Reduction of defects in conductive layers during electroplating

Abstract

The present invention is a method and system to reduce defects in conductive surfaces during electrochemical processes. The system includes a first power supply and a second power supply. The first powers supply is configured to supply a first power between a conductive surface of a workpiece and an electrode of the system. The second power supply is configured to supply a second power between the conductive surface and the electrode when a switching unit switches from the first power from the first power supply to the second power from the second power supply in response to the conductive surface contacting the process solution.

Description

FIELD

The present invention relates to manufacture of semiconductor integrated circuits and, more particularly to a method for electrochemical deposition of conductive layers.

BACKGROUND

Conventional semiconductor devices such as integrated circuits generally include a semiconductor substrate, such as a silicon substrate, and a plurality of sequentially formed dielectric interlayers 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. Interconnects are usually formed by filling copper by a metallization process, into features or cavities etched into the dielectric layers. The preferred method of copper metallization is electrodeposition or 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.

In a typical process, first an insulating layer is formed on the semiconductor substrate. Patterning and etching processes are performed 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. FIG. 1 exemplifies a surface portion of a semiconductor substrate 10 or a wafer having features such as cavities 12, 13 and 14. The cavities are formed in a dielectric layer 16, which is deposited on the substrate 10. Before the electroplating step, cavities 12-14 and the top surface of the dielectric layer 16 are coated with a barrier layer 18 and a seed layer 20. During the copper electrodeposition process, specially formulated plating solutions or electrolytes are used to plate copper onto the seed layer. An exemplary electrolyte contains 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. FIG. 2 shows a simplified schematic of a typical electrodeposition system 50 for processing the wafer 10 in an electroplating solution 52 contained in a chamber 53. The wafer 10 is held by a carrier head 54 so that the front surface 56, which is lined with the seed layer 20 (FIG. 1), is exposed to the electroplating solution 52. During the process, a potential difference is applied between the front surface 56 and an anode 58 by a power supply 60 and material deposition onto the front surface 56 from the solution 52 is achieved.

It is a known fact that process solutions may chemically interact with seed layers at the beginning of an electrochemical process. Thin copper seed layers, for example, are chemically attacked and may be damaged by the process solutions when the work piece is first introduced into the process solution. This is especially a serious problem for wafers with narrow and deep features. In such substrates, the seed layer thickness may be extremely thin especially deep in the narrow features. For example, for 0.15 micrometer wide, 1.0 micrometer deep, the seed layer thickness may be only 20-50 A on the lower portion of the sidewalls of the via, whereas the seed layer thickness at the top surface of the dielectric may be 800 A or more. Thickness of the seed layers and their profiles within the features of the wafers are strong functions of the seed layer deposition equipment and process.

It should be appreciated that electrodeposition solutions, especially those with acidic pH has certain degree of etching rate for the material to be deposited. For example, depending upon the exact formulation, sulfuric acid based copper deposition electrolytes may have a copper etch rate of 5-200 A/min. Therefore, thin seed layers within the features on a wafer may get chemically attacked within a very short period once the wafer surface is wetted by the solution. This period, in some cases, maybe in the order of milliseconds, especially if the seed layer is very thin and it contains oxides which easily dissolve in the solutions used. Etching rate of copper oxide is much higher than etching rate of pure copper in acidic electrolytes.

Hot entry is one way of avoiding this unwanted interaction between the process solution and the seed layer, when wafers with thin or weak seed layers are immersed into the process solutions for electroplating. During hot-entry, a voltage is applied to the seed layer before it is wetted by the process solution. This cathodic voltage protects the seed layer against chemical dissolution and material deposition starts immediately onto the seed layer. However, hot entry has some drawbacks, such as formation of hot spots, which are high current density spots and therefore high deposition locations on the wafer where the solution makes the initial physical contact with the seed layer.

FIGS. 3A-3B exemplify various stages of formation of the hot spots on the seed layer 20 and the effects of hot spots on the plated layer. FIG. 3A illustrates an instant of initial contact between the process solution 52 and the seed layer 20 on the wafer 10 while a plating voltage is applied to the seed layer 20 through the power supply 60. As the wafer 10 is lowered onto the solution, seed layer 20 first may contact ripples 62 on surface of the process solution 52. These ripples may be due to various sources. Vibrations of the various system components or simply movement of the process solution during the process may generate such ripples or small waves. It should be noted that ripples represent specific locations where the solution first makes physical contact with the wafer surface. These locations may not necessarily be due to waves or ripples. For example, in tool designs where the wafer enters the solution at an angle, only one small portion of the substrate surface first touches the solution. The hot-spot problem that we are about to describe takes place at that location in that case.

Referring back to FIG. 3A, as the tips of the ripples touch the seed layer, since a voltage has already been applied between the seed layer and an anode (not shown), current flows from the anode, through the process solution 52 and to the seed layer 20 only through the contact spots 64, depositing copper in the process, preferentially and instantaneously onto the seed layer locations defined by the contact spots 64. Contact spots 64 represent a very small area fraction compared to the total area of the wafer surface. Therefore, during this initial plating the current density at the contact spots 64 is high and as shown in FIG. 3B, it causes almost instantaneous formation of individual copper growths 66 at the location of contact spots 64. These copper growths 66 are also called hot spots. From this point on, if the plating process is continued, a copper layer 68 with a non-uniform thickness is formed on the seed layer, as shown in FIG. 3C. Due to the growths 66, thickness of the layer 68 on the hot spots 64 is thicker than the rest of the layer, which is an unwanted situation in manufacture of interconnects. It should be noted that the sketches of FIGS. 3A through 3C are not drawn to scale. The depth of the features may actually be smaller than the height of the ripples. Therefore, hot spots may form not only on the top surface of the dielectric but also within the features causing defects in the features. Furthermore, the size of the hot spots may change from sub-micron to several millimeters.

To address the problem described above, some prior art methods use cold entry, i.e. entry of the substrate into the solution with no applied voltage and then apply the plating voltage. However, as discussed previously, upon cold entry, thin, oxidized or weak seed layers may get chemically attacked by the process solution within a time period of one second or less unless there is an applied cathodic voltage to protect them.

To this end there is a need for plating methods that provide uniform deposition layers without defects even on substrates with weak seed layers.

SUMMARY OF THE INVENTION

The present invention is a method and apparatus to reduce defects in conductive layers during electrochemical material deposition or electrochemical material removal.

The process of the present invention uses multiple power supplies and multiple process voltages or currents to avoid formation of defects on seed layers and at the same time allow defect-free deposition of a conductor, such as copper, on wafers. In one embodiment, a first power from a first power supply is provided to the seed layer prior to contacting the seed layer to the surface of the electroplating solution. Upon contacting the solution, switching from the first power from the first power supply to a second power from a second power supply automatically takes place.

According to an aspect of the present invention, a system for electroprocessing a conductive surface on a workpiece using a process solution and an electrode while holding the workpiece with a workpiece carrier is disclosed. The system comprises a first power supply configured to supply a first power between the conductive surface and the electrode, a second power supply configured to supply a second power between the conductive surface and the electrode, and a switching unit for switching the first power to the second power in response to the conductive layer contacting the process solution.

According to another aspect of the present invention, method of electroprocessing a conductive surface on a workpiece is provided. The electroprocessing uses a process solution and an electrode wetted by the process solution. The method includes the steps of applying a first power between the conductive surface and the electrode using a first power supply, contacting the conductive surface to the process solution, and applying a second power between the surface and the electrode using a second power supply.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a schematic cross sectional view of a portion of a semiconductor substrate including features and surface of the substrate coated with a conductive layer;

FIG. 2 is a schematic side view of a conventional electrochemical deposition system;

FIGS. 3A-3C are schematic cross-sectional views showing various stages of the formation of the hot spot defects on the conductive layer of the substrate shown in Figure land the effects of the defects on the electroplated layer;

FIG. 4 is a schematic view of the system of the present invention including at least two power supplies; and

FIGS. 5A-5D are schematic cross sectional views showing various stages of a process of an embodiment of the present invention on a semiconductor substrate.

DETAILED DESCRIPTION OF THE INVENTION

The process of the present invention uses multiple power supplies and multiple process voltages or currents to avoid formation of hot spots on the seed layer and at the same time allow defect-free deposition of a conductor, such as copper, on wafers lined with thin seed layers, such as seed layers that are thinner than 30 nm. In one embodiment, a first power such as a contact voltage or current from a first power supply is provided to the seed layer prior to contacting the seed layer to the surface of the electroplating solution. Upon contacting the solution, switching from the contact voltage or current from the first power supply to a second power such as an electroplating voltage or current from a second power supply is automatically performed and is applied to the seed layer. As will be explained more fully below, the contact current is significantly lower than the electroplating current. As a result, as the physical contact is established between the certain spots of the surface and the waves or ripples of the solution, current density at these spots is not high enough to form high-rate deposition or hot spots. In this embodiment, action of switching from the contact current to the electroplating current begins as soon as the seed layer touches the surface of the electroplating solution. Electroplating begins when a full contact between the wafer surface and the solution is established. The invention has the capability to switch from one power supply to the other within 200 milliseconds (ms) or earlier, avoiding formation of hot spots and at the same time preventing chemical dissolution of weak seed layers.

FIG. 4 shows an exemplary system 100 to perform the process of the present invention. The system 100 includes process chamber 102 to contain the process solution 104. Wafer 106 is held by a wafer carrier 108 and rotated. Wafer may additionally be moved vertically and laterally. Front surface 110 of the wafer 106 includes a seed layer, which will be described below. In this embodiment, the surface 110 and anode 111 of the system 100 are configured to be connected to two power supplies, namely a first power supply (FPS) 112 and a second power supply (SPS) 113. The power supplies 112, 113 may preferably be connected to the anode and the surface through a switching unit 114, which allows sequential use of the power supplies. The switching unit may include power switches S₁, S₂, S₃ and S₄. Switch S1 connects the negative terminal of the first power supply 112 to the surface 110 of the wafer 106 when the switch S1 is in closed position. Switch S2 connects the positive terminal of the first power supply 112 to the anode 111 when the switch S2 is in closed position. FIG. 4 shows switches S1 and S2 in closed position (and switches S3 and S4 are in open position) so that surface 110 and the anode 111 are connected to and energized by the first power supply 112.

Switch S3 connects the negative terminal of the second power supply 113 to the surface 110 of the wafer 106 when the switch S3 is in closed position. Similarly, switch S4 connects the positive terminal of the second power supply 113 to the anode 111 when the switch S4 is in closed position. When switches S3 and S4 are in closed position and the switches S1 and S2 are in open position, plating current is connected to and energizes the surface 110 of the wafer and the anode. Power switches S₁-S₄ may be made of solid-state relays and associated circuitry. The system of the present invention may include multiple power supplies and corresponding multiple switch pairs to perform the present invention using multiple powers.

The first power supply 112 includes a monitoring terminal 115 to monitor activity of the first power supply 112. When power supply provides current for the system, the monitoring terminal, in response, generates a signal output. In one embodiment, the signal output of the monitoring terminal 115 is received by a detector 116, preferably an analog detector. A control signal from the analog detector 116 to the switching unit 114 controls the switches S1-S4.

In a sequential use of the power supplies, at a first stage of the process, the first power supply 112 is set to provide a first current. At this time, the switches S1 and S2 are in closed position and the switches S3 and S4 are in open position, and there is no current passing between the surface of the wafer and the anode until a physical contact between the surface and the solution is established. As soon as the physical contact is established between the surface and the solution, the initial small current dictated by FPS 112 flows from the solution to the seed layer. Current flow or sensing the current flow causes an output signal (contact signal) from the monitoring terminal to the analog detector 116. The analog detector sends a control signal (command signal) to the switches S1-S2 of the switching unit 114. Upon receipt of the control signal, the switches S1 and S2 are brought into open position while the switches S3 and S4 are brought into closed position and, thereby allowing a second current from the second power supply 113 to be applied between the front surface of the wafer and the anode. It is understood that, power supplies used in the invention may be on and ready to be switched to the connecting process circuitry. Power is supplied from one or the other by using the switching unit.

In this embodiment, the first current is denoted as contact current and the second current is denoted as electroplating current. It is understood that, in this embodiment, the contact current is significantly lower than the electroplating current and therefore, prevents formation of the hot spots when the wafer surface first touches the solution at certain locations. The contact current may vary depending on the chemistry and the acidity of the process solution. For example, for a low acid chemistry from Enthone, the contact current for a 300 mm diameter wafer may be in the range of 0.1-1.0 A. The electroplating current on the other hand may be 5 A or higher.

FIGS. 5A-5D exemplify stages of the process of the present invention using an exemplary portion of the surface 110 of the wafer 106. The surface 110 of the wafer may include various features, such as vias 120 and trenches 121 formed in a dielectric layer 122. Features and surface of the dielectric layer is coated with a barrier layer 124 and a copper seed layer 126. An electrical contact 128 connects the seed layer 126 to the switches S1 and S3, which are in turn connected to the negative terminals of the first and second power supplies.

As shown in FIG. 5B, as the wafer 106 is lowered onto surface 130 of the process solution 104, the seed layer is connected to the first power supply 112 through switch S1, and the FPS is programmed to apply the contact current. Height of ripples 132 on the surface 130 of the solution 104 may be less than 2 mm. The ripple height may be defined as the distance between surface level of the solution 104 and tip 134 of the ripples 132.

As shown in FIG. 5B, as the wafer 106 is lowered onto the solution with z motion of the carrier head 108, and at one instant, the tips 134 of the ripples 132 touch the seed layer at contact locations 136. This causes low contact current to flow to the contact locations 136 from the process solution 104. As described above, this action generates a signal output from the monitoring terminal for analog detector 116. As described above, upon receipt of the signal, the analog detector 116 controlling the series of switches, switches the connection to the first power supply 112 off and switches the connection to the second power supply 113 on, thus initiating electroplating of the copper onto the seed layer at the plating current density provided by the second power supply. For best results, the time of switching needs to be at least in the range of the travel time of the wafer surface for the ripple height, i.e., the time spent between the initial contact of the tip of the ripples with the seed layer and the time when surface is fully wetted by the solution. The critical importance of the switching time is that the high plating current should not be switched on before the ripples totally disappear from the wafer surface. In other words, contact between the solution and the seed layer should be full rather than local when the high current is switched on. In the present invention, using two power supplies and the z-motion of the carrier head, allows switching from low current to high current conditions in a very short time such as less than 200 milliseconds, preferably less than 100 ms, while preventing problems on the seed layer.

As shown in FIG. 5C, as the wafer is fully submerged into the solution 104, the second power supply 113 is connected between its surface and an anode. As shown in FIG. 5D, as the plating current is applied from the second power supply 113, a copper layer is uniformly plated on the seed layer 126.

In one exemplary process sequence for a 200 mm diameter wafer, the first power supply is set to a small current value of between 0.05 to 0.2 amps. Using the carrier head, the wafer is brought down onto the solution with a speed of 20-40 mm/sec. As soon as the wafer touches the process solution, monitoring terminal output is received by an analog detector having a sampling rate of 1 ms. The analog detector sends a signal to a circuit of solid state relays to switch the anode and wafer connections from the first power supply (contact current) to the second power supply (electroplating current). As solid state relays are very fast, this switching action occurs very fast in a time period of 5-100 ms.

Although various preferred embodiments and the best mode 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.

Claims

1 A method of electroprocessing a conductive surface on a workpiece using a process solution and an electrode wetted by the process solution, the method comprising the steps of:

applying a first power between the conductive surface and the electrode using a first power supply;

contacting the conductive surface to the process solution; and

applying a second power between the surface and the electrode using a second power supply.

2. The method of claim 1 further comprising the step of sensing the contacting the conductive surface and generating a control signal in response to contacting the surface.

3. The method of claim 1, wherein subsequent to the step of contacting, switching from the first power from the first power supply to the second power from the second power supply occurs.

4. The method of claim 3, wherein switching from the first power to the second power supply occurs within a predetermined switching period.

5. The method of claim 4, wherein the predetermined switching period is less than 100 milli seconds.

6. The method of claim 3, wherein the switching begins at a partial contact time, when a partial contact is established between the conductive surface and the solution, and terminates at a full contact time when a full contact is established between the conductive surface and the solution.

7. The method of claim 1 further comprising the step of immersing the conductive surface into the process solution while continuing to apply the second power from the second power supply.

8. The method of claim 1, wherein the first power from the first power supply is a contact power and a second power from the second power supply is a plating power.

9. The method of claim 7, wherein the magnitude of the plating power is higher than the magnitude of the contact power.

10. A semiconductor device manufactured using the method of claim 1.

11. A system for electroprocessing a conductive surface on a workpiece using a process solution and an electrode while holding the workpiece with a workpiece carrier, the system comprising:

a first power supply configured to supply a first power between the conductive surface and the electrode;

a second power supply configured to supply a second power between the conductive layer and the electrode; and

a switching unit configured to switch between the first power and the second power in response to the conductive surface contacting the process solution.

12. The system of claim 11 further comprising a detector coupled to the switching unit configured to receive a contact signal and wherein the first power supply includes a current monitor configured to generate the contact signal in response to the conductive layer contacting the process solution.

13. The system of claim 12, wherein the detector in response to the contact signal generates a command signal to switch from the first power to the second power.

14. The system of claim 11 wherein the switching unit includes solid state relays to switch from the first power to the second power.

15. The system of claim 12, wherein the detector is an analog detector.