Wipe Analysis to Determine Metal Contamination on Critical Surfaces

Transcription

1 By Albert Dato, Ph.D., Warren York, Jennifer Jew, Laarni Huerta, Brice Norton, and Michael Coste On-wafer metallic contamination is detrimental to the fabrication and performance of semiconductor devices. Metals such as Fe, Cr, and Cu can dissolve in silicon and form silicides. Wafer oxidation rates can be affected by Al contamination, while Ni can increase Si etch rates. The presence of mobile ions such as Na + and K + in gate-oxide or oxide-semiconductor interfaces can result in threshold voltage shifting of transistors. Metals can also degrade the insulating and capacitive properties of dielectrics. As the feature sizes of semiconductor devices continue to decrease, so do the maximum allowable levels of on-wafer metallic contamination. On-wafer metallic contamination can originate from critical surfaces, such as components of wafer processing tools, equipment associated with wafer processing tools, and areas used to manufacture wafer processing tools. For example, high Mg levels on the surface of a single component in a tool can contaminate a wafer during processing. In another example, load locks or wafer transfer systems can become contaminated through improper preventative maintenance (PM) procedures (e.g. handling using non-cleanroom gloves with high levels of metals) and can subsequently contaminate wafers moving into the wafer processing tool. There are a myriad of critical surfaces that are potential on-wafer metallic contamination sources. Therefore, the determination of metals on critical surfaces is needed to ensure that metallic contamination on processed wafers remains within specifications. Determining metals on critical surfaces can be a challenge. For instance, when a wafer becomes contaminated in a wafer processing tool, the contamination sources in the tool need to be determined. However, some components of wafer processing tools cannot be shipped to an off-site analytical laboratory for surface metals testing because they are either too large to ship or cannot be removed from the tool. Such is the case for chamber bodies and pedestals. Furthermore, many parts in wafer processing tools exhibit varying surface properties and complex features (e.g. showerheads, electrostatic chucks, and domes). These components are a challenge to analyze using contemporary surface analysis techniques, such as drop scan etch (DSE) inductively coupled plasma mass spectrometry (ICP-MS). This article presents the Balazs Critical Surface Wipe (CSW) testing method, which is a technique that enables the rapid, on-site testing for metals on critical surfaces in a wide range of environmental conditions. The facile method is capable of determining the metals on a variety of surfaces, from heavy components to sensitive parts with complex features. The experimental results presented in this article demonstrate that the technique does not leave any fibers, particles, or metals on materials used in wafer processing tools. A case study of a practical CSW application will be also be discussed. Critical Surface Wipe Testing Method The CSW testing method requires the materials shown in Figure 1. The CSW kit includes four pairs of Low- Metals Gloves, a Glove Cleaning Wiper used to clean the gloves prior to sampling, a Control Wiper, and a Critical Surface Wiper. Each of the wipers in the kit is dampened with ultrapure water, which makes the wipers compatible with a wide range of surfaces.

2 Figure 1. A CSW lit includes four pairs of Low-Metals Gloves, a Glove Cleaning Wiper, a Control Wiper and a Critical Surface Wiper. The testing of a critical surface could be performed by a single operator. For optimal results, a team of two personnel is recommended. As shown in Figure 2a, one operator (hereafter called the Bottle Handler) is assigned the role of handling, opening, closing, and labeling the containers holding the wipers. Another operator (hereafter called the Wipe Tester) is assigned the role of handling the wipers and performing the testing. A team of two operators ensures that contamination from other surfaces (e.g. the outside of the bottle) is prevented during the testing of a critical surface. Figure 2. (a) Two operators are required for best results. (b) Gloves should cover the cuffs of the clean room suit. The first step in the testing of a critical surface involves both operators donning two layers of Low-Metals Gloves. If CSW testing is performed in a clean room, the second layer of gloves should cover the cuff of the clean room suit, as shown in Figure 2b. The second step requires the Wipe Tester to clean the second layer of gloves using the Glove Cleaning Wiper. This is performed to further reduce the metals on the gloves. The Bottle Handler opens the container containing the Glove Cleaning Wiper. The Wipe Tester then removes the Glove Cleaning Wiper from the container and thoroughly cleans the second layer of gloves, as shown in Figure 3. Every surface of the gloves must be wiped, with emphasis on the fingers, palm and cuff. The Glove Cleaning Wiper and container can then be safely discarded in any trash receptacle once the glove cleaning process is completed. Once the gloves are cleaned, the Wipe Tester must only handle the Control Wiper and Critical Surface Wiper to eliminate contamination from any other sources. Figure 3. The cleaning of gloves prior to wipe testing reduces the metals on the gloves.

3 The third step of the CSW testing procedure involves using the Control Wiper to sample the metals contribution of the gloves and the environment around the critical surface to be tested. The Bottle Handler opens the container holding the Control Wiper. As demonstrated in Figure 4, the Wipe Tester removes the Control Wiper, unfolds it completely, and waves it in the air near the surface to be sampled for 30 seconds. This is about the same duration of time it takes to perform a wipe test of the critical surface. After the sampling is completed, the Control Wiper is folded and returned into its original container. The Bottle Handler then tightly closes and labels the container. Figure 4. The Control Wiper is used to determine the metals contribution of wiper handling and the environment around the critical surface. The fourth step involves sampling the critical surface. The Bottle Handler opens the bottle containing the Critical Surface Wiper. As shown in Figure 5a, the Wipe Tester removes the wiper from its container and holds it so that the folded edge is at the fingertips and the unfolded edges are between the thumb and forefinger. A downward force of about 5 psi (35 kpa) is then applied to the wiper as it makes contact with the critical surface. The recommended surface area for sampling is 16 in 2 ( cm 2 ), but larger and smaller areas can also be tested. Figure 5b shows that the sampling of a 16 in 2 ( cm 2 ) area can be accomplished by using three fingers to apply the downward force, which is about 2 in (5.08 cm) wide, then swiping 8 in (20.32 cm) in a straight line across the surface, and lifting off at the end of the swipe. The critical surface is wiped twice, with the wiper re-folded so that a fresh area of the Critical Surface Wiper is exposed for the second pass. Once sampling is completed, the Critical Surface Wiper is returned into its original container. The Bottle Handler then tightly closes and labels the container. Figure 5. The CSW testing procedure for sampling a critical surface. Additional Critical Surface Wipers, Low-Metals Gloves, and Glove Cleaning Wipers may be ordered and used if multiple critical surfaces need to be sampled during the same wipe testing session (e.g. same tool, same time, and same environment). The use of a second control wiper is not required if multiple surfaces are sampled during the same wipe testing session. Once wipe testing is completed, the Control Wiper and Critical Surface Wiper(s) are shipped to Balazs for trace metals analysis by ICP-MS. Each wiper is leached in a dilute acid solution, and each solution is then analyzed by ICP-MS. Metals on critical surfaces are then determined by subtracting the ICP-MS results from the Control Wiper from the ICP-MS results of the Critical Surface Wiper. The ICP-MS reporting limits for a 16 in 2 ( cm 2 ) critical surface are shown in Table 1.

4 Table 1. Reporting limits for the CSW test of a 16 in 2 ( cm 2 ) critical surface. Reporting limits can change based on the size of the area that is tested. Element Reporting Limit (1x10 10 atoms/cm 2 ) Element Reporting Limit (1x10 10 atoms/cm 2 ) Aluminium (Al) 300 Manganese (Mn) 100 Calcium (Ca) 1000 Molybdenum (Mo) 100 Chromium (Cr) 100 Nickel (Ni) 100 Cobalt (Co) 100 Potassium (K) 800 Copper (Cu) 100 Sodium (Na) 300 Iron (Fe) 500 Titanium (Ti) 100 Lithium (Li) 500 Vanadium (V) 100 Magnesium (Mg) 300 Zinc (zn) 100 Critical Surface Wipe Testing Materials The CSW testing method requires wipers with very low levels of metals. Balazs has performed extensive research to determine a commercially-available ISO Class 4 cleanroom wiper with the lowest metallic content. The results of side-by-side testing of several wipers are shown in Table 2. The wiper with the lowest overall metals (Wiper G) is shown in the rightmost column. Table 2. A comparison of metals in commercially-available ISO Class 4 cleanroom wipers. X indicates the lowest level of a metal. Wiper G was determined to have the lowest levels of most metals. Parameter Wiper A Wiper B Wiper C Wiper D Wiper E Wiper F Wiper G 1 Aluminium (Al) x 2 Calcium (Ca) x 3 Chromium (Cr) x x x x 4 Cobalt (Co) x 5 Copper (Cu) x 6 Iron (Fe) x x 7 Lithium (Li) x x x x x x 8 Magnesium (Mg) x 9 Manganese (Mn) x 10 Molybdenum (mo) x 11 Nickel (Ni) x x x x 12 Potassium (K) x 13 Sodium (Na) x 14 Titanium (Ti) x x x x 15 Vanadium (V) x 16 Zinc (Zn) x Despite having the lowest levels of metals, Wiper G exhibits a metallic content that is still too high for the application of determining the metallic contamination on critical surfaces. The metals on Wiper G are reduced through a proprietary Balazs cleaning and packaging process. Figure 6 shows that the process significantly lowers the metals on the wipers. The resulting low-metals background of the processed wipers enables them to be used as Control Wipers and Critical Surface Wipers. Processed wipers are dampened with ultrapure water and are compatible with a wide range of surfaces.

5 Figure 6. The Balazs proprietary cleaning and packaging process significantly lowers the metals on Wiper G. Does Wipe Testing Contaminate Critical with Fibers, Particles, or Residues? Critical Surface Wipers are processed and packaged to achieve a low metallic background, which enables the wipers to be used for the determination of surface metals on critical surfaces. The wipers should not deposit fibers, particles, or residues on surfaces during sampling, particularly when testing sensitive wafer processing tools. Therefore, a series of tests were conducted to show that Critical Surface Wipers do not deposit fibers, particles, or residues on surfaces during wipe testing. Critical Surface Wiper contamination was investigated using the test procedure shown in Figure 7. Three materials that are commonly found in wafer processing tools were analyzed: an alumina plate, an anodized aluminum plate, and a quartz plate. The surface roughness of each sample was determined using an atomic force microscope (AFM). After obtaining AFM data, a Zeiss EV-50 Scanning Electron Microscope (SEM) was utilized to determine if fibers, particles, or residues were deposited on the samples after wipe testing. Samples were imaged using the SEM in a central location prior to undergoing a wipe test. After obtaining pre-wipe images, the SEM sample stage was opened and each plate immediately underwent CSW testing while on the SEM sample stage. Testing the samples directly on the stage ensured minimal contamination from the environment and facilitated the post-wipe imaging of the same location. The SEM sample stage was then closed, and post-wipe SEM images of the same central location were obtained. The Critical Surface Wipers that were used to test each sample were then chemically leached. ICP-MS analysis and subtraction of the Control Wiper ICP-MS results were then performed to determine the surface metals on each of the samples.

6 Figure 7. Procedure for determining wiper contaminants deposited on critical surfaces. The AFM images shown in Figure 8 reveal that the samples had varying surface properties. The alumina plate exhibited an irregular surface with an average roughness of 572 nm. The anodized aluminum sample had an average roughness of 142 nm. The quartz sample exhibited a dimpled surface with an average roughness of 604 nm. Figure 8. AFM images of the alumina plate (Ra = 572 nm), anodized aluminum plate (Ra = 142 nm), and quartz plate (Ra = 604 nm). Pre-wipe and post-wipe SEM images of the alumina, anodized aluminum, and quartz samples are shown in Figures 9, 10, and 11, respectively. Fibers can be defined as having a length 100 μm with an aspect ratio of 10. A comparison of the pre-wipe and post-wipe SEM images at both 500X and 5000X magnification show that fibers, micron-scale particles, or visible residues are not deposited on each of the samples during wipe testing.

7 Figure 9. SEM images show that wipe testing does not leave any fibers, particles,or visible residues on the alumina surface. Figure 10. SEM images show that wipe testing does not leave any fibers, particles, or visible residues on the anodized aluminum surface.

8 Figure 11. SEM images show that wipe testing does not leave any fibers, particles, or visible residues on the quartz surface. ICP-MS analysis of each of the Critical Surface Wipers revealed that each of the samples had unique metallic contamination levels. Figure 12 shows that the anodized aluminum sample exhibited the highest levels of Al, Mg, K and Na. The quartz sample had the highest levels of Ca and Zn. Cu and Ni were highest on the alumina sample. Therefore, the results shown in Figures 9 through 12 demonstrate that Critical Surface Wipers do not leave particles, fibers, or visible residues on critical surfaces, and are capable of determining the varying surface metals on materials commonly used in wafer processing tools. Figure 12. CSW testing was capable of determining the surface metals on each sample.

9 Does Wipe Testing Contaminate Critical with Metals? Critical Surface Wipers undergo a cleaning process to reduce metals on the wipers. However, despite their low levels, some metals are still present on the wipers. Critical Surface Wipers must not deposit metals onto critical surfaces. Thus, tests were conducted to determine if Critical Surface Wipers deposit metals onto a surface during wipe testing. The quartz plate described in the previous section was cleaned and utilized to investigate wiper metallic contamination. DSE ICP-MS testing was performed on the sample in an ISO Class 4 environment to determine the pre-wipe levels of metals on its surface. A CSW test was then performed on the quartz plate. Immediately after wipe testing, a second DSE ICP-MS test was performed on the plate to determine the post-wipe levels on the sample. Pre-wipe and post-wipe DSE ICP-MS results are shown in Figure 13. The data shows that only Cr, Cu, Fe, Ni, Na, Ti, and V were detected on the surface of the sample prior to wipe testing. After wipe testing, Cr, Ni, and V were reduced to below the reporting limits of DSE ICP-MS, while Cu and Na levels were lowered. These results indicate that the wipers remove surface metals during wipe testing. Slight increases in Fe and Ti were observed. Extensive testing of Critical Surface Wipers has shown that Ti is not present on the wipers. Therefore, the increase in Ti could be attributed to the presence of the element in the quartz sample. Furthermore, the pre-wipe and post-wipe levels of Fe are nearly similar, which could also indicate that the Fe detected in the DSE ICP-MS testing is inherent to the quartz sample. Based on the data shown in Figure 13, it could be concluded that CSW testing leaves minimal metallic contamination on a critical surface. Figure 13. Pre-wipe and post-wipe DSE ICP-MS results of the quartz surface. The data shows that CSW testing leaves minimal metallic contamination on surfaces.

10 Critical Surface Wipe Testing Case Study One of the many practical applications for the CSW testing method is the determination of metallic contamination sources in a wafer processing tool. This case study will focus on a wafer processing tool that was not meeting Na specifications on processed wafers. Partition testing using witness wafers was performed, an the investigation ruled out the components connected to the tool (e.g. load locks, wafer transfer systems) as sources of contamination. Therefore, the sources of Na contamination inside the wafer processing tool had to be determined. The tool was operational, and dismantling the tool to ship components for off-site DSE ICP-MS testing was not an option. Therefore, a rapid on-site technique of testing of multiple surfaces on the tool was needed. A CSW kit containing one Control Wiper, ten Critical Surface Wipers, ten Glove Cleaning Wipers, and forty pairs of Low-Metals Gloves was shipped to the site of the wafer processing tool. The tool remained closed to the cleanroom environment until wipe testing was ready to commence. Following the procedures described above, the Control Wiper was used to sample the environment. The tool was then opened, and ten locations inside the tool that were suspected of being sources of Na contamination were tested. The tool was then closed immediately after wipe testing in order to minimize additional contamination from the environment. The ICP-MS results of the CSW testing of the tool are shown in Figure 14. The data shows that the Control Wiper (labeled CTRL) had significantly lower levels of metals compared to each of the ten locations tested. The CTRL results demonstrate that the environment around the tool, as well as the handling of the wiper with cleaned lowmetals gloves, did not significantly contribute to the metals detected on the ten Critical Surface Wipers used to test the various locations on the tool. Furthermore, the CTRL results show that the Critical Surface Wipers exhibited a low metals background, which enabled the detection of metals in the environment and on critical surfaces. The ICP-MS results for the Control Wiper were subtracted from each of the ten Critical Surface Wipers that were used to test locations A through J. Figure 14 shows that the most likely sources of Na contamination in the wafer processing tool were locations F and J. These locations also exhibited the highest K levels in the tool, and higher levels of Ca relative to the other locations tested. Furthermore, locations C and D also had high levels of Na, although the locations had about half the Na levels of locations F and J. Thus, CSW testing enabled the rapid on-site determination of Na sources in the wafer processing tool. The test minimized tool downtime, did not require components to be removed, and maintained the current cleanliness level (e.g. particles, metals) of the tool. Figure 14. CSW testing revealed that locations F and J had the highest levels of Na and K, while location E had the lowest metals.

11 The CSW testing method was not only capable of determining contamination sources in the wafer processing tool, but it could also be utilized to ensure that similar tools meet on-wafer metal specifications. For example, locations F and J could undergo a PM procedure to lower Na and other surface metals. CSW testing could then be performed to verify that the PM did indeed lower the metals on locations F and J. Once the reduction of metals on the surfaces is verified, the tool could then be used to process wafers to determine if the on-wafer metals are within specifications. If the wafers are found to meet specifications, then the surface metals determined through CSW testing could be used to establish a surface cleanliness specification for each of the locations that were tested. Therefore, CSW testing could be performed on similar wafer processing tools (e.g. during the startup of a new tool or after a PM) to determine if the critical locations in the tools meet the surface specifications. This could ensure that the wafers processed in similar tools also meet on-wafer specifications. Conclusion This report presented the CSW testing method and materials. The CSW technique utilizes the cleanest possible wipers, and the two-operator procedure ensures that the metallic contamination from critical surfaces is accurately determined. SEM images showed that the CSW method does not leave any particles, fibers, or visible residues on materials commonly used in wafer processing equipment, even in samples with a high average roughness. DSE ICP-MS verified that CSW testing leaves minimal metallic contamination on critical surfaces. The case study presented in this article showed that CSW testing could be performed on multiple locations in a tool to determine sources of metallic contamination, with minimal tool downtime. Once contamination sources are identified, the CSW test could then be utilized to establish a surface cleanliness specification and ensure that similar tools meet on-wafer specifications. Thus, CSW testing has been demonstrated to be capable of rapid on-site metals testing of a wide range of critical surfaces in various environmental conditions.