ltd | Semiconductor device and method of manufacturing the same |
JP2002110742A (en) * | 2000-10-02 | 2002-04-12 | Hitachi Ltd | Method for manufacturing semiconductor device, and apparatus for manufacturing semiconductor |
KR100580635B1 (en) * | 2003-12-30 | 2006-05-16 | 삼성전자주식회사 | Electronic device and method of manufacturing the same |
JP4467318B2 (en) * | 2004-01-28 | 2010-05-26 | Necエレクトロニクス株式会社 | Semiconductor device, chip alignment method for multi-chip semiconductor device, and method for manufacturing chip for multi-chip semiconductor device |
JP2008109115A (en) * | 2006-09-26 | 2008-05-08 | Sekisui Chem Co Ltd | Semiconductor chip laminate, and manufacturing method thereof |
JP2008177364A (en) * | 2007-01-18 | 2008-07-31 | Denso Corp | Method of manufacturing semiconductor device, and semiconductor device |
JP2010161102A (en) * | 2009-01-06 | 2010-07-22 | Elpida Memory Inc | Semiconductor device |
JP5185885B2 (en) * | 2009-05-21 | 2013-04-17 | 新光電気工業株式会社 | Wiring board and semiconductor device |
JP2011061004A (en) * | 2009-09-10 | 2011-03-24 | Elpida Memory Inc | Semiconductor device, and method of manufacturing the same |
US8796863B2 (en) * | 2010-02-09 | 2014-08-05 | Samsung Electronics Co., Ltd. | Semiconductor memory devices and semiconductor packages |
CN102169875B (en) * | 2010-02-26 | 2013-04-17 | 台湾积体电路制造股份有限公司 | Semiconductor device and producing method thereof |
JP2011187574A (en) | 2010-03-05 | 2011-09-22 | Elpida Memory Inc | Semiconductor device and method of manufacturing the same, and electronic device |
JP5927756B2 (en) * | 2010-12-17 | 2016-06-01 | ソニー株式会社 | Semiconductor device and manufacturing method of semiconductor device |
JP5664392B2 (en) * | 2011-03-23 | 2015-02-04 | ソニー株式会社 | Semiconductor device, method for manufacturing semiconductor device, and method for manufacturing wiring board |
JP2012222161A (en) * | 2011-04-08 | 2012-11-12 | Elpida Memory Inc | Semiconductor device |
JP2013045945A (en) * | 2011-08-25 | 2013-03-04 | Sumitomo Bakelite Co Ltd | Manufacturing method of semiconductor device |
KR101906408B1 (en) * | 2011-10-04 | 2018-10-11 | 삼성전자주식회사 | Semiconductor package and method of manufacturing the same |
JP2013197387A (en) * | 2012-03-21 | 2013-09-30 | Elpida Memory Inc | Semiconductor device |
JP2013033999A (en) * | 2012-10-24 | 2013-02-14 | Hitachi Ltd | Semiconductor device |
JP2014187185A (en) * | 2013-03-22 | 2014-10-02 | Renesas Electronics Corp | Semiconductor device manufacturing method |
-
2013
- 2013-03-22 JP JP2013061087A patent/JP6207190B2/en not_active Expired - Fee Related
-
2014
- 2014-03-03 US US14/194,890 patent/US9117826B2/en active Active
- 2014-03-04 TW TW103107287A patent/TWI596714B/en active
- 2014-03-20 KR KR20140032580A patent/KR20140117285A/en active IP Right Grant
- 2014-03-21 CN CN201410106344.0A patent/CN104064479B/en active Active
- 2014-11-28 HK HK14112039.2A patent/HK1198562A1/en not_active IP Right Cessation
-
2015
- 2015-07-20 US US14/803,486 patent/US20150325528A1/en not_active Abandoned
Patent Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5434745A (en) * | 1994-07-26 | 1995-07-18 | White Microelectronics Div. Of Bowmar Instrument Corp. | Stacked silicon die carrier assembly |
US6972487B2 (en) * | 2001-03-30 | 2005-12-06 | Fujitsu Limited | Multi chip package structure having a plurality of semiconductor chips mounted in the same package |
US20080258287A1 (en) * | 2006-12-12 | 2008-10-23 | Shinko Electric Industries Co., Ltd. | Semiconductor device and method of manufacturing the same |
US8629543B2 (en) * | 2007-06-11 | 2014-01-14 | Invensas Corporation | Electrically interconnected stacked die assemblies |
US8519537B2 (en) * | 2010-02-26 | 2013-08-27 | Taiwan Semiconductor Manufacturing Company, Ltd. | 3D semiconductor package interposer with die cavity |
US8097490B1 (en) * | 2010-08-27 | 2012-01-17 | Stats Chippac, Ltd. | Semiconductor device and method of forming stepped interconnect layer for stacked semiconductor die |
US9040348B2 (en) * | 2011-09-16 | 2015-05-26 | Altera Corporation | Electronic assembly apparatus and associated methods |
US9153520B2 (en) * | 2011-11-14 | 2015-10-06 | Micron Technology, Inc. | Stacked semiconductor die assemblies with multiple thermal paths and associated systems and methods |
US8780600B2 (en) * | 2011-12-07 | 2014-07-15 | Apple Inc. | Systems and methods for stacked semiconductor memory devices |
US9052187B2 (en) * | 2012-03-21 | 2015-06-09 | Kabushiki Kaisha Toshiba | Inspection apparatus and inspection method for semiconductor device |
Also Published As
Publication number | Publication date |
---|---|
JP2014187184A (en) | 2014-10-02 |
TWI596714B (en) | 2017-08-21 |
KR20140117285A (en) | 2014-10-07 |
US9117826B2 (en) | 2015-08-25 |
CN104064479A (en) | 2014-09-24 |
JP6207190B2 (en) | 2017-10-04 |
CN104064479B (en) | 2018-05-15 |
US20140284780A1 (en) | 2014-09-25 |
TW201445681A (en) | 2014-12-01 |
HK1198562A1 (en) | 2015-05-15 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US9117826B2 (en) | Method of manufacturing semiconductor device, and semiconductor device | |
US10312199B2 (en) | Semiconductor device and manufacturing method thereof | |
US11276645B2 (en) | Encapsulation of a substrate electrically connected to a plurality of pin arrays | |
US10651133B2 (en) | Semiconductor package | |
TWI482261B (en) | Three-dimensional system-in-package package-on-package structure | |
US9806061B2 (en) | Bumpless wafer level fan-out package | |
US9129828B2 (en) | Semiconductor device with chip having a different number of front surface electrodes and back surface electrodes | |
KR100868419B1 (en) | Semiconductor device and manufacturing method thereof | |
US20110209908A1 (en) | Conductor package structure and method of the same | |
US8058717B2 (en) | Laminated body of semiconductor chips including pads mutually connected to conductive member | |
US10818637B2 (en) | Thin bonded interposer package | |
US20150179623A1 (en) | Method for manufacturing semiconductor device | |
US9640414B2 (en) | Method of manufacturing semiconductor device | |
US20110031607A1 (en) | Conductor package structure and method of the same | |
US10854576B2 (en) | Semiconductor device and manufacturing method thereof | |
CN112490223B (en) | Semiconductor device with a semiconductor device having a plurality of semiconductor chips | |
US20110031594A1 (en) | Conductor package structure and method of the same | |
CN116487342A (en) | High-density SiP packaging structure and packaging method |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
STCB | Information on status: application discontinuation |
Free format text: ABANDONED -- FAILURE TO PAY ISSUE FEE |