METHOD 314.0

    DETERMINATION OF PERCHLORATE  IN DRINKING WATER USING ION
                            CHROMATOGRAPHY
                                 Revision 1.0

                               November 1999
Daniel P. Hautman and David J. Munch, US EPA, Office of Ground Water and Drinking Water
Andrew D. Eaton and Ali W. Haghani, Montgomery Watson Laboratories
              NATIONAL EXPOSURE RESEARCH LABORATORY
                 OFFICE OF RESEARCH AND DEVELOPMENT
                U.S. ENVIRONMENTAL PROTECTION AGENCY
                          CINCINNATI, OHIO  45268

-------
                                    METHOD 314.0

     DETERMINATION OF PERCHLORATE IN DRINKING WATER USING ION
                                CHROMATOGRAPHY
1.  SCOPE AND APPLICATION

   1.1  This method covers the determination of perchlorate in reagent water, surface water,
        ground water, and finished drinking water using ion chromatography.

   1.2  The single laboratory reagent water Method Detection Limit (MDL, defined in Section
        3.16) for the above analyte is listed in Table 1.  The MDL for a specific matrix may
        differ from those listed, depending upon the nature of the sample and the specific
        instrumentation employed.

        1.2.1   In order to achieve comparable detection limits, an ion chromatographic system
               must utilize suppressed conductivity detection, be properly maintained, and
               must be capable of yielding a baseline with no more than 5 nanosiemen (nS)
               noise/drift per minute of monitored response over the background conductivity.

   1.3  This method is recommended for use only by or under the supervision of analysts
        experienced in the use of ion chromatography and in the interpretation of the resulting
        ion chromatograms.

   1.4  When this method is used to analyze unfamiliar samples for perchlorate, anion
        identification should be supported by the use of a laboratory fortified matrix sample.
        The fortification procedure is described in Section 9.4.1.

   1.5  Users of the method data should identify data quality objectives prior to analysis. Users
        of the method must demonstrate the ability to generate acceptable results, using the
        procedures described in Section 9.0.

   1.6  This method specifies an 1C column and analytical conditions which were determined
        to be the most effective for the widest array of sample matrices. Other 1C procedures
        have been written which incorporate similar columns and conditions, such as hydroxide
        based mobile phases, low hydrophobicity 1C columns, and measurement by suppressed
        conductivity detection.1"5 During the development of this method, these other
        procedures, as well as the columns and conditions outlined in this method, were
        concurrently investigated with comparable results for test matrices with moderate levels
        of common inorganic background anions. These findings were consistent with those of
        the Inter-Agency Perchlorate Steering Committee, Analytical Subcommittee's Report,6
        published in 1998, which reported on the results of an interlaboratory validation  of

-------
        these other Ion Chromatographic Methods. The columns and conditions identified in
        this method were recommended since they bore the greatest tolerance for the highest
        levels of common inorganic anion interference.

2.  SUMMARY OF METHOD

   2.1  A 1.0 mL volume of sample (see Note), is introduced into an ion chromatograph (1C).
        Perchlorate is separated and measured, using a system comprised of an ion
        chromatographic pump, sample injection valve, guard column, analytical column,
        suppressor device, and conductivity detector.

        NOTE: This large sample loop (1.0 mL) can be made using approximately 219 cm (86
                inches) of 0.03  inch i.d. PEEK tubing.  The exact volume is not critical since
                all standards and samples will use the same sample loop. However, the
                volume should be verified to be within 5% of this volume by weighing the
                sample loop empty, filling the loop with deionized water and re-weighing the
                loop. The volume can then be approximated by assuming the density of water
                is 1.0 mg/uL.

3.  DEFINITIONS

   3.1  ANALYSIS BATCH — A sequence of samples, which are analyzed within a 30 hour
        period and include no more than 20 field samples. An Analysis Batch must also
        include all required QC samples, which do not contribute to the maximum field sample
        total of 20. The required QC samples include:
        • Instrument Performance Check Standard (IPC)
        • Laboratory Reagent Blank (LRB)
        • Initial Calibration Check Standard (ICCS)
        • Laboratory Fortified Blank (LFB)
        • Continuing Calibration Check Standard (CCCS), when the batch contains more than
          10 field samples
        • End Calibration Check Standard (ECCS)
        • Laboratory Fortified Matrix (LFM)
        • Either a Field Duplicate, a Laboratory Duplicate or a duplicate of the LFM
        • (if pretreated samples are included in batch) Pretreated LRB
        • (if pretreated samples are included in batch) Pretreated LFB
        • (if pretreated samples are included in batch) Pretreated LFM, for each pretreated
          matrix.

        NOTE: Every field sample analysis, including both diluted and pretreated field
                samples, but excluding any LFM or duplicate field sample analysis which
                qualify as QC samples,  must be applied to the maximum of 20 total field
                samples permitted in an analysis batch.

-------
     3.1.1   A field sample(s), included in the analysis batch, can be reanalyzed following
            the ECCS provided the 30 hr time limit for the analysis batch has not expired.
            The laboratory can reanalyze that sample(s) but must initially conduct a second
            ICCS before the reanalysis and an ECCS after the final reanalysis.  The ECCS
            must be completed within the 30 hr window.

3.2  CALIBRATION STANDARD (CAL) -- A solution prepared from the primary dilution
     standard solution(s) or stock standard solutions.  The CAL solutions are used to
     calibrate the instrument response with respect to analyte concentration.

3.3  INITIAL CALIBRATION STANDARDS --  A series of CAL solutions used to initially
     establish instrument calibration and develop calibration curves for individual target
     anions (Section 10.2).

3.4  INITIAL CALIBRATION CHECK STANDARD (ICCS) -- A CAL solution, which is
     analyzed initially, prior to any field sample analyses, which verifies the previously
     established calibration curve. The concentration for the initial calibration check
     standard MUST be at or below the MRL (Section 3.17) level.

3.5  CONTINUING CALIBRATION CHECK STANDARDS (CCCS) - A CAL solution
     which is analyzed after every tenth field sample analyses, not including QC samples,
     which verifies the previously established calibration curve and confirms accurate
     analyte quantitation for the previous ten field samples analyzed. The concentration for
     the continuing calibration check standards should be either at a middle calibration level
     or at the highest calibration level (Section 10.3.2).

3.6  END CALIBRATION CHECK STANDARD (ECCS)  -- A CAL solution which is
     analyzed after the last field sample analyses which verifies the previously established
     calibration curve  and confirms accurate analyte quantitation for all field samples
     analyzed since the last continuing calibration check. The end  calibration check standard
     should be either the middle or high level continuing calibration check standard (Section
     10.3.2).

3.7  FIELD DUPLICATES (FD) — Two separate  samples collected at the same time and
     place under identical circumstances and treated exactly the same throughout field and
     laboratory procedures. Analyses of field duplicates indicate the precision associated
     with sample collection, preservation and storage, as well as with laboratory procedures.

3.8  INSTRUMENT PERFORMANCE CHECK SOLUTION (IPC) - A solution containing
     a specific concentration of perchlorate and other test substances (namely chloride,
     sulfate and carbonate) used to evaluate the performance of the instrument system with
     respect to a defined set of criteria.

-------
3.9  LABORATORY DUPLICATE (LD) -- Two sample aliquots (LD1 and LD2), taken in
     the laboratory from a single sample bottle, and analyzed separately with identical
     procedures. Analyses of LD1 and LD2 indicate precision associated specifically with
     the laboratory procedures by removing variation contributed from sample collection,
     preservation and storage procedures.

3.10 LABORATORY FORTIFIED BLANK (LFB) - An aliquot of reagent water, or other
     blank matrix, to which a known quantity of perchlorate is added in the laboratory. The
     LFB is analyzed exactly like a sample, and its purpose is to determine whether the
     methodology is in control, and whether the laboratory is capable of making accurate and
     precise measurements.

3.11 LABORATORY FORTIFIED SAMPLE MATRIX (LFM) - An aliquot of an
     environmental field sample to which a known quantity of perchl orate is added in the
     laboratory.  The LFM is analyzed exactly like a sample,  and its purpose is to determine
     whether the sample matrix contributes bias to the analytical result (when compared to
     the result for the LFB). The background concentrations of perchl orate, in the sample
     matrix, must be initially determined in a separate aliquot and the measured value in the
     LFM corrected for this background concentration.

3.12 LABORATORY REAGENT BLANK (LRB) - An aliquot of reagent water or other
     blank matrix that is treated exactly as a sample including exposure to all glassware,
     equipment, solvents, filtration and reagents that are used with other samples. The LRB
     is used to determine if perchl orate or other interferences are present in the laboratory
     environment, the reagents, or the apparatus.

3.13 LINEAR CALIBRATION RANGE (LCR) - The concentration range over which the
     instrument response is linear.

3.14 MATERIAL SAFETY DATA SHEET (MSDS) - Written information provided by
     vendors concerning a chemical's toxicity,  health hazards, physical properties, fire, and
     reactivity data including storage, spill, and handling precautions.

3.15 MATRIX CONDUCTIVITY THRESHOLD (MCT) - The highest permitted
     conductance of an unknown sample matrix, measured prior to conducting the analysis,
     which is used  to determine when sample matrix dilution or pretreatment is required.
     The conductance of a  sample matrix is proportional to the common anions present in
     the matrix (which contribute to the level of total  dissolved solids [TDS]) which can
     greatly affect the integrity of this analysis. The value for this threshold is dependant on
     the conditions, hardware, and state of the hardware employed.   Consequently, this
     threshold is not method defined and must be determined by the individual analytical
     laboratory during the Initial Demonstration of Capability (IDC) and confirmed in each
     analysis batch using the Instrument Performance Check  (IPC) Solution. Matrix

-------
     conductivity is measured in microsiemens/cm (uS/cm) or microMhos/cm (uMhos/cm)
     which are considered equivalent terms.

3.16 METHOD DETECTION LIMIT (MDL) - The minimum concentration of an analyte
     that can be identified, measured and reported with 99% confidence that the analyte
     concentration is greater than zero.7'8

3.17 MINIMUM REPORTING LEVEL (MRL) - The minimum concentration that can be
     reported as a quantitated value for a target analyte in a sample following analysis. This
     defined concentration can be no lower than the concentration of the lowest calibration
     standard and can only be used if acceptable quality control criteria for this standard are
     met.

3.18 PEAK AREA TO HEIGHT RATIO (A/H) - The ratio of the peak area divided by the
     peak height which is used as a tool to monitor analytical performance. This ratio is
     used to establish and monitor the MCT and represents an objective means of assessing
     analytical performance when analyzing high conductivity matrices. A gradual
     distortion of the baseline is typically observed in the retention time window for
     perchlorate as the matrix conductivity increases (consistent with elevated levels of
     common  anions) which will  more significantly influence peak height relative to the
     influence on peak area. As the distortion of the baseline increases, this ratio increases,
     and the integrity of the measured perchlorate will be compromised.

3.19 PROFICIENCY TESTING (PT) or PERFORMANCE EVALUATION (PE) SAMPLE -
     - A certified solution of method analytes whose concentration is unknown to the
     analyst. Often, an aliquot of this solution is added to a known volume of reagent water
     and analyzed with procedures used for samples. Often, results of these analyses  are
     used as part of a laboratory certification program to objectively determine the
     capabilities of a laboratory to achieve high quality results.

3.20 QUALITY CONTROL SAMPLE (QCS) - A solution of method analytes of known
     concentrations that is obtained from a source external to the laboratory and different
     from the source of calibration standards.  It is used to check laboratory performance
     with externally prepared test materials.

3.21 STOCK STANDARD  SOLUTION (SSS) - A concentrated solution containing
     perchlorate which is either prepared in the laboratory using assayed reference materials
     or purchased from a reputable commercial source.

3.22 TOTAL DISSOLVED  SOLIDS (TDS) - Both organic and inorganic constituent which
     are dissolved in a sample matrix and are not removed by  particulate filtration.

-------
4.  INTERFERENCES

   4.1   Method interferences may be caused by contaminants in the reagent water, reagents,
         glassware, and other sample processing apparatus that lead to discrete artifacts or
         elevated baselines in an ion chromatogram. These interferences can lead to false
         positive results for the target analyte as well as reduced detection limits as a
         consequence of elevated baseline noise.

   4.2   Interferences can be divided into three different categories:  direct chromatographic
         coelution, where an analyte response is observed at very nearly the same retention time
         as the target anion; concentration dependant coelution, which is observed when the
         response of higher than typical concentrations of the neighboring peak overlap into the
         retention window of the target anion; and, ionic character displacement, where retention
         times may significantly shift due to the influence of high ionic strength matrices (high
         mineral content or hardness) overloading the exchange sites in the column and
         significantly shortening target analyte's retention times.

         4.2.1   A direct chromatographic coelution may be solved by changing columns, eluent
                strength, modifying the eluent with organic solvents (if compatible with 1C
                columns), changing the detection systems, or selective removal of the
                interference with pretreatment. Sample dilution will have little to no effect. The
                analyst MUST verify that these changes do not induce any negative affects on
                method performance by repeating and passing all the QC criteria as described in
                Section 9.

         4.2.2   Sample dilution may resolve some of the difficulties if the interference is the
                result of either concentration dependant coelution or ionic character
                displacement, but it must be clarified that sample dilution will alter your
                Minimum Reporting Limit (MRL) by a proportion equivalent to that of the
                dilution. Therefore, careful consideration of project objectives should be given
                prior to performing such a dilution. An alternative to sample dilution, may be
                dilution of the eluent as outlined in Section 11.2.6.

         4.2.3   Pretreatment cartridges can be effective as a means to eliminate certain matrix
                interferences.  With any proposed pretreatment, the analyst must verify that the
                target analyte is not affected by monitoring recovery after pretreatment
                (additional pretreated LFM requirement see Section 11.1.4.6) and that no
                background contaminants are introduced by the pretreatment (additional
                pretreated LRB requirement see Sections 9.3.1.1  and 11.1.4.2). With advances
                in analytical separator column technology which employ higher capacity anion
                exchange resins, the need for these cartridges has been greatly reduced.

-------
            4.2.3.1  Extreme caution should be exercised in using these pretreatment
                    cartridges. Artifacts are known to leach from certain cartridges which
                    can foul the guard and analytical columns causing loss of column
                    capacity indicated by shortened retention times and irreproducible
                    results. Frequently compare your calibration standard chromatograms
                    to those of the column test chromatogram (received when the column
                    was purchased) or use calibration chromatograms generated when the
                    column was initially installed, to insure proper separation and similar
                    response ratios between the target analytes are observed.

            4.2.3.2  If LRB background problems are encountered in the retention time
                    window for perchlorate when these  pretreatment cartridges have been
                    employed, increase the initial reagent water rinse of the cartridge to
                    approximately five times the volume specified by the manufacturer.

4.3  Sample matrices with high concentrations of common anions such as chloride, sulfate
     and carbonate can make the analysis problematic by destabilizing the baseline in the
     retention time window for perchlorate. This is evidenced by observing a protracted
     tailing following the initial elution of the more weakly retained anions (chloride,
     carbonate, and sulfate) which extends into the perchlorate retention time window.
     These common anion levels can be indirectly assessed by monitoring the conductivity
     of the matrix. Consequently, all sample matrices must be monitored for conductivity
     (Section  11.1.2) prior to analysis. When the laboratory determined Matrix Conductivity
     Threshold (MCT,  see Section 9.2.8) is exceeded, procedures incorporating sample
     dilution and/or pretreatment must be performed as specified in Sections 11.1.3 and
     11.1.4, respectively.

4.4  All reagent solutions (eluents, external water for ASRS suppressor,  etc...) used by the
     instrument must be filtered through no larger than a 0.45 um nominal pore size
     membrane or frit to remove particulates and prevent damage to the instrument, columns
     and flow systems.  Sample filtration must also be employed on every sample prior to
     analysis.  This applies not only to field samples but also to  the laboratory reagent blank
     (LRB) and laboratory fortified blank (LFB).  The LRB and LFB samples function as
     controls and must be filtered to confirm no bias is attributable to the filtration.5  Filter
     the samples through a membrane or frit with no larger than a 0.45 um nominal pore
     size. Syringe mounted, cartridge type, filters work well. Filters specifically designed
     for 1C applications should be used.

4.5  Close attention  should be given to the potential for carry over peaks from one analysis
     which will effect the proper detection of perchlorate in a second, subsequent analysis.
     It is the responsibility of the user to confirm  that no late eluting peaks have carried  over
     into a subsequent analysis thereby compromising the integrity of the analytical results.

-------
5.  SAFETY

   5.1  The toxicity or carcinogen!city of each reagent used in this method have not been fully
        established. Each chemical should be regarded as a potential health hazard and
        exposure should be as low as reasonably achievable. Cautions are specifically listed
        below in Section 5.3 for hazardous materials.

   5.2  Each laboratory  is responsible for maintaining a current awareness file of OSHA
        regulations regarding the safe handling of the chemicals specified in this method. A
        reference file of Material Safety Data Sheets (MSDS) should be made available to all
        personnel involved in the chemical analysis. The preparation of a formal safety plan is
        also advisable. Additional references on laboratory safety are available.9"12

   5.3  The following chemicals have the potential to be highly toxic or hazardous, consult
        MSDS.

        5.3.1  Sodium Hydroxide (NaOH), used in the preparation of the eluent is considered
               caustic.

6.  EQUIPMENT AND  SUPPLIES

   6.1  Ion chromatograph (1C) - Analytical system complete with eluent reservoirs, an ion
        chromatographic pump, injection valves, both guard and analytical separator columns,
        suppressor, conductivity detector, and computer based data acquisition system.

        6.1.1  Anion guard column — Dionex AG16 4 mm (P/N 55377), or equivalent. This
               column functions as a protector of the separator column. If omitted from the
               system, the retention times will be shorter.

        6.1.2  Anion separator column — Dionex AS16, 4 mm (P/N 55376), or equivalent (see
               Sections  6.1.2.1 - 6.1.2.2). The AS 16, 4 mm column using the conditions
               outlined in Table 1 produced the separations shown in Figures 1 through 4.

               6.1.2.1   The development of this method included investigations into the
                        performance of alternate 4 mm 1C guard and analytical separator
                        columns which have been used for the 1C  analysis of perchlorate and
                        are specified in procedures external to the U.S.EPA.1"5  These alternate
                        guard /separator columns included the Dionex AG5 / ASS and the
                        Dionex AG11 / AS 11.  The AG5 / ASS  is currently specified in the
                        standard operating procedure (SOP) for the 1C analysis of perchl orate
                        by the State of California,  Department of Health Services.1'5  The
                        AG11 / ASH is used by several commercial labs conducting 1C
                        analysis for perchlorate and is recognized by California as an

-------
               acceptable alternate to the AG5 / ASS.2"4 A multilab validation study
               included both of these analytical columns and indicated comparable
               results could be attained.6  In U. S.EPA studies, both the AG5 / ASS
               and the AG11 / AS11  performed well for reagent water and simulated
               drinking water samples with low to moderate common anion levels but
               as these levels increased, performance began to diminish for both
               columns. The AG16 / AS 16 columns could tolerate much higher
               levels of these common anions and therefore it is recommended in this
               method as the column of choice. A summary of the results of
               examining these three columns for simulated matrices with various
               common anion levels is presented in  Table 4.

       6.1.2.2  Any alternate, equivalent column must be characterized as hydrophilic
               or conversely, must be rated as having low to very low
               hydrophobicity.4 This is one characteristic that is consistent for the
               ASS, ASH and AS 16 analytical separator columns. This requirement
               for low hydrophobicity is to allow the efficient, reproducible and
               symmetrical band elution of polarizable anions, such as perchlorate.
               If the perchlorate analysis is attempted on a hydrophobic column, such
               as those typically used for the analysis of common anions,13 poor
               performance will result due to very asymmetric, tailing peaks.  Using a
               middle to high calibration standard, conduct a typical analysis. Any
               alternate column must be capable of yielding symmetrical peak elution
               for this perchlorate response as demonstrated by yielding a  Peak
               Gaussian Factor of between 0.80 and 1.15 using the following
               equation,

                              1.83 x WO/2)
                       PGF =	
                                 W (V10)
               where,
               WO/2) is the peak width at half height, and
               W (V10) is the peak width at tenth height.

               NOTE: Values for WO/2) and W (V10) can be attained through most
                       data acquisition software.

6.1.3   Anion suppressor device —  The data presented in this method were generated
       using a Dionex Anion Self Regenerating Suppressor (4 mm ASRS, ULTRA,
       P/N 53946).  An equivalent suppressor device may be utilized provided
       comparable conductivity detection limits are achieved and adequate baseline
       stability is attained as measured by a combined baseline drift/noise of no more
       than 5 nS per minute over the background conductivity.  Proper suppressor
                                  10

-------
            performance is essential to analytical data reproducibility and sensitivity of the
            conductivity detector.

            6.1.3.1  The ASRS was set to perform electrolytic suppression at a current
                    setting of 300 mA using the external water mode. External water was
                    delivered to the suppressor directly from a pressurized source at a flow
                    rate of 5 mL/min

            6.1.3.2  If pretreated samples (Section 11.1.4), or sample matrices which
                    contain appreciable concentrations of transition metal cations (e.g., Fe
                    or Al) are frequently analyzed, cationic components may bind to the
                    suppressor membrane and over time effect suppressor performance. If
                    the instrument begins to have problems with reduced peak response or
                    asymmetrical perchlorate peaks, the suppressor membranes should be
                    cleaned. As a quick and easy cleaning step, the manufacturer's ASRS
                    "Quickstart" procedure for installing a new ASRS should be
                    followed.14 If this procedure does not correct the problem, follow the
                    manufacturer's recommended cleaning procedure for removing metal
                    contaminants.15

     6.1.4   Detector - Conductivity cell (Dionex CD20, or equivalent) capable of providing
            data as required in Section 9.2.

6.2  Data Acquisition System — The Dionex Peaknet Data Chromatography Software was
     used to generate all the data in Tables 1  through 4.  Other computer based data systems
     may achieve approximately the same performance but the user should demonstrate this
     by the procedures outlined in Section 9.

6.3  Conductivity Meter - Used to monitor sample matrix conductance which is directly
     related to the common anion levels in a matrix and used to determine if sample
     pretreatment is required.  At a minimum, this meter should be capable of measuring
     matrix conductance over a range of 1 - 10,000 uS/cm.

6.4  Analytical balance — Used to accurately weigh target analyte salt for stock standard
     preparation (±0.1 mg sensitivity).

6.5  Top loading balance —  Used to accurately weigh reagents such as sodium hydroxide
     solution in the preparation of eluents (±10 mg sensitivity).

6.6  Weigh boats — Plastic, disposable - for weighing eluent reagents.

6.7  Micro beakers — Plastic, disposable - used during sample preparation.
                                       11

-------
   6.8   Syringes — Plastic, disposable, 10 mL - used during sample preparation.

   6.9   Pipets — Pasteur, plastic or glass, disposable, graduated, 5 mL and 10 mL.

   6.10  Bottles — High density polyethylene (HDPE) or glass, amber or clear, 30 mL, 125 mL,
         250 mL. For sampling and storage of calibration solutions. Stability studies presented
         by the Interagency Perchlorate Steering Committee for Analytical Methods 6 and
         confirmed at the EPA (see TableS A), indicate perchlorate is neither photoreactive nor
         prone to adsorption to the walls of either HDPE plastic or glass bottles.

   6.11  Particulate filters — 0.45 micron syringe filters, specifically designed for 1C applications
         (Gelman 1C Acrodisc, PN 4485, or equivalent). These cartridges are used to remove
         particulates from the sample matrix while loading the sample manually or if the
         autosampler employed does not filter the sample during loading.

   6.12  Matrix pretreatment cartridges in the barium form — (Dionex OnGuard-Ba cartridges,
         PN 046072, or equivalent.) These cartridges are conditioned according to the
         manufacturer's directions and are used to reduce the matrix levels of sulfate.

   6.13  Matrix pretreatment cartridges in the silver form -  (Dionex OnGuard-Ag cartridges
         PN 039637, or equivalent.) These cartridges are conditioned according to the
         manufacturer's directions and are used to reduce the matrix levels of chloride.

   6.14  Matrix pretreatment cartridges in the hydrogen form — Dionex OnGuard-H cartridges
         (PN 039596) or equivalent. These cartridges are conditioned according to the
         manufacturer's directions and are used to reduce cations in the sample matrix.  This
         protects the analytical column by removing silver which has leached from the Ag
         cartridge and may indirectly minimize the effect of carbonate by removing the cationic
         counter ion.

7.  REAGENTS AND STANDARDS

   7.1   Reagent water — Distilled or deionized water 17.8 Mohm or better, free of the  anions of
         interest.  Water should contain particles no larger than 0.20 microns.

   7.2   Eluent solution -- 50 mM sodium hydroxide (NaOH, [CASRN 1310-73-2]), dissolve
         8.0 grams of 50% (WAV) sodium hydroxide in reagent water to a final volume of 2.0 L.
         NOTE: This eluent solution is specific to the columns listed in Table 1.  Any alternate
         columns will likely have unique and specific conditions identified by the manufacturer.

         7.2.1   Solutions of NaOH are very susceptible to carbonate contamination resulting
               from adsorption of carbon dioxide from the atmosphere. This contamination
               will result in poor reproducibility of perchlorate retention times, elevated
                                           12

-------
            instrument background conductivity, and increased baseline noise/drift.
            Consequently, exposure to the atmosphere should be minimized by storing these
            eluent solutions in sealed reservoirs under low pressure (3 to 5 psi) helium. In
            addition, these solutions should be regularly prepared and held for no more than
            5 days.  When refilling the eluent reservoir, completely replace old eluent
            solution by emptying the old eluent, rinsing the reservoir with reagent water,
            and refilling with the freshly prepared eluent solution. With this eluent, the
            suppressed conductivity detector background signal should be between 2 - 5 uS.

     7.2.2   This eluent solution must be purged for 10 minutes with helium prior to use.
            This effectively removes dissolved gases which may form micro bubbles in the
            1C, compromising system performance and adversely effecting the integrity of
            the data. Alternatively, an in-line degas apparatus may be employed.

     7.2.3   A system or apparatus which automatically generates the hydroxide eluent
            (Dionex EG40, or equivalent) is an acceptable alternative to physically
            preparing this hydroxide eluent.

7.3  Perchlorate stock standard solution, 1000 mg/L (1 mg/mL) - A stock standard solution
     may be purchased as a certified solution or prepared from ACS reagent grade, sodium
     salt as listed below. (NOTE:  Sodium perchlorate represents a molar weight fraction of
     81.2 % perchlorate anion)

     7.3.1   Perchlorate (C1O4") 1000 mg/L --  Dissolve 0.1231 g sodium perchlorate
            (NaClO4, CASRN [7601-89-0] in reagent water and dilute to 100  mL in a
            volumetric flask.

     NOTE:  Stability of standards — Perchlorate stock standards, stored at room
              temperature, appear to be very stable and may be  stable for an extended period
              of time.  However, specified expiration dates should be marked  on each
              prepared stock standard as part of any laboratory's quality  control program.  In
              this regard, it is recommended that stock standards for perchlorate be held for
              no more than 12 months and an expiration date should be clearly specified on
              the label.

7.4  Mixed Common Anion Stock Solution - containing the anions chloride, sulfate and
     carbonate each  at 25 mg/mL anion concentration.  This solution is used to prepare
     simulated common anion samples in the determination of the MCT  (Section 9.2.8).

     7.4.1   Dissolve the following salts in reagent water to a final volume of 25.0 mL:
            1.0 g sodium chloride  (NaCl, CASRN [7647-14-5]) = 0.61 g Cl'
            0.93 g sodium sulfate (Na2SO4, CASRN [7757-82-6]) = 0.63 g  SO4=
            1.1 g sodium carbonate (Na2CO3,  CASRN [497-19-8]) =  0.62 g CO3=
                                       13

-------
   7.5  Conductivity Meter Calibration Solution

        7.5.1  Potassium Chloride (KC1), 745 mg/L (total salt weight) -- Dissolve 0.745 g
               potassium chloride (KC1, [CASRN 7447-40-7]) in reagent water and dilute to a
               final volume of 1.00 L in a volumetric flask.  On a properly functioning and
               calibrated conductivity meter, the reference conductance for this solution is
               1410uS/cmat25°C.16

8.  SAMPLE COLLECTION. PRESERVATION AND STORAGE

   8.1  Samples may be collected in plastic or glass bottles.  All bottles must be thoroughly
        cleaned and rinsed with reagent water. The volume collected should be sufficient to
        insure a representative sample, allow for replicate analysis and laboratory fortified
        matrix analysis, if required, and minimize waste disposal.

   8.2  Samples do not need to be shipped iced or stored cold in a refrigerator but every effort
        should be taken to protect the  samples from temperature extremes. A thermally
        insulated sampling kit, designed to fit sampling bottles securely during shipment,
        should be used to protect the samples from these temperature extremes.
   8.3  Sample preservation and holding times for the anions are as follows:
        Analyte       Preservation       Holding Time
        Perchlorate    None required     28 days

        NOTE:  Perchlorate has been shown to be stable for more than 28 days6 but extended
                 holding time studies (beyond 35 days) were not conducted by EPA.
                 Typically, when analytes are believed to be stable, a 28 day holding time is
                 established as a sufficient time period to permit a laboratory to conduct the
                 analysis.

9.  QUALITY CONTROL

   9.1  Each laboratory using this method is required to operate a formal quality control (QC)
        program.  The requirements of this program consist of an initial demonstration of
        laboratory capability, and subsequent analysis in each analysis batch (Section  3.1) of an
        Instrument Performance Check Standard (IPC), Laboratory Reagent Blank (LRB),
        Initial Calibration Check Standard (ICCS), Laboratory Fortified Blank (LFB),
        Continuing and End Calibration Check Standards (CCCS/ECCS), Laboratory Fortified
        Sample Matrix (LFM) and either a Field, Laboratory or LFM duplicate sample analysis.
        This section details the specific requirements  for each of these QC parameters. The QC
        criteria discussed in the following  sections are summarized in Section 17, Table 5 and
                                          14

-------
     6.  The laboratory is required to maintain performance records that define the quality of
     the data that are generated.

9.2  INITIAL DEMONSTRATION OF CAPABILITY

     9.2. 1   The Initial Demonstration of Capability (IDC) — This is used to characterize
            instrument and laboratory performance prior to performing analyses by this
            method. The QC requirements for the IDC discussed in the following section
            are summarized in Section 17, Table 5.

     9.2.2   Initial demonstration of low system background — See Section 9.3.1.

     9.2.3   Initial Demonstration of Accuracy (IDA) — Prepare and analyze 7 replicate
            LFBs fortified at 25.0 ug/L.  Calculate the mean measured concentration (Cx) of
            the replicate values as follows.

                                (C1 + C2+C3+....CJ
                                         n
            where,
                   Cx = Mean recovered concentration of the replicate analysis.
                   Cj C2 ...Cn= Recovered concentrations of the replicate 1,2. ..n.
                    n =7

            To pass the IDA, the value derived for Cx must be within ± 10% of the true
            value or between 22.5 ug/L and 27.5 ug/L.

     9.2.4   Initial Demonstration of Precision (IDP) — Using the data generated for Section
            9.2.3, calculate the percent relative standard deviation (%RSD) of the replicate
            analysis, as indicated below.  To pass the IDP, the %RSD must be less than
            10%.
                                        (Sn.!)
                              %RSD = ----------  x 100
                                        (Cx)
            where,
                   Sn.! =  sample standard deviation (n-1) of the replicate analyses.
                   Cx = mean recovered concentration of the replicate analysis.

     9.2.5   Quality Control Sample (QCS) - After calibration curves have initially been
            established or have been re-established, or as required to meet data quality
            needs, verify both the calibration and acceptable instrument performance with
            the preparation and analyses of an external/second source  QCS.  If the
            determined concentrations are not within ± 10% of the stated values,
                                       15

-------
       performance of the determinative step of the method is unacceptable. The
       source of the problem must be identified and corrected before either proceeding
       with the IDC or continuing with on-going analyses.

9.2.6   Method Detection Limit (MDL) - An MDL must be established using reagent
       water (blank) fortified at a concentration of three to five times the estimated
       instrument detection limit.7'8  To determine MDL values, take seven replicate
       aliquots of the fortified reagent water and process through the entire analytical
       method over a three day period. These seven MDL replicate analyses may be
       performed gradually over three days or may represent data that has been
       collected, at a consistent MDL estimated concentration, over a series of more
       than three days. Perform all calculations defined in the method and report the
       concentration values in the appropriate units. Calculate the MDL as follows:

                           MDL = (t)x(Sn.1)

       where,
             t = student's t value for a 99% confidence level and a standard deviation
                 estimate with n-1 degrees of freedom [t = 3.14 for seven replicates]
            Sn_! =  sample standard deviation (n-1) of the seven replicate analyses.

       9.2.6.1   MDLs should be periodically verified, but MUST be initially
               determined when a new operator begins work or whenever there is a
               significant change in the background, or instrument response.

       NOTE: Do not subtract blank values when performing MDL calculations.

9.2.7   Minimum Reporting Level (MRL) - The MRL is the threshold concentration of
       an analyte that  a laboratory can expect to accurately quantitate in an unknown
       sample.  The MRL should be established at an analyte concentration either
       greater than three times the MDL or at a concentration which would yield a
       response greater than a signal to noise ratio of five.  Setting the MRL too low
       may cause repeated QC failure upon analysis of  the ICCS.  Although the
       lowest calibration standard may be below the MRL, the MRL must never
       be established at a concentration lower than the lowest calibration
       standard

9.2.8   Matrix Conductivity Threshold (MCT) - The MCT is an individual laboratory
       defined value which must be determined by preparing a series of sequentially
       increasing, common anion fortified,  reagent water samples each contain a
       constant perchlorate  concentration. Initially, a reagent water prepared LFB,
       containing no common anions, must be analyzed which contains perchlorate at a
       suggested concentration of 25 ug/L perchlorate.  Next, the series of sequentially
                                  16

-------
increasing anionic solutions are prepared, each containing perchlorate at a
suggested concentration of 25 ug/L, which also containing the individual
common anions of chloride, sulfate and carbonate, all included at uniform
increasing concentrations of 200, 300, 400, 500, 600, 800, and 1000 mg/L for
each anion. A concentration of 25 ug/L perchlorate has been suggested
assuming the MRL has been set in the range of 3.0 ug/L to 5.0 ug/L.  If a
laboratory's MRL is higher, choose a perchlorate concentration for this exercise
at approximately 5 times that MRL.

9.2.8.1   Prepare the mixed common anion stock solution (see Section 7.4)
         containing chloride, sulfate and carbonate, each at 25 mg/mL.

9.2.8.2   Prepare a perchlorate secondary stock dilution standard at 1.00 mg/L
         from the 1000 mg/L perchlorate stock standard (Section 7.3) by
         diluting 0.50 mL of the  stock solution to a final volume of 500 mL.

9.2.8.3   Prepare the LFB at suggested perchlorate concentration of 25 ug/L by
         diluting 0.625 mL of the perchlorate secondary stock dilution standard
         (Section 9.2.8.2) to a final volume of 25.0 mL.

9.2.8.4   Next, prepare the series of common anion fortified reagent water
         samples by adding 0.20 mL, 0.30 mL, 0.40 mL, 0.50 mL, 0.60 mL,
         0.80 mL, and 1.00 mL of the mixed common anion stock solution
         (Section 7.4) into separate 25 mL volumetric flasks. Next, add 0.625
         mL of the perchlorate secondary stock dilution standard (Section
         9.2.8.2) to each 25 mL volumetric flask and dilute to volume with
         reagent water to yield a final perchlorate concentration of 25.0 ug/L.

9.2.8.5   Measure and record the conductance of each of these prepared
         solutions on a calibrated conductivity meter (This meter must be
         calibrated as described in Section 10.4 prior to measuring
         conductance). To use as a relative reference conductance, the 400
         mg/L mixed anion sample,  which contains chloride at 400 mg/L,
         sulfate at 400 mg/L and carbonate at 400  mg/L,  should display a
         conductance of between 3200 uS/cm and 3700 uS/cm.

9.2.8.6   Analyze each solution, recording the peak area to height (A/H) ratio
         and the quantified concentration of perchlorate.  In many data
         acquisition and instrument control software, the peak area to height
         ratio  is a definable parameter which can be specified for printout on
         the analysis report.
                            17

-------
9.2.8.7  Both the A/H ratio and quantified perchlorate concentration for the
        LFB and the 200 mg/L mixed common anion solution should be
        reproducibly consistent but as the common anion levels increase, the
        A/H ratio will also begin to increase as the peak height is distorted and
        reduced.  As the peak is distorted, the area will also eventually begin to
        be distorted and the quantitated concentration will be reduced, but this
        is typically secondary, with the ratio of peak area to height initially
        predicting this pending quantitation problem.
9.2.8.8   Calculate the A/H ratio percent difference (PD^) between the average
         A/H ratio for the LFB (A/H^) and the average A/H ratios for each
         mixed common anion solutions (A/H^) using the following equation.
                                          X 100
9.2.8.9  As the conductivity of the matrices increase, the PD^ will increase.
        The MCT is the matrix conductance where the PD^ exceeds 20%.
        To derive the MCT, perform a linear regression on these data by
        plotting PD^yjj (as the independent variable, x) versus the matrix
        conductance (as the dependent variable, y). The resulting regression
        data should yield an r2 value of > 0.95.  (See Figure 5) Record the
        "constant" (intercept value) and the "X-coefficient" (slope) and
        calculate the MCT as follows,

        MCT = (20%) x (X-coefficient) + (constant)

        NOTE:  Be careful to consistently apply percentages as either whole
        numbers or as fractional values (20% = 0.20) for both the regression
        analysis and the MCT calculation.

9.2.8.10 As an alternate to the regression analysis, the laboratory can choose to
        establish their MCT at the conductance level of the highest mixed
        anion solution which yielded a PD^ value below the 20 % threshold.
9.2.8.1 1 As a final procedure, the laboratory should confirm their perchlorate
        MRL in a mixed common anion solution which reflects a conductance
        near (within +/- 10%) that specified as the MCT. This solution must
        contain perchlorate, at the laboratory determined MRL, as well as the
        common anions chloride, sulfate and carbonate, prepared consistent
        with the instruction for the mixed anion solutions in this section and at
        a concentration estimated to generate a conductance near the MCT.
                            18

-------
                    The conductance of this solution must be measured at within ±10% of
                    the MCT and following the analysis, the recovered perchlorate must be
                    between 70 - 130% of the MRL concentration.  If the MRL recovery
                    fails this criteria, the MCT should be lowered by 10% and this MRL
                    verification must be repeated.

            9.2.8.12 Prior to conducting any field sample analysis, the conductivity of that
                    matrix must be determined.  When the conductance of a field sample is
                    above the MCT, sample dilution or pretreatment, as described in
                    respective Sections 11.1.3 and 11.1.4 must be performed.

9.3   ASSESSING LABORATORY PERFORMANCE - The following items must be
     included in every analysis batch (Section 3.1).

     9.3.1   Laboratory Reagent Blank (LRB) - An LRB must be prepared and treated
            exactly as a typical field sample including exposure to all glassware, equipment,
            solvents, filtration and reagents that are used with field samples. Data produced
            are used to assess instrument performance of a blank sample and evaluate
            contamination from the laboratory environment.  Values that exceed /^ the MRL
            indicate a laboratory or reagent contamination is present. The source of the
            contamination must be determined prior to conducting any sample analysis.
            Any sample included in an automated analysis batch which has an invalid LRB,
            indicated by a quantitated perchlorate that exceeds /^ the MRL, must be
            reanalyzed in a subsequent analysis batch after the contamination problem is
            resolved.

            9.3.1.1   When sample matrices have been pretreated to reduce the risk of high
                    common anion interference (Section 11.1.4), a second LRB must be
                    prepared, pretreated in exactly the same manner, and analyzed to
                    confirm no background effects from the pretreatment process are
                    present. If an analysis batch only contains pretreated samples, then
                    only a pretreated LRB is required.

     9.3.2   Instrument Performance Check (IPC) — The MCT, which was determined as
            part of the IDC in Section 9.2.8, must be verified through the analysis  of an IPC.
            The IPC is three tiered and is used to verify the state of the 1C system, over time,
            to quantitate perchlorate in highly ionic matrices. This must be conducted with
            each analysis batch since over time, column performance can change.

            9.3.2.1   Prepare a mixed common anion solution which reflects a conductance
                    near (within +/- 10%) that specified as the MCT.  This solution must
                    be prepared consistent with the instruction in Section 9.2.8, and
                    containing the common anions chloride, sulfate and carbonate as well
                                       19

-------
         as perchlorate at a suggested concentration of 25 ug/L. This
         perchlorate concentration has been specified assuming the MRL has
         been set in the range of 3.0 ug/L to 5.0 ug/L. If a laboratory's MRL is
         higher, chose a perchlorate concentration for this exercise at
         approximately 5 times that MRL.

9.3.2.2   Confirm the conductance of the IPC and analyze it as the initial sample
         in the analysis batch.  If, after several weeks of storage, the measured
         conductance of this solution has shifted by more than 10% from the
         original measured value, prepare a fresh IPC solution.  Following the
         analysis, calculate the PD^ (Section 9.2.8.8), by comparing the peak
         area to height ratio of this IPC mixed anion standard (A/H,^) for this
         analysis batch to the value that was derived for the LFB (A/FL^) either
         in the original IDC or in the previous analysis batch. As the first tier
         criteria, the value for PD^ must be less than 25% before proceeding
         with the analysis batch.

9.3.2.3   At the second tier criteria, the measured recovery for perchlorate in this
         IPC must fall between 80% and 120 % (20.0 ug/L to 30.0 ug/L for a 25
         ug/L fortification).

9.3.2.4   As a third tier and final  criteria for the IPC, the laboratory must closely
         monitor the perchlorate retention time for this analysis. Small
         variations in retention time can be anticipated when a new solution of
         eluent is prepared but if sudden shifts of more than 5% are observed in
         the perchlorate retention time, some type of instrument problem may
         be present. Potential problems include improperly prepared eluent,
         erroneous method parameters programmed such as flow rate or some
         other system problem. The observed retention time for perchlorate
         should closely replicate the times established when the column was
         originally installed. As a column ages, it is normal to see a gradual
         shift and shortening of retention times, but if after several years of use,
         extensive use over less than a year, or use with harsh samples, this
         retention time has noticeably shifted to any less than 80% of the
         original recorded value, the column requires cleaning (according to
         manufacturer's instructions) or replacement. A laboratory should
         retain a historic record of retention times for perchlorate to provide
         evidence  of an analytical column's continued performance.

9.3.2.5   If any of the conditions defined in Section 9.3.2.2 through 9.3.2.4 are
         not met, the MCT must be repeated and revised to a more appropriate
         lower matrix conductivity threshold or the source of the problem must
         be determined and the IPC reanalyzed.
                            20

-------
     9.3.3   Laboratory Fortified Blank (LFB) - Prepare a secondary dilution stock using the
            same stock solution used to prepare the calibration standards.  This separate,
            secondary dilution stock is used as a concentrate to fortify the LFB and the
            LFMs (Section 9.4.1). An external source stock or QCS, which is used to verify
            the accuracy of the calibration curve when it was initially prepared (Section
            10.2.5), should not be used to prepare this secondary dilution stock.
            Laboratories are required to analyze a LFB (filtered as if it were a field sample)
            with each analysis batch immediately  following the ICCS.  The LFB must be
            prepared with the same solution used to prepare the LFM and should be
            prepared at concentrations no greater than ten times the highest concentration
            observed in any field sample  and should be varied to reflect the range  of
            concentrations observed in field samples. By  analyzing the LFB initially, a
            control check is performed on the concentrated solution used to prepare the
            LFM. If any deviations in the perchlorate concentration are present, it will be
            reflected in the LFB and not exclusively attributed to a matrix upon analysis of
            the LFM. Calculate accuracy as percent recovery (Section 9.4.1.3).  The
            recovery for perchlorate must fall in the range of 85 - 115% prior to analyzing
            samples. If the LFB recovery for an analysis batch does not meet these recovery
            criteria the data are considered invalid, and the source  of the problem  should be
            identified and resolved before continuing analyses.

            9.3.3.1   When sample matrices have  been pretreated to reduce the risk of high
                     common anion interference (Section 11.1.4),  a  second LFB must be
                     prepared, pretreated in exactly the same manner, and analyzed to
                     confirm no background effects or recovery bias induced by the
                     pretreatment are present.  If an analysis batch only contains pretreated
                     samples, then only a pretreated LFB is required.

9.4  ASSESSING ANALYTE RECOVERY AND DATA QUALITY - The following must
     be included in every analysis batch (Section 3.1).

     9.4.1   Laboratory Fortified Sample  Matrix (LFM) - The laboratory must add a known
            amount of each target analyte to a minimum of 5% of the collected field samples
            or at least one with every analysis batch, whichever is  greater.  Samples which
            exceed the MCT must either  be diluted (Section 11.1.3) or pretreated to reduce
            the common anion levels (Section  11.1.3).  Samples which are pretreated have
            additional LFM requirements described in Section 11.1.4.6, and must  be
            fortified before pretreatment.  For a LFM to be valid, the target analyte
            concentrations must be greater than the native level and should adhere to the
            requirement outlined in Section 9.4.1.2. It is recommended that the solutions
            used to fortify the LFM be prepared from the same stocks used to prepare the
            calibration standards and not  from external source stocks. This will remove the
                                       21

-------
       bias contributed by an externally prepared stock and focus on any potential bias
       introduced by the field sample matrix.

       9.4.1.1  The fortified concentration must be equal to or greater than the native
               sample concentration. Fortified samples that exceed the calibration
               range must be diluted to be within the linear range. In the event that
               the fortified level is less than the observed native level of the
               unfortified matrix, the recovery should not be calculated.  This is due
               to the difficulty in calculating accurate recoveries of the fortified
               concentration when the native sample concentration to fortified
               concentration ratio is greater than one.

       9.4.1.2  For normal drinking waters, the LFM typically should be prepared in
               the range of 20 - 50 ug/L. The LFM should not be prepared at
               concentration greater than ten times the highest concentration observed
               in any field sample and should be varied to reflect the range of
               concentrations expected in field samples.

       9.4.1.3  Calculate the percent recovery for each target analyte, corrected for
               concentrations measured in the unfortified sample.  Percent recovery
               should be calculated using the following equation:

                                  (C, - C)
                         %REC =	 x 100
               where,
               %REC  = percent recovery,
               Cs = measured perchlorate in the fortified sample,
               C = measured native perchlorate sample concentration, and
               s  = concentration equivalent of analyte added to sample.

       9.4.1.4  Recoveries may exhibit a matrix dependence. If the recovery for
               perchlorate falls outside 80 - 120%, and the laboratory's performance
               for all other QC performance criteria is acceptable, the accuracy
               problem encountered with the fortified sample is judged to be matrix
               related, not system related. The result for that analyte in the unfortified
               sample and the LFM must be labeled  suspect/matrix to inform the data
               user that the result is suspect due to matrix effects. Repeated failure to
               meet suggested recovery criteria indicates potential problems with the
               procedure and should be investigated.

9.4.2   FIELD, LABORATORY DUPLICATES OR DUPLICATE LFM -  The
       laboratory must analyze either a field duplicate, a laboratory duplicate, or a
                                  22

-------
               duplicate LFM for a minimum of 5% of the collected field samples or at least
               one with every analysis batch, whichever is greater.  The sample matrix selected
               for this duplicate analysis must contain measurable concentrations of the target
               anions in order to establish the precision of the analysis set and ensure the
               quality of the data.  Without prior knowledge or strong suspicion that an
               unknown sample has measurable perchlorate concentrations, the best alternative
               is to analyze a duplicate LFM.

               9.4.2. 1   Calculate the relative percent difference (RPD) of the initial
                        quantitated concentration (Ic) and duplicate quantitated concentration
                        (Dc) using the following formula.
                                 RPD = -------------- XI 00
                                       (Pc + Dc]/2)

               9.4.2.2   Duplicate analysis may exhibit a matrix dependance. If the RPD for
                        the duplicate measurements of perchlorate falls outside ± 15% and if
                        all other QC performance criteria are met, laboratory precision is out
                        of control for the sample and perhaps the analytical batch. The result
                        for the sample and duplicate should be labeled as suspect/matrix to
                        inform the data user that the result is suspect due to a potential matrix
                        effect, which led to poor precision. This should not be a chronic
                        problem and if it frequently recurs (>20% of duplicate analyses), it
                        indicates a problem with the instrument or individual technique that
                        must be corrected.

         9.4.3  In recognition of the rapid advances occurring in chromatography, the analyst is
               permitted certain options, such as the use of different columns (which meet the
               criteria in Section 6.1.2.2), injection volumes, and/or eluents, to improve the
               separations or lower the cost of measurements. Each time such modifications to
               the method are made,  the analyst is required to repeat the procedure in Section
               9.2 and adhere to the condition of conductivity baseline stability found in
               Section 1.2.1.

         9.4.4  It is recommended that the laboratory adopt additional quality assurance
               practices for use with  this method.  The specific practices that are most
               productive depend upon the needs of the laboratory and the nature of the
               samples. Whenever possible, the laboratory should perform analysis of quality
               control check samples and participate in relevant proficiency testing (PT) or
               performance evaluation (PE) sample studies.

10. CALIBRATION AND STANDARDIZATION
                                           23

-------
10.1  Demonstration and documentation of acceptable initial calibration is required prior to
     the IDC and before any samples are analyzed, and is required intermittently throughout
     sample analysis to meet required QC performance criteria outlined in this method and
     summarized in Table 6.  Initial calibration verification is performed using a QCS as
     well as with each analysis batch using an initial, continuing (when more than 10 field
     samples are analyzed), and end calibration check standards.  The procedures for
     establishing the initial calibration curve are described in Section 10.2. The procedures
     to verify the calibration with each analysis batch is described in Section 10.3.

10.2  INITIAL CALIBRATION CURVE

     10.2.1  Establish ion chromatographic operating parameters equivalent to those
            indicated in Table 1.

     10.2.2  Estimate the Linear Calibration Range (LCR) — The LCR should cover the
            expected concentration range of the field samples and should not extend over
            more than two orders of magnitude in concentration.   The restriction of two
            orders of magnitude is prescribed since beyond this it is difficult to maintain
            linearity throughout the entire calibration range.

            10.2.2.1 If quantification is desired over a larger range, then two separate
                    calibration curves should be prepared.

            10.2.2.2 A minimum of three calibration standards are required for a curve that
                    extends over a single order of magnitude and a minimum of five
                    calibration standards are required if the curve covers two orders of
                    magnitude.

            10.2.2.3  Since the anticipated concentration range for perchlorate in actual field
                    samples is expected to cover two orders of magnitude, the use of at
                    least five calibration standards in the range 4 - 400  jig/L is
                    recommended.

     10.2.3  Prepare the calibration standards by carefully adding measured volumes of the
            stock standard (Section 7.3) to a volumetric flask and diluting to volume with
            reagent water.

     10.2.4  Inject 1.0 mL of each calibration standard. Tabulate peak area responses against
            the perchlorate concentration. The results are used to prepare a calibration
            curve. Acceptable calibration is confirmed after reviewing the curve for
            linearity (second order fits are also acceptable) and passing the criteria for the
            initial calibration check standard in Section 10.3.1. Alternately, if the ratio of
            area to concentration (response factor) is constant over the LCR (indicated by <
                                        24

-------
               15% relative standard deviation), linearity through the origin can be assumed
               and the average ratio or response factor can be used in place of a calibration
               curve.

               10.2.4.1 Peak areas must be used as a measure of response since they have been
                       found to be more consistent, in terms of quantitation, than peak
                       heights. Peak height can tend to be suppressed as a result of high
                       levels of common anions in a given matrix which can compete for
                       exchange  sites leading to peak broadening. Using peak areas, it is the
                       analyst's responsibility to review all chromatograms to insure accurate
                       baseline integration of target analyte peaks, since poorly drawn
                       baselines will significantly influence peak areas.

        10.2.5 After establishing or reestablishing calibration curves, the accuracy of this
               calibration must be verified through the analysis of a QCS or externally
               prepared second source.  The QCS should be prepared at a concentration near
               the middle of the calibration curve. As specified in Section 9.2.5, determined
               concentrations must fall within ± 10% of the stated values.

10.3  CONTINUING CALIBRATION VERIFICATION - Initial calibrations may be stable for
     extended periods of time. Once the calibration curve has been established it MUST be
     verified for each analysis batch, prior to conducting any field sample analysis using an
     Initial Calibration Check Standard. Continuing Calibration Check Standards and End
     Calibration Check Standards are also required as described in the sections below.

        10.3.1 INITIAL CALIBRATION CHECK STANDARD (ICCS) - For each analysis
               batch the calibration must initially be verified prior to analyzing any samples.
               The lowest level  standard used to prepare the linear calibration curve must be
               used. In cases where the analyst has chosen to set the MRL above the lowest
               standard, a standard at a concentration equal to the MRL is acceptable. Percent
               recovery for the ICCS must be in the range or 75 - 125% before continuing the
               analysis batch and conducting any sample analyses.

        10.3.2 CONTINUING CALIBRATION CHECK/END CALIBRATION CHECK
               STANDARDS (CCCS/ECCS) -- Continuing calibration check standards MUST
               be analyzed after every tenth field sample analysis and at the end of the analysis
               batch as an end calibration check standard. If more than 10 field samples are
               included in an analysis batch, the analyst must alternate  between the middle and
               high continuing calibration check standard levels.

               10.3.2.1 The percent recovery for perchlorate in the CCCS/ECCS must be
                       between 85- 115%.

               10.3.2.2 If during the analysis batch, the measured concentration for perchlorate

                                          25

-------
                       in the CCCS or ECCS differs by more than the calibration verification
                       criteria shown above, or if the perchlorate peak retention time shifts
                       outside the retention time window (as defined in Section 11.2.4), all
                       samples analyzed after the last acceptable check standard are
                       considered invalid and must be reanalyzed. The source of the problem
                       must be identified and resolved before reanalyzing the samples or
                       continuing analyses.

               10.3.2.3 In the case where the end calibration fails to meet performance criteria,
                       but the initial and middle calibration checks are acceptable, the
                       samples bracketed by the acceptable calibrations may be reported.
                       However, all field samples between the middle and end calibration
                       checks MUST be reanalyzed.

   10.4 CONDUCTIVITY METER CALIBRATION -- Prior to conducting the MCT and
        coinciding with each analysis batch, conductivity meter calibration must be verified or
        established using a standard KC1 solution (Section 7.5).

        10.4.1 Thoroughly rinse the conductivity electrode with reagent water. Place the
               electrode in the reagent water, turn on the meter and confirm the conductance of
               this blank is < 1 uS/cm.

        10.4.2 Pour approximately 15 mL of the standard KC1 solution (Section 7.5) into a
               plastic disposable micro beaker (Section 6.7) and place the electrode into the
               solution. The reference conductance for this solution is 1410 uS/cm at 25 °C.16
               The conductivity meter must yield a conductance between 1380 uS/cm and 1440
               uS/cm to be in calibration.

        10.4.3 If the conductivity meter fails calibration, recalibrate the unit per manufacture's
               instruction and repeat the procedure in 10.4.2 as if the standard solution were an
               unknown matrix.

11. PROCEDURE

   11.1 SAMPLE PREPARATION

        11.1.1 Samples do not need to be refrigerated but if samples are held refrigerated as a
               standard practice for sample control, ensure the samples have come to room
               temperature prior to conducting sample analysis.

        11.1.2 MATRIX CONDUCTANCE VERIFICATION - Prior to conducting the
               analysis of a field sample matrix,  the conductance of that matrix must be
               measured. Matrix conductivity is directly related to the common anion levels
                                          26

-------
       which, at high concentrations, can influence the integrity of the perchlorate
       analysis.

       11.1.2.1  Verify conductivity detector calibration by following the procedure
               outlined in Section 10.4.

       11.1.2.2  Pour approximately 15 mL of sample into a plastic disposable micro
               beaker (Section 6.7) and reseal the sample bottle to protect the sample
               integrity.

       11.1.2.3  Place the electrode into the matrix and measure the conductivity.

       11.1.2.4  If the conductance is less than the MCT, continue to Section 11.1.5.

       11.1.2.5  If the conductance is greater than the MCT, the matrix requires
               dilution or pretreatment prior to analysis.  The dilution procedure is
               found in Section 11.1.3.  Pretreatment is described in Section 11.1.4.

       11.1.2.6  Discard this aliquot of sample and be certain to thoroughly rinse the
               electrode with reagent water between each matrix conductivity
               measurement.

11.1.3  MATRIX DILUTION - If matrix conductivity is less than the MCT, go to
       Section 11.1.5.

       11.1.3.1  A sample  can be analyzed once diluted with reagent water to a
               conductance below the MCT.  The exact magnitude of this dilution
               will adversely increase the MRL by an equivalent proportion.

       11.1.3.2  Knowing the matrix conductance exceeds the MCT, estimate the
               proportion required for the dilution by dividing the measured matrix
               conductance by the MCT. Round up to the next whole number and
               dilute the  sample by a proportion equivalent to this value. For
               example, if the established MCT is 6100 uS/cm and a sample
               reflecting  a conductance of 8000 uS/cm was measured, dilute the
               sample with reagent water by a factor of 2.

       11.1.3.3  Measure the conductance of the diluted sample to confirm it is now
               below the MCT. Analyze the sample as specified in Section 11.1.5
               with the understanding that the MRL has now been elevated by a
               proportion equivalent to the dilution.

       11.1.3.4  If perchlorate is measured above the elevated MRL, back calculate
                                  27

-------
               actual field sample concentration and report. If no perchlorate is
               measured above the elevated MRL and analysis or project objectives
               required monitoring below the concentration of the elevated MRL,
               proceed to Section 11.1.4 and pretreat the matrix.

11.1.4  PRETREATMENT FOR MATRICES WHICH EXCEED THE MCT -  If
       matrix conductivity is less than the MCT, go to Section 11.1.5.  If sample
       dilution did not yield the required results, sample pretreatment should be
       employed. When the MCT is exceeded, it is most often due to a high levels of
       common anions (chloride, sulfate, and carbonate) in a particular matrix. If the
       analyst were to attempt the 1C analysis of this  particular matrix, the common
       anions present in the sample would distort the baseline and negatively affect the
       accurate quantitation of perchl orate. To effectively reduce a significant amount
       of these anions which contribute to the high conductivity reading, a series of
       pretreatment cartridges must be employed. For this pretreatment, three
       cartridges are attached in series in the following order: Ba, Ag, and H. It is
       recommended that all three cartridges be employed unless the analyst has
       specific knowledge that a matrix primarily has high levels of a specific common
       anion.

       11.1.4.1  Individually and thoroughly rinse each pretreatment cartridge with
               reagent water in order to insure all residual background contaminants
               are removed from the cartridge.  Perform this rinse per manufacturer's
               instructions.

       11.1.4.2  Prior to pretreating any field  samples, prepare and pretreat both an
               LRB and an LFB. These pretreated quality control samples are
               required when an analysis batch contains a matrix which must be
               pretreated.  This pretreatment is conducted by placing the cartridges in
               the following prescribed series (->Ba->Ag->H). The pretreated
               LRB and LFB are used to verify that no background interference or
               bias is  contributed by the pretreatment.  If a response is observed in
               the pretreated LRB, triple or quadruple the volume of reagent water
               rinse suggested by the manufacturer in Section 11.1.4.1 and repeat
               until a  blank measures no more than  /^ the MRL.  If this additional
               rinsing procedure is required, it must be consistently applied to all the
               cartridges prior to conducting any matrix pretreatment.

       11.1.4.3  Filter 3 mL of sample through the series of rinsed, stacked cartridges
               as an initial sample rinse (Ba, Ag and H) at a flow rate of 1.0 mL/ min
               or less  (approximately one drop  every 3 to 4 seconds). This flow  rate
               is critical to the pretreatment and must be carefully followed. Discard
               this fraction and begin collecting the pretreated sample aliquot of
                                  28

-------
               collected sample.

       11.1.4.4 When sufficient volume has been collected, measure the conductance
               of the pretreated sample aliquot being certain the conductivity meter's
               probe has been thoroughly rinsed and excess water has been shaken
               from the tip.  If the conductance is now below the MCT, the sample is
               ready for analysis. If the conductance is still above the MCT, the flow
               rate through the pretreatment cartridge is likely too fast and the
               pretreatment should be repeated with new cartridges. In some
               instances, double pretreatment cartridges may need to be applied.
               When this pretreatment is performed properly, U.S.EPA has found
               70% to 95% reduction in matrix conductance with good recoveries for
               perchlorate.

       11.1.4.5 Place this aliquot of pretreated sample into an autosampler vial as
               described in Section  11.1.3.

       11.1.4.6 In order to  ensure data quality, all samples which fail the MCT and
               have been selected for pretreatment, as described in Section 11.1.4,
               must also be used to prepare an LFM.  This LFM must be fortified
               with perchlorate at concentrations close to, but greater than, the level
               determined in the native sample prior to the pretreatment. Initially, the
               pretreated sample is analyzed and perchlorate  level is determined.
               Then, a second aliquot of sample must be fortified with perchlorate,
               pretreated to reduce the high common anion levels, and analyzed to
               assess perchlorate recovery from that matrix.  This additional QC is
               required to rule out matrix effects and to confirm that the laboratory
               performed the pretreatment step appropriately. If the perchlorate
               recovery falls outside the acceptance range of 80 -120% (Section
               9.4.1.4), that particular sample should be reported as
               suspect/matrix.

       11.1.4.7 The pretreatments prescribed above are effective at reducing the
               chloride and sulfate content of a sample matrix but will not reduce
               matrix concentrations of other anions such as nitrate or phosphate.

11.1.5  Pour approximately  15 mL of sample into a micro beaker (Section 6.7) and
       reseal the sample bottle to protect the sample integrity.  Using a Luer lock,
       plastic 10 mL syringe, withdraw approximately 10 mL of sample from the
       micro beaker and attach a 0.45  jam particulate filter (Section 6.11), which has
       been demonstrated to be free of ionic contaminants, directly to the syringe.
       Filter the sample into an autosampler vial or manually load the injection loop
       injecting a fixed amount of filtered, well mixed sample. If using a manually
                                  29

-------
            loaded injection loop, flush the loop thoroughly between sample analysis using
            sufficient volumes of each new sample matrix.

            11.1.5.1 If the autosampler vials or vial caps are designed to automatically filter
                    the sample matrix as the sample is loaded on the 1C system, this
                    filtration procedure can be omitted and the sample can be directly
                    transferred to the autosampler vial.
11.2 SAMPLE ANALYSIS

     11.2.1  Table 1 summarizes the recommended operating conditions for the ion
            chromatograph.  Included in this table is the estimated retention time for
            perchlorate which has been achieved by this method. Other columns,
            chromatographic conditions or detectors may be used if the requirements of
            Sections 1.2.1, 6.1.2.2 and 9.2 are met.

     11.2.2  Establish a valid initial calibration and verify this calibration by conducting a
            QCS as described in Section 10.2 and complete the IDC (Section 9.2). Initially,
            analyze the IPC  solution, followed by the LRB. Then confirm the 1C system
            calibration by analyzing an ICCS (Section 10.3.1) and,  if required, recalibrate as
            described in Section 10.2. Lastly, analyze the LFB.

     11.2.3  Inject 1.0 mL of each filtered sample. Use the same size loop for standards and
            samples. An automated constant volume injection system may also be used.
            Record the resulting peak size in area units and retention time for each analyte.

     11.2.4  The width of the retention time window used to make identifications should be
            based upon measurements of actual retention time variations of standards
            measured over several days. Three times the standard deviation of retention
            time may be used as a suggested window size but the retention time window
            should not extend beyond ± 5% of the retention time for perchlorate. The
            experience of the analyst should weigh heavily in the interpretation of these
            chromatograms.

     11.2.5  If the response of a sample analyte exceeds the calibration range, the sample
            must be diluted with an appropriate amount of reagent water and  reanalyzed.  If
            this is not possible then three new calibration concentrations must be employed
            to create a separate high concentration calibration curve, one standard near the
            estimated concentration and the other two bracketing around an interval
            equivalent to approximately ± 25% the estimated concentration.  The response
            generated by these three new high concentration calibration standards must not
            exceed the upper linear range for the conductivity detector. The latter procedure
                                       30

-------
            involves significantly more time than a simple sample dilution therefore, it is
            advisable to collect sufficient sample to allow for sample dilution and sample
            reanalysis, if required.

     11.2.6  Should more complete resolution be needed between perchlorate and a
            coeluting, shoulder peak, the eluent (Section 7.2) may be diluted.  This will
            spread out the peaks, causing later elution of perchlorate. Analysts are advised
            to carefully evaluate any of these eluent dilutions since when these eluent
            changes are incorporated, other coelutions may be encountered which were not
            initially evident.  Additionally, the analyst must verify that this dilution does not
            negatively affect performance by repeating and passing all the QC criteria in
            Section 9, and by reestablishing a valid initial calibration curve (Section 10.2).

            11.2.6.1 Eluent dilution will reduce the overall response of an anion due to
                    chromatographic band broadening which will be evident by shortened
                    and broadened peaks.  This will adversely effect the MDLs for each
                    analyte.

11.3  AUTOMATED ANALYSIS WITH METHOD 314.0

     11.3.1  Laboratories conducting analyses on large numbers of samples often prepare
            large analysis batches that are run in an automated manner. When conducting
            automated analyses, careful  attention must be paid to ensure sufficient volume
            of eluent in the reservoir is available to sustain extended operation.  In order to
            ensure their data are of acceptable quality, laboratories must ensure that all QC
            performance criteria are met throughout the analysis batch through subsequent
            careful inspection of the data.

     11.3.2  Analysis sequences must be carefully constructed to meet required QC
            specifications and frequency (Table 6). To help with this task, an acceptable
            sequence for a sample analysis batch, with all the method-required QC, is shown
            in Table 7.  This schedule is included only as an example of a hypothetical
            analysis batch which contains normal sample matrices as well  as samples which
            have failed the MCT. Within this analysis batch, references to exact
            concentrations for the ICCS, CCCS and ECCS are for illustrative purposes only.

     11.3.3  Table 7 may be used as a guide when preparing analysis batches.  Additional
            batches may be added sequentially on to the end of these types of schedules as
            long as all QC samples, which define an individual batch (IPC, LRB, ICCS,
            LFB, LFM, etc.) are individually reanalyzed with each successive serial batch
            and the QC criteria for these analyses are continually met (from the IPC through
            ECCS).
                                       31

-------
12. DATA ANALYSIS AND CALCULATIONS

   12.1 Identify perchlorate in the sample chromatogram by comparing the retention time of a
        suspect peak within the retention time window to the actual retention time of a known
        analyte peak in a calibration standard. If the perchlorate retention time has slightly
        shifted (generally towards shorter times) since the initial calibration, but is still within
        acceptance criteria and are reproducible during the analysis batch, the analyst should
        use the retention time in the daily calibration check standards to confirm the presence or
        absence of perchlorate anion.

        12.1.1 If a low concentration of perchlorate is suspected in an unknown sample, but the
               retention time has drifted to the edge of the retention time window, a low level
               perchlorate LFM, prepared at nearly the same concentration as the suspect peak,
               should be prepared from this sample matrix to confirm the matrix induced
               retention time shift.  If the fortified sample reveals a split or shouldering peak
               response, the low concentration in the unfortified sample is likely an interferant
               and should not be reported as perchlorate.

   12.2 Compute sample concentration using the initial calibration curve generated in Section
        10.2.

   12.3 Report ONLY those values that fall between the MRL and the highest calibration
        standards.  Samples with a perchlorate response which exceeds the highest calibration
        standard concentration must be diluted and reanalyzed.  When this is not possible the
        alternate calibration procedures described in Section 11.2.5 must be followed.  Samples
        with perchlorate identified but quantitated below the concentration established by the
        lowest calibration standard, may be reported as "trace present" above the MDL but
        below the minimum reporting limit (MRL) and therefore not reported as a quantitated
        concentration.

   12.4 Report results in |ig/L.

13. METHODS PERFORMANCE

   13.1 Table 1 gives the standard conditions, typical  retention time, single laboratory MCT and
        single laboratory MDL in reagent water, as determined for perchlorate. This retention
        time is graphically indicated in the chromatograms in Figures 1 through 4.

   13.2 Table 2 shows the precision and accuracy of the perchlorate measurement at two
        fortified  concentrations, in reagent water, simulated high ionic strength water (HIW),
        simulated high organic content water (HOW), ground water, untreated surface water
        and treated surface water. The mean perchlorate recovered concentration (accuracy
        relative to the fortified level) and the precision (expressed as %RSD of the replicate
        analysis) are tabulated. The HIW was designed to simulate a high ionic strength field
                                           32

-------
        sample and the HOW designed to simulate a high organic content field sample.  The
        HIW was prepared from reagent water which was fortified with the common anions of
        chloride at 400 mg/L, carbonate at 600 mg/L, and sulfate at 500 mg/L.  The HOW was
        prepared from reagent water fortified with 10.0 mg/L fulvic acid.

   13.3 Table 3 shows the stability data for perchlorate held for 35 days and stored under
        various conditions.  Conditions investigated included sample bottle construction
        (HDPE plastic vs. glass), storage condition (refrigerated vs. held at room temperature)
        and various matrices including some with a measured perchlorate concentration
        assumed to contain microbiological constituents acclimated to the presence of the
        anion.  Matrices without perchlorate were fortified at 25 ug/L. Each data point in this
        table represents the mean percent recovery following triplicate analyses. These  data
        were used to formulate the holding times shown in Section 8.3.

   13.4 Table 4, in conjunction with the chromatograms overlaid in Figure 4 as well as the
        linear regression plots in Figure 5, show the results of the single  laboratory MCT
        determination. The data presented in Table 4 and graphically illustrated in Figure 5,
        show results for not only the AS 16 but also the ASH and ASS.  The chromatogram
        shown in Figure 4 were generated using the AS 16 column.

14. POLLUTION PREVENTION

   14.1 Pollution prevention encompasses any technique that reduces or eliminates the quantity
        or toxicity of waste at the point of generation.  Numerous opportunities for pollution
        prevention exist in laboratory operation.  The EPA has established a preferred hierarchy
        of environmental management techniques that places pollution prevention as the
        management option of first choice.  Whenever feasible, laboratory personnel should use
        pollution prevention techniques to address their waste generation.  When wastes cannot
        be feasiblely reduced at the source, the Agency recommends recycling as the next best
        option.

   14.2 Quantity of chemicals purchased should be based on expected usage during its shelf-
        life and the disposal cost of unused material. Actual reagent preparation volumes
        should reflect anticipated usage and reagent stability.

   14.3 For information about pollution prevention that may be applicable to laboratories and
        research institutions, consult "Less is Better: Laboratory Chemical Management for
        Waste Reduction," available from the American Chemical Society's Department of
        Government Regulations and Science Policy, 1155 16th Street N.W., Washington D.C.
        20036, (202) 872-4477.

15. WASTE MANAGEMENT

   15.1 The Environmental Protection Agency requires that laboratory waste management
        practices be conducted consistent with all applicable rules and regulations.  Excess
                                          33

-------
        reagents, samples and method process wastes should be characterized and disposed of
        in an acceptable manner. The Agency urges laboratories to protect the air, water, and
        land by minimizing and controlling all releases from hoods and bench operations,
        complying with the letter and spirit of any waste discharge permit and regulations, and
        by complying with all solid and hazardous waste regulations, particularly the hazardous
        waste identification rules and land disposal restrictions. For further information on
        waste management consult the "Waste Management Manual for Laboratory Personnel,"
        available from the American Chemical Society at the address listed in Section 14.3.
16. REFERENCES

1.   "Determination of Perchlorate by Ion Chromatography." State of California, Department of
    Health Services, Sanitation and Radiation Laboratories Branch, Rev. No. 0 (June 3, 1997).

2.   "Analysis of Low Concentrations of Perchl orate in Drinking Water and Ground Water by
    Ion Chromatography." Application Note 121, Dionex Corporation,  Sunnyvale, CA (1998).

3.   "Perchlorate by Ion Chromatography, Modified EPA 300.0 Using lonPac ASH." Standard
    Operating Procedure, Montgomery Watson Laboratories (March 17, 1998).

4.   Jackson, P.E.; Laikhtman, M.; and Rohrer, J.S.  "Determination of Trace Level Perchlorate
    in Drinking Water and Ground Water by Ion Chromatography,"  Journal of Chromatography
    A, 850(1999), 131-135.

5.   Okamoto, H.S.; Rishi, O.K.; Steeber,  W.R.; Baumann, F.J.; and Perera, S.K. "Using Ion
    Chromatography to Detect Perchlorate," Journal AWWA. Vol. 91 (October 1999), 73-84.

6.   Inter-Agency Perchlorate Steering Committee, Analytical Subcommittee Report (1998).
    Report on the interlaboratory validation of 1C methods for perchlorate.

7.   Glaser, J.A.; Foerst, D.L.;  McKee, G.D..; Quave, S.A. andBudde, W.L.  "Trace Analyses
    for Wastewater," Environmental Science and Technology. Vol. 15,  Number 12, page 1426,
    December, 1981.

8.   Code of Federal Regulations 40, Pt. 136, Appendix B (July 1,  1998).

9.   "OSHA Safety and Health Standards, General Industry," (29CFR1910). Occupational Safety
    and health Administration, OSHA 2206, (Revised, Jan. 1976).

10. ASTM Annual Book of Standards,  Part II, Volume 11.01, D3370-82, "Standard Practice for
    Sampling Water," American Society for Testing and Materials, Philadelphia, PA, 1986.

11. "Carcinogens-Working with Carcinogens," Publication No. 77-206, Department of Health,


                                          34

-------
    Education, and Welfare, Public Health Service, Center for Disease control, National Institute
    of Occupational Safety and health, Atlanta, Georgia, August 1977.

12.  "Safety In Academic Chemistry laboratories," 3rd Edition, American Chemical Society
    Publication, Committee on Chemical Safety, Washington, D.C., 1979.

13.  U.S. EPA Method 300.1. EPA Document number: EPA/600/R-98/118.  NTIS number
    PB98-169196INZ.

14.  "Anion Self-Regenerating Suppressor (ASRS) Quickstart Procedure", Document Number
    031368-01, Dionex Corporation, Sunnyvale, CA, March,1988.

15.  "Installation Instructions and Troubleshooting Guide for the Anion Self-Regenerating
    Suppressor-Ultra", Document Number 031367, Rev. 03, Section 5.1, Dionex Corporation,
    Sunnyvale, CA, December, 1988.

16.  CRC Handbook of Chemistry and Physics.  Standard Solutions for Calibrating Conductivity
    Cells, p. D-166, 70th Ed., 1989-1990, CRC Press, Boca Raton, Florida.
                                         35

-------
17. TABLES. DIAGRAMS. FLOWCHARTS AND VALIDATION DATA
TABLE 1.    CHROMATOGRAPHIC CONDITIONS AND METHOD DETECTION
             LIMITS IN REAGENT WATER FOR PERCHLORATE.
Standard Conditions and Equipment(a):
Ion Chromatograph:
Sample Loop:
Eluent:
Eluent Flow:
Columns :
Typical System Backpressure:
Suppressor:
Detectors:
Determined MCT®:
Dionex DX500
1000 (A
SOmMNaOH
1.5 mL/min
Dionex AG 16,4 mm / AS 16, 4 mm
2600 psi
ASRS ULTRA (P/N 53946), external water mode, 300 mA current
Suppressed Conductivity Detector, Dionex CD20
Background Conductivity: 2 - 3 (iS

6100uS/cm
Recommended method total analysis time:     15 minutes (may be shortened to 12 minutes)
Analvte Retention Times and Method Detection Limits (MDLs):
Analyte
Perchlorate
Retention Time (c)
(min.)
10.1 ±0.2
MDL DETERMINATION
Fortified Cone. # of MDL
(Hg/L) Reps. (jjg/L)
2.0 7 0.53
(a)  Mention of trade names or commercial products does not necessarily constitute endorsement or
    recommendation for use.
(b)  This was the single laboratory MCT determined for these conditions listed (See Table 4 and Figure
    5 for more detail as well as data pertaining to the AS 11 and AS5).
(c)  Reference to chromatograms in Figure 1 through 4.
                                        36

-------
TABLE 2.    SINGLE LABORATORY PRECISION AND RECOVERY FOR
              PERCHLORATE IN VARIOUS MATRICES
Matrix Unfortified
Conductivity Cone.
Matrix uS/cm (Mg/L)
Reagent Water ~ 1

Synthetic High 4200
Inorganic Water ^
Synthetic High 5.0
Organic Water (c)
Ground Water 710
(high TDS)
Untreated Surface 460
Water
Chlorinated 460
Surface Water

-------
TABLE 3.   STABILITY STUDY RESULTS FOR PERCHLORATE IN VARIOUS MATRICES
IA. Stability when
Matrix
Reagent Water
Reagent Water
Reagent Water
Reagent Water
stored in various sampling bottles - All stored at room temperature
Bottle type
Clear Glass
Amber Glass
Opaque HDPE Plastic
Translucent HDPE
Plastic
Unfortified Fortified
Conc.(Mg/L) Conc.(ng/L) Day 0

-------
TABLE 4.      SINGLE LABORATORY RESULTS FOR THE DETERMINATION OF
                MCT - Determination on the AS16, ASH and the ASS.
4S16 Studies - Per chlorate fortified at 25 ug/L
Sample
LFB
MA(50)(a)
MA(IOO)
MA(200)
MA(400)
MA(600)
MA(800)
MAnnnm
Conductivity
uS/cm
<1
540
932
1770
3570
SOIO^
6450
7870
RT
min.
10.3
10.3
10.3
10.2
10.2
10.2
10.1
10?
Measured
C1O4-, ug/L
25.3
26.0
26.3
26.2
25.2
24.2
25.1
743
%Rec
101%
104%
105%
105%
101%
97%
100%
97%
Area
20268
20799
21060
20998
20170
19307
20038
19400
Height
1151
1135
1144
1112
1028
954
932
878
A/H ratio
17.6
18.3
18.4
18.9
19.6
20.2
21.5
77 1
PDA/H
0.00%
4.07%
4.54%
7.24%
11.4%
14.9%
22.1%
75 5%
4S11(C) Studies - Perchlorate fortified at 25 ug/L
Sample
LFB
MA(50)(a)
MA(IOO)
MA(200)
MA(400)
MA(600)
MA(800)
MAHOOO^
Conductivity
uS/cm
<1
540
932
1770 ^
3570
5010
6450
7870
RT
min.
8.9
8.9
9.0
9.0
9.0
9.0
8.9
8 8
Measured
C1O4-, ug/L
25.0
25.2
25.0
24.1
23.6
22.7
19.9
170
%Rec
100%
101%
100%
96%
94%
91%
80%
68%
Area
25213
25445
25192
24340
23855
22922
20243
17407
Height
1591
1515
1486
1384
1243
1101
870
678
A/H ratio
15.8
16.8
17.0
17.6
19.2
20.8
23.3
75 7
PDA/H
0.00%
5.98%
6.98%
11.0%
21.1%
31.4%
46.8%
67 0%
4S5(d) Studies - Perchlorate fortified at 25 ug/L
Sample
LFB
MA(50)(a)
MA(IOO)
MA(200)
MA(400)
MA(600)
MA(800)
MAnnnm
Conductivity
uS/cm
<1
540
932
1770 ^
3570
5010
6450
7870
RT
min.
9.7
9.7
9.7
9.7
9.6
9.6
9.6
96
Measured
C1O4-, ug/L
22.75
24.89
23.72
22.99
23.51
23.84
21.01
77 95
%Rec
91.0%
99.6%
94.9%
92.0%
94.0%
95.4%
84.0%
91 8%
Area
30348
33505
31776
30704
31474
31948
27792
30650
Height
1780
1751
1721
1591
1478
1441
1214
1183
A/H ratio
17.0
19.1
18.5
19.3
21.3
22.2
22.9
75 9
PDA/H
0.00%
12.2%
8.30%
13.2%
24.9%
30.0%
34.3%
57 0%
(a) "MA" indicates mixed common anion solution with each anion (chloride, sulfate and carbonate)
   included in the sample matrix at the parenthetical mg/L concentration for each anion.
(b) If the regression analysis is not performed on these data, 5010 uS/cm, 1770 uS/cm and 1770 uS/cm
   would be the default MCT for the AS 16, ASH and AS5, respectively, as described in Section 9.2.8.10.
   See Figure 5 for a graphical representation of this data, applying a regression analysis of PD^ vs
   matrix conductivity for the AS 16, AS 11 and AS5.
(c) ASH conditions: See reference #2 and #3.
(d) AS5 conditions: See reference #1.
                                            39

-------
 TABLE 5.     INITIAL DEMONSTRATION OF CAPABILITY QC REQUIREMENTS
	Requirements prior to beginning any analysis batch	
 Reference
  Requirement
     Specification and Frequency
   Acceptance Criteria
Sect. 9.2.2
     9.3.1
Initial
Demonstration of
Low System
Background
Analyze a method blank (LRB) and
determine that all target analytes are
below !/2 of the proposed MRL prior to
performing the IDC.
The LRB concentration
must be <1A of the
proposed MRL.
 Sect. 9.2.3
Initial
Demonstration of
Accuracy (IDA)
Analyze 7 replicate LFBs fortified with
perchlorate at 25 ug/L. Calculate the
mean recovered concentration (Cx)
See Equation in Section 9.2.3.
The Cx must be ± 10% of
true value.
 Sect. 9.2.4
Initial
Demonstration of
Precision (IDP)
Calculate percent relative standard
deviation (%RSD)of IDA replicates.
See Equation in Section 9.2.4.
The %RSD must be < 10%
 Sect. 9.2.5
Quality Control
Sample (QCS)
Initially, upon reestablishing calibration
or at least quarterly analyze a QCS from
an external/second source.
The QCS must be ± 10%
of the true value.
 Sect. 9.2.6
Method
Detection Limit
(MDL)
Determination
Select a fortifying level at 3-5 times the
estimated instrument detection limit.
Analyze 7 replicate LFBs over multiple
days and calculate MDL using equation in
Section 9.2.6 - do not subtract blank
 Sect. 9.2.7
Minimum
Reporting Level
(MRL)
An MRL should be established for
perchlorate during the IDC.
The low CAL standard can
be lower than the MRL,
but the MRL MUST be no
lower than the low CAL
standard
 Sect. 9.2.8
    Sect.
  9.2.8.11
Matrix
Conductivity
Threshold (MCT)
MRL verification
Prepare a series of LFB samples, each
containing a suggested perchlorate
concentration of 25 ug/L, at sequentially
increasing fortified levels of common
anions. Measure sample conductance and
analyze each,  calculate average A/H ratios
and PDA/H (using  equation in Section
9.2.8.8). Perform linear regression to
calculate MCT (using equation in Section
9.2.8.9) or follow step outlined in Section
9.2.8.10.

Verify the MRL in a solution prepared at
the MCT.
MCT, based upon linear
regression, is point where
PD^ equals 20%.

Alternatively, the MCT is
set at the highest measured
conductance observed in
the last fortified MCT
sample to yield a PD^
value below 20%.
Prepared within ±10% of
the MCT.
Perchlorate recovery must
be 70- 130% of the MRL.
 TABLE 6.     QUALITY CONTROL REQUIREMENTS (SUMMARY)
                                               40

-------
                Requirements specific for each analysis batch
 Reference
 Requirement
      Specification and Frequency
   Acceptance Criteria
  Sect. 8.3
Sample
Holding Time /
Preservation /
Storage
Perchlorate   28 days
No Preservation technique required.
Room Temperature adequate for shipping
and storage.
Holding time must not be
exceeded.
 Sect. 10.2
Initial
Calibration
Generate calibration curve. At least 5
calibration  standards are recommended.
MRL MUST be no lower
than the lowest calibration
standard
 Sect. 9.3.2
Instrument
Performance
Check (IPC)
Designed to verify Matrix Conductivity
Threshold (MCT). Prepare mixed common
anion solution at the MCT (prepared
consistent with procedures in Section
9.2.8). Confirm the sample's conductance
and analyze at the beginning of each
analysis batch.
Prepared within ±10% of
the MCT.

IPC solution conductance
verified to within ± 10%
of original measured value
(when originally prepared)

PDA/JJ, (when compared to
the A/HLFB) must be <
25%.

Perchlorate quantitated
between 80-120% of
fortified level.

<5% shift in perchlorate
retention time.
 Sect. 10.3.1
Initial
Calibration
Check (ICCS)
With each analysis batch, initially verify
calibration at the MRL by analyzing an
initial low-level continuing calibration
check standard (ICCS).
Recovery must be 75-
125% of the true value.
 Sect. 10.3.2
Continuing
Calibration
(CCCS) and
End Calibration
Checks (ECCS)
Alternately analyze separate mid and high
level CCCS/ECCS after every 10 samples
and after the last sample in an analysis
batch.
Recoveries must fall
 between 85- 115%
 Sect. 9.3.1
Laboratory
Reagent Blank
(LRB)
Include LRB with every analysis batch (up
to 20 samples)
Analyze prior to analyzing field samples
Perchlorate must be
< !/2 MRL
Sect. 9.3.1.1
PRETREATED
Laboratory
Reagent Blank
(LRB)
REQUIRED in any analysis batch which
includes samples which have exceeded the
MCT and have been pretreated in any way
to reduce the common anion levels.
Perchlorate must be
< !/2 MRL
                                               41

-------
   TABLE 6.     QUALITY CONTROL REQUIREMENTS (SUMMARY CONTINUED).
                  Requirements specific for each analysis batch
  Reference
  Requirement
     Specification and Frequency
  Acceptance Criteria
  Sect. 9.3.3
Laboratory
Fortified Blank
(LFB)
Laboratory must analyze LFB in each
analysis batch following the ICCS.
Calculate %REC prior to analyzing
samples.
The concentration selected for the LFB
in subsequent analysis batches should
be varied throughout the calibration
range.
Recovery for LFB MUST
be 85 - 115% prior to
analyzing samples.
Sample results from
batches that fail LFB are
invalid.
  Sect. 9.3.3.1
PRETREATED
Laboratory
Fortified Blank
(LFB)
REQUIRED in any analysis batch
which includes samples which have
exceeded the MCT and have been
pretreated in any way to reduce the
common anion levels.
Fortification must be made prior to
pretreatment
Recovery for pretreated
LFB MUST be 85-115%
prior to analyzing samples.
Sample results from
batches that fail a
pretreated LFB are invalid.
  Sect. 9.4.1
Sect. 11.1.4.6
Laboratory
Fortified Sample
Matrix (LFM)
SPECIAL LFM
for matrices
requiring
pretreatment
Must add known amount of perchlorate
to a minimum of 5% of field samples or
at least one within each analysis batch.

LFM must be fortified above the native
level and at no greater than 10 x the
highest field  sample concentration.
Calculate target analyte recovery using
formula (Sect. 9.4.1.3).

When a sample exceeds the MCT and
pretreatment is employed to reduce the
common anion levels, an additional
LFM must be prepared from this matrix
and subsequently pretreated exactly as
the unfortified matrix.
Recovery must be
80 - 120%


If fortified sample fails the
recovery criteria, label
both as suspect/matrix.
Same criteria, recoveries
must be 80-120%.
  Sect. 9.4.2
Field or
Laboratory
Duplicates or
LFM Duplicate
Analyze either a field, laboratory or
LFM duplicate for a minimum of 5% of
field samples or at least one within each
analysis batch.
Calculate the relative percent difference
(RPD) using formula in Section 9.4.2.1.
RPDmustbei 15%.
  Sect. 6.1.2.2
ALTERNATE 1C
analytical column
performance
criteria
If a laboratory chooses an alternate
analytical column for this analysis, it
must be hydrophilic and pass the
criteria for Peak Gaussian Factor (PGF)
using equation (Sect. 6.1.2.2).
PGF must fall between
0.80 and 1.15.
                                               42

-------
TABLE 7.     EXAMPLE SAMPLE ANALYSIS BATCH WITH QUALITY CONTROL
            REQUIREMENTS
Injection
#
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Sample
Description
Instrument Performance Check Standard at MCT
Laboratory Reagent Blank (LRB)
ICCS at the MRL (4.0 Mg/L)
Laboratory Fortified Blank (LFB)
Sample 1
Sample 1 - Laboratory Duplicate (LD) (a)
Sample 2
Sample 2 - Laboratory Fortified Matrix (LFM) (a)
Sample 3
Sample 4
Sample 5
Sample 6
Sample 7
Sample 8
Sample 9
Sample 10
CCCS(25.0ug/L)
Sample 1 1 (failed MCT, matrix conductance = 8000 uS/cm)
- Analyzed diluted (Section 11.1.3) by factor of 2 or by
50% with reagent water (diluted matrix conductance =
3800uS/cm).
Sample 12
Sample 13
Acceptance
Criteria
PD^H forIPC<25%
< !/2 MRL
3.00to5.00Mg/L
Recovery of 85 - 115%
normal analysis
± 15%RPD
normal analysis
Recovery of 80 - 120%
normal analysis
normal analysis
normal analysis
normal analysis
normal analysis
normal analysis
normal analysis
normal analysis
21.3to28.8ug/L
MRL increases from 4 to 8
ug/L, noted in analysis report -
sample found to contain 50 ug/L
(measured at 25 ug/L in diluted
sample)
normal analysis
normal analysis
CONTINUED TO NEXT PAGE
                                     43

-------
Injection
#
21
22
23
24
25
26
27
28
29
30
31
Sample
Description
Sample 14 (failed MCT, matrix conductance=15000 uS/cm)
Analyzed diluted (Section 1 1 . 1 .3) by a factor of 3 or by
33% with reagent water (Diluted matrix conductance =
4600 uS/cm)
Ba/Ag/H Pretreated LRB (Section 9.3.1.1)
Ba/Ag/H Pretreated LFB (Section 9.3.3.1)
Sample 14 - Ba/Ag/H pretreated (Section 11.1.4), following
pretreatment the matrix conductance = 230 uS/cm.
Sample 14 ^ - pretreated LFM (Section 11.1.4.6)
Sample 15
Sample 16
Sample 17
Sample 18
Sample 19W
ECCS (100Mg/L)
Acceptance
Criteria
MRL increases from 4 to 12
ug/L, noted in analysis report -
No perchlorate > 12ug/L
measured - project required
monitoring to MRL - sample
pretreatment is therefore
required
< !/2 MRL
Recovery of 85 - 115%
normal pretreated analysis
perchlorate < MRL of 4.0 ug/L
Recovery of 80 - 120%
normal analysis
normal analysis
normal analysis
normal analysis
normal analysis
85.0 to 125Mg/L
(a)   If no analytes are observed above the MRL for a sample, an alternate sample which contains reportable
    values should be selected as the laboratory duplicate. Alternately, the LFM can be selected and reanalyzed
    as the laboratory duplicate ensuring the collection of QC data for precision.

^   Sample #19 (inj #30) was the final field sample permitted in this batch but 20 total field samples were
    analyzed. Sample #14 (inj #21 and #24) was analyzed both initially as a diluted sample and subsequently
    as a pretreated sample, therefore it accounted for two "field sample analyses" toward the maximum of
    twenty in an analysis batch (Section 3.1).

Note:   Sample #11 and #14 illustrate examples of proper ways to handle sample matrices which exceed the
        MCT.
                                                 44

-------
FIGURE 1.   CHROMATOGRAM OF LOW LEVEL PERCHLORATE (4.0 ug/L) IN REAGENT WATER

              (Conditions as indicated in Table 1)
      0)
      a:
      O

      O
           0.200
           0.150-
           0.100-
           0.050
           -0.100
                           2.00
                                            Perchlorate at 4.0 ug/L in Reagent Water
4.00
                                                                               CIO4-
6.00         8.00

      Minutes
10.00
12.00
14.00
                                                            45

-------
FIGURE 2.   CHROMATOGRAM OF 25 ug/L PERCHLORATE IN REAGENT WATER
             (Conditions as indicated in Table 1)
=3
CL

or

Q
c
O
0.500


0.400


0.300 —


0.200 —


0.100
           -0.100
                            2.00
                                             Perchlorate at 25 ug/L in Reagent Water
                                  4.00
                                                                               CIO4-
6.00
                                                      8.00
10.00
12.00
14.00
                                                           Minutes
                                                             46

-------
FIGURE 3.    STACKED CHROMATOGRAMS INDICATING INFLUENCE OF HIGH CONCENTRATIONS OF COMMON ANIONS ON
             LOW CONCENTRATION MEASUREMENT OF PERCHLORATE AT 4.0 ug/L (Conditions as indicated in Table 1)
                                            Common Anion Effect on trace CIO4-
         0.500  -j
         0.400
      (D
      ^  0.300  -\
      Q_
      (/)
      0)
      a:
      ~  0.200  -i
      Q
      O  0.100
      o
        -0.100
                                                 Above MCT
                                                 Below MCT
                                 	 MA(1000), matrix cond=7820 uS/cm
                                 	 MA(800), matrix cond=6450 uS/cm
                                 	 MA(600), matrix cond=5010 uS/cm
                                 	 MA(200), matrix cond=1770 uS/cm
                                 	 LFB at4.0ug/L, matrix cond <1 uS/cm
                                                     Perchlorate at 4.0ug/L
                          2.00
4.00
6.00
      8.00
Minutes
10.00
12.00
14.00
                                                          47

-------
FIGURE 4.    STACKED CHROMATOGRAMS INDICATING INFLUENCE OF HIGH CONCENTRATIONS OF COMMON ANIONS ON
             PERCHLORATE AT 25 ug/L DURING THE MCT DETERMINATION (Conditions as indicated in Table 1)
                                           Common Anion Effect for MCT analyses
         0.500
         0.400
         0.300

         0.200
       o 0.100
        -0.100
                                                                                         Above MCT
                                                                                         Below MCT
                                  	 MA(1000), matrix cond=7820 uS/cm
                                  	 MA(800), matrix cond=6450 uS/cm
                                  	 MA(600), matrix cond=5010 uS/cm
                                  	 MA(200), matrix cond=1770 uS/cm
                                  .— LFB at 25ug/L, matrix cond <1 uS/cm
                                                     Perchlorate at 25.0ug/L
                           2.00
4.00
6.00
      8.00
Minutes
10.00
12.00
14.00
                                                           48

-------
FIGURE 5.  REGRESSION ANALYSIS OF THE MCT DETERMINATION DATA


8000

o
3
01
o
c 4nnn
s 4000
O
3
•c
c onnn
o ^uuu
O


0°
AS16
MCT Regression analysis

Regression Data
R-square = 0.986 # pts = 8
y = 32076x-338 6100
600 _^****'
^^^^
V^_
20>--* 	
JJP'* 	 "
D ^^^^^^M
\**^^
k 5% 10% 15% 20
PD (A/H)


1000
UJ^-****<*a
jS/crf|f


respective mixed anion
samples
O Regression Line
A MCTindicated at 6100
i iS/cm

X 25% 30'












/o
9000
8000
g 7000
;§ eooo
rf 5000
| 4000
3 3000
•c
0 2000
O
1000
0
0
AS11
MCT Regression analysis
Regression DC
R-square = 0.977 #
y= 13136x + 2



^^2850 '
^^**^
^r
ta ^^*
-------