US 20060009033 A1
The process of the present invention forms copper interconnects in a semiconductor wafer surface. During the process, initially, narrow and large features are provided in the top surface of the wafer, and then a primary copper layer is deposited by employing an electrochemical deposition process. The primary copper layer completely fills the features and forms a planar surface over the narrow feature and a non-planar surface over the large feature. By employing an electrochemical mechanical deposition process, a secondary copper layer is deposited onto the primary copper layer to form a planar copper layer over the narrow and large features. After this process step, the thickness of the planar copper layer is reduced using an electropolishing process.
1. A system for forming conductive structures on a semiconductor wafer having a top surface with a first feature and a second feature disposed therein, the first feature having a first width and the second feature having the widest width, the system comprising:
an electrochemical deposition module configured to completely fill the first feature and the second feature with a first conductive layer; and
an electrochemical mechanical deposition module configured to deposit a second conductive layer over the first conductive layer and form a planar conductive layer.
2. The system of
3. The system of
4. The system of
5. The system of
6. The system of
7. The system of
8. The system of
9. The system of
10. A system for forming conductive structures on a semiconductor wafer having a top surface with a first feature and a second feature disposed therein, the first feature having a first width and the second feature having the widest width, the system comprising:
a deposition module configured to completely fill the first feature and the second feature with a first conductive layer and to deposit a second conductive layer over the first conductive layer to form a planar conductive layer,
11. The system of
12. The system of
13. The system of
14. The system of
This application is a divisional application of and claims priority to co-pending U.S. patent application Ser. No. 10/379,265, filed Mar. 3, 2003, which is a continuation-in-part of U.S. patent application Ser. No. 10/201,606, filed Jul. 22, 2002, now U.S. Pat. No. 6,867,136, issued Mar. 15, 2005 and which in turn claims priority to U.S. Provisional Patent Application 60/384,643, filed May 31, 2002. The disclosure of each of the priority application is hereby incorporated by reference in its entirety.
1. Field of the Invention
The present invention relates to manufacture of semiconductor integrated circuits and, more particularly to a method for planar deposition and polishing of conductive layers.
2. Description of the Related Art
Conventional semiconductor devices generally include a semiconductor substrate, such as a silicon substrate, and a plurality of sequentially formed dielectric interlayers such as silicon dioxide and conductive paths or interconnects made of conductive materials. Copper and copper-alloys have recently received considerable attention as interconnect materials because of their superior electro-migration and low resistivity characteristics. The interconnects are usually formed by filling copper in features or cavities etched into the dielectric layers by a metallization process. The preferred method of copper metallization is electroplating. In an integrated circuit, multiple levels of interconnect networks laterally extend with respect to the substrate surface. Interconnects formed in sequential layers can be electrically connected using vias or contacts.
In a typical process, first an insulating layer is formed on the semiconductor substrate. Patterning and etching processes are porformed to form features or cavities such as trenches and vias in the insulating layer. Then, a barrier/glue layer and a seed layer are deposited over the patterned surface and a conductor such as copper is electroplated to fill all the features. However, the plating process, in addition to filling the features with copper, also deposits excess copper over the top surface of the substrate. This excess copper is called an “overburden” and needs to be removed during a subsequent process step. In standard plating processes this overburden copper has a large topography since the Electrochemical Deposition (ECD) process coats large features on the wafer in a conformal manner. For example, a wafer with 0.5 micron deep features may be coated with 0.8 micron thick copper by the standard ECD process, to assure complete, defect-free filling of all the features, including those that are wider than about 5 microns. The resulting copper surface then may have a topography that has a step of about 0.5 microns over the large features. Conventionally, after the copper plating, CMP process is employed to first globally planarize this topographic surface and then to reduce the thickness of the overburden copper layer down to the level of the surface of the barrier layer, which is also later removed leaving conductors only in the cavities. CMP is a costly and time consuming process. High pressures used in the CMP processes also damage low-k dielectrics, which are mechanically weaker than the silicon oxide. Therefore, minimizing CMP step in an integration process is a goal for all IC manufacturers. The topography on the wafers also causes problems for the CMP process. Specifically, the large steps such as the 0.5 micron step of the above example over the large features such as 100 micron wide trenches or bond-pads, cause dishing defects after CMP. Therefore, both in terms of cost and enabling features, it is very attractive to have processes that have the ability to yield thinner copper deposits with reduced surface topography on wafers.
During the copper electrodeposition process, specially formulated plating solutions or electrolytes are used. These electrolytes typically contain water, acid (such as sulfuric acid), ionic species of copper, chloride ions and certain additives which affect the properties and the plating behavior of the deposited material. Typical electroplating baths contain at least two of the three types of commercially available additives such as accelerators, suppressors and levelers. It should be noted that these additives are sometimes called different names. For example, the accelerator may be referred to as a brightener and the suppressor as a carrier in the literature. Functions of these additives in the electrolyte and the role of the chloride ion are widely known in the field (see for example, Z. W. Sun and G. Dixit, “Optimized bath control for void-free copper deposition”, Solid State Technology, November 2001, page 97), although the details of the mechanisms involved may not be fully understood or agreed upon.
Electrodeposition process needs to fill all the features, small and large, on the substrate.
Several developments have been provided by prior art methods to improve the copper topography depicted by the solid line 15 in
As the above review demonstrates, some of the prior art techniques aiming to obtain relatively flat copper topography on patterned wafer surfaces may be applicable for the class of wafers with large or medium aspect-ratio features. However, many 1C interconnect designs contain features with a wide variety of aspect ratios on a given wafer surface. Especially in multi-level interconnect structures the width of the high-current-carrying lines increase while their aspect ratios decrease at the higher wiring levels. Therefore, an approach that has the capability of reducing or eliminating surface topography of copper over features with a large range of aspect ratios is needed.
A technique that can reduce or totally eliminate copper surface topography for all feature sizes is the Electrochemical Mechanical Processing (ECMPR). This technique has the ability to eliminate steps D1, D2 and overfill 0 shown in the example of
Descriptions of various ECMPR approaches and apparatus, can be found in the following patents, published applications and pending applications, all commonly owned by the assignee of the present invention: U.S. Pat. No. 6,126,992 entitled “Method and Apparatus for Electrochemical Mechanical Deposition,” U.S. application Ser. No. 09/740,701 entitled “Plating Method and Apparatus that Creates a Differential Between Additive Disposed on a Top Surface and a Cavity Surface of a Workpiece Using an External Influence,” filed on Dec. 18, 2001 and published as U.S. patent application on Feb. 21, 2002 with patent application No. 20020020628, U.S. application filed on Sep. 20, 2001 with Ser. No. 09/961,193 entitled “Plating Method and Apparatus for Controlling Deposition on Predetermined Portions of a Workpiece”, U.S. application with Ser. No. 09/960,236 filed on Sep. 20, 2001, entitled “Mask Plate Design”, and U.S. application No. 0/155,828 filed on May 23, 2002 entitled “Low Force Electrochemical Mechanical Processing Method and Apparatus.” These methods can deposit metals in and over cavity sections on a workpiece in a planar manner.
The present invention relates to a process for forming a near-planar or planar layer of a conducting material, such as copper, on a surface of a workpiece using ECD and ECMPR techniques. The process preferably uses at least two separate plating solution chemistries to form a near-planar or planar copper layer on a semiconductor substrate that has features or cavities on its surface.
In one aspect, the present invention provides a process for forming conductive structures on a semiconductor wafer having a top surface with first and second cavities formed in the top surface. The first cavity of the wafer has a narrower width than the second cavity.
The process includes the step of depositing a first conductive layer by employing an electrochemical deposition process. The deposition process uses a first electrolyte solution having a first additive composition. The conductive layer completely fills the cavities while forming a planar surface over the first cavity and a non-planar surface over the second cavity.
The process further includes the step of depositing a second conductive layer on the first conductive layer to form a planar conductive layer over the first and second cavities. The planar conductive layer comprises the first and the second conductive layers and the second conductive layer is deposited using an electrochemical mechanical deposition process using a second electrolyte solution having a second additive composition.
The above and other features and advantages of the invention are further described in the detailed description which follows, with reference to the drawings by way of non-limiting exemplary embodiments of the invention, wherein like reference numerals represent similar parts of the invention throughout several views and wherein:
The process of the present invention relates to a process for forming a near-planar or planar layer of a conducting material, such as copper, on a surface of a workpiece using an ECMPR technique such as ECMD. The process of the present invention preferably uses at least two separate plating solution chemistries to form a near-planar or planar copper layer on a semiconductor substrate that has features or cavities on its surface. The unique feature of the present invention is the fact that it can be applied to substrates with features or cavities of a wide variety shapes and dimensions. For example the substrate may have high-aspect ratio small size features, such as sub-micron size vias or trenches, as well as very low aspect-ratio features such as large than 10 micron wide trenches and larger than 20 micron size pads or channels. There may even be larger than 500 micron features on this exemplary wafer. At a first stage of the present invention, a plating process that is optimized for the defect-free filling of the small features is carried out. During this stage, plating is done in non-touch mode with no mechanical sweeping on the wafer surface. The no-touch process uses a first plating solution with first additive chemistry that is optimized for the best gap-fill performance. This chemistry may contain accelerators and suppressors and optionally levelers. At this stage the small and mid size features are entirely filled with copper while the larger size features are partially or completely conformally coated with a copper layer. An example of a copper layer profile resulting from the first stage of the present invention is shown in
A second stage of the present invention uses an ECMPR touch-process, typically including at least an ECMD process, with a second plating solution that has a second additive chemistry, which is optimized specifically for this planar deposition or planarization step. This second chemistry, for example, may contain only one or two of the three additives that may be used for the first stage. Depending upon the duration of the touch-step a copper profile that is substantially flat (shown as profile 31) or absolutely flat (shown as profile 32) can be obtained. As the duration of the touch-step is increased the surface becomes flatter and flatter. In any case the reduced, flatter topographies of profiles 31 and 32 in
In the first stage of the process, the WSID acts as a shaping plate. It is important that the channels of the WSID not only allow the process solution to flow to the surface of the wafer, but also shapes the plating current density and thus the resulting thickness profile of the deposited copper layer. The distribution, shape and size of the openings may introduce regions of low, intermediate, and high deposition rates above the WSID. By moving the wafer over these regions during the process a desired thickness profile, e.g., a uniform thickness profile, of the depositing“ ” layer is obtained. An exemplary process of thickness profile control is disclosed in U.S. patent application Ser. No. 09/760,757 filed on Jan. 17, 2001, entitled “Method and Apparatus for Electrodeposition of Uniform Film on Substrate” which is commonly owned by the assignee of the present invention. If desired, this plating step can be carried out without the WSID if other means are provided to yield a uniform deposit.
In this embodiment, the first electroplating solution 120 may comprise at least two, additives to enhance bottom up filling of the small features without any voids, seams and other defects. For example a high-acid plating electrolyte containing 0.8-2 ml/l of Cubath® ViaForm™ Accelerator available from Enthone-OMI®, West Haven, Conn. and 6-12 ml/l of Cubath®ViaForm™ Suppressor marketed by the same company may be used in the basic plating bath chemistry which contains sulfuric acid, copper sulfate, water and chloride ions. Low acid versions of the plating chemistry may require quite different concentrations of accelerators and suppressors (such as am accelerator concentration of about 4-8 ml/l, and suppressor concentration of 2-4 ml/l for the case of Enthone low acid chemistry). During the process, the accelerators allow bottom up growth of copper in the small features that are filled quickly. The suppressor molecules which are adsorbed at the top of the openings of the small features slow down the copper growth therein, thus avoiding both premature closing of these passages and formation of voids. In addition to the accelerator and suppressor species, levelers may also be added in the formulation to reduce or eliminate the over-fill phenomenon” discussed before in this application. Levelers get adsorbed preferentially on high-current density regions of the plated surface and help reduce this current density and thus protrusions that would have resulted. A leveler concentration of 0.5-2 ml/l may be used for this purpose in addition to the accelerator and suppressor species in the exemplary high acid chemistry mentioned above. An exemplary leveler which is sold under the brand name Cubath® ViaForm™ Leveler is available from Enthone-OMI®.
As illustrated in
The present invention allows optimization of the chemistry for the first stage of the process separate from the optimization of the chemistry for stage 2. This is important because although it is possible to use the same chemistry for both the first and second stage of the ECMPR as described in the above mentioned patents and patent applications, it is attractive and beneficial to have the ability to optimize chemistry for the two stages independent from each other as will be further described below.
Concentrations and types of additives used in the first stage of the subject process may vary depending upon the nature of the additives, nature of the small features, nature of the barrier/seed layers, etc. For example, some accelerator species containing sulfur react with weak seed layers. If the seed layer is very thin on the side-walls of the vias for a specific wafer, it may be necessary to reduce the accelerator-to-suppressor ratio in the electrolyte in the first stage of the process that is employed to coat this specific wafer. For other wafers with other types of seed layers it may be necessary to fullher adjust the relative concentrations of additives to obtain the best gap fill performance. If the over-fill over the dense array of small features presents a problem, it may be necessary to include leveler into the formulation in addition to the accelerator and suppressors.
The electrolyte chemistry, which is adjusted for best gap fill in the first stage may not be the best for the second stage of the process when planarization takes place. For example, levelers are known to get attracted to high current density reunions on the wafer surface. However, ECMD process is known to accelerate growth in cavities on the substrate surface by increasing the deposition current density in such cavities compared to the top surface of the wafer swept by 1:he WSID. Therefore, levelers in an electrolyte used for the second stage of the process may lower the planarization efficiency. This is an example of how presence of an additive (such as a leveler) in the process electrolyte may be favored during the first stage (no-touch step) of the process, whereas its presence may not be necessary or desired during the second stage (touch plating step). Similarly, a higher accelerator-to-suppressor ratio may be desirable in the second stage than for the first stage of the process as described before. The experiment of
In accordance with the principles of the present invention the first and the second stages of the process may be either performed in the same process module or multiple process modules. If the same process module approach is used to perform both stages, the first and second stages are performed sequentially, and using a first solution for the first stage and a second solution for the second stage. As fully described above, the first solution chemistry includes additives enhancing bottom up filling of the features on the wafer. And, the second solution chemistry includes only one or two of the three additives that may be used for the first stage and is optimized specifically to obtain a planar copper layer. If there are multiple process modules, the first stage may be carried out in a first or first group of process modules with the first solution chemistry, and the second stage may be carried out with a second or second group of process modules using the second solution chemistry. Following the two-stage deposition, the wafer is cleaned and planar or near-planar copper overburden is removed using the CMP or other (e.g., electropolishing) removal methods. The copper overburden may be removed before or after an annealing step.
It should be noted that after the second stage of the present process, optionally a third and even a fourth step can be used to reduce the thicloless of the copper overburden. After the second stage of the process, the copper removal process may be performed as a third stage of the process that employs either an electrochemical etching or polishing stage or an ECME (electrochemical mechanical etching or polishing) stage. The removal process may be also performed using both steps sequentially as a third and fourth stage, for example, a no-touch electroetching step that is followed by a touch ECME; or a touch ECME followed by a no-touch electroetching. Thin planar deposits can be obtained by first depositing a planar layer using the ECMD technique at the first and second stages, and then electroetching or electropolishing this planar film in the same electrolyte or in an electroetching solution by reversing the applied voltage. This way the thickness of the deposit may also be reduced in a planar manner. In fact, etching may be continued until all the metal on the field regions is removed. These techniques may be performed subsequent to second stage and using the second solution as an electro etching solution while reversing the polarity of the applied voltage and rendering the workpiece surface more anodic compared to the electrode. Alternatively, a third solution comprising an electro etching solution may replace the second solution at the third stage and a fourth stage of the process, for example a no-touch electroetc4ing (third stage) followed by a touch ECME (fourth stage). In this respect, a fourth stage such as a touch ECME process stage may be performed using a fourth solution.
As shown in
In an exemplary process sequence, at the first stage of the process, the first process solution from the first solution tank 224 of the PSM1 is delivered, via the valve 218, to the process solution container 206 of the PM1 and circulated back through return line 236. After the wafer 214 is processed the valve 218 is turned to ill supply and DI water from the rinse tank 228 is delivered to the process container 206 via the valve 218 to clean the process container from the residues of the first solution. During the cleaning, the valve 218 may be periodically turned off and the valve 220 is turned onto direct the used rinsing solution to the drain 222. After the rinsing the second stage of the process is performed similar to the first stage but using the fresh second solution from the second tank 226 and delivering the used second solution back to second solution tank 226 for the replenishment and storage purposes. After the second stage of the process, the process container is once again rinsed for the following wafer to be processed with the process of the present invention. It should be noted that there are many ways of feeding the various solutions to the process module. The example given here is just one of many possibilities. In case the two solutions used in this example are compatible, the rinse steps in between may be skipped and small amount of intermixing between the solutions may be allowed.
A second stage of the process is a planarization stage and uses an ECMPR touch-process, typically including an ECMD process, with a second plating solution that has a second additive chemistry, which is optimized specifically for this planar deposition or planarization step. The second stage of this embodiment may be performed either in the same process module that the first stage is performed or in a different process module. If it is performed in the same ECMPR system, the first plating solution may be replaced with the second solution. Further, the second stage of the process can be performed in a different ECMPR system using the second solution. With the touch process, deposited copper fills the steps on the copper layer formed at the first stage, and forms a planar layer. Since the copper deposited at the first stage has already filled the features; the copper deposited in the second stage is for forming a planar copper layer on the surface i.e. planarizing the wafer surface. The second chemistry used in this planarization stage may contain only one or two of the three additives that may be used for the first stage. The copper quality resulting from the second chemistry may not be as good as the quality of the copper from the first chemistry using three components. This may be acceptable since the copper deposited in the second step is sacrificial and will be removed later.
It should be noted that there may be an optional annealing step, between the first and second stage of the present process. In other words, since the first step fills all the features with good quality copper, the resulting film depicted in
As a result of the second step and depending upon the duration, a copper profile that is substantially flat can be obtained. As described in the previous embodiments, as the duration of the touch-step is increased the surface becomes flatter.
Flatter topography provided by the second stage of the process increases the efficiency of the following third stage or the material removal stage using for example an electropolishing or electroetching stage. The electropolishing or electroetching may be performed in the same ECMPR system by reversing the polarities between the substrate and the electrode. The second solution used in the previous planarizalion stage may be used as the etching solution under a the reversed P9larity conditions. However, the second solution may also be replaced with a specially formulated etching solution to perform the electropolishing process. Alternatively, the electropolishing process may be performed in a separate electropolishing module using a separate etching solution. This alternative electropolishing system may be an electrochemical mechanical etching or polishing module or an electrochemical polishing module that does not introduce mechanical action on the wafer surface during the electropolishing process. It should be noted that throughout this application electrochemical etching or electrochemical polishing refer to the same process of removing material by electrochemical means. Similarly electrocherilical mechanical etching (ECME) and electrochemical mechanical polishing (ECMP) refer to the same group of processes.
When applied to the plated substrate, the electropolishing process may entirely remove the copper layer deposited during the planarization step using the second chemistry while partially removing the copper layer deposited during the ECD step using the first chemistry. The partially removed part of the copper deposited with first chemistry is the copper portion that is outside the features. Although in this embodiment, the step following the second stage of the process is an electropolishing process, a CMP process can also be used to remove the planar layer and is within the scope of this invention. As used herein, features include cavities, vias, trenches, and pads.
Referring now to
In this embodiment, the first electroplating solution 320 may comprise at least two additives to enhance filling of the all size features without any voids, seams arid other defects. The chemistry of the solution 320 forms a flat film on the small or mid size features. When the flat film thickness ‘t’ gets larger than approximately the depth ‘d’ of the large feature, the first stage of the plating process is halted. The t>d condition indicates that the thickness of the copper layer within the large feature is also larger than about “d” and thus the large feature is completely filled with copper. The primary copper is a good quality copper that would, after annealing and other post-deposition procedures, provide good stress migration and electromigration properties. For example a high acid plating electrolyte containing 1.5 ml/l of Cubath® ViaForm™ Accelerator, 8 ml/l of Cubath® ViaForm™ Suppressor and 2 ml/l Cubath@ ViaForm™ Leveler, all available from Enthone-OMI®, West Haven, Conn., may be used in the first solution chemistry which contains sulfuric acid, copper sulfate, water, and chloride ions. During the process, the accelerators allow bottom up growth of copper in the small features that are filled quickly. The suppressor molecules which are adsorbed at the top of the openings of the small features slow down the copper growth therein, thus avoiding both premature closing of these passages and formation of voids. Levelers reduce or eliminate the over-fill phenomenon and are absorbed preferentially on high-current density regions of the plated surface and help reduce this current density and thus minimize the protrusions or bumps that would have resulted. It should be noted that two-component chemistries without levelers may also be used for this process as long as a near-uniform thickness of copper is obtained over the small and medium size features, without excessive bumping or overfill.
As illustrated in
The second plating solution 336 preferably does not contain any levelers although it may contain some small aIllount (such as 0.5 ml/l) and it may employ one component chemistry such as a chemistry containing only an accelerator, or two component chemistry such as one containing accelerator and suppressor. If the second solution is an one-component solution, the accelerator concentration may be in the range of 1-10 ml/l. If the second solution is a two-component solution, the accelerator concentration in the second solution may be in the range of 1-10 ml/l, whereas the suppressor concentration may be in the range of 1-10 ml/l. In other words, the process window can be very wide. This can be done because all the small features have already been filled on the wafer by the first plating step and therefore there is no danger of causing void formation in such features by these new additive concentrations, which, if used during the first st8;ge process, would result in non-optimal filling of the small features. It should be noted that the touch-plating stage of the process may be terminated before total planarization is achieved if the degree of planarization achieved at that point is adequate for simplifying the thickness reduction process such as an electropolishing or CMP process, which follows the second stage of the deposition process.
It should be noted that, although use of two different chemistries is described here,- the invention may also be used employing the same chemistry for the first and second stages of the process. In this case, however, chemistry needs to be selected to provide good quality copper for the first step and good planarization for the second step. Electrolytes with two additives (accelerator and suppressor) or electrolytes with three additives (accelerator, suppressor and leveler) where leveler concentration is low may be used for this purpose. For example, for commercially available low acid ViAForm chemistry supplied by Enthone, the standard ECD additive concentrations are accelerator of about 6 ml/l, suppressor of about 2 ml/l and leveler of about 2 ml/l. If a single 3-component chemistry is to be used for the present invention, the leveler concentration may be cut to below 1 ml/l, for example to 0.5 ml/l, not to affect negatively the planarization capability of the second step.
As an alternative, a two-component chemistry may be used in the first step using the invention to obtain a flat profile over the small features while filling completely the large features. An example of such process may be found in U.S. Provisional Application No. 60/444,355 entitled “Method of Electroplating Copper Layers with Flat Topography” filed Jan. 30, 2003, which is assigned to the assignee of the present invention. In this alternative approach, during the first stage, after filling the features having less than 10 microns width, or preferably less than 5 microns width, the filling of the wider features (>10 microns) with copper to a thickness more than “d” indicated in
As shown in
By employing a fourth ECME or electrochemical mechanical polishing (ECMP) step, process window for copper residue removal may be opened up. In other words after the bulk of the copper is removed by electropolishing, a new solution that forms a protective layer on copper surface may be used for the electrochemical mechanical polishing step. Under applied anodic voltage this protective layer increases polarization at the copper surface, therefore decreasing electro etching or polishing. In areas where a pad makes contact with the copper surface and sweeps the protective layer, the polarization is reduced and etching/polishing is enhanced. Since pad touches the upper-most surface of the copper layer, the top of the copper layer is removed, copper surfaces that are not swept by the pad in the ECME or ECMP step gets protected by the protective layer, avoiding over-etching of these areas, and reducing dishing. These techniques may also be performed subsequent to second stage and using the second solution as an electro etching solution while reversing the polarity of the applied voltage and rendering the workpiece surface more anodic compared to the electrode.
After copper removal from the top surfaces, the barrier layer is also removed by at least one of CMP, Reactive Ion Etching (RIE) and electrochemical mechanical polishing processes to complete the formation of copper structures that are electrically isolated from each other.
Process sequence of the present invention may also be carried out in modules that are placed in separate systems unlike one preferred embodiment shown in
It should be understood, of course, that the foregoing relates to preferred embodiments of the invention and that modifications may be made without departing from the spirit and scope of the invention.