US 20050279641 A1
The present invention relates to methods and apparatus for plating a conductive material on a substrate surface in a highly desirable manner. The invention removes at least one additive adsorbed on the top portion of the workpiece more than at least one additive disposed on a cavity portion, thereby allowing plating of the conductive material take place before the additive fully re-adsorbs onto the top portion and causing greater plating of the cavity portion relative to the top portion.
1. A method of establishing a differential in the concentration of an additive on a surface of a workpiece, comprising:
having the additive adsorbed on the surface of the workpiece; and
applying an external influence to a portion of the surface without applying the external influence to an other portion of the surface,
wherein applying the external influence establishes a differential between a concentration of the additive adsorbed on the portion of the surface and a concentration of the additive adsorbed on the other portion of the surface.
2. The method of
3. The method of
4. The method of
5. The method of
6. The method of
7. The method of
8. The method of
9. The method of
10. The method of
11. The method of
12. The method of
13. The method of
14. The method of
15. The method of
16. The method of
17. The method of
18. The method of
19. The method of
20. A method of establishing a differential in surface concentration of at least one additive adsorbed on a surface of a workpiece, the surface of the workpiece including a top portion and a cavity portion, the method comprising:
placing the surface of the workpiece in a solution;
applying an external influence to the top portion of the surface without applying the external influence to the cavity portion of the surface; and
altering relative concentrations of adsorbed additives to establish a differential between a concentration of the at least one additive adsorbed on the top portion of the surface and a concentration of the at least one additive adsorbed on the cavity portion of the surface.
21. The method of
22. The method of
23. The method of
24. The method of
25. The method of
26. The method of
27. A method of selectively electroplating copper on a cavity portion of a conductive surface comprising a top portion, the method comprising:
having at least one additive species uniformly adsorbed on the cavity portion of the conductive surface and on the top portion of the conductive surface;
at least partially clearing off the at least one additive species adsorbed on the top portion;
delivering an electroplating solution to the conductive surface;
applying power to the conductive surface; and
depositing copper from the electroplating solution so that more copper deposits in the cavity portion than on the top portion.
28. The method of
29. The method of
30. The method of
31. The method of
32. The method of
This application is a continuation of and claims priority to co-pending U.S. patent application Ser. No. 10/358,925, filed Feb. 4, 2003, which is a divisional of U.S. patent application Ser. No. 09/740,701, filed Dec. 18, 2000, now U.S. Pat. No. 6,534,116, which claims priority to U.S. Provisional Patent Application No. 60/224,739, filed Aug. 10, 2000. The disclosures of all of the foregoing patents and applications are hereby incorporated herein by reference in their entireties.
The present invention relates generally to a semiconductor plating method and apparatus. More particularly, the present invention is directed to a method and apparatus that creates a differential between additive adsorbed on a top surface of a workpiece and additive adsorbed within a cavity portion of the workpiece using an external influence to enhance plating of a conductive material in the cavity portion of the workpiece.
There are many steps required in manufacturing multi-level integrated circuits (IC). Such steps include depositing conductive and insulator materials on a semiconductor wafer or substrate followed by full or partly removal of these materials using photo-resist patterning, etching, and the like. After photolithography, patterning and etching steps, the resulting surface is generally non-planar as it contains many cavities or features such as vias, lines, trenches, channels, bond-pads, and the like that come in a wide variety of dimensions and shapes. These features are typically filled with a highly conductive metal material before additional processing steps such as etching and/or chemical mechanical polishing (CMP) is/are performed. Accordingly, a low resistance interconnection structure is formed between the various levels/sections of the IC.
Copper (Cu) is quickly becoming the preferred material for interconnections in ICs because of its low electrical resistively and high resistance to electromigration. Electrodeposition is one of the most popular methods for depositing Cu into the features on the substrate surface.
As can be expected, there are many different designs of Cu plating systems that have been used in this industry. For example, U.S. Pat. No. 5,516,412 issued on May 14, 1996, to Andricacos et al. discloses a vertical paddle plating cell that is designed to electrodeposit a film on a flat article. Next, U.S. Pat. No. 5,985,123 issued on Nov. 16, 1999, to Koon discloses yet another vertical electroplating apparatus, which purports to overcome the non-uniform deposition problems associated with varying substrate sizes. Further, U.S. Pat. No. 5,853,559 issued on Dec. 29, 1998, to Tamaki et al. discloses an electroplating apparatus that minimizes waste of the plating electrolyte and accomplishes high recovery of the electrolyte.
During the Cu electrodeposition process, specially formulated plating solutions or electrolyte are used. These solutions or electrolyte contain ionic species of Cu and additives to control the texture, morphology, and the plating behavior of the deposited material. Additives are needed to make the deposited layers smooth and somewhat shiny.
There are many types of Cu plating solution formulations, some of which are commercially available. One such formulation includes Cu-sulfate (CuSO4) as the copper source (see James Kelly et al., Journal of The Electrochemical Society, Vol. 146, pages 2540-2545, (1999)) and includes water, sulfuric acid (H2SO4), and a small amount of chloride ions. As is well known, other chemicals can be added to the Cu plating solution to achieve desired properties of the deposited material.
The additives in the Cu plating solution can be classified under several categories such as suppressors, levelers, brighteners, grain refiners, wetting agents, stress-reducing agents, accelerators, etc. In many instances, different classifications are often used to describe similar functions of these additives. Today, solutions used in electronic applications, particularly in manufacturing ICs, contain simpler additives consisting of two-component two-ingredient packages (e.g., see Robert Mikkola and Linlin Chen, “Investigation of the Roles of the Additive Components for Second Generation Copper Electroplating Chemistries used for Advanced Interconnect Metallization”, Proceedings of the International Interconnect Technology Conference, pages 117-119, Jun. 5-7, 2000). These formulations are generically known as suppressors and accelerators.
Suppressors are typically polymer formulated from polyethylene glycol-PEG or polypropylene glycol-PPG and is believed to attach themselves to the substrate surface at high current density regions, thereby forming a high resistance film and suppressing the material deposited thereon. Accelerators are typically organic disulfides that enhance Cu deposition on portions of the substrate surface where they are adsorbed. The interplay between these two additives and possibly the chloride ions determines the nature of the Cu deposit.
The following figures are used to more fully describe the conventional electrodeposition method and apparatus.
The Cu material 7 completely fills the via 4 a and is generally uniform in the large trench 4 b, but does not completely fill the trench 4 b because the additives that are used are not operative in large features. For example, it is believed that the bottom up deposition into the via 4 a occurs because the suppressor/inhibitor molecules attach themselves to the top of the via 4 a to suppress the material growth thereabouts. These molecules can not effectively diffuse to the bottom surface of the via 4 a through the narrow opening. Preferential adsorption of the accelerator on the bottom surface of the via 4 a results in faster growth in that region, resulting in bottom-up growth and the Cu deposit profile as shown in
Adsorption characteristics of the suppressor and accelerator additives on the bottom surface of the large trench 4 b is not expected to be any different than the adsorption characteristics on the top surface of the field regions 8 of the substrate. Therefore, the Cu thickness t1 at the bottom surface of the trench 4 b is about the same as the Cu thickness t2 over the field regions 8.
As can be expected, to completely fill the trench 4 b with the Cu material 7, further plating is required.
Thus far, much attention has been focused on the development of Cu plating chemistries and plating techniques that yield bottom-up filling of small features on substrates. This is necessary because, as mentioned above, lack of bottom-up filling can cause defects in the small features. As part of these development efforts, it was discovered that the filling behavior of the small features could be affected not only by the solution chemistry, but also by the type of the power supply used for electrodeposition.
Recent studies suggest that it might be preferable to use pulse or pulse-reverse plating methods to deposit defect free Cu into the small vias (e.g., U.S. Pat. No. 5,972,192 issued to Dubin et al. on Oct. 26, 1999, and Gandikota et al. “Extension of Copper Plating to 0.13 um Nodes by Pulse-Modulated Plating”, Proceedings of the International Interconnect Technology Conference, pages 239-241, Jun. 5-7, 2000). In the pulse-reverse plating process, a cathodic voltage pulse rather than a cathodic DC voltage is applied to the substrate surface. After a short period of plating during the cathodic pulse, the polarity of the voltage is reversed for a brief period causing electrochemical etching from the deposited material. Plating and etching cycles are then repeated until the small features are filled with high quality Cu. A recent study (e.g., C. H. Hsieh et al., “Film Properties and Surface Profile after Gap Fill of Electrochemically Deposited Cu Films by DC and Pulse Reverse Processes”, Proceedings of the International Interconnect Technology Conference, pages 182-184, Jun. 5-7, 2000), shows that the filling of the vias is controlled mainly by the additive diffusion when the DC process is used, whereas it is mainly controlled by additive adsorption when a pulse-reverse process is used.
As described above, the attention in the semiconductor industry has mainly been concentrated on filling the various features on semiconductor wafers with Cu. Both DC and pulsed power supplies have been used in the deposition of these Cu films. Filling properties of Cu into small features was found to be a strong function of the type of the power supply used. Although the exact roles of the plating solution additives and their interaction with the applied voltage waveforms are not well understood, it is clear that the kinetics of the additive adsorption and diffusion processes influence the way metals deposit on non-planar substrate surfaces.
As mentioned above, special bath formulations and pulse plating processes have been developed to obtain bottom-up filling of the small features. However, these techniques have not been found effective in filling the large features. In large features, the additives can freely diffuse in and out of them. The use of standard pulse plating techniques in conjunction with the commonly used additive systems containing chloride ions, accelerators and suppressors/inhibitors do not yield accelerated growth from the bottom surface of the features where the width of the feature is considerably larger than its depth. The growth of Cu in such features is conformal and the film thickness deposited on the bottom surface of the large features is approximately the same as that deposited on the field regions.
Methods and apparatus to achieve accelerated bottom-up plating in small as well as large features on a substrate would be invaluable in terms of process efficiency and cost since such a process would yield a Cu deposit that is generally planar as illustrated in
Others have previously recognized attractive features of a plated Cu structure such as the one shown in
In the co-pending U.S. application Ser. No. 09/201,928, entitled “Method and apparatus for electrochemical mechanical deposition”, commonly owned by the assignee of the present invention, a technique is disclosed that achieves deposition of the conductive material into the cavities on the substrate surface while minimizing deposition on the field regions by polishing the field regions with a pad as the conductive material is deposited. The plating electrolyte in this application is supplied to the small gap between the pad and the substrate surface through a porous pad or through asparities in the pad.
The anode assembly 19 can also be rotated around a second axis 10 c at controlled speeds in both the clockwise and counter-clockwise directions. It is also understood that axes 10 b and 10 c are substantially parallel to each other. The gap between the wafer 16 and the pad 18 is adjustable by moving the carrier head 10 in the z direction.
When the wafer 16 surface and the pad 18 are in contact, the pressure that is exerted on the two surfaces can also be adjusted. The co-pending U.S. application Ser. No. 09/511,278, entitled “Pad Designs and Structures for a Versatile Materials Processing Apparatus”, filed Feb. 23, 2000, describes various shapes and forms of the holes in the pad 8, through which the electrolyte flows to the wafer surface.
During operation, a potential is applied between the electrical lead 17 to the wafer 16 and the electrical lead 24 to the anode assembly 19, making the wafer 16 surface more negative than the anode assembly 19. The electrolyte 20 can be introduced to the pad 18 from a reservoir (not shown) located in proximity to the anode assembly 19. The anode assembly 19 can have an in-channel and holes that are made therein, which together provide a path for the electrolyte 20 to be fed to the gap between the pad 18 and the wafer 16.
Under applied potential, Cu plates out of the electrolyte 20 onto the wafer 16 surface. The moving pad 18 that is pushed against the wafer 16 surface at a controlled pressure minimizes accumulation of Cu over certain portions of the wafer 16 surface by polishing the same.
The pad 18 is preferably nonconductive, hard, porous, or perforated type material so that an electric field can pass through it, while preventing shorting between the anode assembly 19 and the cathode wafer 16. The spacing or gap between the pad 18 and the cathode wafer 16 may range from less than 1 micron up to 2 millimeter. The diameter or cross sectional length of the pad 18 and the wafer 16 may range from about 5 millimeter to over 300 millimeter. The larger the wafer 16 diameter, the larger the pad 18 diameter.
It is an object of the present invention to provide a method and apparatus that plates a conductive material on a substrate surface in a highly desirable manner.
It is another object of the present invention to provide a method and apparatus that plates a conductive material in both small and large features of a substrate surface with greater efficiency, cost-savings, and superior quality than prior art methods and apparatus.
It is a further object of the present invention to provide a method and apparatus that plates a conductive material in small and large features on a substrate surface using a mask having one or more opening therein.
It is yet another object of the present invention to provide a method and apparatus that plates a conductive material in small and large features while electrical power is locally pulsed on the substrate surface due to the movement of the opening(s) in the mask with respect to the substrate surface.
It is a further object of the present invention to provide a method and apparatus that removes additive that has been previously adsorbed onto the top surface portion of a workpiece in order to enhance the plating of a conductive material on a cavity feature surface portion of the workpiece that did not have previously adsorbed additive removed.
It is yet another object of the present invention to create a differential between additive adsorbed on a top surface of a workpiece and additive adsorbed within a cavity portion of the workpiece to enhance plating of a conductive material in the cavity portion of the workpiece.
The above objects of the invention, among others, taken alone or in combination, are achieved by the present invention, which provides an apparatus for, and a method of, plating a conductive material on the surface of a workpiece.
In one aspect of the method, an electrolyte solution with at least one additive disposed therein is applied over the workpiece, such that the additive becomes adsorbed onto the top portion and the cavity portion of the workpiece. An external influence is applied so that the additive adsorbed onto the top surface is removed, and plating of the conductive material take place before the additive fully re-adsorbs onto the top portion, thereby causing greater plating of the cavity portion relative to the top portion.
In another aspect of the method, an electrolyte solution with at least one additive disposed therein is applied over the workpiece, such that the additive becomes adsorbed onto the top portion and the cavity portion of the workpiece. An external influence is applied so that a differential in the amount of additive that is adsorbed on the top portion relative to the cavity portion is achieved. Plating takes place while the differential still exists, thereby causing greater plating of the cavity portion relative to the top portion.
In one aspect of the apparatus, a mask disposed between an anode and the workpiece and is movable with respect to the workpiece physically sweeps a top portion of the workpiece, thereby reducing additive adsorbed thereon, while additive adsorbed on a cavity portion remains. An anode that assists in creating an electric field between it and the workpiece is used to promote plating of the conductor within the electrolyte that is disposed over the workpiece.
In another aspect of the apparatus, the mask contains an open area that is used to help define where the electric field will exist, thus allowing greater control over where plating will occur on the workpiece.
These and other objects and advantages of the present invention will become apparent and more readily appreciated from the following detailed description of the presently preferred exemplary embodiments of the invention taken in conjunction with the accompanying drawings, of which:
The preferred embodiments of the present invention will now be described with reference to the following figures. The inventors of the present invention have found that by mask-pulse plating the conductive material on the substrate surface, a more desirable and high quality conductive material can be deposited in the various features therein.
The present invention can be used with any substrate such as a semiconductor wafer, flat panel, magnetic film head, packaging substrate, and the like. Further, specific processing parameters such as time, pressure, mask designs, and the like are provided herein, which specific parameters are intended to be explanatory rather than limiting.
The plating method described herein is called “mask-pulsed” plating. The present invention describes a method and apparatus for mask-pulse plating the conductive material onto the substrate by intermittently moving the mask to make contact with the substrate surface and applying power between an anode and the substrate, the mask being positioned in between the anode and the substrate. Furthermore, the present invention is directed to novel plating method and apparatus that provide enhanced electrodeposition of conductive materials into the various features on the substrate surface.
Referring back to
As can be seen from the above example, a DC power supply can be used for this plating technique. By moving the solid insulating mask 40 that makes physical contact with the wafer 32, any section on the wafer surface can be suddenly and briefly exposed to the electrolyte and to the applied plating current. This is quite different from the prior art techniques defined above. For instance, in the present invention, certain sections of the wafer surface are substantially free from the electrolyte. The electrolyte is applied to a section of the wafer only when that section is exposed to the electrolyte and a pulse of current is simultaneously applied.
If the current mask-pulsed plating method is used with simple metal deposition electrolytes with no additives (i.e., inhibitors and accelerators), it would not be expected to be much different than conventional plating. This is because the size of the openings 42 in the mask 40 is much larger than the feature size on the wafer 32 surface. Therefore, when a section is exposed through the opening 42 to the electrolyte, regular plating would commence. However, if additives are added that influence polarization, then the mask-pulsed plating method can offer advantages that are not existent in conventional pulsed plating techniques.
For example, consider a Cu plating bath containing conventional solutions/chemicals (Cu sulfate, water, sulfuric acid and chloride ions) and an additive A. The additive A enhances deposition when it is adsorbed on the wafer surface. When this electrolyte is used in a conventional plating cell such as the one depicted in
If, however, the mask-pulsed plating technique is used with the same electrolyte, the mask would clear away the additive A from the field regions since it makes physical contact with these regions. Both the small and large features, however, will still contain the adsorbed additive A since these features are not in direct physical contact with the mask. When a section of the wafer is suddenly exposed to the electrolyte, the bottom and side surfaces of the features with the previously adsorbed additive A would immediately start plating at a higher rate than the field regions. If the time period At is less than the adsorption period required for the additive A to attach itself to the substrate surface, the applied plating current preferentially flows through the features to be filled, thereby yielding an enhanced deposition rate within the features in relation to the deposition rate on the field regions.
The mask-pulsed plating method of the present invention utilizes the differences between response times of various additives to achieve enhanced plating into the various features of the substrate surface. The mechanism involves “sweeping” of the top surface of the substrate (field regions) by the mask, which does not make physical contact with the regions inside the features. The sweeping on the field regions establishes a differential between the concentration of the adsorbed species in those regions that are swept away and the regions that are within the features. When the surface is then suddenly exposed to the electrolyte and the electric field, the features with the adsorbed species attracts most of the plating current from the field regions.
This present method works equally well using multiple additives. For example, if the plating solution contains an inhibitor B and an accelerator C with the adsorption kinetics of the inhibitor being much faster than that of accelerator, the following mechanism can be used by the mask-pulsed plating method. Both the inhibitor B and the accelerator C would be partially or wholly swept off the field regions of the substrate by the mask. However, both species would still be present in the features. When the substrate is exposed to the electrolyte and the electric field, the inhibitor B would readily adsorb onto the field regions introducing a high resistance path for the plating current. The accelerator C, which is already present within the features, compensates for the action of the inhibitor in those regions and the current can easily flow through these features. Therefore, until the accelerator C is properly adsorbed onto the field regions, the film growth rate within the features will be higher.
This same result can also be expected from yet another chemistry where an inhibitor D has the property of strong adsorption and the accelerator E is weakly bonded to the field regions. In this case, the mask can readily remove the weakly bonded accelerator E from the field regions whereas, the accelerator E remains attached to the surfaces within the features. Upon exposure to the electrolyte and electric field, the plating current flows through the features preferentially until the accelerator E begins to get adsorbed again onto the field regions.
It should be noted that the above descriptions are just some examples of the mechanisms involved in the present invention and are not meant to be limiting. The present invention utilizes differences between adsorption/de-sorption kinetic of various electrolyte additives. The present invention accomplishes this by applying a solution and power suddenly and simultaneously to a specific section of the substrate surface that has been previously cleared off, partially or wholly, of one or more of the additive species.
The geometry of the plating system shown in
The power supply used in the present invention may be pulsed or DC power supply, but preferably it is a DC power supply. The power supply can be used in the current controlled or voltage controlled mode, i.e., it either keeps the applied current constant or applied voltage constant. For the case of using a current controlled mode, it is important that the size of the opening(s) in the mask be large enough to cover portions of the field regions as well as portions of the features simultaneously. In other words, when the wafer surface is exposed to the electrolyte through the opening(s), there should not be just the field regions that are exposed to the electrolyte at any given time. For example, if the opening is very small or the number of features on the wafer surface is low (low density features), the field regions are exposed to the electrolyte. In this case, since the power supply pushes through a fixed current, all the current would flow through the field regions and the Cu will be plated on the field region without discrimination. But if both field regions and features are exposed simultaneously, then the current would preferentially flow through the features and more Cu would be plated into the features and less on the field regions. This situation can be assured by increasing the number of openings in the mask so that there are always portions of the both regions (field and feature) exposed through some of the holes simultaneously.
If a constant voltage power supply is used, then the current automatically adjusts itself depending upon the resistance on the wafer surface. Therefore, if the mask hole exposes only the field regions of the wafer, less current is supplied to that surface and the plating amount is smaller. When features are exposed to the solution, more current flows into the feature and thus preferential plating takes place into the features. Therefore, it is more appropriate to use voltage controlled mode of the power supply if wafers with low feature density is coated and/or the number of holes in the mask is limited.
This invention can be used to fill both small and large features. However, a serial process can also be utilized. In that approach, there are two processing steps. During the first step the mask is pulled away from the wafer surface allowing substantial amount of plating solution between the mask and the wafer surface. In this position, the system acts just like a traditional plating cell. With the help of the additives in the plating solution, the small features are filled during this step and the situation as shown in
Along with using copper and its alloys as the conductive material, other conductive materials such as copper alloys, iron, nickel, chromium, indium, lead, tin, lead-tin alloys, nonleaded solderable alloys, silver, zinc, cadmium, ruthenium, their respective alloys may be used in the present invention. The present invention is especially suited for the fabrication of high performance and highly reliable chip interconnect, packaging, magnetic, flat panel and opto-electronic applications.
In the previous descriptions, numerous specific details are set forth, such as specific materials, mask designs, pressures, chemicals, processes, etc., to provide a thorough understanding of the present invention. However, as one having ordinary skill in the art would recognize, the present invention can be practiced without resorting to the details specifically set forth.
Although various preferred embodiments have been described in detail above, those skilled in the art will readily appreciate that many modifications of the exemplary embodiment are possible without materially departing from the novel teachings and advantages of this invention.