US20050279964A1

US20050279964A1 - Chemical mechanical polishing slurry for polishing copper layer on a wafer

Info

Publication number: US20050279964A1
Application number: US10/868,969
Authority: US
Inventors: Ming-Tseh Tsay; Zong-Whie Shih; Wen-Chueh Pan; Yang-Jiann Fann
Original assignee: National Chung Shan Institute of Science and Technology NCSIST
Current assignee: National Chung Shan Institute of Science and Technology NCSIST
Priority date: 2004-06-17
Filing date: 2004-06-17
Publication date: 2005-12-22

Abstract

A chemical mechanical polishing slurry is provided for the copper layer on a wafer. The slurry contains colloidal silica and a chemical etching agent composed of hydrogen peroxide, acetic acid, and phthalic acid. The hydrogen peroxide oxides the surface of the copper layer. The acetic acid then reacts with the copper oxide to form copper acetate. This selective and functional chemical reaction mechanism can speed up the polishing removal rate and reduce scratches. The phthalic acid functions as both a pH buffering agent and a complexing agent to make the reaction concentration at each point of the wafer surface more homogeneous. Therefore, the copper layer during the chemical mechanical polishing process has a high removal rate and uniformity.

Description

BACKGROUND OF THE INVENTION

1. Field of Invention
The invention relates to a chemical mechanical polishing (CMP) slurry and, in particular, to a slurry with a special chemical etching agent and colloidal silica for the CMP process of the copper layer on a wafer.
2. Related Art
As IC devices become smaller in size, the metal wires on chips are also miniaturized. However, smaller wire widths result in larger resistance and narrower separations result in larger capacitance. As the sizes of devices continue to shrink, there will be more signal delays. Therefore, copper wires with better conductivity have replaced conventional aluminum wires in order to increase the signal transmission speed.
For a high-density multiple-layer copper wire production process, one can employ chemical mechanical polishing (CMP) to satisfy the requirement of a uniform wafer. The basic principle is to use a chemical etching agent in the slurry to react with the wafer surface. Along with the abrasive mechanical actions of polishing particles, the dielectric layers, barriers, or metal layers on the wafer are removed in order to achieve global planarity of the wafer. In the abrasive mechanism, the chemical and mechanical forces have to form an optimized dynamical interaction in order to reach high-quality polishing results. Simply using chemical reactions cannot achieve the global planarity. Likewise, considering only the mechanical actions will reduce the removal rate and are likely to scratch the wafer surface. To have appropriate material removal energy, it is desirable to have a slurry that can reduce the structural strength on the surface of the material through chemical reactions. This can reduce the stress (including normal forces and shear stress) required to take away polishing particles in the material polishing and removal process. It can further achieve the goals of no scratch, high removal rate, and global planarity.
The CMP slurry used in wafers is comprised of a slurry (chemical etching agent) and polishing particles. The slurry is responsible for chemical etching, while the polishing particles are responsible for mechanical polishing. Tests of the polishing slurry include the polishing removal rate, uniformity, surface scratches, and stability. Moreover, the slurry has to be cheap and safe for industrial uses.
As a polishing slurry for the copper layer, most of the polishing particles in the prior art are fumed alumina which have a high rigidity and irregular shapes (as in the U.S. Pat. Nos. 6,217,416 and 6,432,828) or fumed silica (as in the U.S. Pat. No. 6,309,560). This type of polishing particles synthesized by burning is expensive. Since they are hard to become dispersed, one usually has to add a separating agent. Although they have a higher removal rate in polishing, scratches are often left on the wafer surface.
The polishing removal rate of colloidal silica is not as good as fumed alumina or fumed silica; nonetheless, the particles have a uniform size, a better separating property, and a cheaper price. They have a spherical shape and a lower rigidity. Therefore, scratches can be avoided. The only problem in the prior art is that no appropriate chemical etching agent has been found for the colloidal silica.

SUMMARY OF THE INVENTION

In view of the foregoing, an objective of the invention is to provide a low-cost and highly efficient polishing slurry for the chemical mechanical polishing (CMP) of the copper layer on a wafer.
To achieve the above objective, the disclosed CMP slurry comprises colloidal silica and a chemical etching agent composed of hydrogen peroxide, acetic acid, and phthalic acid. The chemical mechanical removal mechanism is to have the hydrogen peroxide and the copper layer surface react to form copper oxide. The acetic acid then reacts with the copper oxide to form copper acetate to be readily removed (the acetic acid does not react with copper). Afterwards, the colloidal silica polishing particles remove the copper acetate from the wafer surface. The phthalic acid functions as both a pH buffering agent and a complexing agent of the polishing slurry, making the reaction concentration at each point of the wafer surface more homogeneous. Moreover, the required concentration of the chemical etching agent according to the invention is very low. The cost can thus be greatly reduced.

BRIEF DESCRIPTION OF THE DRAWINGS

The invention will become more fully understood from the detailed description given hereinbelow illustration only, and thus are not limitative of the present invention, and wherein:

- Table 1 shows the solid content and pH value of the colloidal silica in various particle diameters;
- Table 2 shows the effects of polishing particles of various abrasive diameters on the removal rate and the non-uniformity;
- Table 3 shows the effects of different solid contents of the colloidal silica on the removal rate and the non-uniformity;
- Table 4 shows the effects of slurries with the same composition but different pH values on the removal rate and the non-uniformity; and
- Table 5 shows the effects of the composition and concentration of the chemical etching agent in the slurry on the removal rate and the non-uniformity.

DETAILED DESCRIPTION OF THE INVENTION

According to a preferred embodiment of the invention, the chemical mechanical polishing (CMP) slurry contains:

- (1) nano-scale colloidal silica with a concentration of 0.5˜10 wt %, the average particle diameter being 10˜100 nm;
- (2) hydrogen peroxide with a concentration of 0.6˜2.5 V %;
- (3) acetic acid with a concentration of 0.1˜1 V %; and
- (4) phthalic acid with a concentration of 0.1˜0.8 wt %.

Moreover, the concentrations of hydrogen peroxide, acetic acid, and phthalic acid satisfy (hydrogen peroxide)/(acetic acid)=0.6˜25 and (acetic acid)/(phthalic acid)=0.125˜10. The pH value of the polishing slurry is 2.5˜5.5. The above configuration can achieve better polishing efficiency.
The polishing machine used in subsequent analyses is the CMP-300P machine made by Chung-Sun Institute of Science & Technology. We also use an 8-inch floating polishing head. Other details are as follows. The down force is 3 psi. The platen speed is 47 rpm. The head speed is 49 rpm. The slurry supply rate is 150 ml/min. The polishing pad is Rodel IC-1400. The polishing moisture time is 30 sec. The removal rate and uniformity are measured using NAPSON RT-80/RG-120 surface resistance meter.
Test 1
The current test evaluates the effects of colloidal silica particle sizes on polishing results. The solid content of colloidal silica in various sizes and their pH values are indicated in Table 1. The concentrations of the hydrogen peroxide, acetic acid, and phthalic acid are fixed to 1.2 V %, 0.375 V %, and 0.2 wt %, respectively. The concentration of the polishing particles is 3.0 wt %. The pH values and polishing results of the slurries are given in Table 2. The experiments show that the disclosed slurry has a high removal rate and achieves a high uniformity. The removal rate is maximized when the particle diameter is 35.7 nm; however, its uniformity is not as good as those with an abrasive diameter of 19.7 nm and 90.5 nm.

TABLE 1

Abrasive diameter (nm) Solid content (wt %) pH

19.7 14.78 11.59

35.7 22.87 3.71

69.8 30.4 11.16

90.5 28.22 11.09
TABLE 2

Cu removal rate

Abrasive diameter (nm) pH (Å/min) Non-uniformity (%)

19.7 4.80 4196 4.05

35.7 2.90 6383 7.81

69.8 3.25 5457 7.95

90.5 3.35 4991 5.93

Test 2
This test evaluates the effects of the solid content of the colloidal silica on the polishing removal rate and the uniformity. The abrasive particle diameter of all slurries is fixed to 90.5 nm. The concentrations of the hydrogen peroxide, acetic acid, and phthalic acid are fixed to 1.2 V %, 0.375 V %, and 0.2 wt %, respectively. The pH value and polishing results are shown in Table 3. It can be seen that the removal rate does not increase significantly when the abrasive slid content is 1˜3 wt %. Both the removal rate and the non-uniformity increase simultaneously when the solid content is greater than 3 wt %.

TABLE 3

Abrasive solid Cu removal rate

content (wt %) PH (Å/min) Non-uniformity (%)

1 3.04 4900 2.78

3 3.35 4991 5.93

6 3.35 6334 6.57

Test 3
In this test, we want to see the effects by changing the pH value of the slurry in the same composition. The polishing particles in all slurries have a concentration of 3 wt % and an abrasive diameter of 90.5 nm. The concentrations of the hydrogen peroxide, acetic acid, and phthalic acid are fixed to 1.2 V %, 0.375 V %, and 0.2 wt %, respectively. The slurries are then added with diluted ammonia to adjust their pH values, as listed in Table 4. The results show that the removal rate decreases as the pH value increases. The non-uniformity, on the other hand, increases with the pH value.

TABLE 4

pH Cu removal rate (Å/min) Non-uniformity (%)

3.35 4991 5.93

4.00 4693 7.91

5.00 3974 7.83

Test 4

This test checks the effects of changing the composition and concentration of the chemical etching agent. The polishing particles in all slurries have a concentration of 3 wt % and an abrasive diameter of 90.5 nm. The pH values and polishing results of all prepared slurries are shown in Table 5. It shows that the uniformity is not good when only acetic acid is added. The removal rate is greatly enhanced while the non-uniformity is reduced for slurries that have both acetic acid and phthalic acid. When the concentration ratios (hydrogen peroxide)/(acetic acid) and (acetic acid)/(phthalic acid) are fixed, simultaneously increasing the concentrations of the hydrogen peroxide, acetic acid, phthalic acid can increase the removal rate and the uniformity. With both the concentration ratio (hydrogen peroxide)/(acetic acid) and the phthalic acid concentration being fixed, increasing the concentration ratio (acetic acid)/(phthalic acid) can increase the removal rate without affecting the uniformity to a good approximation. With the same concentrations of hydrogen peroxide and acetic acid, decreasing the concentration ratio (acetic acid)/(phthalic acid) (or increasing the phthalic acid concentration at the same time) can increase the removal rate and the uniformity.

TABLE 5


		(hydrogen		(acetic
hydrogen	acetic	peroxide)/	phthalic	acid)/		Cu removal	Non-
peroxide	acid	(acetic	acid	(phthalic		rate	uniformity
(V %)	(V %)	acid)	(wt %)	acid)	pH	(Å/min)	(%)

0.8	0.25	3.2	0	—	4.01	2475	11.44
1.2	0.375	3.2	0	—	3.82	3476	11.66
1.6	0.5	3.2	0	—	3.69	4207	8.73
0.8	0.35	2.28	0	—	3.87	3444	10.83
0.8	0.45	1.78	0	—	3.76	4261	8.42
0.8	0.25	3.2	0.2	1.25	3.20	3825	4.12
1.2	0.375	3.2	0.2	1.875	3.35	4991	5.93
1.6	0.5	3.2	0.2	2.5	3.24	6011	5.29
1.2	0.375	3.2	0.3	1.25	2.98	5661	3.92
1.6	0.5	3.2	0.4	1.25	3.07	6522	3.70

From the above tests, one can readily see that the disclosed slurry composition can achieve the goals of high removal rates and uniformity.
It should be emphasized that the CMP slurry for the copper layer on a wafer first have the hydrogen peroxide interact with the copper surface to form copper oxide. The acetic acid then reacts with the copper oxide to form copper acetate, which is then removed by colloidal silica polishing particles. This can achieve a high removal rate. The invention further proposes to use the phthalic acid as the pH buffering agent and complexing agent in the slurry to increase the polishing uniformity at the same time.
Although the disclosed embodiment uses colloidal silica as the polishing particles in the slurry, other appropriate polishing particles can be employed in the chemical etching agent to enhance the removal rate and the uniformity.
Certain variations would be apparent to those skilled in the art, which variations are considered within the spirit and scope of the claimed invention.

Claims

1. A chemical mechanical polishing (CMP) slurry for polishing the copper layer on a wafer, comprising polishing particles, hydrogen peroxide, acetic acid, and phthalic acid;

wherein the polishing particles are of a particle diameter of 10˜100 nm and include a concentration of 0.5˜10 wt % colloidal silica;

wherein the concentration of the hydrogen peroxide is 0.6˜2.5 V %, and

wherein the pH value is 2.5-5.5.

2-4. (canceled)

5. The CMP slurry of claim 1, wherein the concentration of the acetic acid is 0.1˜1.0 V %.

6. The CMP slurry of claim 1, wherein the concentration of the phthalic acid is 0.1˜0.8 wt %.

7. (canceled)

8. The CMP slurry of claim 1, wherein the concentration ratio of hydrogen peroxide to acetic acid is 0.6˜25.

9. The CMP slurry of claim 1, wherein the concentration ratio of acetic acid to phthalic acid is 0.125˜10.

10. The CMP slurry of claim 5, wherein the concentration of the phthalic acid is 0.1˜0.8 wt %.

11. The CMP slurry of claim 5, wherein the concentration ratio of hydrogen peroxide to acetic acid is 0.6˜25.

12. The CMP slurry of claim 6, wherein the concentration ratio of hydrogen peroxide to acetic acid is 0.6˜25.

13. The CMP slurry of claim 10, wherein the concentration ratio of hydrogen peroxide to acetic acid is 0.6˜25.

14. The CMP slurry of claim 5, wherein the concentration ratio of acetic acid to phthalic acid is 0.125˜10.

15. The CMP slurry of claim 6, wherein the concentration ratio of acetic acid to phthalic acid is 0.125˜10.

16. The CMP slurry of claim 10, wherein the concentration ratio of acetic acid to phthalic acid is 0.125˜10.

17. The CMP slurry of claim 11, wherein the concentration ratio of acetic acid to phthalic acid is 0.125˜10.

18. The CMP slurry of claim 12, wherein the concentration ratio of acetic acid to phthalic acid is 0.125˜10.

19. The CMP slurry of claim 13, wherein the concentration ratio of acetic acid to phthalic acid is 0.125˜10.