Although collision-cell technology initially showed great promise, it does not effectively reduce background levels enough to improve detection limits over the cool-plasma approach for problematic elements such as iron, potassium, and calcium. It appears to work well for some types of spectral interferences, but it does not perform significantly better than traditional quadrupole instrumentation in semiconductor applications.

DRC Technology

The need for ICP-MS to provide good detection limits for semiconductor-related elements led to the development of DRC technology.^7,8 Similar in appearance to the collision cell, the dynamic reaction cell is a pressurized multipole positioned in front of the analyzer quadrupole. However, in DRC technology, a quadrupole is used instead of a hexapole or octapole.

The technology performs a series of steps. First, a reactive gas, such as ammonia, is bled into the cell, which produces many ion-molecule interactions. Several different mechanisms cause the gaseous molecules to react with the interfering ions to convert them into species that will not interfere with the analyte. The analyte mass then emerges from the dynamic reaction cell free of interferences and is steered into the analyzer quadrupole for conventional mass separation.

One benefit of using a quadrupole in the reaction cell is that the quadrupole's stability regions are much better defined than those of a hexapole or octapole, allowing for precise control of ion-molecule interactions in the cell. By carefully optimizing the quadrupole's electrical fields, unwanted reaction by-products, which can potentially lead to new interferences, are prevented. This cleansing process, which is known as chemical resolution, ensures that every time an analyte and interfering ions enter the DRC, the bandpass of the quadrupole can be optimized for the specific materials and then changed on the fly for the next analyte.

Figure 1: Schematic diagram showing how the ion-molecule chemistry in the DRC reduces the polyatomic interference of 40Ar16O on the determination of 56Fe.

The process of ion-molecule chemistry is shown schematically in Figure 1, which presents one of the most common interferences in ICP-MS: the polyatomic interference of ⁴⁰Ar¹⁶O on the determination of ⁵⁶Fe. In this example, the reaction gas (NH₃) is bled into the DRC, which contains analyte ions (⁵⁶Fe) and the ⁴⁰Ar¹⁶O interference. The NH₃ gas reacts with the ⁴⁰Ar¹⁶O interference to form new species that do not interfere with ⁵⁶Fe. Moreover, the NH₃ gas does not react significantly with ⁵⁶Fe. Under optimum gas-flow conditions, the chemical cleansing process is so thorough that the background from the ⁴⁰Ar¹⁶O is virtually eliminated.

Figure 2: Spectral scans of varying concentrations of iron. Virtually no signal intensity can be seen at mass 56 for doubly deionized water (DDIW), indicating that there is little background interference.

Cleansing efficiency is illustrated in Figure 2, which presents a spectral scan of various iron concentrations. The figure demonstrates that the signal contribution for DI-water at mass 56 is <10 counts per second, which is significantly lower than that achievable with a collision cell–based ICP-MS.⁹ If no reaction gas is used in the cell, the background would be 5 to 6 orders of magnitude higher.

The DRC approach can effectively reduce background levels for other common polyatomic spectral interferences, including ⁴⁰Ar on the determination of ⁴⁰Ca and ³⁸ArH on the determination of ³⁹K. For those elements, detection limits are <0.5 ppt.

Another benefit of DRC technology is that it uses a high-temperature plasma. Unlike a cool plasma, a high-temperature plasma can determine elements with high ionization potential without sacrificing sensitivity. Thus, many problematic interferences, such as ⁴⁰Ar⁴⁰Ar on ⁸⁰Se, ⁴⁰Ar³⁵Cl on ⁷⁵As, ³²S¹⁶O¹⁶O on ⁶⁴Zn, and ⁴⁰Ar¹²C on ⁵²Cr, can be overcome easily by using the optimum reaction gas. DRC technology can also be used without a reaction gas to determine elements that do not suffer from polyatomic spectral interferences.

Element DetectionLimit (ppt) BEC(ppt) Li 0.26 0.22 As 0.48 1.60 B 1.93 1.50 Na 0.14 0.22 Mg 0.08 0.18 Al 0.05 0.09 K 0.27 2.60 Ca 0.10 0.10 Ti 0.92 1.70 V 0.12 0.04 Cr 0.12 0.12 Mn 0.17 0.54 Fe 0.12 0.40 Ni 0.10 0.20 Co 0.04 0.04 Cu 0.05 0.10 Zn 0.45 1.20 Sn 0.12 0.88 Sb 0.08 0.08 Ba 0.06 0.04 Pb 0.07 0.09

Table II: Detection limits and BEC achieved with the DRC ICP-MS, generated in multielement mode using 1-second integration time. (The data for the elements in red were obtained in DRC mode using NH3 reaction gas, while the elements in black were obtained in standard mode.)

Table II lists multielement detection limits and background equivalent concentrations (BEC) achieved with DRC technology for the 21 elements considered most important by SEMI. BEC is defined as the apparent concentration for the background signal based on the sensitivity of the element at a specified mass. The lower the BEC value, the more easily a signal generated by an element can be discerned from the background. Many experts in the semiconductor industry believe that BEC is a more accurate indicator of ICP-MS performance than detection limit.

Applying DRC Technology to High-Purity Chemicals

While DRC technology can help chemical manufacturers to achieve SEMI's next-generation purity standards because of its ultralow detection limits for difficult elements, it can also analyze complex semiconductor chemicals such as concentrated acids and organic solvents with very few or no sample-preparation steps.^10,11

For example, before the use of DRC technology, the determination of chromium in the organic solvent isopropyl alcohol (IPA) was problematic, because an elevated baseline caused by the polyatomic species ⁴⁰Ar¹²C interfered with the major isotope of chromium at mass 52, producing very questionable data. To reduce ⁴⁰Ar¹²C interference, the sample had to be evaporated to dryness to remove the organic matrix and then redissolved in nitric acid to eliminate carbon-based polyatomic spectral interferences. By using DRC technology to minimize ⁴⁰Ar¹²C polyatomic interference, ⁵²Cr can be determined successfully in the neat solvent without the sample having to undergo the lengthy pretreatment steps.

Element DetectionLimit (ppt) BEC(ppt) 10-ppt Spike Recovery (%) 23Na 3.90 32.00 101 24Mg 2.30 2.40 90 27Al 5.90 36.00 105 39K 1.70 49.00 106 40Ca 0.83 4.90 98 48Ti 5.60 9.60 100 51V 1.90 3.20 96 52Cr 3.60 19.00 109 55Mn 1.70 4.10 99 56Fe 5.80 21.00 105 59Co 0.31 0.69 91 60Ni 3.50 13.00 101 63Cu 3.60 43.00 91 64Zn 2.90 15.00 98 75As 6.10 15.00 94 114Cd 1.70 1.20 100 120Sn 1.20 3.60 106 138Ba 0.70 1.30 103 208Pb 0.15 0.72 91

Table III: Detection limits, BEC, and spike recoveries achievable in neat IPA, generated in multielement mode. (Data for elements in red were obtained in DRC mode.)

Another element that has benefited from the use of DRC technology is arsenic. The high-power, high-temperature plasma used in DRC technology produces favorable ionization conditions that result in a far lower arsenic detection limit than the cool plasma used in standard ICP-MS instrumentation. Table III lists detection limits, BEC, and 10-ppt spike recoveries for a group of critical semiconductor elements determined in neat IPA. The analysis was carried out using a high RF power of 1600 W and a concentric quartz nebulizer with narrow sample tubing to reduce the flow of organic solvent entering the mass spectrometer. In addition, a small amount of oxygen was bled into the nebulizer gas flow to minimize the build-up of carbon deposits on the interface cones. This procedure enabled the neat IPA to be aspirated into the instrument and minimized the negative effect of carbon-based spectral interferences.

Another problematic chemical is highly corrosive concentrated sulfuric acid (97% weight/vol). Because its extremely high sulfur concentration causes many sulfur-based polyatomic spectral interferences, such as ³²S¹⁶O, ³⁴S¹⁶OH, ³⁴S¹⁸O, and ³²S³²S, the determination of elements such as ⁴⁸Ti, ⁵¹V, ⁵²Cr, and ⁶⁴Zn is virtually impossible using traditional ICP-MS. A low-temperature cool plasma cannot be used to analyze 97% sulfuric acid because even with dilution, the high sulfuric acid content causes severe matrix suppression on the analyte masses. Therefore, before trace elements in sulfuric acid can be determined using conventional ICP-MS, the sample typically must be evaporated to dryness and then redissolved in dilute nitric acid. However, besides being very time-consuming, the evaporation and redissolving procedure compromises the elements' detection limits in the original concentrated sample.

Using DRC technology, the sulfuric acid matrix has virtually no effect on the determination of the trace elements, because harmful sulfur-based spectral interferences can be reduced significantly by careful optimization of the reaction cell's chemical resolution parameters. In addition, because a high-temperature plasma is used with DRC technology, ionization conditions are more suited to determining the full suite of elements in the sulfuric acid matrix.

Figure 3: Comparison between standard ICP-MS and DRC calibration curves for the determination of 51V in sulfuric acid.

Figure 3 illustrates the effect of ³⁴S¹⁶OH interference on the calibration of ⁵¹V in 10% sulfuric acid. Figure 3a shows a calibration curve generated using a standard ICP-MS process for 0–200 ppt of ⁵¹V, while Figure 3b shows a calibration curve generated by DRC technology for 0–200 ppt of ⁵¹V with 1 ml/min of NH₃ gas in the reaction cell. In standard ICP-MS mode, the polyatomic ion ³⁴S¹⁶OH elevated the background at mass 51 to more than 60,000 counts per second, making the determination of ultralow levels of vanadium virtually impossible. When the ³⁴S¹⁶OH was removed using the DRC's chemical-resolution capability, the background at mass 51 was reduced to about 30 counts per second.

Element DetectionLimit (ppt) BEC(ppt) 50-ppt Spike Recovery (%) 11B 5.50 6.40 106 23Na 0.31 0.42 99 24Mg 0.24 0.42 96 27Al 0.73 0.80 95 39K 2.70 35.00 100 40Ca 0.67 1.60 100 51V 0.16 0.13 99 52Cr 0.76 1.30 103 55Mn 0.13 1.10 99 56Fe 1.40 4.00 103 59Co 0.67 0.82 100 60Ni 0.52 3.50 104 63Cu 1.10 2.10 98 74Ge 2.30 1.50 99 75As 1.30 5.30 98 115Zn(NH3)3 16.00 14.00 94 120Sn 0.57 1.80 94 121Sb 0.20 0.40 99 208Pb 0.18 0.24 100

Table IV: Detection limits, BEC, and spike recoveries for a group of elements in 10% sulfuric acid, generated in multielement mode. (Data for elements in red were obtained in DRC mode.)

These calibration curve data translate into a detection limit of 0.7 ppt for ⁵¹V in 10% sulfuric acid, as confirmed in Table IV, which shows detection limits, BEC, and 50-ppt spike recoveries for a group of elements. Although zinc can be determined in standard ICP-MS mode using one of the zinc isotopes at 64, 66, 67, 68, or 70 amu, its detection limit is severely compromised because of the spectral background produced by sulfur-based polyatomic ions. In contrast, the DRC process forms a brand new molecular ion complex between the zinc and the NH₃ gas. By optimizing the reaction cell parameters, ⁶⁴Zn and ¹⁴NH₃ can be made to react to generate Zn(NH₃)₃ ions at mass 115, which is then used for quantitation.

Conclusion

As the IC industry moves toward <200-nm lithography, all semiconductor-related materials will have to achieve lower trace element contamination levels than is currently the case. By investing in cutting-edge technologies such as DRC ICP-MS, chemical suppliers will acquire the analytical capabilities necessary to meet and exceed changing industry demands. In experiments to determine ultratrace contamination levels in propanol and sulfuric acid, it has been demonstrated that DRC technology has the potential to analyze difficult semiconductor materials and keep pace with increasingly stringent industry requirements.

References

1. The International Technology Roadmap for Semiconductors (San Jose: SIA, 2001); available from Internet: http://public. itrs.net.

2. Book of SEMI Standards (BOSS) (San Jose: SEMI, updated periodically).

3. SD Tanner et al., "Use of a Novel Reaction Cell to Reduce Polyatomic Spectral Interferences by ICP-MS," in Proceedings of the 17th Annual Semiconductor Pure Water and Chemicals Conference (Santa Clara, CA: SPWCC, 1998).

4. SJ Jiang, RS Houk, and MA Stevens, "Alleviation of Overlap Interferences for Determination of Potassium Isotope Ratios by Inductively Coupled Plasma Mass Spectrometry," Analytical Chemistry 60, no. 11 (1988): 1217–1221.

5. SD Tanner et al., "The Application of Cold Plasma Conditions for the Determination of Trace Levels of Fe, Ca, K, Na, and Li by ICP-MS," Atomic Spectroscopy 16, no. 1 (1995): 16–18.

6. P Turner et al., "Plasma Source Mass Spectrometry: Developments and Applications," in Proceedings of the 5th International Conference on Plasma Source Mass Spectrometry (Cambridge, UK: Royal Society of Chemistry, 1996), 28–34.

7. SD Tanner and VI Baranov, "Theory, Design, and Operation of a Dynamic Reaction Cell for ICP-MS," Atomic Spectroscopy 20, no. 2 (1999): 45–52.

8. U.S. Pat. 6,140,638.

9. I Feldman et al., "Application of a Hexapole Collision and Reaction Cell in ICP-MS. Part 2. Analytical Figures of Merit and First Applications," Fresnius Journal of Analytical Chemistry 365 (1999): 422–428.

10. J-M Collard et al., "Implementing Enhanced ICP-MS Technology to Attain SEMI Grade 5 Purity Levels," MICRO 20, no. 1 (2002): 39–46.

11. DS Bollinger and AJ Schleisman, "Analysis of High Purity Acids Using a Dynamic Reaction Cell ICP-MS," Atomic Spectroscopy 20, no. 2 (1999): 60–63.

Chia Mui Ping is laboratory manager at Merck Singapore, where she leads a team of chemists and technologists who provide chemical management services to the IC industry. She develops new methods and standard operating procedures for her laboratory, and coordinates the company's production of ultrapure chemicals. Before her present position, Ping worked in the materials and process laboratory at Seagate Technology, where she was involved with the microcontamination and failure analysis of hard-disk drives. She received a BS in chemistry from the National University of Singapore. (Ping can be reached at +65 68631939 or mercklab@pacific.net.sg.)

 

Yoko Kishi is the ICP-MS product specialist for the semiconductor business unit of PerkinElmer Instruments (Shelton, CT) and is responsible for all semiconductor applications development using DRC technology. She joined the company in 2000 and is based at the SCIEX facility in Toronto. For 14 years, she held various ICP-MS applications and marketing positions at Yokogawa Analytical Systems and has extensive knowledge of the electronics industry. She received a BA in analytical chemistry from Keio University in Tokyo. (Kishi can be reached at 905/660-9005 or yoko.kishi@perkinelmer.com.)

Katsu Kawabata is the semiconductor business development specialist for PerkinElmer Instruments, where he is responsible for developing DRC ICP-MS business in the electronics industry. He joined the company in 2000 and is based at the SCIEX facility in Toronto. Before joining PerkinElmer, he served as the ICP-MS R&D applications manager for Yokogawa Analytical Systems, where he worked on the shield torch system and its application to the analysis of semiconductor-related materials. He received a chemistry degree from Kobe Technical College, Japan. (Kawabata can be reached at 905/660-9005 or katsu.kawabata@perkinelmer.com.)

 

Robert Thomas is principal of Scientific Solutions (Gaithersburg, MD), a consulting company that serves the technical writing and marketing needs of the scientific community. He has worked in the field of trace-element analysis for more than 30 years and has more than 15 years of experience in ICP-MS technology. He has published more than 40 technical articles, papers, and publications covering a wide variety of scientific topics. He received a degree in analytical chemistry from Gwent College in Wales, UK. (Thomas can be reached at 301/570-2811 or thomasrj@bellatlantic.net.)