Determination of Trace Anions in Concentrated Hydrofluoric Acid

Transcription

1 Technical Note Determination of Trace Anions in Concentrated Hydrofluoric Acid INTRODUCTION There is a need for a reliable method to determine trace anions in hydrofluoric acid. The large excess of fluoride ion hampers this analysis. To determine chloride at µg/l (ppb) in.% (v/v) hydrofluoric acid represents a concentration ratio of : (chloride to fluoride). Diluting the concentrated sample overcomes the problem of the large concentration of the interfering matrix ion but results in a lack of sensitivity for the contaminant ions of interest. An improved method for determining trace anions in concentrated hydrofluoric acid has been developed to overcome this problem., Trace inorganic anions of strong acids are separated from the high concentration of fluoride by an ion exclusion separation prior to an ion chromatographic separation. This Technical Note describes the theory, set up, and analytical procedure for the determination of chloride, nitrate, sulfate, and phosphate at sub mg/l (ppm) levels in.% (v/v) hydrofluoric acid. SUMMARY OF THE METHOD An IonPac ICE-AS ion exclusion column is used on-line for sample pretreatment to separate the analyte ions from an excess of fluoride matrix ions. A selected fraction from the ion exclusion separation is cut and sent to a -mm IonPac AS9-HC, anion exchange concentrator column. The concentrated ions are then eluted onto a -mm IonPac AS9-HC column where the anions of interest are separated and detected by suppressed conductivity. EQUIPMENT DX- Ion Chromatography System consisting of: GP Gradient pump, microbore configuration CD Conductivity detector with a temperature controlled conductivity cell (DS) LC Enclosure, () Rheodyne valves, PEEK, rear loading Dionex RP- single piston pump Pressurizable Reservoir Chamber (P/N 7) AC Power Control Accessory CAM (Controlled Air Module) Low Pressure -Way, - fittings (P/N ) () Plastic bottle assemblies, L, [() for external water, () for rinse solution] (P/N 9) () O-ring, Teflon encapsulated (for rinse solution bottle) (P/N ) () O-rings Teflon encapsulated for reservoir chamber (P/N 7) () Air pressure gauge to psi (7 kpa) (for external water) cm ( in.) of green PEEK tubing, diameter of.7- mm (. in.) to connect columns and make a 7-µL sample loop Teflon bottles as sample containers (VWR P/N 7- or Nalge P/N - for ml narrow mouth bottles) Disposable chemical resistant gloves for hydrofluoric acid (Lab Safety P/N 8A-79) PeakNet Chromatography Workstation Technical Note

2 COLUMNS IonPac AG9-HC guard column, mm (P/N 8) IonPac AS9-HC analytical column, mm (P/N ) IonPac AG9-HC concentrator column, mm (P/N 79) IonPac AG as trap column, mm (P/N 9) IonPac ICE-AS pretreatment column (P/N ) Anion Self Regenerating Suppressor (ASRS -ULTRA), mm (P/N 97) REAGENTS AND STANDARDS Deionized water (DI H O) Type I reagent grade, 7.8 MΩ-cm resistance or better. Sodium hydroxide % (w/w) aqueous solution (Fisher Scientific). M Carbonate Anion Eluent Concentrate (Dionex P/N 7) Chloride standard mg/l, ml (Dionex P/N 79) Sulfate standard mg/l, ml (Dionex P/N 7) Nitrate standard mg/l, ml (Ultra Scientific, VWR P/N ULICC-) Phosphate standard mg/l, ml (Ultra Scientific, VWR P/N ULICC-) CONDITIONS: Ion Exclusion : Trap column: Eluent: Flow rate: IonPac ICE-AS IonPac AG, mm Deionized water. ml/min Ion Chromatography Analytical column: IonPac AS9-HC, mm Guard column: IonPac AG9-HC, mm Concentrator column: IonPac AG9-HC, mm Eluent: 8 mm Sodium carbonate. mm Sodium hydroxide Flow rate:. ml/min volume: 7 µl Detection: Suppressed conductivity, ASRS, AutoSuppression external water mode Suppresor Current setting: Expected system Backpressure: ma Expected Background Conductivity:.8 MPa ( psi) (with concentrator column in line) PREPARATION OF SOLUTIONS AND REAGENTS Eluent Solutions 8. mm Sodium Carbonate /. mm Sodium Hydroxide Add.8 g of. M sodium carbonate and. g of % sodium hydroxide directly to 9 g degassed, deionized water (having a specific resistance of 7.8 MΩ-cm or greater) in an eluent bottle. Dilute to g. Transfer this solution to an eluent container and vacuum degas for minutes. mm Sodium Hydroxide (AG trap column regeneration) preparation: Dilute. g of % (w/w) sodium hydroxide with degassed, deionized water (having a specific resistance of 7.8 MΩ-cm or greater) to a final weight of g in an eluent bottle. Avoid the introduction of carbon dioxide from the air into the aliquot of % (w/w) sodium hydroxide and the deionized water being used to make the eluent. Preparation Pour ml of deionized water into a clean Teflon sample container. While in a fume hood, very slowly and carefully add ml (.8 g) of 9% concentrated hydrofluoric acid to the deionized water. Gently mix the acid with the water. Always add acid to water not water to acid. Allow the solution to cool to room temperature. Caution: Use extreme care when handling hydrofluoric acid. Avoid all contact with exposed skin. The use of eye protection, proper gloves, and a fume hood is an absolute necessity. Consult the Material Safety Data Sheet (MSDS) for more specific details about protective equipment reactivity and health effects. STANDARD SOLUTIONS Stock standard solution ( mg/l) Use Dionex or commercially available mg/l ion standard solution. Determination of Trace Anions in Concentrated Hydrofluoric Acid

3 Working standard solution ( µg/l) To prepare a mixed working standard solution, combine. ml of each anion stock solution with deionized water and dilute to a final volume of ml. CALIBRATION Prepare at least three different calibration standards at different concentrations of the working standard. Select a range similar to the expected concentrations in the samples. The method of standard additions (adding one or more increments of a standard solution to sample aliquots of the same size) should be used to minimize the effect of the concentrated acid matrix on the measured conductivity of the analytes of interest. Calibration standards are added to the deionized water used to dilute the concentrated hydrofluoric acid from 9% (w/w) to.% (v/v). The following formula can be used to calculate concentrations for dilutions: (Conc. of stock solution, mg/l) (Vol. of stock solution, ml) = (Conc. of dilute standard, mg/l) (Vol. of dilute standard, ml) The procedure for preparing standards in ml of.% (v/v) HF is as follows:. Prepare aqueous standards that are twice as concentrated (X) as desired by diluting µg/l anion stock standard solutions using the volumes listed in Table.. Carefully add. ml 9% (w/w) HF to. ml with each of the aqueous standards. This will generate.% (v/v) HF standards containing,, and µg/l anions. Table Method of Standard Additions for a Final Volume of.-ml Aliquots of.% (v/v) HF IONPAC AG TRAP COLUMN REGENERATION The AG trap column must first be regenerated. Monitoring the blank will indicate when regeneration is necessary. Typically, monthly regeneration is necessary, but it will depend upon the quality of the deionized water and usage rate of the instrument. creased contamination in the water blank indicates that the IonPac AG needs to be regenerated. The procedure is as follows:. Pump mm sodium hydroxide through the AG at. ml/min for minutes.. Follow with a rinse of deionized water at the same flow rate for minutes. DISCUSSION OF THE METHOD This method addresses the challenge of determining trace concentrations of contaminant ions such as chloride, nitrate, sulfate, and phosphate in a matrix composed of a high concentration of fluoride ion. This is accomplished in two steps: an ion exclusion (ICE) preseparation followed by injecting a portion of the ICE separation to an ion chromatographic (IC) separation. The ion exclusion mechanism separates ionized species from nonionized or weakly ionized species. This occurs because of a negatively charged hydration shell on the stationary phase surface called the Donnan membrane. Figure illustrates the application of this ICE mechanism to the separation of mg/l of chloride from mg/l : IonPac ICE-AS Eluent: Deionized water Flow Rate:. ml/min Volume: 7 µl Detection: Conductivity Peaks:. Chloride mg/l. Fluoride Concentration of X anion stock (µg/l) Volume of µg/l anion stock (ml) Aqueous Volume (ml) To To Concentrator To 9 Figure. Ion exclusion separation of mg/l chloride from mg/l fluoride. Technical Note

4 fluoride using an ICE-AS ion exclusion column. This chromatogram is a measurement only of the unsuppressed conductivity response for the ICE separation. The strong acid ions, such as chloride and sulfate, are excluded and elute as a small peak at approximately nine minutes. The weakly ionized fluoride matrix ions are retained and elute as a large peak. A series of schematics (Figures ) illustrates the operation of the chromatographic hardware. The concentrated hydrofluoric acid sample is loaded via a pressurized reservoir into the 7-µL sample loop (Figure ). Helium or nitrogen at. kpa ( psi) is used to push sample from the sample container into the sample loop at a flow rate of ~ ml/min (.8 g/min). This technique ensures that a representative sample of the concentrated hydrofluoric acid sample is loaded into the sample loop. It is important to pass at least four loop volumes through the sample loop to ensure reproducible sampling. The concentrated hydrofluoric acid in the sample loop is then delivered with the high purity water carrier stream to the IonPac ICE-AS. An AG is placed after the RP- pump to act as an anion trap column for the deionized water. Any contaminants present in this water will impact the quality of the blank. The first portion of the ICE separation from. to 7. min is sent to waste, (Figure ). Then, the concentrator column is placed in line with the ICE column and the portion from 7. to. min is captured on the concentrator column (Figure ). After. minutes the -mm AG-9HC concentrator column is placed in-line with the -mm AS9-HC analytical column set and the concentrated ions are separated (Figure ). This time window should ensure the preconcentration of all the chloride, nitrate, sulfate, and phosphate with a minimal amount of fluoride. The IC separation uses an AS9-HC with an isocratic eluent of 8 mm sodium carbonate. mm sodium hydroxide. The high capacity of the AS9-HC column allows injection of these concentrated hydrofluoric acid samples without overloading. Figure shows a separation of the common anions in deionized water with the -mm IonPac AS9-HC column. These chromatographic conditions enable the separation of chloride from fluoride as well as carbonate. A -mm microbore analytical column set was chosen because it has a four fold increase in mass sensitivity over the standard bore column. This facilitates faster loop loading because smaller sample volumes are required. The microbore format also offers low eluent consumption as well as less waste generation. An IonPac AG9-HC ion exhange column was used as the concentrator column in the x mm format instead of x mm because the -mm column had four times more capacity than the -mm column and lower back pressure at the microbore flow rate. No significant degradation in separation efficiency was observed when coupling a -mm concentrator column with a -mm analytical column set. Pump DI H O AG Trap Eluent Loading Helium Out Helium Loop Reservoir ICE-AS Concentrator AG9-HC, mm jection AG9-HC AS9-HC, mm Analytical s 9 Figure. Loading the sample loop. Determination of Trace Anions in Concentrated Hydrofluoric Acid

5 Pump DI H O AG Trap Eluent Loading Helium Out Helium Loop Reservoir ICE-AS Concentrator jection AG9-HC AS9-HC, mm Analytical s AG9-HC, mm 97 Figure. First stage of ICE separation (time. 7. min). Pump DI H O AG Trap Eluent Loading Helium Out Helium Loop Reservoir ICE-AS Load Concentrator AG9-HC, mm jection AG9-HC AS9-HC, mm Analytical s 98 Figure. Concentrating the "cut" portion from the ICE separation (time 7.. min). Pump DI H O AG Trap Eluent Loading Helium Out Helium Loop Reservoir ICE-AS Concentrator jection AG9-HC AS9-HC, mm Analytical s AG9-HC, mm 99 Figure. Separating the retained ions. Technical Note

6 During the IC separation, the deionized water rinses the ICE-AS and associated tubing to ensure that there is no contamination from the previous sample. To prevent sample depletion, the PeakNet method stops the flow of helium pressure at.9 minutes to the reagent reservoir after enough sample has been loaded. SYSTEM PREPARATION AND TEST Refer to the system configuration schematics in Figures - and Table, that summarize the types and lengths of tubing required for system configuration. The chromatographic hardware is divided into two parts: the ion exclusion pretreatment portion with the ICE-AS and the IC analysis portion with the AS9-HC. IC system. Prepare the ASRS by following the Quickstart instructions (Dionex Document 8-) included with the structions and Troubleshooting Guide for the ASRS.. stall the -mm IonPac AG9-HC and AS9-HC column set using the red.-mm (. in.) tubing. To minimize dead volume, use the smallest possible lengths of tubing and ensure that the ends of the tubing are smooth and level.. Construct a -µl sample loop by cutting a 9.9-cm (.9 in.) portion of the black.-mm (. in.) PEEK tubing.. stall this sample loop between ports and of the injection valve in the IC analysis system.. stall the ASRS and configure it in the external water mode as described in the ASRS manual.. Establish eluent flow through the -mm IonPac AG9-HC and AS9-HC analytical column set. The expected background conductivity is ~. (Note: for trace analysis it will take at least five hours for the system to achieve a stable background conductivity.) 7. Verify proper operation of the IC portion of the system by injecting a low level ppm standard to replicate the column test chromatogram. 8. Remove the -µl sample loop and install the -mm IonPac AG9-HC column. Make sure that the arrow indicating flow on the column is pointed from port to port and that the tubing length connecting the outlet of this column and port is as short as possible. 9. Configure the IC valve so that the -mm IonPac AG9-HC column is in-line with the -mm IonPac AG9-HC and AS9-HC analytical column set. Check for leaks. The expected system back pressure for these three columns at. ml/min is ~.8 MPa (, psi). Ion Exclusion Pretreatment System This section describes the preparation of the pretreatment portion of the system. It is important that the same type and length of tubing as described in Table be used to Table Details of Tubing Configuration for Trace Anions in Hydrofluoric Acid Connection Tubing Length Remarks Points Description (cm) ICE outlet => Port Green.7-mm (. in.) ICE inlet => Port Green.7-mm 7 (. in.) Port => Port Green.7-mm 7-µL (. in.) sample loop AG9-HC, mm => Port Red.-mm Should be as (. in.) short as possible AG9-HC, mm => Port Black.-mm (. in.) Port => Analytical Red.-mm Should be as (. in.) short as possible Pretreatment : IonPac ICE-AS ICE Eluent: Deionized water ICE Flow Rate:. ml/min Volume: 7 µl : IonPac AG9-HC, AS9-HC, -mm Concentrator: IonPac AG9-HC, mm Eluent: 8. mm Sodium carbonate. mm Sodium hydroxide IC Flow Rate:. ml/min Detection; Suppressed conductivity, ASRS-ULTRA, AutoSuppression, external water mode Peaks:. Fluoride µg/l (ppb). Chloride. Carbonate. Nitrate. Unidentified. Sulfate Phosphate 7 Figure. Trace anion standard in deionized water. Determination of Trace Anions in Concentrated Hydrofluoric Acid

7 successfully perform this analysis. Changes in tubing length will result in a different cut from the ICE-AS column being delivered to the AS9-HC concentrator column.. Cut a -cm ( in.) portion of the.7-mm (. in.) i.d. green PEEK tubing to make a 7-µL sample loop and install this loop between ports and of the sample valve.. Prepare the AG trap column according to the directions in the section entitled IonPac AG Trap Regeneration. (Caution: Before the AG is installed in the system it is important that the sodium hydroxide solution used for storage or cleaning be completely rinsed away. The ICE-AS is not compatible with sodium hydroxide eluents.). The entire pathway from the RP- pump to port of the IC valve is plumbed with the green PEEK.7-mm (. in.) tubing. stall the AG at the exit port of the RP- pump.. stall the ICE-AS column using a -cm piece of the green PEEK.7-mm (. in.) tubing between the outlet of the ICE-AS and port of the injection valve. Use a 7-cm portion of green tubing between port of the sample valve and the inlet of the ICE-AS.. Check to see that there is about. kpa ( psi) head pressure on the incoming deionized water that feeds the RP- pump.. Connect the exit port of the reagent reservoir to port of the sample valve with.7-mm (. in.) green tubing. 7. Configure a waste line from port of the sample valve with the green PEEK tubing. 8. Connect the reagent reservoir to helium pressure of about. kpa ( psi). 9. Begin with a container filled with deionized water as a sample to rinse the sample lines of any trace contamination.. Set a flow rate of. ±. ml/min for the RP- pump by adjusting the dial on the pump. This should be measured with the -mm AG9-HC concentrator column out of line. Measure the flow rate by collecting the waste coming out of port of the IC valve. It is critical for the success of this method that this flow rate be consistent.. Pump deionized water at. ml/min through the ICE-AS to waste without the -mm AG9-HC in line for h. This will remove the. mm heptafluorobutyric acid storage solution. It is typical with this method to detect trace amounts of sulfate in the deionized water blank.. The ICE-AS can be further conditioned by rinsing it with mm HF for h at. ml/min followed by a -h rinse with 7.8 MΩ-cm deionized water. This will reduce the sulfate blank to µg/l or less. Continue to monitor the blank, especially when starting up the system after it has been idle for more than two days.. Connect the wires between the GP and the AC according to the wiring diagram in Figure 7. This wiring scheme will ensure that in the event of a power outage, that the units will be in the OFF position when power is restored.. Plug the RP- pump into outlet of the AC and the Controlled Air Module (CAM) into outlet.. Connect the color-coded air lines of the CAM (see Figure 7). The CAM directs 89 kpa ( psi) air pressure to the low pressure slider valve to regulate the flow of. kpa ( psi) helium to the reagent reservoir. a. Red tubing to a source of air pressure b. Yellow tubing to the barbed fitting at the top of the slider valve (orientation groove marks the top of the valve) c. Green tubing to the barbed fitting at the bottom of the slider valve. Configure the valve ports on the slider valve as follows: a. Port - Helium Out to the reagent reservoir b. Port - Helium. kpa ( psi) c. Port - Closed d. Port - Open 7. Confirm that by actuating Relay that the RP- pump can be turned on and off. Likewise, confirm that by actuating Relay that the flow of helium to the reagent reservoir can be turned on and off. 8. To reduce the sulfate blank continue pumping deionized water through the ICE-AS overnight. Technical Note 7

8 GP CPU/RLY PCB Rly Outputs Rly TTL TTL TTL puts TTL AC TTL put Relay controls AC TTL Relay controls AC TTL RP- Pump AC Outlet CAM Compressed Air Out { RED YELLOW GREEN Slider Compressed Air Open Barbed Fitting Top of Closed To Reservoir Helium TTL TTL + + Barbed Fitting Bottom of Loading Figure 7. Wiring diagram for AC and CAM. SYSTEM OPERATION After all aspects of the instrumentation have been prepared and the system is ready for sample analysis.. Load the PeakNet method shown in Table.. Fill the 7-µL sample loop with deionized water. Use helium gas pressure to push sample using the reagent reservoir (Figure ).. Analyze a blank by loading deionized water as the sample. It may take several runs until the system has been rinsed of contamination.. After an acceptable blank has been established the system is ready for analysis.. Caution: Exercise extreme care when handling concentrated hydrofluoric acid. Consult the applicable Material Safety Data Sheet (MSDS) for specific details about protective equipment, reactivity, storage, disposal and health effects.. Concentrated hydrofluoric acid, 9.% (w/w) is diluted to.% (v/v) with the procedure described in the Preparation section. Table PeakNet Method for the Analysis of Concentrated Hydrofluoric Acid Total Time (min) ICE Time (min) IC Time (min) jection Relay %A Figure Comments it ject A Open. ject A Closed Begin loading the sample loop.9 ject A Open End loading the sample loop.. ject B Open Begin ICE separation. 7. Load B Open Send cut portion from ICE separation to Concentrator column... ject B Open Begin IC separation. Concentrator column in-line.. ject B Open End IC Separation A = ject, B = Load 8 Determination of Trace Anions in Concentrated Hydrofluoric Acid

9 7. The diluted.% (v/v) hydrofluoric acid can be loaded directly into the 7-µL loop by the reagent delivery module as discussed earlier. Ensure that the loop has had enough sample pass through by collecting the liquid that exits port of the sample pretreatment valve. A good practice is to load at least four loop volumes. The method is set up so that the reagent reservoir pushes sample into the sample loop prior to the IC separation. By applying helium pressure at. kpa ( psi) for.9 minutes ~ ml will have been passed through the sample loop. 8. Ensure that the RP- pump is delivering at a consistent rate of. ±. ml/min. Figure 8 illustrates what happens when the flow rate is faster or slower. At the slower flow rate, not enough of the sample is cut from the ICE separation, resulting in lower recovery of the analytes. Changing the flow rate from. ml/min to. ml/min in this example significantly reduced chloride response. At the faster flow rate, however, more of the ICE separation was cut, resulting in a less separation between chloride and carbonate. 9. Other factors will also affect the quality and consistency of the ICE preseparation. Changing the cut time window (7. to. minutes) specified in the method will impact the amount of analyte and matrix ions that are delivered to the concentrator column. Varying the sample volume will also affect the character of the ICE separation. Method development will be needed to ascertain the impact of any changes from the specified method.. Quantitating the levels of anions in hydrofluoric acid is best accomplished by the method of standard additions. This involves adding one or more increments of a standard solution to sample aliquots of the same size. (See Calibration section.) RESULTS AND DISCUSSION: A blank was determined by using deionized water as a sample. This is used to establish a background level of contamination present from the chromatographic pathway. The only anion present in significant concentration from the initial series of deionized water blanks was sulfate at approximately µg/l, quantified based on sulfate standards in deionized water. A rinse of the column with deionized water followed by replicate injections of concentrated.% HF brought the blank down to µg/l sulfate as shown in Figure 9. This deionized water blank was subtracted from the levels found in the concentrated hydrofluoric acid samples. A chromatogram for the analysis of trace anions in semiconductor grade.% (v/v) hydrofluoric acid is shown in Figure. The large fluoride peak beginning at minutes is well separated from chloride. Using a 8. mm carbonate /. mm sodium hydroxide eluent allows chloride to be well resolved from carbonate. Using an eluent gradient on a hydroxide selective column such as the IonPac AS or AS could have enhanced this separation between fluoride and chloride. The drawback of this approach is that the analysis time increases to 7 minutes. Thus the use of the AS9-HC represents the best compromise in separation efficiency and analysis time. To verify proper quantification of analytes in the hydrofluoric acid matrix, increasing concentrations of chloride, nitrate, phosphate and sulfate were added into the deionized water used to dilute the 9% hydrofluoric acid to.%. Spikes of,, and µg/l of chloride, nitrate, and phosphate and,, and µg/l of sulfate yielded coefficients of determination (r ) values of greater than.99. Based on this calibration curve, values of µg/l nitrate, µg/l phosphate, µg/l chloride, and µg/l sulfate were obtained. These values are spikes at % of the specified levels for the Semiconductor Equipment and Materials ternational (SEMI) guidelines established for the best grade of concentrated hydrofluoric acid. Recoveries were found to be in the target range specified by SEMI between 7% and %, as shown in Table. To determine precision for the method, a sample of semiconductor purity grade hydrofluoric acid diluted to.% (v/v) was analyzed by this method. For n = 7 replicate injections of the same sample, a relative standard deviation (RSD) of less than % was obtained for an average value of 8. µg/l chloride,.9 µg/l nitrate and. µg/l phosphate. Sulfate levels were determined to Table Spike Recovery of Trace Anions in.% (v/v) Hydrofluoric Acid Anion * Spike Found-blank Recovery (µg.l ± S.D.) (µg/l) µg/l ± S.D.) (%) Chloride 8. ±.7 ±. 8 Nitrate.9 ±.9.7 ±. 87 Sulfate. ±..9 ±.9 Phosphate. ±..7 ±.9 87 * Corrected for system blank n = 7 Technical Note 9

10 A Peaks:. Fluoride. Chloride. Carbonate. Sulfate. ml/min Pretreatment : IonPac ICE-AS ICE Eluent: Deionized water ICE Flow Rate:. ml/min Volume: 7 µl : IonPac AG9-HC, AS9-HC, -mm Concentrator: IonPac AG9-HC, mm IC Eluent: 8. mm Sodium carbonate. mm Sodium hydroxide IC Flow Rate:. ml/min Detection; Suppressed conductivity, ASRS-ULTRA, AutoSuppression, External water mode Peaks:. Fluoride µg/l (ppb). Carbonate. Unidentified. Sulfate B. ml/min Figure 9. Representative system blank. C. ml/min have an average concentration of. µg/l with an RSD of.7. This higher RSD is due to the level of sulfate present in the blank. Method detection limits (MDLs) were calculated using the standard deviation of the anions of interest found in seven replicate injections of unspiked semiconductor grade.% HF multiplied by the Student s t value for the 99.% confidence level. MDLs for chloride, sulfate, and nitrate ranged from. to. µg/l (ppb). The MDLs for this method compare favorably with the maximum limit of impurity guidelines for hydrofluoric acid established by SEMI for the purest grade of hydrofluoric acid, as shown in Table. D. ml/min Table Method Detection Limits and SEMI Specifications for Trace Anions in Concentrated High Purity Hydrofluoric Acid Anion Method SEMI Method detection limits Limits (µg/l) C8.-9 Specification (µg/l) HF concentration.% (v/v) 9% (w/w) Chloride. Nitrate. Sulfate. Phosphate. Method Detection Limit = (SD) x (t s ) 99.% where (t s ) is for a single sided Student s t-test distribution for n = 7. Figure 8. Effect of flow rate on ICE separation. Determination of Trace Anions in Concentrated Hydrofluoric Acid

11 Pretreatment : IonPac ICE-AS ICE Eluent: Deionized water ICE Flow Rate:. ml/min Volume: 7 µl : IonPac AG9-HC, AS9-HC, mm Concentrator: IonPac AG9-HC, -mm ICE Eluent: 8. mm Sodium carbonate. mm Sodium hydroxide IC Flow Rate:. ml/min Detection; Suppressed conductivity, ASRS-ULTRA, AutoSuppression, external water mode Peaks:. Fluoride µg/l (ppb). Chloride 7.9. Carbonate. Nitrate.89. Unidentified. Sulfate. 7. Phosphate. Figure. Determination of trace anions in high purity.% hydrofluoric acid. PRECAUTIONS Exercise extreme caution when handling concentrated hydrofluoric acid. Use only the highest quality deionized water for the preparation of standards and eluents. Any ionic contamination present in the deionized water will adversely affect results. Teflon containers are recommended for holding the concentrated acid samples for delivery to the sample loop. Containers should be soaked for at least h with 7.8 MΩ-cm deionized water prior to use. It is good practice to dedicate all containers for trace analysis and keep them filled with deionized water when not in use. The level of sulfate in the deionized water blank should be monitored on a regular basis. Method success depends on maintaining a consistent flow rate of deionized water from the RP- so that the proper fraction is cut from the ICE-AS. Verify that the flow is. ±. ml/min. If the deionized water container feeding the RP- pump is not pressurized to at least. kpa ( psi) the pump will be prone to losing prime. Periodically rinse the Rheodyne valve used for loading sample. Residual hydrofluoric acid can crystallize and block injector passages. Also, impurities present in the concentrated acid samples can lead to a loss of column capacity. This causes an increase in back pressure and shorter retention times. Consult the IonPac AS9-HC column stallation and Troubleshooting Guide for more information. Do not leave concentrated HF in the sample loop and sample inlet lines for more than hours. The PEEK tubing can degrade after extended contact time with the concentrated acidic sample. 7 REFERENCES. Watanabe, K. Presented at the ternational Ion Chromatography Symposium, Dallas, TX, October 99; Poster.. Wu, M.; Chen, J. Micro 997, (), 7.. Bader, M. J. Chem Educ. 98, 7, 7.. Weiss, J. Ion Chromatography, nd ed., VCH, Weinheim, Germany, 99, 9.. Troubleshooting Guide for HPLC jection Problems, Rheodyne: Cotati, CA, 99.. SEMI ternational Standards: Semiconductor Equipment and Materials ternational, Mountain View, CA, Chemical/Reagents Volume, 997. LIST OF SUPPLIERS Fisher Scientific, 7 Forbes Ave., Pittsburgh, PA 9-78, USA. Tel: (8) 7-7 VWR Scientific, P.O. Box 79, San Francisco, CA 9, USA. Tel: (8) 9- Lab Safety Supply c., PO Box 8, Janesville, WI 7-8, USA. Tel: (8) -78 Technical Note

12 IonPac, ASRS, and AutoSuppression are registered trademarks of Dionex Corporation. Printed on recycled & recyclable paper with soy-based inks. Dionex Corporation Dionex Corporation Dionex U.S. Regional Offices Dionex ternational Subsidiaries 8 Titan Way Salt Lake City Technical Center Sunnyvale, CA (8) 77-8 Austria () Belgium () 8 Canada (9) 8-9 France 9 8 Germany -99- P.O. Box West South, Suite A Westmont, IL () 789- Italy () Japan () 88- The Netherlands () Switzerland () 99 United Kingdom (7) 97 Sunnyvale, CA Salt Lake City, UT Houston, TX (8) 87- * Designed, developed, and manufactured under an NSAI registered ISO 9 Quality System Atlanta, GA (77) -8 Determination of Trace Anions in Concentrated Hydrofluoric Acid (8) 77-7 (8) Marlton, NJ (8) 9- LPN -.M / Dionex Corporation