COATED SEMICONDUCTOR PROCESSING MEMBERS HAVING CHLORINE AND FLUORINE PLASMA EROSION RESISTANCE AND COMPLEX OXIDE COATINGS THEREFOR

Abstract

A semiconductor processing member is provided, including a body and a plasma spray coating provided on the body. The coating is an ABO or ABCO complex oxide solid solution composition, where A, B and C are selected from the group consisting of La, Zr, Ce, Gd, Y, Yb and Si, and O is an oxide. The coating imparts both chlorine and fluorine plasma erosion resistance, reduces particle generation during plasma etching, and prevents spalling of the coating during wet cleaning of the semiconductor processing member.

Claims

1. A semiconductor processing member comprising: a body; a plasma spray coating provided on said body; wherein said coating is an ABO or ABCO complex oxide solid solution composition, where A, B and C are selected from the group consisting of La, Zr, Ce, Gd, Y, Yb and Si, and O is an oxide, whereby said coating imparts both chlorine and fluorine plasma erosion resistance, reduces particle generation during plasma etching, and prevents spalling of said coating during wet cleaning of said member.

2. The semiconductor processing member according to claim 1, wherein said coating is selected from the group consisting of La.sub.2Zr.sub.2O.sub.7, La.sub.1.5Ce.sub.0.5Zr.sub.2O.sub.7, Ce.sub.0.25Zr.sub.0.75O.sub.2, Ce.sub.0.7Gd.sub.0.3O.sub.2, Y.sub.0.15Zr.sub.0.85O.sub.1.93 and Y.sub.2Si.sub.2O.sub.7.

3. The semiconductor processing member according to claim 1, wherein said coating contains less than 50 ppm of any trace element.

4. The semiconductor processing member according to claim 3, wherein the coating contains less than 20 ppm of any heavy metal, less than 10 ppm of any alkaline metal, and less than 2 ppm of each of Fe, Ni and Cu.

5. The semiconductor processing member according to claim 1, wherein the body is a bare aluminum alloy body.

6. The semiconductor processing tool according to claim 1, wherein the body is an anodized aluminum alloy body.

7. The semiconductor processing member according to claim 1, wherein an outermost surface of said coating has a surface roughness in a range of 80 to 150 μin Ra, wherein a peak to valley ratio, Rp/Rv, of said surface is in a range of 0.30 to 0.60.

8. The semiconductor processing member according to claim 1, further comprising a bonding layer between the coating and the body to improve a tensile bond strength of said coating.

9. The semiconductor processing member according to claim 8, wherein said bond layer is selected from the group consisting of (1) Si and (2) co-phased fully stabilized zirconia and Y.sub.2O.sub.3.

10. The semiconductor processing member according to claim 1, wherein said coating reduces contamination of a working wafer in the semiconductor tool so that an amount of Na measured on the wafer is less than 1×10.sup.14 atoms/cm.sup.2.

11. A plasma sprayed coating for semiconductor processing members, said coating being selected from the group consisting of solid solutions of La.sub.2Zr.sub.2O.sub.7, La.sub.1.5Ce.sub.0.5Zr.sub.2O.sub.7, Ce.sub.0.25Zr.sub.0.75O.sub.2, Ce.sub.0.7Gd.sub.0.3O.sub.2, Y.sub.0.15Zr.sub.0.85O.sub.1.93 and Y.sub.2Si.sub.2O.sub.7, whereby said coating imparts both chlorine and fluorine plasma erosion resistance, reduces particle generation during plasma etching, and prevents spalling of said coating during wet cleaning.

12. The coating according to claim 11, wherein the coating contains less than 50 ppm of any trace element.

13. The coating according to claim 12, wherein the coating contains less than 20 ppm of any heavy metal, less than 10 ppm of any alkaline metal, and less than 2 ppm of each of Fe, Ni and Cu.

14. The coating according to claim 11, wherein an outermost surface of the coating has a surface roughness in a range of 80 to 150 μin Ra, where a peak to valley ratio, Rp/Rv, of the surface is in a range of 0.30 to 0.60.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The present invention is described in detail herein below in connection with the associated drawing figures, in which:

[0019] FIG. 1 is a graph showing the acid solubility for the test and comparative sample materials in HCL and H.sub.2SO.sub.4 measured over 24 hours as a function of weight % loss;

[0020] FIG. 2 is a graph showing the plasma erosion rate as a function of thickness loss for the test and comparative materials (Yttria);

[0021] FIG. 3 is a schematic view of a coated cylindrical shell (body) made from 6061 Al alloy representing a commonly used plasma etch chamber shape;

[0022] FIG. 4 is a schematic cross-sectional view of a coated chamber component body post-testing;

[0023] FIG. 5 is a schematic cross-sectional view of a coated component member (body) including a bonding layer; and

[0024] FIG. 6 is a graph showing the helium leak rate at 20 torr versus the life cycle of coated electrostatic chucks.

DETAILED DESCRIPTION OF THE INVENTION

[0025] Specific materials that are resistant to fluorine and chlorine chemistries were initially examined to understand their thermodynamic response to the reactions that take place in a semiconductor processing chamber, such as a reactive ion etch (RIE).

[0026] Pyrochlores, fluorites and perovskites were analyzed. More specifically, materials from the general families of LaZrO, LaCeZrO, LaGdZrO, YSiO, GdZrO, GdCeO, LaCeO, HfGdO, YbSiO, CeZrO, and CeGdO were examined. In addition, AO-type materials, like Al.sub.2O.sub.3, Y.sub.2O.sub.3, Yb.sub.2O.sub.3 and HfO.sub.2 and ZrO.sub.2, were also examined as comparative references. It was found that the ABO or ABCO materials, or complex binary or ternary oxides, provided more stable compounds compared to the AO family of simple oxides.

[0027] In order to determine which compositions would be the best to use for eventual plasma spay applications, the initial evaluation of materials involved the use of sintered ceramic samples using conventional sintering methods in consideration of the low cost and ease of manufacturing. Sintered ceramic reference or comparative (AO-type oxide material) samples were made, including Y.sub.2O.sub.3, Yb.sub.2O.sub.3 and HfO.sub.2. In addition, sintered ABO-type material test samples were also made, including Y.sub.0.15Zr.sub.0.85O.sub.1.93, Y.sub.2Si.sub.2O.sub.7, Yb.sub.2Si.sub.2O.sub.7, Ce.sub.0.25Zr.sub.0.75O.sub.2, Ce.sub.0.7Gd.sub.0.3O.sub.2 and La.sub.2Zr.sub.2O.sub.7.

[0028] Commercially available, high purity feedstock powders for each sample composition were mechanically mixed with 3-5 wt % of a polymeric binder, then pressed in the form of a 1 inch diameter samples that were 0.30 inch thick. Each of the pressed samples was fired at 600° C. to burn off the polymeric binder. These samples were then sintered from 1400-1600° C. for 10 to 15 to achieve 95-98% of the theoretical density of each composition, and then cooled in the furnace over a period of 24 hrs. The specific sample compositions were sintered under the following conditions:

[0029] Y.sub.2O.sub.3 at 1400° C. for 12 hours;

[0030] Yb.sub.2O.sub.3 at 1400° C. for 10 hours;

[0031] HfO.sub.2 at 1400° C. for 10 hours;

[0032] Y.sub.0.15Zr.sub.0.85O.sub.1.93 at 1500° C. for 14 hours;

[0033] Y.sub.2Si.sub.2O.sub.7 at 1400° C. for 10 hours;

[0034] Yb.sub.2Si.sub.2O.sub.7 at 1400° C. for 12 hours;

[0035] Ce.sub.0.25Zr.sub.0.75O.sub.2 at 1600° C. for 14 hours;

[0036] Ce.sub.0.7Gd.sub.0.3O.sub.2 at 1600° C. for 14 hours; and

[0037] La.sub.2Zr.sub.2O.sub.7 at 1600° C. for 15 hours.

[0038] The sintered samples were then ground and polished on both sides to achieve a thickness of 0.24 inch and the surface roughness was at least 2 μin Ra. The samples were cleaned for 10 minutes in an ultrasonic bath filled with deionized (DI) water and then dried in an oven at 85° C. for 2 hrs. All samples were then cooled to room temperature prior to any subsequent testing.

[0039] All of the test and comparative samples were then evaluated for acid solubility in HCL and H.sub.2SO.sub.4, which are acids that are commonly used for cleaning semiconductor chamber components. The test and comparative samples were fully immersed in 5 wt % HCl and also in 5 wt % H.sub.2SO.sub.4 at room temperature (at 21° C.) for 24 hours, after which time the material loss (in wt %) was measured. FIG. 1 shows the results of the comparative and test materials with respect to the acid solubility. As shown in FIG. 1, this data translates into poor wet cleaning resistance for Y.sub.2O.sub.3, the solubility of which in HCl was 6 wt %, and the other AO-type oxides. On the other hand, the ABO-type complex oxides did not show any measurable acid solubility, which translates into good wet cleaning resistance.

[0040] Further experiments were done to determine the plasma etch resistance of the test and comparative sample materials. Table 1 shows fluorine chemistries that are widely used for dielectric and conductor etching. CF.sub.4 and oxygen plasma were then selected as the etchants for this comparative evaluation of the plasma erosion rate of the select materials. Table 2 shows the experimental conditions that were used according to the present experiment.

TABLE-US-00001 TABLE 1 Commonly used Plasma Etch conditions in semiconductor tools. Bottom Bias Top RF Power, Power, Watts @ Process Process Gases Watts @ 13.56 MHz 13.56 MHZ Pressure, mtorr Dielectric Etch CF4/CHF3/SF6/O2 100 to 1200 200 to 1000  30 to 300 Via Etch C4F6/CHF3/O2 400 to 1000 100 to 1700 30 to 80 Conductor Etch Cl2/BCL3/O2 500 to 1000 100 to 200  10 to 20

TABLE-US-00002 TABLE 2 Experimental Plasma Etch Conditions for the Present Experiments Top RF Power Bottom Bias Power Process Gasses Watts @13.56 MHZ Watts @13.56 MHZ Pressure, mtorr CF.sub.4/O.sub.2 2500 250 40

[0041] As described in Table 2, the etch chamber was set to operate at 2500 watts of top RF coil power and a bottom capacitive bias of 250 Watts. RF generators were set to operate at a frequency of 13.56 MHz and the chamber was maintained at a vacuum pressure of 40 mtorr. A flow rate of the ratio of 10 to 1 sccm was maintained for CF.sub.4 to O.sub.2 to strike the plasma. Each test sample was etched for 30 min. These harsh conditions were selected to in order to accelerate the plasma erosion of the test and comparative sample materials for rapid testing purposes.

[0042] After the plasma exposure, the erosion rates of the test and comparative sample materials were measured by recording the thickness loss from the exposed surfaces of the samples. FIG. 2 shows relative plasma erosion rate results of the test and comparative materials, where the plasma erosion of Y.sub.2O.sub.3 of 1 μm/Rf hour is used as a reference.

[0043] The plasma erosion data shows that the ABO or complex oxide materials have lower plasma erosion rates compared to that of the AO or single oxide-type materials.

EXAMPLES

[0044] Further wet corrosion studies and plasma erosion studies identified that ABO-type complex oxide materials provide a better solution for semiconductor processing members with respect to reducing particle generation and improving the number of wet cleaning cycles compared to AO-type single oxides materials that have been used thus far. While the preferred materials according to the present invention can be used as sintered ceramics, where chamber components are made by sintering ceramics, it should be noted that such sintered ceramic components are more expensive and brittle, and that such sintered ceramic parts can break if impacted. In contrast, thermal sprayed coatings made of the complex oxides according to the present invention offer a more robust and economical solution to reduce particles and improve wet cleaning resistance.

[0045] The examples below involve thermal spray technology to form protective coatings on substrates (semiconductor chamber component bodies) utilizing the preferred materials according to the present invention.

Comparative Example 1

[0046] A mechanically blended powder was made using La.sub.2O.sub.3 and ZrO.sub.2 in blend ratios to target a pyrochlore phase at 56 wt % La.sub.2O.sub.3 with a balance of ZrO.sub.2 in a range of 50-60 wt % La.sub.2O.sub.3. Coatings were sprayed with a DC plasma arc spray system using Argon and Hydrogen plasma gasses. The formed coatings showed a La.sub.2Zr.sub.2O.sub.7 phase and the presence of free La.sub.2O.sub.3 and ZrO.sub.2. These coatings were tested for plasma erosion, but did not perform as well as compared to the Y.sub.2O.sub.3 coatings. These results were undesirable.

Example 1

[0047] In view of the undesirable results in Comparative Example 1, powders were then prepared where a solid solution of La.sub.2O.sub.3 and ZrO.sub.2 was made with La.sub.2Zr.sub.2O.sub.7 phase, as verified by X-ray diffraction. Powder samples were prepared using conventional methods of crushing, sieving and spray dry agglomeration to provide an average particle size of 30 μm. The purity of the powder was maintained using powder manufacturing systems lined with plasma sprayed zirconia coatings or polymerics such that all trace elements were less than 50 ppm, all heavy metals were less than 20 ppm, and alkaline metals were less than 10 ppm, as measured by Glow Discharge Mass Spectroscopy (GDMS).

[0048] Coatings were made from this La.sub.2Zr.sub.2O.sub.7 powder using a DC plasma arc system using argon and hydrogen with an envelope of argon around the plasma plume to form La.sub.2Zr.sub.2O.sub.7 coatings. These La.sub.2Zr.sub.2O.sub.7 coatings did not show any weight loss when immersed in HCl for 24 hours. These La.sub.2Zr.sub.2O.sub.7 coatings were then subjected to an ultrasonic bath of deionized (DI) water at 40 KHz for 15 min, dried in an oven at 85° C. for at least 4 hours, and then tested in the plasma etch chamber and exposed to the fluorine plasma conditions as shown in Table 2. The finish on these La.sub.2Zr.sub.2O.sub.7 coatings did not change, and the measured plasma erosion rate was 66% less than that that of the reference Y.sub.2O.sub.3 coating.

[0049] In view of the success of these test samples, a large part was made to represent a shape commonly used in a plasma etch chamber. The part (1) was a cylindrical shell made from 6061 Al alloy and having an inner diameter (ID) of 14 inch and 5 inch height (H), as shown in FIG. 3. The thickness (T) of the cylindrical shell wall was 0.125 inch. The part (1) was grit blasted on the inner diameter surface and then the entire part was hard anodized. The anodization film (2) thickness was about 0.002 inch. A La.sub.2Zr.sub.2O.sub.7 coating (3) having a coating thickness of 0.006 inch and made from the high purity, solid solution-formed La.sub.2Zr.sub.2O.sub.7 powder described above was then deposited on the inner diameter surface without any further grit blasting. The part was then cleaned in an ultrasonic bath filled with deionized (DI) water for 15 min and then dried in an oven at 85° C. for at least 4 hours. The coated part was placed in a plasma etch chamber to simulate a fluorine dielectric etch environment, and then a chlorine conductor etch environment according to the conditions shown in Table 1. The part was exposed to each of the plasma etch conditions for 15 min for 5 cycles.

[0050] When the part was removed from the etch chamber, no erosion was observed on the coated surface. The inner surface of the removed part was wiped with 5 wt % HF and then 5 wt % HCl. The part was then submerged in a DI water ultrasonic cleaner for 10 min and the dried in an oven at 85° C. for at least 4 hours. This cycle of cleaning was repeated 10 times, and no coating delamination from the substrate occurred.

Example 2

[0051] The coating from the coated shell used in Example 1 was stripped off using a grit blast method while preserving the anodization. Then the La.sub.2Zr.sub.2O.sub.7 coating according to Example 1 was reapplied to restore dimensions of the part. The part was again tested in the plasma etch chamber under the same conditions for 5 times with a cleaning cycle in between the process. The part showed no erosion or delamination of the coating from the substrate.

[0052] During the plasma etching processes, Si wafers were also processed, and the contamination levels on the wafers were measured by Mass Spectroscopy using inductively coupled Plasma (MS-ICP). It was found that trace elements on the wafers were less than 10×10.sup.10 atoms/cm.sup.2. More importantly, heavy metal elements were less than 5×10.sup.10 atoms/cm.sup.2, and alkaline metals were less than 2×10.sup.10 atoms/cm.sup.2.

[0053] After the above-described testing, the coated part was sectioned and the coating microstructure was revealed, as schematically shown in FIG. 4. The substrate (body) (1) made of 6061T6 Aluminum has an anodized film (2) on the outer surface thereof. The inner side of the substrate (body) surface was roughened and then anodized; this rough anodized interface (2R) was used to deposit the La.sub.2Zr.sub.2O.sub.7 coating (3) without grit blasting. The coating (3) showed no signs of voids within its microstructure, and there was no delamination of the coating from the substrate at interface between the rough anodized interface (2R) and the coating (3), indicating that the coating did not corrode in the chamber or during the wet cleaning process.

[0054] The hardness of the coating in cross-section was measured to be in a range of 500 to 600 kg/mm.sup.2, and the porosity was measured to be less than 1%.

Example 3

[0055] A part having the shape of a cylindrical shell (14 inch ID and 5 inch height) made of 6061 Al alloy was coated in the same manner as described above in connection with Example 1. This time, the surface of the La.sub.2Zr.sub.2O.sub.7 coating was textured using a series of flexible pads with backing for diamond coated abrasives. Specifically, 60 μm and then 30 μm diamonds were used as abrasives to texture the coated surface to remove only high spots from the coated surface. This flexible finishing method provided the coated surface with a roughness Ra in a range of 80 to 120 pin. The removal of high spots also provided a peak to valley ratio, Rp/Rv of the surface roughness in a range 0.3 to 0.6.

[0056] The part was then cleaned in an ultrasonic bath filled with DI water at 40 KHz at 60° C. for 15 min. Measurement of the particles entrained in the cleaning fluids by a laser particle counter showed a drop of 50% in the total particle count compared to that of Example 1. The part was then dried in an oven at 85° C. for at least 4 hours. The coated part was then placed in a plasma etch chamber to simulate a fluorine plasma etch environment for 30 min. This time, the 200 mm wafers that were also simultaneously processed were examined to determine the number of particles landing thereon. Only one particle was detected in a run where 10 wafers were processed.

Example 4

[0057] To further confirm the validity of using complex oxides for plasma erosion application in semiconductor tools, a powder of the composition Ce.sub.0.25Zr.sub.0.75O.sub.2 as a solid solution was made using high purity materials, where the content of heavy metals was less than 20 ppm, the content of alkaline metals was less than 5 ppm, and the content of Fe, Cu or Ni was less than 5 ppm each. The powders of the composition had an average particle size of 20 μm. The powder was sprayed with a DC plasma spray system, using argon and hydrogen, with an envelope of argon around the plasma plume to minimize entrainment of air into the effluent. Free standing Ce.sub.0.25Zr.sub.0.75O.sub.2 coatings were made and then different samples were immersed in one of HCL and H.sub.2SO.sub.4 for 24 hours. These Ce.sub.0.25Zr.sub.0.75O.sub.2 coatings did not show any significant weight loss in either acid exposure test. Samples of Ce.sub.0.25Zr.sub.0.75O.sub.2 coatings exposed to fluorine plasma for 30 min showed 60% less erosion as compared to the reference Y.sub.2O.sub.3 coating.

Example 5

[0058] Powder was also made of Ce.sub.0.7Gd.sub.0.3O.sub.2 as a solid solution (hereinafter GDO powder), where the average particle size was 30 μm. The GDO powder was a high purity content to maintain all trace elements to less than 50 ppm, as verified by GDMS. Free standing coatings were made with the CGO powder, and when the coatings were immersed in HCL and H.sub.2SO.sub.4 for 24 hours, no weight loss was measured. Samples of the GDO coatings exposed to fluorine plasma for 30 min showed 58% less plasma erosion compared to the reference Y.sub.2O.sub.3 coating.

Example 6

[0059] A first set of 3 cylindrical test samples having a diameter of 1 inch and a length of 1.5 inch were made of 6061 Al alloy. Bare Al alloy coupons were grit blasted on the 1 inch diameter face, and a La.sub.2Zr.sub.2O.sub.7 solid solution coating of 0.012 inch thickness was deposited on the grit blasted surface by the plasma spray coating process described in connection with Example 1. The tensile bond strength of the coating was measured according to the ASTM C633 standard method, and was determined to be an average of 8,900 psi.

[0060] Another set of 3 cylindrical samples were grit blasted and then anodized (hereinafter referred to as anodized Al alloy coupons). A La.sub.2Zr.sub.2O.sub.7 solid solution coating of 0.012 inch thickness was deposited on the 1 inch diameter face. The tensile bond strength of the coating was measured to be an average of 7,900 psi.

[0061] The bare Al alloy and anodized Al alloy samples having the coatings thereon were cycled to 300° C. 10 times. It was found that the tensile bond strength of the coating on the anodized Al alloy surfaces was reduced by 5%, whereas for bare Al alloy surfaces, the tensile bond strength of the coating was reduced by 30% under the same conditions.

[0062] Another set of bare and anodized Al alloy test samples were first coated with an Si bond coating layer, and then with the La.sub.2Zr.sub.2O.sub.7 coating. The test samples were then cycled 10 times from room temperature to 300° C., and then the tensile bond strength of the test samples was measured. These samples did not show any significant decrease in tensile bond strength of the coating.

[0063] Another set of samples were prepared, both with and without bond layers, as described in more detail below. A cross-section of a sample including a bonding layer 4 is shown schematically in FIG. 5. This sample (1) was first roughened on one side and then anodized on all surfaces. The anodization on the non-roughened surface is shown as (2) and the roughened interface is shown as (2R). Then a bond layer (4) was deposited. This time, the bond layer coating (4) was made of a co-phase of fully stabilized zirconia (FSZ) and Y.sub.2O.sub.3 coating, where Y.sub.2O.sub.3 was 18 wt. %. Then a top layer coating (3) of La.sub.2Zr.sub.2O.sub.7 was deposited on the bond layer (4). Both of these samples were cleaned with HCl and then immersed in DI water in an ultrasonic tank for 5 minutes. These samples were then tested for tensile bond strength. It was found that the test samples with bond layer showed no decrease in bond strength, whereas the test samples without a bond layer showed a 25% decrease in bond strength.

Example 7

[0064] It is critical to maintain the purity of the coatings in semiconductor chambers in order to maintain the functionality of the transistor devices that are formed in such processing chambers. To understand the purity of the formed coatings, a set of 4×4×0.125 inch coupons were coated with a La.sub.2Zr.sub.2O.sub.7 solid solution coating. The surfaces were examined by GDMS after laser ablation of the surface in successive steps, where the coating was removed by rastering a laser beam on the coated surface and the coated surface was ablated in steps to a depth of 0.001 inches and then the trace elements of each exposed coating surface layer were measured by GDMS. This measurement method showed that through the thickness of these coatings, that there were no trace metal elements in excess of 20 ppm (i.e., less than 10 ppm, and for critical elements of Na, K, Mg, Fe, Cr and Ni, and less than 5 ppm or even less than 2 ppm for Fe, Ni and Cu).

Example 8

[0065] A solid solution powder was made with a La.sub.1.5Ce.sub.0.5Zr.sub.2O.sub.7 phase, where the average particle size of the powder was 20 μm. Coatings were made using the plasma spraying methods according to Example 1. A set of samples, hereinafter referred to as bend strips, being 6 inch in length, 1 inch in width and 0.125 inch in thickness, made of 6061T6 Aluminum alloy, were coated with La.sub.1.5Ce.sub.0.5Zr.sub.2O.sub.7 coatings of 0.004 inch thick. Reference samples were also made having the same dimensions and coated with La.sub.2Zr.sub.2O.sub.7 solid solution coatings of the same thickness. These bend strips were bent in a fixture around a mandrel of 0.5 inch diameter. It was found that the La.sub.2Zr.sub.2O.sub.7 coatings started cracking and spalling from the substrate earlier than the La.sub.1.5Ce.sub.0.5Zr.sub.2O.sub.7 coatings during this bend test, which indicated an increase in ductility and adhesion strength of the La.sub.1.5Ce.sub.0.5Zr.sub.2O.sub.7 coating due to the ternary phase formation in the presence of CeO.sub.2 as verified by x-ray diffraction.

Example 9

[0066] An electrostatic chuck (ESC) coated with Y.sub.2O.sub.3 that showed surface erosion and higher helium leak rates when used in a chlorine plasma etch environments was obtained. The chuck surface was provided with a La.sub.2Zr.sub.2O.sub.7 solid solution coating on the surfaces that are exposed to plasma conditions for dielectric etching. The chuck was used again in the plasma chamber under the same conditions, and showed no erosion of the surfaces that were protected with the La.sub.2Zr.sub.2O.sub.7 coating, and the electrostatic chuck continued to provide electrostatic clamping to a working substrate with very stable helium leak rates. FIG. 6 shows that there was no increase in backside gas leak rates due to higher erosion resistance of the sealing surfaces of the ESC coated with La.sub.2Zr.sub.2O.sub.7 as compared to the one coated with Y.sub.2O.sub.3. Use of the high purity, solid solution La.sub.2Zr.sub.2O.sub.7 coatings also resulted in very little transfer of trace elements to the wafer. Trace elements of heavy metals like Cr, Fe, Ni and Cu were less than 1×10.sup.12 atoms/cm.sup.2, and Na was less than 1×10.sup.14 atoms/cm.sup.2, as measured by Mass Spectroscopy using Inductively Coupled Plasma instrumentation (MS-ICP).

Example 10

[0067] A sintered ceramic part made from Al.sub.2O.sub.3 that showed an excessive formation of AlF.sub.3 particles in a plasma chamber was coated with 0.004 inches of a Ce.sub.0.7Gd.sub.0.3O.sub.2 solid solution coating. The coated part, when installed in a semiconductor etch tool in the plasma zone did not show any formation of AlF.sub.3 particles when exposed to fluorine plasma. A polymer film, which is a byproduct of the fluorine etching process, begins to build up as soon as Rf exposure begins and was deposited on the part. After about 100 Rf hours, the buildup is measurable and requires removal. The part was removed from the chamber, the Ce.sub.0.7Gd.sub.0.3O.sub.2 coating was removed by grit blasting, and then a new coating layer was applied to restore the part to full functionality.

[0068] The various examples above illustrate the use of complex oxide coatings to prevent plasma erosion of various components used in semiconductor processing members.

[0069] While the present invention has been particularly shown and described with reference to the preferred mode as illustrated in the drawings, it will be understood by one skilled in the art that various changes in detail may be effected therein without departing from the spirit and scope of the invention as defined by the claims.