Determining critical trace elements in high-purity phosphoric acid

Robert A. Aleksejczyk and Donald Gibilisco, FMC Corp.

There has been much attention paid in recent years to the quest to produce and certify high-purity process chemicals for the manufacture of integrated circuit components. Phosphoric acid is one of the many process chemicals to undergo scrutiny in this regard. Yet the 1997 Book of SEMI Standards (BOSS) does not specify a method for determining trace elements in high-purity phosphoric acid at the Tier B level.¹ Suggested analytical methods for other high-purity chemicals at this level include graphite furnace atomic absorption spectrophotometry (GFAAS) for the determination of Group 1 elements and inductively coupled plasma mass spectrometry (ICP-MS) for all others.

A more recent study has reported the applicability of ICP-MS under standard plasma conditions to determine trace element levels in high-purity 37.5% hydrochloric acid at the 1-ppb level required by SEMI Tier B guidelines. Cold plasma conditions were used only for those elements (calcium, potassium, sodium, iron) affected by argon-based molecular interferences (Ar⁺, ArH⁺, ArO⁺).² The study thus eliminated the need for GFAAS or expensive high-resolution ICP-MS for Group 1 elements such as sodium, potassium, and calcium as well as iron. The work reported in this article details a method for achieving spike recoveries at the levels required by SEMI Tier B guidelines for phosphoric acid. The procedure involves a single sample preparation using an ICP-MS equipped with a microconcentric nebulizer (MCN) under standard plasma conditions.

Element Level (ppb max)
Aluminum 50
Antimony 1000
Arsenic 50
Barium 50
Boron 50
Cadmium 50
Calcium 150
Chromium 50
Cobalt 50
Copper 50
Gold 50
Iron 100
Lead 50
Lithium 10
Magnesium 50
Manganese 50
Nickel 50
Potassium 150
Silicon 50
Sodium 250
Strontium 10
Titanium 50
Zinc 50

Table I: SEMI Tier B guideline for cations in phosphoric acid.

Experimental Procedures

The study reported here conformed to the SEMI guideline for recovery, which deems a test method valid if there is a recovery of 75—125% of the spiked element at 50% of the specified level. Twenty-three elements defined by the SEMI Tier B guideline for phosphoric acid (listed in Table I) were measured using an Elan 6000 ICP-MS (Perkin-Elmer SCIEX Instruments, Concord, Ontario, Canada). The instrument was equipped with nickel interface cones to minimize signal degradation caused by phosphoric acid corrosion and an MCN equipped with a mass-flow controller. The instrument also featured an ion optic system that provided computer-controlled ion lens tuning by automatically optimizing the lens voltage for a given element. The auto lens capability (or dynamic voltage scanning) allowed for maximum analyte ion transmission with minimum matrix interferences and was used for each element measurement. Other operating conditions are summarized in Table II.

Parameter Operating Conditions
RF power 1500 W
Nebulizer Microconcentric nebulizer
Nebulizer flow rate 1.12 L/min
Plasma argon flow 15 L/min
Auxiliary argon flow 1.20 L/min
Sample uptake rate 50 µL/min
Quadrupole scan mode Peak hopping
MCA channels/amu 1
Ion lens voltage Dynamic
Dwell time Varied (1050 msec)
Number of sweeps 200
Integration time/mass 1000 msec
Number of replicates 5
Detector mode Dual

Table II: Operating conditions for ICP-MS equipped with microconcentric nebulizer.

The calibration procedure for measuring the 23 elements in phosphoric acid using an ICP-MS consisted of a matrix-matched blank plus external standards at the Tier B level as well as at double the Tier B level. The water used for the standards and sample preparation met the quality defined as "attainable" in SEMI's Suggested Guidelines for Pure Water for Semiconductor Processing. The phosphoric acid specified in this work had an ultralow metal ion content. The external multielement standard solution, prepared with Spex certified standards, contained the trace metals aluminum, antimony, arsenic, barium, boron, cadmium, calcium, chromium, cobalt, copper, gold, iron, lead, lithium, magnesium, manganese, nickel, potassium, silicon, sodium, strontium, titanium, and zinc. For the study reported here, the blank, standards, and samples contained the following levels of internal standards: 10 ppb beryllium, 20 ppb germanium, and 5 ppb indium, lutetium, and scandium.

The external standards were prepared by weighing 5g of 85% ultralow-metal-ion phosphoric acid into the sample tubes. An aliquot of the internal and external multielement stock solutions were added to the acid sample and brought to a final volume of 10 ml, with water producing the required concentrations of internal and external standard elements. The test samples were prepared in a similar manner by weighing the phosphoric acid to be analyzed and omitting the aliquot of external standards to the sample. The blank sample was prepared similarly to the test sample but using the ultralow-metal-ion phosphoric acid.

Results and Discussion

Detection Limits. Detection limits were based on three standard deviations of the five replicates of the blank and can be found in Table III. A three-point external calibration curve (blank and two standards) was constructed to demonstrate linearity within the calibration range. Each calibration was linear in excess of an .999 correlation coefficient.

Element Spike
Concentration
(ng/L) % Recovery %RSD Detection
Limit
(ppb)
⁷Lithium 5 103 4.6 0.06
¹¹Boron 25 91 12 5.9
²³Sodium 125 107 2.6 1.3
²⁴Magnesium 25 96 1.8 0.3
²⁷Aluminum 25 95 1.8 0.6
²⁹Silicon 25 82 38 28.1
³⁹Potassium 75 86 3.1 11.0
⁴³Calcium 75 99 6.0 41
⁵⁰Titanium 25 83 5.8 9.3
⁵²Chromium 25 98 1.7 0.3
⁵⁵Manganese 25 100 1.6 0.1
⁵⁷Iron 50 88 17 22
⁵⁹Cobalt 25 99 2.0 0.02
⁶⁰Nickel 25 95 1.4 1.0
⁶⁵Copper 25 96 22 12
⁶⁶Zinc 25 84 7.3 4.3
⁷⁵Arsenic 25 96 6.5 3.7
⁸⁸Strontium 5 100 2.4 0.05
¹¹¹Cadmium 25 96 2.6 1.2
¹²¹Antimony 500 97 1.2 12
¹³⁸Barium 25 103 0.5 0.05
¹⁹⁷Gold 25 81 12 1.1
²⁰⁸Lead 25 98 0.7 0.3

Table III: Spike recovery for trace elements in 85% phosphoric acid using standard operating conditions.

Spike Recoveries. Spike recoveries were measured at concentrations consistent with the SEMI Tier B recovery guideline, i.e., 50% of the element specification, yielding the results shown in Table III. Spikes for all elements were recovered within the 75—125% range required by the SEMI Tier B guideline. In addition, all the elements were recovered within the required SEMI specifications. Silicon, iron, and copper yielded higher percent relative standard deviations (%RSD) compared to the remaining Tier B elements while meeting the required SEMI recovery guideline. It may be possible to improve the spike recoveries and %RSD for these elements by using extended measurement conditions for them.

Sample Introduction System. Previous work at our laboratory has shown that a cross-flow nebulizer can also provide similar Tier B recovery results for phosphoric acid. However, some of the elements required enhanced instrument conditions. For example, analyzing calcium, copper, iron, potassium, sodium, and silicon separately from the other SEMI elements using an extended measurement condition resulted in longer dwell time and increased sweeps per reading. The study reported here employed an MCN with the expressed purpose of analyzing the Tier B elements using a single operating condition. The instrument should be preconditioned with a blank phosphoric acid solution for several minutes prior to taking actual measurements. A new or recently cleaned ion optic is believed to require this preconditioning process to promote signal stability.

Conclusion

This study showed the successful measurement of SEMI Tier B level trace elements in a single sample preparation of phosphoric acid using ICP-MS in compliance with the 75—125% SEMI recovery guideline. This was accomplished while also demonstrating good spike recovery reproducibility. In addition, all elements can be measured from one sample preparation using a single operating condition with this procedure.

Acknowledgments

This project would not have been possible without the advice and counsel of many people, particularly Thomas J. Gluodenis, Jr., of Perkin Elmer, Wilton, CT.

References

1. Book of SEMI Standards (BOSS), Mountain View, CA, Semiconductor Equipment and Materials International, 1996.

2. Jacksier T, Gluodenis TJ Jr., and Thomas RJ, "Determining Critical Trace Elements in High-Purity Hydrochloric Acid by ICP-MS Alone," MICRO, 14(3):63—68, 1996.

Robert A. Aleksejczyk, PhD, is a senior research associate in the phosphorus chemicals division at FMC Corp., Princeton, NJ. He is responsible for applications development and process improvements for phosphoric and polyphosphoric acid products. Aleksejczyk has an MS and PhD in chemistry from Rensselaer Polytechnic Institute (Troy, NY). He also did postdoctoral studies in chemistry at Massachusetts Institute of Technology (Cambridge). He is an active participant in the SEMI Standards process chemicals division. He has authored 12 articles and technical presentations and has 5 patents. (Aleksejczyk can be reached at 609/951-3653.)

Donald Gibilisco is a senior analytical specialist in FMC's analytical technology department. As laboratory supervisor he is responsible for methods development and routine elemental/ion analysis using FLMAAS, GFAAS, ICP, and ICP-MS techniques. Gibilisco has a BA from the College of New Jersey with more than 20 years' experience in the area of trace element analysis and a patent in this area. He is an active participant in SEMI Standards' process chemicals division. (Gibilisco can be reached at 609/951-3147.)

MicroHome | Search | Current Issue | MicroArchives
Buyers Guide | Media Kit

Questions/comments about MICRO Magazine? E-mail us at cheynman@gmail.com.