US20140176174A1

US20140176174A1 - Designed asperity contactors, including nanospikes for semiconductor test, and associated systems and methods

Info

Publication number: US20140176174A1
Application number: US13/842,830
Authority: US
Inventors: Douglas A. Preston; Morgan T. Johnson
Original assignee: Advanced Inquiry Systems Inc
Current assignee: TRANSLARITY Inc
Priority date: 2012-12-26
Filing date: 2013-03-15
Publication date: 2014-06-26
Also published as: WO2014105896A1; WO2014105899A1; US20180003737A1; US20140179031A1; US20170074904A1; US9733272B2; US9494618B2

Abstract

Nano spike contactors suitable for semiconductor device test, and associated systems and methods are disclosed. A representative apparatus includes a translator having a wafer side positioned to face toward a device under test and an inquiry side facing away from the wafer side. A plurality of wafer-side sites are carried by the translator at the wafer side of the translator. The nanospikes can be attached to nanospike sites on a wafer side of a translator. Because of their small size, multiple nanospikes make contact with a single pad/solderball on the semiconductor device. In some embodiments, the nanospikes can be formed by sputtering over a metal carrier with a photoresist mask. In particular embodiments, the nanospikes have generally conical cross-section.

Description

CROSS-REFERENCE TO RELATED APPLICATION

The present application claims priority to U.S. Provisional Application No. 61/746,000 filed on Dec. 26, 2012 and incorporated herein by references. To the extent the foregoing application and/or any other materials incorporated here by reference conflict with the present disclosure, the present disclosure controls.

TECHNICAL FIELD

The present technology is directed generally to designed asperity contactors, including nanospikes, capable of contacting pads or solderballs on a semiconductor device, and associated systems and methods.

BACKGROUND

Integrated circuits are used in a wide variety of products. Integrated circuits have continuously decreased in price and increased in performance, becoming ubiquitous in modern electronic devices. These improvements in the performance/cost ratio result, at least in part, from miniaturization, which enables more semiconductor dies to be produced from a wafer with each new generation of the integrated circuit manufacturing technology. Furthermore, the total number of the signal and power/ground contacts on a die generally increases with new, more complex die designs. An increased number of contacts on a die (e.g., pads or solderballs) over a decreased size of the die necessitates smaller contacts.
Prior to shipping an integrated circuit die to a customer, the performance of the integrated circuit is tested, either on a statistical sample basis or by testing each die. An electrical test of a semiconductor die typically includes powering the die through the power/ground connectors, transmitting signals to the die input (I) connectors, and measuring the resulting signals at the die output (O) connectors. Therefore, during the integrated circuit test at least some connectors on the die must be electrically contacted to connect the die to a source of power and a source of the test signals.
FIG. 1A is a side view of a conventional test contactor 10 a in contact with a semiconductor device 10 b (e.g., a device under test). The semiconductor device 10 b can include a packaged die 12 attached to a die substrate 18 with a die attach material 16. Wirebonds 14 provide electrical connections between circuits in the die 12 and solderballs 20 (or pads or other contact structures) that provide communication between the die 12 and other devices. The solderballs 20 also provide communication between the die 12 and the test contactor 10 a. As shown in FIG. 1A, the test contactor 10 a has an array of spring-loaded pins 22 between a contactor substrate 26 and the solderballs 20. As the contactor substrate 26 moves downwardly toward the semiconductor device 10 b, the pins 22 make contact with the solderballs 20. When the contactor substrate 26 moves further toward the semiconductor device 10 b or the semiconductor device 10 b moves toward the contactor substrate 26, the springs of the pins 22 compress, producing increased contact forces between the pins 22 of the contactor 10 a and the solderballs 20 of the semiconductor device 10 b. In general, the contact forces should be high enough to allow the pins 22 to break through an oxide layer on the pads/solderballs (e.g., the contact structures), but not so high as to damage the pads/solderballs. This may be a difficult requirement for the conventional spring loaded pins 22 because they tend to produce high contact forces, thus penetrating through the oxide layer on the pads/solderballs 20, but also possibly damaging the contacts on the die. Furthermore, for a large number of pads/solderballs 20, the total contactor force can quickly become very high, thus requiring powerful mechanisms to compress the contactor 10A.
FIG. 1B illustrates several designs for commercially available spring loaded pins positioned in a compressed state. In one example, a pin 22 a has a pair of cylindrical, male/female pin segments 40 a/41 a that can slide relative to each other along a common centerline. As the contactor substrate (shown in FIG. 1A) pushes the female pin segment 41 a toward the solderball 20 a, a spring 32 a becomes compressed between a shoulder 38 a on the female pin segment 41 a and a shoulder 36 a on the male pin segment 40 a. The contact force between a crown tip 34 a and the solderball 20 a increases generally linearly with the compression of spring 32 a.
In another spring loaded pin design, also shown in FIG. 1B, a pin 22 b includes a spring 32 b that is compressed between a shoulder 36 b on a pin segment 40 b and the connector substrate (shown in FIG. 1A). The pins 22 a/22 b typically have either a crown tip 34 a or a pointed tip to facilitate breaking through the oxide layer over the solderballs 20 a or pads 20 b, respectively. The range of compression for the springs 32 a/32 b is ultimately limited either by a full compression of the springs or by the force provided by a compression mechanism. The individual pins are arranged in a contactor that keeps the pins aligned in a proper layout, as explained below with reference to FIG. 1C.
FIG. 1C is a bottom view of the contactor 10 a used for contacting the semiconductor device 10 b (shown in FIG. 1A). The contactor 10 a has a two dimensional array of pins 22 corresponding to the array of contact structures on the packaged device under test. The pins 22 protrude through a perforated mask 46. The contactor substrate 26 mechanically supports one end of the pins 22 while the opposite ends of the pins 22 engage with the device under test. Alignment features 44 a-b can align the device under test and the contactor 10 a. The contactor substrate 26 also provides electrical signals to the pins 22. A characteristic diameter of the pins 22 generally scales with a characteristic dimension of the contact structures on the semiconductor die or the package. Therefore, as the contact structures on the die become smaller and/or have a smaller pitch, the pins must become smaller, too. It is difficult to significantly reduce the diameter and pitch of the spring loaded pins, however, because of the difficulties in machining and assembling such small parts, which in turn can cause inconsistent performance from one assembly to another.
FIG. 2A is an isometric view of a test contactor 10 c suitable for a bare die test (i.e., suitable for testing un-packaged dies). The test contactor 10 c has a two-dimensional array of flexible needles 52 carrying needle blades 54 and corresponding to the layout of the pads/solderballs on the die under test. The flexible needles 52 can be curved to provide springiness when the needle blades 54 engage with the corresponding pads/solderballs on the die. The opposite ends of the flexible needles 52 are mechanically and electrically connected with a substrate 56 and further to a source of signals and/or power/ground. As the needle blades 54 engage with the corresponding pads/solderballs, the needle blades may slide over the surface of the pads/solderballs. This sliding action, coupled with a relatively high stiffness of the flexible needles 52, typically works well to break the oxide layer on the pads/solderballs, but it can also cause penetration damage, as illustrated in FIG. 2B described below.
FIG. 2B is a top view of a solderball 22 after undergoing the test described with reference to FIG. 2A. As shown in FIG. 2B, the pressure and/or sliding action of the needle blade 54 produced a depression 62 in the solderball 22. Such an undesired depression can be exacerbated by repetitive tests requiring multiple contacts or “touchdowns” between the needles 52 and the contact structures on the device. Repetitive tests are common with, for example, devices that marginally pass or fail an initial test (therefore requiring additional testing), devices placed close to a wafer edge (therefore undergoing multiple contacts due to a contactor design that does not allow for an overhang over the wafer edge), devices subjected to multiple test suites using different testers (therefore also requiring multiple contacts), etc. Furthermore, conventional test contactors have relatively high electrical resistances, which limits their ability to test high current devices. Accordingly, there remains a need for cost effective test contactors that do not damage the contact structures on the die and that can scale down in size with the contact structure size and pitch, while being capable of delivering high electrical currents.

BRIEF DESCRIPTION OF THE DRAWINGS

Many aspects of the present disclosure can be better understood with reference to the following drawings. The components in the drawings are not necessarily to scale. Instead, emphasis is placed on clearly illustrating the principles of the present disclosure. Furthermore, in the drawings, like reference numerals designate corresponding parts throughout the several views.

FIG. 1A is a partially schematic, side view of a packaged semiconductor device undergoing a test in accordance with the prior art.

FIG. 1B is a partially schematic, side view of several representative pin designs in accordance with the prior art.

FIG. 1C is a top view of a prior art contactor.

FIG. 2A is a partially schematic, isometric view of a bare die contactor in accordance with the prior art.

FIG. 2B is an isometric illustration of solderball damage caused by a prior art contactor.

FIG. 3 is a schematic view of a translator stack in accordance with an embodiment of the presently disclosed technology.

FIG. 4 is a partially schematic, side view of a translator stack in accordance with another embodiment of the presently disclosed technology.

FIG. 5 is a partially schematic, side view of a nanospikes-based contactor in accordance with an embodiment of the presently disclosed technology.

FIG. 6 is a bottom view of the nanospikes-based contactor in accordance with another embodiment of the presently disclosed technology.

FIGS. 7A-7C include several partially schematic views illustrating the nanospikes-based contactor in accordance with the presently disclosed technology.

FIG. 8 is a side view of the nanospikes-based contactor when vacuum is applied in accordance with an embodiment of the presently disclosed technology.

FIG. 9 is a side view of a nanospikes-based translator stack in accordance with the presently disclosed technology.

FIGS. 10A-10G are schematic illustrations of a manufacturing process for a nanospike contactor in accordance with the presently disclosed technology.

FIG. 11 is a schematic diagram of sputtering-based nanospike manufacturing process.

FIGS. 12A-12D are schematic diagrams of a nanospike manufacturing process in accordance with an embodiment of the presently disclosed technology.

FIGS. 13A-13H are schematic diagrams of a nanospike manufacturing process in accordance with another embodiment of the presently disclosed technology.

FIGS. 14A-14F illustrate several nanospike shapes in accordance with the present technology.

FIGS. 15A-15B illustrate a star-shaped nanospike in accordance with the present technology.

FIGS. 16A-16B illustrate a blade-shaped nanospike in accordance with the present technology.

FIGS. 17A-17B illustrate a cross-shaped nanospike in accordance with the present technology.

FIGS. 18A-18E are schematic diagrams of a nanospike manufacturing process in accordance with an embodiment of the presently disclosed technology.

FIG. 19 is a graph of electrical resistivity and hardness of representative contact pad and nanospike materials.

DETAILED DESCRIPTION

Specific details of several embodiments of representative designed asperity contactors and associated systems and methods for manufacture and use are described below. The contactors can be used for testing bare semiconductor dies on a wafer, and/or packaged semiconductor dies. The contactors can be used for testing different types of semiconductor devices including, for example, memory devices, logic devices, light emitting diodes, micro-electro-mechanical-systems, and/or combinations of these devices. A person skilled in the relevant art will also understand that the technology may have additional embodiments, and that the technology may be practiced without several of the details of the embodiments described below with reference to FIGS. 3-19.
Briefly described, methods and devices for testing bare and packaged semiconductor dies are disclosed. The disclosed methods and systems enable operators to test devices having pads, solderballs and/or other contact structures that have a small size and/or pitch. Solderballs, pads and/or other suitable conductive elements are collectively referred to herein as “contact structures.” In many embodiments, the technology described in the context of one or more type of contact structure can also be applied to others. In contrast with the conventional contact structures, the designed asperities can be significantly smaller than the corresponding contact structures on the device under test. Unless otherwise noted, the term designed asperities is used herein to encompass designed non-uniformities (e.g., deliberately formed structures that are small, but are not simply surface roughness that may form as a fallout from a manufacturing process). Accordingly, designed asperities have a scale that is substantially smaller than the scale of the corresponding pad/solderball of the device it contacts (e.g., device under test). Designed asperities can be formed in a controlled repeatable process, e.g., to maintain consistent size/performance from one batch to another. Designed asperities, e.g., nanospikes, subscale penetrating structures and/or other structures performing similar functions can be tailored for the metal, surface finish, and/or pitch of the contact structures contacted to improve (e.g., optimize) electrical contact and/or reduce (e.g., minimize) damage to the contact structures. Accordingly, while representative examples of the present technology are described below in the context of nanospikes, the technology is not limited to nanospikes, and can instead involve other designed asperities. The designed asperities, e.g., the nanospikes, are engineered and manufactured to possess a high degree of regularity, as opposed to naturally occurring roughness on the surface of solid materials. In some embodiments, a high degree of size uniformity and a smoothness of surface of the nanospikes can reduce or minimize the tendency for contact materials to stick to the nanospikes even when the nanospikes repeatedly penetrate into the contacts. Because the nanospikes are much smaller than conventional contact structures, the damage to the contact structures is reduced or eliminated, even if the device is contacted repeatedly. In general, the contact force can be smaller because nanospikes displace a smaller amount of material by a smaller distance than do conventional test devices. In at least some embodiments, the multiple nanospikes transmit electrical current to a pad/solderball with less resistance and therefore less current constriction than does a single, larger contact point associated with the conventional technology. Since the contact resistance between the nanospikes and corresponding pad/solderball is, at least in the first order of approximation, inversely proportional to the total contact area, the nanospikes can produce a relatively low contact resistance due to the large combined contact area provided by the multiple nanospikes contacting corresponding pad/solderball. Similarly, the electrical resistance of the nano-spike-based contactors is proportional to the ratio of the cross sectional area of the nanospikes and their electrical conductivity, and is inversely proportional to the length of the nanospikes. Due to the small length of the nanospikes and high total cross-sectional area provided by multiple nanospikes, the electrical resistance of the nanospike-based contactors can be acceptable even with the nanospike material having a relatively low intrinsic electrical conductivity. Consequently, some materials that are not suitable for the conventional contactors due to their poor electrical conductivity may be suitable as a nanospike material. In some embodiments, the nanospikes can be about 0.3-2 μm long. In other embodiments, the nanospikes can be about 50-200 nm long, making them suitable for the manufacturing process based on e-beam lithography. Other sizes of the nanospikes, for example smaller than 50 nm and bigger than 2 μm are also possible. In at least some embodiments, the relatively small length of the nanospikes limits the penetration of the nanospikes through the pads on a device under test, thus protecting the dielectric layer that is typically located under the pad. Much greater pressure would be required for the nanospikes to further penetrate through the dielectric layer than to simply penetrate into the solder and/or the pad material. Since multiple nanospikes are distributed over a wafer-side contact pad of a translator, the alignment between the pad/solderball of a device under test and the wafer-side contact pad of the translator is more robust (e.g., more tolerant of mis-alignments and/or temperature excursions) than is the alignment associated with conventional methods that use a single, larger spring-loaded pin for contacting the pad/solderball. In some embodiments, the nanospikes can be used on the inquiry-side contact pads of the wafer translator in addition or in lieu of using the nanospikes on the wafer-side contact pads. Consequently, in at least some embodiments, the current carrying capability, alignment and/or repeatability of the tests conducted with contactors that include nanospike technology are improved when compared with conventional technology.
FIGS. 3 and 4 illustrate a translator system 100 configured in accordance with embodiments of the presently disclosed technology. FIG. 3 is a schematic side view of the translator system 100. The overall system 100 can include a tester 102 that transmits signals and power to one or more dies on a wafer 150. The tester 102 can also measure, analyze and store the return signals from the device under test to determine its quality. The wafer 150 is supported by a wafer chuck 160, which can also provide thermal control and precise positioning for the wafer. The signals and power from the tester 102 pass through a tester cable 104 to a device interface board 120, which may be a printed circuit board with an appropriate routing arrangement to distribute the signals and power to required locations on a wafer-facing side of the printed circuit board. In addition or in lieu of the tester cable 104, the tester 102 (or some parts of it) can dock directly to the device interface board 120 through suitably distributed pairs of connectors. The device interface board 120 can electrically and mechanically connect to an interposer board 130, which, in turn, can precisely dock with a translator 140 to distribute the signals and power to the translator. In some embodiments of the present technology, the translator 140 has larger contact pads facing the interposer board 130 (e.g., inquiry-side contact pads of the translator) and smaller contact pads facing the wafer 150 (e.g., wafer-side contact pads of the translator). In at least some embodiments of the present technology, larger inquiry-side contact pads on the translator 140 facilitate contact with the pins of the interposer board 130. The smaller wafer-side contact pads and/or pins of the translator 140 correspond to the finer pitch and size of the pads/solderballs on the wafer 150.
FIG. 4 is a side view of selected elements of the translator system 100 of FIG. 3. Many features of FIG. 4 are not shown to scale for purposes of illustration. The system 100 can include a translator stack 400 that in turn includes the device interface board 120, the interposer board 130 and the translator 140 connected to an active side of the wafer 150. The wafer chuck 160 supports the opposite, non-active side of the wafer 150. Starting from the top of the stack 400, the device interface board 120 can have interface board connectors 406 for interfacing with the tester or tester cable (not shown). The signals and power from the tester can be distributed to interface board pins 428 facing corresponding contacts 438 on the interposer board 130. In other embodiments, the interposer board 130 can have pins facing corresponding contacts of the device interface board 120, or other suitable connection arrangements. In any of these embodiments, the interposer board 130 can be aligned relative to the translator 140, e.g., using a support ring 480. In some embodiments of the present technology, a space 486 between the interposer board 130 and the translator 140 can be at least partially evacuated, e.g., through openings 482 to provide good mechanical contact between interposer pins 436 of the interposer board 130 and the inquiry-side contact pads 460 of the translator 140. Seals 484 can be configured to maintain the vacuum between the interposer board 130 and the translator 140. In other embodiments, the contact between the interposer board 130 and the translator 140 can be provided by, for example, mechanical clamps or other suitable devices.
The inquiry-side contact pads 460 on the inquiry side of the translator 140 are electrically connected to corresponding wafer-side contact pads 430 using traces 445. The relatively large size of the inquiry-side contact pads 460 facilitates alignment with the interposer pins 436, while on the opposite side of the translator 140, the relatively small size of the wafer-side contact pads 430 corresponds to the smaller pads/solderballs (e.g., contact structures 423) on the wafer. Traces 445 in a translator board 435 electrically connect the inquiry-side pads to the wafer-side contact pads.
In some embodiments of the present technology, the wafer side contact pads 430 include nanospike elements (e.g., nanospikes 410) that can be carried by nanospike pads 420. In other embodiments, the nanospikes 410 can be attached to or formed integrally with the contact pads 430 without the intermediary nanospike pads 420. In any of these embodiments, the nanospikes can improve the electrical and/or mechanical contact between the translator 140 and the wafer 150 by having a vacuum in a space 496 between the translator 140 and the wafer 150, e.g., via openings 492. Seals 494 can be positioned around the space 496 to maintain the vacuum in the space 496 between the translator 140 and the wafer 150.
The translator 140 can be aligned with the wafer 150 via the wafer chuck 160, since the wafer is aligned and secured against the wafer chuck. In certain embodiments, the wafer 150 and the wafer translator 140 can be removeably attached by systems and methods described in U.S. patent application Ser. No. 12/547,418, assigned to the assignee of the present application, filed on Aug. 25, 2009, and entitled “Maintaining a Wafer/Wafer Translator Pair in an Attached State Free of a Gasket Disposed Therebetween,” which is hereby incorporated by reference in its entirety. In other embodiments, the translator and wafer can be kept in contact by a mechanical clamping device or other suitable devices. The alignment and vacuum between the inquiry-side of the translator 140 and the wafer 150 pulls the nanospikes in contact with contact structures of the wafer 150, as described further below with reference to FIG. 5.
FIG. 5 is an enlarged view of a portion of the translator 140 and the wafer 150, illustrating the contact between the nanospikes 410 of the translator 140 and the contact structure 423 (e.g., the solderball 422). The nanospikes 410 can be distributed over the wafer-side nanospike pad 420 which is attached to the translator board 435. In some embodiments of the present technology, several nanospikes 410 contact the contact structure 423, and in other embodiments hundreds or more nanospikes may contact the contact structure. The nanospikes 410 can have a length (L) of about 1.0-1.5 μm and a base (B) of the same scale (e.g., about a μm scale), resulting in a nanospike cross-section (C) ranging from about several μm²at the base B to sub-μm²closer to the tip of the nanospike. In some embodiments of the present technology, the nanospikes are approximately 0.3-2 μm long and in other embodiments the nanospikes, have other suitable lengths. Since the nanospikes 410 are significantly smaller than the solderball 422, the contact does not mechanically damage the solderball. If a vacuum is used in the space 496 between the translator board 435 and the pad 422, the translator board 435 and the nanospike pads 420 may bend (as illustrated in dashed lines), thus further reducing the electrical contact resistance by bringing even more nanospikes 410 in contact with the solderball 422.
FIG. 6 is a bottom view of a wafer-side contact pad array 600 of the wafer translator 140. The contact pad array 600 may include multiple nanospike pads 420 arranged to correspond to the die contacts of one or more dies on a wafer. For example, the pad array 600 can include a 14×14 array of nanospike pads 420 in the illustrated embodiment, or a variety of other suitable configurations in other embodiments. In some embodiments of the present technology, the nanospike pads 420 on the wafer-side of the translator 140 may be formed even where there are no corresponding contact pads/solderballs on the die, for example to simplify the manufacturing process of the nanospikes. Detail A of FIG. 6 is an enlarged view of one nanospike pad 420. Detail A illustrates a generally rectangular layout of the nanospikes 410. In other embodiments, the nanospikes 410 can be distributed in concentric circles, in staggered rows, randomly, and/or in other arrangements. Many suitable layouts can produce good mechanical and electrical contact between a sufficient number of the nanospikes 410 on a nanospike pad 420 and a corresponding pad/solderball of the device under test because of the small size of the nanospikes in comparison with the larger pads/solderballs.
FIGS. 7A-7C are a side view, a top view cross-section and a bottom view cross-section, respectively, of a row of the solderballs 422 in contact with the corresponding nanospikes 410 described above. The solderballs 422 are electrically connected to a die 712 via the corresponding pads 421. In some other embodiments, for example, when the device under test is a packaged semiconductor device, the pads 421 may be connected to a substrate which carries the die. In either of these embodiments the nanospikes 410 can provide suitable electrical contact with the corresponding solderballs 422 even with some misalignment between the nanospike pads 420 and the solderballs 422. Because the multiple current-carrying nanospikes 410 are distributed over a suitable area, the nanospikes contact corresponding solderballs even if the nanospike pads 420 are not perfectly centered on the corresponding solderballs 422. Accordingly, the alignment requirements for the overall system can be relaxed, thus potentially reducing system complexity and/or cost. In some embodiments, the alignment between the translator and the device under test can be purposely offset from one touchdown to another so as to contact different regions of nanospikes with the corresponding device contact structures. An advantage of this arrangement is that it can, for example, more evenly wear the nanospikes or require less frequent cleaning of the nanospikes. In other embodiments, the device under test or/and the translator may move, at least in part, laterally (i.e., in generally parallel planes) after establishing a contact to further improve the contact between the nanospikes and corresponding contact structures of the device under test
FIG. 8 is a side view of the solderballs 422 in contact with the nanospikes 410. In this embodiment, a vacuum (represented by reference number 855) is applied to the space between the wafer translator and the wafer. The vacuum 855 can improve the contact between the nanospikes 410 and the solderballs 422. In some embodiments of the present technology, the vacuum 855 may also cause the wafer translator board 435 to flex (e.g., as described above with reference to FIG. 5), which can further improve the contact between the nanospikes 410 and the solderballs 422. For example, flexing the translator board 435 can drive more nanospikes 410 into the solderballs 422 than can a flat translator board because the flexed translator board conforms, at least in part, to the surface of the solderballs 422. The flexing of the wafer translator board illustrated in FIG. 8 is not necessarily drawn to scale, and may be smaller and/or distributed in other manners in some embodiments of the present technology.
FIG. 9 is a side view of a test assembly 900 having a stack of (from bottom-to-top) the wafer 150, the translator 140 and the interposer board 130. The wafer 150 contains multiple dies 712, which can be contacted and tested one die at a time or simultaneously. The nanospikes 410 on the wafer-side of the translator 140 are electrically routed through the translator board 435 to the inquiry-side pads 960. In some embodiments, inquiry-side solderballs 970 may be disposed on the inquiry-side pads 960. Since the inquiry-side solderballs 970 are relatively large (in comparison with the curved top surface of the die solderball 422), the interposer board 130 can include relatively large interposer pins 990. In some embodiments, the interposer pins 990 can have inquiry-side penetrators 980 for breaking the oxide layer over the inquiry-side solderballs 970. The opposite sides of the interposer pins 990 can be connected to the interposer board and further to the tester.
FIGS. 10A-10G illustrate a representative process for manufacturing the nanospikes in accordance with an embodiment of the present technology. FIG. 10A shows a nanospike manufacturing assembly 1000 having carriers 1030 attached to a manufacturing substrate 1010 by carrier attach elements 1020. The carriers 1030 can have a diameter (or, for a non-circular shape, another characteristic cross-sectional dimension) that is generally close to the size of the nanospike pad. The carriers 1030 can be made of metal or another electrically conductive material including, for example, heavily doped semiconductors or graphite. As explained in relation to FIGS. 10B-10G below, the nanospikes form on top sides 1031 of the carriers 1030.
FIG. 10B shows the nanospike manufacturing assembly 1000 with the carriers 1030 in an encapsulation material 1040. The encapsulation material 1040 can include, for example, silicon epoxy, and/or other suitable constituents. The encapsulation material 1040 can maintain a fixed distance between the carriers 430 and can also protect the overall assembly from damage during the manufacturing process.
FIG. 10C illustrates the nanospike manufacturing assembly 1000 with a suitable photoresist material 1050 applied over the top sides 1031 of the carriers 1030. FIG. 10D (including Detail A) illustrates the nanospike manufacturing assembly 1000 after openings 1055 have been formed in the photoresist material 1050. The locations of the openings 1055 generally corresponds to the locations of the nanospikes which will be manufactured on the top sides 1031 of the carriers 1030. In at least some embodiments, the diameter of each opening 1055 approximates a desired diameter of a corresponding nanospike. Therefore, the nanospikes having similar dimensions, i.e., high regularity, can be manufactured when the openings 1055 have similar size. The thickness of the photoresist material 1050 can, at least in part, determine the length of the nanospikes.
FIG. 10E (including Detail A) illustrates the nanospike manufacturing assembly 1000 after a nanospike material has been disposed on the photoresist material 1050 and in the openings 1055 in the photoresist material 1050, e.g., by chemical vapor deposition, sputtering, or other suitable methods. Accordingly, disposing the nanospike material creates both a layer 1060 over the photoresist material 1050, and the nanospikes 410 in the openings 1055 of the photoresist material 1050.
FIG. 10F (including Detail A) illustrates the nanospike manufacturing assembly 1000 after the photoresist material and the encapsulation material are removed. Suitable methods for removing the photoresist including, for example, applying a liquid resist-stripper, which chemically alters the resist so that it no longer adheres to the underlying carriers 1030. In some other embodiments, the photoresist material may be removed by a plasma containing oxygen (i.e., by “ashing”), which oxidizes and removes the photoresist material. After the photoresist material is removed from the carriers 1030, the nanospikes 410 are exposed.
FIG. 10G illustrates a process for thinning the carriers 1030, which forms the nanospike pads 420 that support the nanospikes 410. The removal process is indicated schematically by arrows 1065. Suitable methods for thinning the carriers 1030 include, for example, chemical etching or mechanical grinding. The carriers 1030 may be singulated from the manufacturing substrate 410 (not shown) prior to thinning Suitable handlers (e.g., pick-and-place devices) can manipulate and/or hold in place the carriers 1030 during the thinning process. In at least some embodiments of the present technology, the manufacturing process described with reference to FIGS. 10A-10G can produce the nanospikes 410 attached to the nanospike pad 420 that is suitable for attachment to the translator.
FIG. 11 (including Detail A) illustrates a sputtering process for manufacturing the nanospikes. In the illustrated embodiment, a sputtering material 1110 is fed past an electron source 1120 which can ionize the sputtering material. The ionized sputtering material 1170 is directed through a nozzle 1130 to a sputtering chamber 540 and further toward a sputtering chuck 1160 that supports the manufacturing substrate 1010, which in turn includes the carriers 1030. Detail A of FIG. 11 shows a carrier 1030 covered with a patterned photoresist material 1050. The sputtering material 1110 can accumulate in the openings 1055 of the photoresist material to form the nanospikes 410. The sputtering chuck 1160 can rotate about an axis R, thus exposing the openings 1055 to sputtering material arriving from different angles of incidence. The resulting build-up of the nanospikes 410 is discussed further below with reference to FIGS. 12A-13H.
FIGS. 12A-12D illustrate a sequence of steps for building up the nanospikes. Each of FIGS. 12A-12D illustrates a plan top view (the lower portion of the figure) and a cross-sectional side view (the upper portion of the figure) at a different angle of incidence relative to the ionized sputtering material 1170. Each figure corresponds to a 90° rotation of the sputtering chuck relative to the neighboring figure(s). At each rotation increment, different portions of the opening 1055 in the photoresist material 1050 are exposed to the incident of sputtering material 1170, as illustrated in cross-section in the upper portions of the figures. The illustrated rotations uniformly or generally uniformly expose the opening 1055 to the ionized sputtering material 1170, despite the non-zero incidence angle of the ionized sputtering material 1170. During the illustrated process, the accumulation of the ionized sputtering material 1170 in the opening 1055 causes the nanospike to grow, as shown by the progression of accumulated material 1210 a-1210 d. Ultimately, the process can produce a generally symmetric and conical nanospike 410.
FIGS. 13A-13H illustrate a sequence of steps for building up a nanospike in accordance with another embodiment of the nanospike manufacturing process. Representative cross-sections and corresponding plan views are shown one above the other for each step. In this embodiment the 90° rotation increments occur more frequently than in the embodiment shown in FIGS. 12A-12D. Therefore, the sputtering wafer spends less time at each position, resulting in smaller increments of the vertical build-up of the nanospike at a given position. As a result, in at least some embodiments, the nanospike 410 can be made in pyramid-like layers, for example layers 1310 a-1310 d. As the sputtering chuck spends ever shorter times atany given position, the rotation approximate a continuous rotation, producing a nanospike shaped as a cone with amore continuous outer surface.
FIGS. 14A-14F illustrate several nanospikes in accordance with still further embodiments of the present technology. Combinations of the sputtering wafer rotation, the sputtering source strength and/or the sputtering source angle can produce nanospikes with different cross-sectional shapes. Some representative shapes are shown in FIGS. 14A-14F. For example, FIG. 14A illustrates a nanospike 410 a having a concave cross-section with a tip diameter D and height H. In some embodiments the height H can be about 1-1.5 μm. The nanospike 410 a can be covered by a cover material 1410 a, which is an overlay layer generally thinner than the bulk material forming the nanospike 410 a. The nanospike 410 a and the cover material 1410 a can include different materials. For example, the nanospike material can be good electrical conductor, e.g., aluminum or copper, while the cover material 1410 a can be hard and/or abrasion resistant material, e.g., a carbide compound or titanium. In other embodiments, the cover material can resist or inhibit sticking to the solder or pad material, thus reducing the likelihood for contaminating the nanospike contactors. As will be described further below with reference to FIG. 18, hard materials may provide for suitable coatings, even if they are not optimally electrically conductive. FIG. 14B illustrates a nanospike 410 b with a concave cross-section, similar to the nanospike 410 a shown in FIG. 14A, but without the cover material 1410 a. FIG. 14C illustrates a nanospike 410 c with a generally triangular cross-sectional shape having a height H and a base width D. FIG. 14D illustrates a nanospike 410 d with a convex cross-section. FIGS. 14E and 14F illustrate nanospikes 410 e and 410 f with angled tips, which may improve the contact when the wafer translator (or other contact structure that carries the nanospikes) moves sideways during the contact with the contactor structure of the device under the test. The particular nanospike shape used for a given device under test can be selected based, for example, on the type, shape, and/or location of the contact structures carried by the device.
FIGS. 15A-17B illustrate the photoresist openings and the corresponding nanospike shapes in accordance with yet further embodiments of the present technology. FIG. 15A is a plan view of a star-shaped photoresist opening 1550, which, when used in conjunction with suitable sputtering chuck rotation increments, can produce a star-shaped nanospike 410 g shown in FIG. 15B. Such a nanospike can be advantageous in some embodiments of the disclosed technology, for example to create a larger contact area between the nanospikes and the corresponding pads/solderballs. FIGS. 16A-16B illustrate a blade-shaped photoresist opening 1650 and the corresponding nanospike 410 h. The blade-shaped nanospike 410 h can be advantageous when, for example, a sliding of the nanospike over the pad/solderball is expected or desired. FIGS. 17A-17B illustrate a cross-shaped photoresist opening 1650 and the corresponding nanospike 410 h. The cross-shaped nanospike 410 h can be used when, for example, it is difficult to lithographically manufacture corners of the mask for the star-shaped nanospikes. In other embodiments, the photoresist opening and the corresponding nanospikes can have other suitable shapes to meet particular application requirements.
FIGS. 18A-18E illustrate a representative process for molding the nanospikes in accordance with an embodiment of the present technology. FIG. 18A shows a nanospike manufacturing assembly 1800 having a molding substrate 1810 and a mask 1820. The molding substrate 1810 can be, for example, a silicon wafer having a <100> crystal orientation or other molding substrate material that is suitable for anisotropic etching.
FIG. 18B shows the nanospike manufacturing assembly 1800 after a pattern of mask openings 1825 has been created in the mask 1820. The location of the openings 1825 generally corresponds to the location of the nanospikes which will be manufactured on or in the molding substrate 1810. In at least some embodiments, the diameter (or other characteristic dimension) of each opening 1825 approximates a desired diameter of a base of corresponding nanospike.
FIG. 18C shows the nanospike manufacturing assembly 1800 after a pattern of mold openings 1830 has been etched in the molding substrate 1810. In some embodiments, the molding substrate 1810 can be anisotropically etched using potassium hydroxide (KOH) or tetramethylammonium hydroxide (TMAH) and in other embodiments other anisotropic etchants are used. The shape of mold openings 1830 can be different depending at least in part on the shape of the openings 1825. For example, if the molding substrate 1810 is formed from <100> silicon, a circular mash opening 1825 results in a generally conical mold opening 1830, while a rectangular mash opening 1825 results in a generally pyramidal mold opening 1830. Furthermore, combinations of different shapes of mash openings 1825 are used in other embodiments, resulting in corresponding combinations of shapes of the mold openings 1830.
FIG. 18D shows the nanospike manufacturing assembly after the mask 1820 has been removed and a nanospike material 1840 has been applied over the molding substrate 1810. In some embodiments, the nanospike material 1840 can include several materials. For example a seed conductive material can be applied first, followed by the remainder of the nanospike material 1840.
FIG. 18E shows a process for separating the nanospike material 1840 from the molding substrate 1810. In some embodiments, the nanospike material 1840 can be mechanically peeled from the molding substrate 1810 as generally indicated by an arrow 1845. The molding substrate 1810 can be reused to mold more nanospikes. In other embodiments, the molding substrate 1810 can be etched away leaving the nanospike material 1840 with the nanospikes 410 attached to the nanospike pad 420. After separation from the molding substrate 1810, the nanospike pad 420 carrying nanospikes 410 can be attached to a suitable board, for example the translator board 435 (FIG. 5) for contacting a device under test.
In a sample embodiment of a translator manufacturing method, the method can include: applying the photoresist material 1820 over a first surface of the molding substrate 1810, patterning the photoresist material 1820 to create apertures 1825 exposing portions of the first surface of the molding substrate, etching the molding substrate 1810 to form mold openings 1830 corresponding to the apertures 1825 of the photoresist material 1820, removing the photoresist material 1820 from the first surface of the molding substrate 1810, disposing the nanospike material 1840 into the mold openings to form the nanospikes 410, removing the nanospikes from the molding substrate, and attaching the nanospikes 410 to the wafer side of the translator. The sample embodiment can further include removing the nanospikes 410 from the molding substrate 1810 by etching away the molding substrate using an etchant selected from a group consisting of KOH and TMAH. The molds 1830 can have a shape selected from a group consisting of a cone, a pyramid, and a combination thereof. An intermediate material can be applied over wafer-side contact pads on the translator, and the nanospikes 410 can be distributed over the intermediate material of the wafer-side contact pad. The nanospikes 410 can be joined with the wafer-side contact pads by heating the nanospikes, the wafer-side contact pad, or both.
FIG. 19 is a graph of electrical resistivity and Vickers hardness for several materials, e.g., materials suitable for nanospikes (the materials toward the left side of the graph) and materials suitable for representative contact pads (the materials toward the right side of the graph). The nanospike materials are generally selected to be harder than the contact pad materials to assure penetration of the pad materials (e.g., aluminum illustrated in the graph), and/or copper and tin based solders, which are softer than aluminum. In many applications, a layer of alumina (i.e., aluminum oxide) is the hardest material that can be expected on the contact pad. The nanospike material is also generally selected to be harder than the mating contact pad/solderball material on the device under test to increase the durability of the nanospikes. Since many nanospikes make contact with the corresponding pad/solderball, thus resulting in a relatively high cumulative contact area, the electrical conductivity of the nanospike material can be relatively low, yet can still produce acceptably high electrical conductivity of the nanospike based contactor. Therefore, in at least some embodiments, the nanospikes with relatively high electrical resistivity can perform well for as long as they are sufficiently hard. For example, tungsten carbide or hafnium carbide can be suitable nanospike materials based on their high hardness even though the electrical resistivity of these materials is higher than the electrical resistivity of commonly used electrical conductors like, for example, aluminum or copper.
From the foregoing, it will be appreciated that specific embodiments of the technology have been described herein for purposes of illustration, but that various modifications may be made without deviating from the disclosure. For example, in some embodiments, the nanospikes can be made by chemical etching, in addition to or in lieu of an additive process. The etching process may start by attaching wires having diameter of several microns to a metal substrate. Next, the substrate can be inverted and the wires repeatedly exposed to a chemical etchant by, for example, dipping the wires into an etchant pool. Because the etchant drains along the length of the wires, the tips of the wires tend to etch more than the bases of the wires, thus producing tapered (e.g., pointy) cones. The process can stop when the desired size of the nanospikes is achieved. In other embodiments, nanospikes having different sizes and/or shapes can be configured on a nanospike pad. Furthermore, in some embodiments of present technology, the nanospike pads 420 can be carried by the interposer board 130, which can directly contact the device under test, without the intermediate translator. The test stackup without the translator may be particularly applicable in cases for which the contact structures of the device under test are relatively large, thus not requiring a translator having smaller wafer-side pads and larger inquiry-side pads. In some embodiments of the present technology, the nanospikes can be made by applying a metal coating over a plastic substrate. Even a relatively small thickness of the metal coating can result in an acceptably low overall contact resistance because of the large surface area of the nanospikes in contact with the device under test. In some embodiments, the tip of the nanospikes can be planarized to create a blunted nanospike that may be more resistant to tip breakage. In some other embodiments, the nanospikes of different sizes can be used on same contactor. Furthermore, larger nanospikes can have smaller nanospikes on the side surfaces to, for example, decrease contact resistance by increasing the contact area. Furthermore, the smaller nanospikes can have different mechanical/electrical properties than the larger nanospikes. In some embodiments, the nanospikes can be used for contacting biological material, for example, tissue cells. The size of the nanospikes can generally correspond to the size of tissue cell or elements of the cell, for example proteins.
Moreover, while various advantages and features associated with certain embodiments have been described above in the context of those embodiments, other embodiments may also exhibit such advantages and/or features, and not all embodiments need necessarily exhibit such advantages and/or features to fall within the scope of the technology. Accordingly, the disclosure can encompass other embodiments not expressly shown or described herein.

Claims

1. An apparatus for testing semiconductor dies, comprising:

a translator having a wafer side positioned to face toward a device under test, and an inquiry side facing away from the wafer side;

a plurality of wafer-side contact sites carried by the translator at the wafer side of the translator; and

a plurality of nanospikes carried by at least one wafer-side contact site.

2. The apparatus of claim 1 wherein the nanospikes are approximately 0.3-2 μm long.

3. The apparatus of claim 1 wherein the nanospikes are arranged in a grid with a spacing of approximately 0.3-2 μm from one nanospike to another.

4. (canceled)

5. The apparatus of claim 1 wherein a cross-section of the nanospikes is selected from a group consisting of generally concave, triangular, convex, and a combination thereof.

6. The apparatus of claim 1 wherein the nanospikes are generally perpendicular to the at least one wafer-side contact pad.

7. The apparatus of claim 1 wherein a shape of the nanospikes is selected from a group consisting of a star, a blade, a cross, a spike, and a combination thereof.

8. The apparatus of claim 1 wherein the nanospikes comprise a cover material.

9. (canceled)

10. The apparatus of claim 1, further comprising:

a plurality of inquiry side contact sites on the inquiry side of the translator; and

an interposer having a plurality of interposer contacts in electrical contact with the inquiry side contact sites.

11. (canceled)

12. The apparatus of claim 10 wherein the inquiry side contact sites of the translator are larger than the contact sites at the wafer-side of the translator.

13-16. (canceled)

17. A method for establishing electrical communication with a device under test, comprising:

positioning the device under test proximate to a translator, wherein the device under test carries a plurality of contact structures, wherein the translator has a wafer side facing the device under test and an inquiry side facing away from the wafer-side, wherein the wafer side of the translator has a plurality of wafer-side contact sites; and

contacting one or more contact structures of the device under test with nanospikes carried by the wafer-side contact sites.

18. The method of claim 17, further comprising:

positioning an interposer against the inquiry-side of the translator, wherein the interposer has a first side facing the translator and a second side facing away from the first side, wherein the first side of the interposer carries a plurality of interposer pins; and

contacting the inquiry side of the translator with the interposer pins.

19. (canceled)

20. The method of claim 18 wherein the device under test is a semiconductor wafer, the method further comprising applying a vacuum between the semiconductor wafer and the translator.

21. The method of claim 18 wherein the device under test is a packaged device.

22. (canceled)

23. The method of claim 18, further comprising causing the nanospikes to move in a generally parallel direction with respect to the device under test.

24. A method for manufacturing a translator having a wafer side configured to face toward a device under test, and an inquiry side facing away from the wafer side, comprising:

applying a photoresist material over a first surface of a metal carrier;

patterning the photoresist material to create apertures exposing portions of the first surface of the metal carrier;

disposing nanospike material in the apertures of the photoresist material to form the nanospikes;

removing the photoresist material, at least in part, from the first surface of the metal carrier; and

attaching the nanospikes to the wafer side of the translator.

25. The method of claim 24, further comprising thinning the metal carrier by removing a portion of the metal carrier facing away from the nanospikes.

26. The method of claim 24, further comprising rotating the metal carrier to shape the nanospikes by controlling the disposing of the nanospike material in the apertures of the photoresist material.

27. The method of claim 24 wherein the nanospikes include a material selected from a group consisting of tantalum nitride, tungsten carbice, hafnium carbide, titanium carbide, titanium diboride, molybdenum carbide, alumina, rhenium diboride, and a combination thereof.

28. The method of claim 24, further comprising separating the nanospikes from the metal carrier prior to attaching the nanospikes to the wafer-side of the translator.

29. The method of claim 24, further comprising:

applying an intermediate material over wafer-side contact sites on the translator; and

distributing the nanospikes over the intermediate material of the wafer-side contact pad, wherein joining the nanospikes with a wafer-side contact pad includes heating the nanospikes, the wafer-side contact pad, or both.