APPLICATION NOTE ICP-Optical Emission Spectroscopy Author: Kenneth Ong PerkinElmer, Inc. Shelton, CT Determination of Various Elements in SemiconductorGrade Phosphoric Acid by ICP-OES Introduction In the manufacture of semiconductor products, the purity of the reagents is of utmost importance since the presence of contaminants can affect the performance of the final products. Phosphoric acid is commonly used in the production of various semiconductor materials and, therefore, requires trace levels to be measured as specified in the SEMI C36-1107 Grade 3 requirements. Phosphoric acid presents an analytical challenge due to its viscosity, composition, and concentration. The combination of matrix composition and measurement levels make phosphoric acid an ideal candidate for analysis by ICP-OES. This work describes the analysis of phosphoric acid with ICP-OES to meet the SEMI requirements. Experimental Results and Discussion All analyses were carried out on a PerkinElmer Optima 8300 ICP-OES which uses flat induction plates (Flat Plate™ plasma technology) instead of the traditional helical load coil. This innovative technology delivers a plasma generation system that does not require coil cooling and is capable of operating at a plasma argon flow as low as 8 L/min. The operating conditions are listed in Table 1. ® To avoid contamination from the lab environment, samples were prepared in a clean hood, and the autosampler was covered. All sample preparation was done in PFA bottles which were pre-soaked in 5% (v/v) nitric acid for 24 hours, then filled with ultra-pure water for storage before use. In order to eliminate contamination from the pump tubing, self-aspiration was used. All measurements were made against calibration curves using the Method of Additions Calibrations, with 5 ppb, 10 ppb and 20 ppb standards. For analysis, ultra-pure grade phosphoric acid (H3PO4) was diluted ten times with deionized water. Table 1. Sample introduction system and plasma parameters Parameter Injector: Spray chamber: Cyclonic Nebulizer: Meinhard Sample tubing: Capillary Drain tubing: 1.14 mm i.d. Quartz torch: Single slot Sample capillary: PTFE 1 mm i.d. Sample vials: Polypropylene Equilibrium delay: Plasma aerosol type: RF power: 15 sec Wet 1300 W Nebulizer flow: 0.55 L/min Auxiliary flow: 0.2 L/min Plasma flow: 8 L/min Sample uptake: Self aspiration Plasma viewing: Axial Processing mode: Peak area Auto integration: 5 - 10 sec Replicates: Background correction: 2 Condition Alumina 2 mm i.d. 3 1 or 2-point Table 2 shows the elements, wavelengths, correlation coefficients, and detection limits of the method. The detection limits, determined from 25 replicate measurements of the sample, are within the SEMI 36-1107 criteria. The low detection limits are an indication of the system’s short-term stability. The elevated Pb detection limit results from the higher level of Pb in the acid. Table 2. Elements, wavelengths, correlation coefficients, and detection limits Analytes and Wavelengths (nm) Al 396.153 Correllation Coefficients 0.997 Detection Limits (ppb) 1.86 Ba 233.527 0.999 0.529 Be 313.107 0.999 0.119 Ca 317.933 0.999 0.640 Cd 228.802 0.996 0.909 Co 228.616 0.998 0.794 Cr 205.560 0.999 0.838 Cu 327.393 0.999 0.953 Fe 239.562 0.999 0.736 K 766.490 1.000 8.71 Li 670.784 0.997 0.205 Mg 285.213 0.998 0.253 Mn 257.610 0.999 0.176 Mo 202.031 0.999 1.86 Na 589.592 0.999 0.523 Ni 221.648 0.999 1.22 Pb 220.353 1.000 8.56 Sn 189.927 0.999 3.83 Ti 334.940 0.999 0.172 V 290.880 0.999 0.690 Zn 202.548 0.999 0.527 To ascertain the accuracy and robustness of the method, a spike recovery test was carried out. The H3PO4 was divided into three containers, and each analyzed. A 5 µg/L spike was then added to each container and re-analyzed. The spike recoveries, shown in Table 3 (Page 3), ranged from 90-111%, well within the SEMI requirement of 75%-125%. Table 3. Repeatability and recovery of 5 µg/L spikes Analyte and Wavelength Al Sample 1 Sample 2 Sample 3 Sample Avg Spike 1 Spike 2 Spike 3 Spike Avg Recovery 396.153 0.161 0.586 -0.039 0.236 5.58 5.07 6.09 5.59 107% Ba 233.527 1.01 0.660 0.869 0.848 5.74 5.82 5.88 5.81 99.3% Be 313.107 0.076 0.071 0.067 0.071 5.00 4.99 4.99 4.99 98.4% Ca 317.933 0.604 0.776 0.719 0.700 5.58 5.21 5.80 5.53 96.6% Cd 228.802 11.7 11.6 11.8 11.7 16.6 16.6 16.8 16.7 99.6% Co 228.616 6.11 6.06 6.11 6.09 11.1 11.1 11.2 11.1 101% Cr 205.560 5.41 5.38 5.47 5.42 10.9 10.4 10.4 10.6 103% Cu 327.393 3.08 2.87 2.56 2.84 7.77 7.76 7.69 7.74 98.1% 97.7% Fe 239.562 1.36 1.62 1.71 1.56 6.17 6.60 6.57 6.45 K 766.490 -9.72 -9.50 -9.24 -9.49 -4.14 -3.67 -3.98 -3.93 111% Li 670.784 1.31 1.25 1.16 1.24 6.15 6.10 6.11 6.12 97.6% Mg 285.213 -0.071 0.069 -0.013 -0.005 5.06 5.11 5.14 5.10 102% Mn 257.610 0.581 0.591 0.602 0.591 5.53 5.49 5.51 5.51 98.4% Mo 202.031 -1.39 -1.68 -1.56 -1.54 4.11 3.61 4.03 3.92 109% Na 589.592 0.424 0.590 0.498 0.504 5.49 5.46 5.44 5.46 99.2% Ni 221.648 -1.41 -1.06 -1.11 -1.19 3.71 3.63 3.86 3.74 98.5% Pb 220.353 21.0 21.3 23.1 21.8 26.2 25.9 26.8 26.3 89.5% Sn 189.927 -0.667 0.175 -0.088 -0.193 4.66 4.20 4.49 4.45 92.9% Ti 334.940 0.276 0.354 0.292 0.307 5.26 5.28 5.30 5.28 99.4% V 292.402 0.218 0.008 0.124 0.117 5.08 5.18 5.04 5.10 99.6% Zn 202.548 -0.695 -0.806 -0.605 -0.702 4.33 4.129 4.20 4.22 98.4% To verify the stability of the method, a 5 µg/L spike solution was aspirated continuously for an hour, with measurements being made consecutively. Figure 1 shows the resulting stability plot, where results were normalized to the average reading. The variations were generally less than ± 10%, indicating no significant drift. This result demonstrates the robustness of the Optima 8300’s Flat Plate plasma to handle a complex matrix at a low plasma gas flow. Conclusion This work has demonstrated that the Optima 8300 ICP-OES can effectively analyze phosphoric acid to meet the specifications of SEMI 36-1107. Utilizing Flat Plate plasma technology, only 8 L/min of argon gas are needed, significantly reducing operating costs. Even with low argon consumption, a robust plasma is produced, which results in signal stability in complex matrices and low detection limits. The combination of low argon consumption and robustness makes PerkinElmer’s Optima 8300 ICP-OES ideally suited for the analysis of complex samples, while minimizing operating costs. References SEMI C10-1109–Guide for Determination of Method Detection Limits SEMI C1-0310 – Guide for the Analysis of Liquid Chemicals SEMI C36-1107– Specification for Phosphoric Acid Figure 1. Stability plot for continuous aspiration of a 5 µg/L spike in H3PO4 measured over one hour. PerkinElmer, Inc. 940 Winter Street Waltham, MA 02451 USA P: (800) 762-4000 or (+1) 203-925-4602 www.perkinelmer.com For a complete listing of our global offices, visit www.perkinelmer.com/ContactUs Copyright ©2014, PerkinElmer, Inc. All rights reserved. PerkinElmer® is a registered trademark of PerkinElmer, Inc. All other trademarks are the property of their respective owners. 011820_01