PB82-257817
Speciation of Arsenic Compounds in Water Supplies
Kurt J. Irgolic
Texas A&M University
College Station, Texas
June 1982
                       U.S.  DEPARTMENT OF COMMERCE
                    National Technical Information Service

-------
                                            PBS2-257817


                                               EPA 600/1-82-010
                                               June 1982
       SPECIATION OF ARSENIC COMPOUNDS
              IN WATER SUPPLIES
                      by

               Kurt J. Irgolic
Department of Chemistry, Texas ASM University
        College Station, Texas  77843
                R 8047 740 10
               Project Officer

            Frederick C. Kopfler
    Toxicology and Microbiology Division
     Health Effects Research Laboratory
          Cincinnati, Ohio  45268
     HEALTH EFFECTS RESEARCH LABORATORY
     OFFICE OF RESEARCH AND DEVELOPMENT
    U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA  27711
                           . • •
            NATIONAL  TECHNICAL
            INFORMATION SERVICE
               IS. OEPAimiEfir OF COMMERCE
                 SFtMGFIUO. VA. 22161

-------
                                     TECHNICAL REPORT DATA
                              (Please read Instructions on the reverse before completing)
   1. REPORT NO.
    EPA-600/1-82-010
                                 ORD  Report
3. RECIPIENT'S ACCESSION NO.
 PBB7  '  9 g -7 a 1 7
  4. TITLE AND SUBTITLE
    Speciation of.Arsenic  Compounds in Water Supplies
                                                           5. REPORT DATE '
                                                                June  1QS?
                                                             6. PERFORMING ORGANIZATION CODE
  7. AUTHOR(S)
    Kurt J. Irgolic
                                                             8. PERFORMING ORGANIZATION REPORT NO.
  9. PERFORMING ORGANIZATION NAME AND ADDRESS
    Texas A&M University
    Department of Chemistry
    College Station, Texas  77843
                                                              10. PROGRAM ELEMENT NO.
                                                           11. CONTRACT/GRANT NO.
                                                                     R-804774010
   12. SPONSORING AGENCY NAME AND ADDRESS
   USEPA
   Toxicology & Microbiology Division
   Health Effects Research Laboratory
   Research Triangle Park,  North Carolina 27711
                                                              13. TYPE OF REPORT AND PERIOD COVERED
                                                           14. SPONSORING AGENCY CODE
   13. SUPPLEMENTARY NOTES .
\
16. ABSTRACT                                         .''•'.-••'
The objectives of this project  were to develop  and test analytical, methods that;would
allow the qhemical form (i.e. valence state or compound) of arsenic in drinking waters to
be determined,  and to use the methods to analyze samples  of  drinking water from.sources
where.adverse health effects  in  consumers had been .attributed to arsenic.   Analytical
techniques  were   developed  for   the  determination  of  arsenate:  (differential  pulse
polarography),  for  inorganic  and  organic  arsenic  compounds  (high  pressure 'liquid
chromatography with  graphite furnace atomic absorption  spectrometry as element-specific
detector) and for the detection of arsenocholine,  arsenobetaine, and  iodoarsines (mass
spectrometry).   These  techniques, -inductively coupled argon  plasms  emission  spectro-
metry, and hydride generation/DC-helium ar« emission were used  for  the characterization
of water samples  from Utah,  Alaska,  Antofagasta,  Taiwan and  Nova Scotia.  The. total
arsenic  concentrations were in the range  18 ppb to 8  ppra with arsenic/arseriate ratios
between  0.007 and 3.4.   No organic arsenic  compounds  were detected  in any of the water
samples.  The1 trace  elements A1;,B, Ba,  Ca,  Cu, Fe,  Li.,  Mg, Mn,  Na,. P,. S, Si and Sr were
present  in most of the water samples..  The  results  show that, the various  physiological
effects,  observed  in  populations exposed to  the arsenic-containing water supplies could
not be caused by arsenic compounds other than arsenite or arsenate. . Other, trace elements
acting in concert  with  arsenite  and/or arsenate might,produce these  symptoms.   However,
sufficient data are  not yet available to  evaluate  th4sg hypothesis.          ..	
  17.
                                 . KEY WORDS AND DOCUMENT ANALYSIS
                    DESCRIPTORS
                                                b.IDENTIFIERS/OPEN ENDED TERMS
                                                                        c.. COSATI Field/Group
   18. DISTRIBUTION STATEMENT

     Release to Public.
                                              19. SECURITY CLASS /Tliis Report!
                                                    Unclassified
              21. NO. OF PAGES
                   124
                                                20. SECURITY CLASS (This page)
                                                      Unclassified
                                                                           22. PRICE
   EPA Form 2220-1 (Rev. 4-77)   PREVIOUS EDITION is OBSOLETE

-------
               NOTICE

THIS. DOCUMENT" HAS BEEN  REPRODUCED
FROM, THE; BEST COPY  FURNISHED us.BY
THE SPONSORING AGENCY,,/A.LTH,OUGH: IT
IS RECOGNIZED' THAT. CE.RTAIN PORTIONS
ARE: ILLEGIBLE;,, ix is; BEING RELEASED
IN THE. IN.TE.REST: OF..MAKING; AVAILABLE
AS MUCH INFORMATION AS  POSSIBLE,

-------
                     NOTICE
Mention of trade names or commercial products does not
constitute endorsement or recommendation for use.

-------
                                 FOREWORD
      The  many benefits of our modern,  developing,  industrial society are
 accompanied by certain hazards.  Careful assessment  of  the relative risk of
 existing  and  new man-made  environmental  hazards   is necessary  for the
 establishment of  sound  regulatory policy.   These  regulations  serve  to
 enhance the quality of our environment in order to promote the  public health
 and welfare and-the  productive capacity of our Nation's population.

      The-complexities  of the environmental problems originate in the deep
 interdependent  relationships between the various physical  arid biological
 segments  of  man's  natural  and  social world.   Solutions  to  these en-
 vironmental problems require  an  integrated program of  research  and deve
 lopment  using input from a  number of  disciplines.  The  Health Effects
 Research  Laboratory  conducts a coordinated  environmental  health research
 program  in inhalation toxicology,  genetic toxicology,  neurotoxicology,
 developmental and experimental biology  and clinical  studies using human
 volunteer subjects.  These studies address  problems  in  air pollution, water
 pollution,  non-ionizing radiation,'environmental- carcinogenesis  and the
 toxicology  of pesticide-s  and  other chemical pollutants.   The Laboratory
 participates  in  and provides data  for the development and  revision  of
 criteria documents on pollutants for which  national  ambient air quality and
 water quality  standards exist  or are  proposed,  provides  the  data for
 registration  of  new  pesticides or proposed suspension of those already in
 use,  conducts research oh hazardous and toxic  materials, and is primarily
 responsible   for  providing the  health  basis  for  non-ionizing  radiation
 standards.   Direct  support  to the regulatory  function of the  Agency  is
 provided  in the  form of expert testimony and preparation of affidavits as
 well  as expert  advice  to  the : Administrator  to- assure the  adequacy  of
.environmental regulatory decisions involving the protection of the health
 and welfare of  all U.;;S.: inhabitants.   ;.,,..;:'.,,•/ '••"•     :  .-•.•--. .•-,  ..••.".

      This  report contains  a description  of methods  that can  be used  to
 determine  the  specific arsenic  ions and compounds  that may be present in
 drinking waters and have heretofore been reported as  total arsenic.  It also
 contains  the  results of  the application  of these  methods to  samples  of
 several drinking  waters  known  to contain arsenic, which  in earlier studies
 has been  implicated  as the cause of disease .in:consumers of these^waters.

                                    F.  G. Hueter, Ph.D.        •"".'.•'•"
                       ''..,. .    , ' '  :      Director ••'.'' .':• .•  " -  ". •••'"•••••'.•-
                            Health Effects. Research  Laboratory    ;
                                    ill

-------
                                  ABSTRACT
     The objective of this project was to determine qualitatively and
quantitatively the arsenic compounds present in drinking water supplies which
are known to have caused adverse health .effects..

     Water samples from Hinckley, Utah; Delta, Utah; the Barefoot Site,
Alaska; the Mauer Site, Alaska; Antofagasta, .Chile; Yenshei, Taiwan; and two
sites in Nova Scotia were; analyzed for total arsenic-, arsenite, arsenate,
arsenic compounds reducible to methylarsines and other organic arsenic
compounds.  Graphite furnace atomic absorption spectrometry, differential
pulse polarography, hydride generation/DC-helium arc emissionj high pressure
liquid chromatpgraphy with a. graphite furnace atomic absorption spectrometer
as an element-specific detector  (HPLC-GFAA) and simultaneous inductively
coupled argon plasma emission spectroscopy  (ICP) were employed as analytical
techniques for the determination of. total arsenic and arsenic compounds.  ICP
was also used to determine other trace elements in the water samples.  The
total arsenic concentrations ranged from 18 ppb to 8 ppm with arsenite/
arsenate ratios between 0.007 and 3.4.  No organic arsenic compounds reduci-
ble, to methylarsine or dimethylarsine were, detected by the hydride generation
technique, which has a detection limit of 1 ppb for these compounds.  The
HPLC-GFAA system gave no indication of the presence of methylated or other
organic compounds. , The following trace elements were found in almost all
water samples:  Al, B, Ba, Ca, Cu, Fe, K, Li, Mg, Mn, Na, P, S (probably
present as sulfate), Si and Sr.

     Various physiological effec'ts (melanosis, keratosis, skin cancer,
vascular ailments,, pneumonia) have been attributed to arsenic.  These effects
cannot be caused by arsenic compounds other than arsenite and/or arsenate.
The presence of arsenite and arsenate in different ratios in the water
samples or one or more of the other trace elements acting in concert with
arsenite and/or arsenate might produce the various symptoms.  Sufficient
data are not yet available to evaluate these hypotheses.

     The water samples from Yenshei, Taiwan, were investigated to determine
the nature of-the dissolved fluorescent compounds.  High pressure liquid
chromatography employing UV- and fluorescence-detectors indicated that
ID-lysergic acid and/or ergometrine, ergotamine and perhaps calciferol might
be present in these water samples.  Several unknown, UV-detectable and
fluorescent compounds were also  found in these samples.  These results call
for a reevaluation of the long-held belief  that arsenic alone caused the
adverse health effects in the population in Taiwan exposed to these waters.

     As part of this project the following  techniques were developed for the
                                     iv

-------
determination of arsenic compounds:  differential pulse polarography for
arsenate; a high pressure liquid chromatography system with a Hitachi Zeeman
graphite furnace atomic absorption spectrometer as an element-specific
detector for arsenate, arsenite, methylarsonic acid, dimethylarsinic acid,
arsenocholine and arsenobetaine; mass spectrometry for organylarsonic acids,
diorganylarsinic acids, organyldiiodoarsines, diorganyliodoarsines,
arsenocholine and arsenobetaine.
                              X
     This report was submitted in fulfillment of Grant No. R 8047 740 10 by
Texas A&M University under the sponsorship of the U.S.. Environmental
Protection Agency.  This report covers the period'November 1, 1976 to
February 28, 1981, and work, was completed as of March 31, 1981.
                                      v

-------
                                  CONTENTS
Foreword	. .	iii
Abstract .	iv
Figures	vii
Tables	   x
Abbreviations and Symbols	xiii
Acknowledgments  ........... 	  xv

   1.   Introduction  .»•  .  .  .  . . . .  . .  .  .-•'•.•  . ...  .  .  .  .,  ....   1
   2.   Conclusions	   2
   3.   Recommendations  .;....	   3
   4.   Arsenic Compounds  in Natural Water Samples   ...  .  .  .  •  .  •  •   4
   5.   Preservation of Arsenic Compounds  in Aqueous Solutions   ....   7
   6.   Development of Analytical Techniques for the Determination
         of Arsenic Compounds	   12
            Hitachi Zeeman Graphite Furnace Atomic
              Absorption  Spectrometer as  an Element-
              Specific Detector for High  Pressure
              Liquid Chromatography   . ...  .	   14
            The Speciation of Arsenite, Arsenate,
              Methylarsonic Acid, Dimethylarsinic
              Acid,. Arsenocholine and Arsenobetaine
           .   Using the HPLC-Hitachi  Zeeman GFAA
              System  ................	   24
            Differential  Pulse Polarograhic         /       .     .
              Determination of Arsenate and
              Arsenate  .  .  ...  •	   32
            Mass Spectrometry of Organylarsonic
              Acids, Diorganylarsinic Acids,
              Organyliodoarsines, Arsenocholine
              and Arsenobetaine	   53
   7.   Analysis .of Water  Samples  .	   65
            Water Samples from Hinckley and
              Delta, Utah	   68
            Water Samples from Antofagasta, Chile   	   74
            Water Samples from Alaska	   80
            Water Samples from Taiwan  .	   84
            Water Samples from Nova Scotia	   98
            Discussion	   100

References	   103
                                      vi

-------
                                  FIGURES

Number                                                                    Page

  1   Transformation of arsenic compounds in  the
        environment .	    5

  2   Block diagram for the HPLC-GFAA system	   15

  3.  Schematics for the HPLC-GFAA interface	   17

  4   The sampling sub-system of the HPLC-GFAA
        interface . . . ....... . . .  . . .	.  .  .  ...  .   19

  5   The injection system for the HPLC-GFAA
  ;    .interface ••.,,. . . . . . . . . ..... .  . ..  .  ...  .  .  .  ...   21

  6   Separation of inorganic arsenic (arsenite                             r
        and arsenate), arsenocholine and  :
        arsenobetaine by HPLC-GFAA.on.a  C-18       .
        reversed-phase column . . . . .  . . .  . .  .  ...  .  .	   25

  7   Separation of arsenite and arsenate by                        .
        HPLC-GFAA on a C-18 reversed-phase       .
        column in the absence and presence of
        tetrabutylammonium phosphate  .  . . .  ....  .  ..,..-. .  .  .  ...   27

  8   The influence of phosphate on the  GFAA
        arsenic signals ......	   28

  9   The separation of arsenite, arsenate, ........             ,  ,
        dimethylarsinic acid and. methylarsonic
        acid by HPLC-GFAA on a C-18 reversed-:phased
        column in the presence of tetraheptylammonium
        nitrate . . . .	 . .  . .  .  .  .  .  .  .  .  .  ...   29

 10   HPLC-GFAA calibration curve for sodium  arsenite  .  .  .  .  .  .  .  /  .   30

 11   HPLC-GFAA calibration curves for sodium arsenite,
        dimethylarsinic acid, methylarsonic acid and
        disodium hydrogen arsenate  . .  . . . :.: .  .  ....  .  .  .  .  .  .   31

 12   Differential pulse polarograins of  arsenate
        solutions two molar with respect to perchloric
        acid in the presence of ^D-mannitol  ...  '. ••;......  ........   34

                                      vii

-------
13   The dependence of the height of the differential
       pulse polarographic curve of arsenate at
       -0.55 V on the concentration of I>-mannitol	   35

14   Differential pulse polarographic curves for
       10 ppm As (arsenate)in solutions 0;5 M with
       respect to &-mannitol at various perchloric
       acid concentrations . .	. .	   37

15   Calibration curves for the determination of
       arsenate by differential pulse polarography
       in aqueous 20 M perchloric acid in the
       presence of IKmannitol	   39

16   Differential pulse polarograins for arsenite
       in 2.0 M aqueous perchloric acid in the
       presence and absence of D-mannitol	   40

17   Calibration curve for the determination of
       arsenite by differential pulse polarography
       in aqueous 2.0 M perchloric acid in the
       presence of J>-mannitol  . .	   41

18   Differential pulse polarograms of mixtures
       of arsenite and arsenate in 2.0 M
       perchloric acid in the presence of D-mannitol	   42.

19   Differential pulse polarograms of solutions
       containing 20 ppb As(arsenite) and 40 to
       100 ppb As(arsenate) in 2.0 M perchloric
       acid in the presence of D-mannitol	   43

20   Differential pulse polarograms of arsenite,
       and of arsenate obtained by Ce(IV)
       oxidation of arsenite 	   45

21   The dependence of the peak height of the
       differential pulse polarographic maximum
       at -0.34 V (arsenite reduction) as a
       function of the amount of cerium(IV)
       ammonium nitrate added  ........ 	 	   46

22   Differential pulse polarograms of 500 ppb
       As(arsenate) obtained by cerium(IV)
       oxidation of arsenite in 2.0 M perchloric
       acid, followed by addition of hydroxylamine
       hydrochloride, addition of J>mannitol and
       standard addition of arsenate 	   47
                                      viii

-------
23   Calibration curves for the determination of
       arsenate obtained by oxidation of arsenite
       with cerium(IV)	   48

24   Flow chart for the determination of arsenite
       and arsenate by differential pulse polarography 	   51

25   The mass spectrum of methylarsonic acid . .	   54

26   The mass spectrum of dimethylarsinic acid	   55

27   The mass spectrum of butyldiiodoarsine	   61

28   The mass spectrum of dioctyliodoarsine  .............   62


29   The mass spectrum of arsenocholine chloride .	 .63

30   The mass spectrum of arsenobetaine chloride .....  .  .  ....... .  . .   64

31   Cation exchange liquid chromatogram of the          ...,,.,..         .
       concentrated Yenshei'water' sampled .  .  .'f; .  .  ...  .....   94

32   Ion-pair reversed phase chromatogram of the
       concentrated Yenshei water sample and
       alkaloid standards employing an UV-detector ....  	   95

33   Ion-pair reversed phase chromatogram of the
       concentrated Yenshei water sample and
       alkaloid standards employing a fluorescence
       detector	  .  .  .  .  .  . .   96

34   Ion-pair reversed phase chromatograms of the
       aged, concentrated Yenshei water sample                  :
       employing a UV-detector at 254 nm and a
       fluorescence detector	-'.,	   97

35   Rapid scanning fluorescence analysis of
       unconcentrated Yenshei well water ...............   98
                                   ";-  ix

-------
                                   TABLES

Number                                                                    Pa8e

1    Specifications and Sources for Electronic
       Components for the HPLC-GFAA Interface 	    18

2    Peak Heights of the Differential Pulse
       Polarographic Curves at -0.55 V Obtained
       with 2.0 M Aqueous Perchloric Acid
       Solutions of 100 ppm As(Arsenate) in the
       Presence of Various Polyhydroxy Compounds  	    33

3    Heights of the First Differential Pulse
       Polarographic Maximum in the Reduction of
       9.5 ppm As(Arsenate) in Aqueous Acidic Media
       in the Presence of D-Mannitol at 0.5 Molar
       Concentration	    36

4    Relative Abundances of Ions in the Electron
       Impact Mass Spectra of Organylarsonic Acids  .....  	    57

5    Relative Abundances of Ions in the Electron
       Impact Mass Spectra of Diorganylarsinic Acids	.  ,    58

6    Relative Abundances of Ions in the Electron
       Impact Mass Spectra of Organyldiiodoarsines -... .........    59

 7    Relative Abundances of Ions in the Electron
       Impact Mass Spectra of Diorganyliodoarsines      	    60

 8    Total Arsenic Concentrations in ppb in the
       Hlnckley and Delta Water Samples 	  .....    69

 9    Concentrations of Arsenite and Arsenate in
       ppb in the Hinckley and Delta Water Samples	    71

10    Trace Element Concentrations in the Hinckley
       Water Samples Determined by .ICP	    72

11   Trace Element Concentrations in the Delta Water
       Samples Determined by ICP	    73

-------
12   Total Arsenic Concentrations in ppm in the
       Antofagasta Water Samples  	  	    75

13   Concentrations of Arsenite and Arsenate in ppm
       in the Antofagasta Water Samples	    76

14   Trace Element Concentrations in the Untreated
       Antofagasta Water Samples Determined by ICP  	    77

15   Trace Element Concentrations in the Treated
       Antofagasta Water Samples Determined by ICP  	    78

16  • Results of Total Element Analyses for Water
       Samples from Antofagasta, Chile, Collected
       and Analyzed in 1977	    79

17   Results of Total. Arsenic Analyses and Arsenite
       and Arsenate Determinations in Alaska Water
       Samples  . . 	 ...........  	    81

18   Trace Element Concentrations Determined by ICP
       in the Barefoot Water Samples from Alaska                    .
       (Collected in 1979).  . . . . . .... ......	  . .  .   ,82,

19   Trace Element Concentrations Determined by ICP
       in the Barefoot Water Samples from Alaska.
       (Collected in 1980)  . . . ... . ,.'	    83

20   Trace Element Concentrations Determined by ICP
       in the Mauer Water Samples from Alaska
       (Collected in 1979)  . . ". . . . .v:. ..".-. .  .  .".  .  . ';' . .  .    85

21   Trace Element Concentrations Determined by ICP
       in the Mauer Water Samples from Alaska
    '   (Collected in 1980)  . ....... ..'..............  .",...86

22   Analysis Results for the Pei-Men and Pu-Tai
       Water Samples	    87

23   Results of Total Arsenic Analyses and Arsenite
       and Arsenate Determinations in the Yenshei
       Water Samples	    89

24   Trace Element Concentrations in the Yenshei  <
       Water Samples, Well. I, Determined by ICP ............   .90

25   Trace Element Concentrations in the Yenshei
       Water Samples, Well II, Determined by ICP   ...........    91
                                     xi

-------
26   Analysis Results for the Nova Scotia Water Samples	     99

27   Summary of Total Arsenic, Arsenite, Arsenate and
       Trace Element Concentrations in Drinking Water
       Samples	    101

28   Arsenic-Containing Water Supplies and Their
     Physiological Manifestations in Man	    102
                                     xii

-------
                       LIST OF ABBREVIATIONS AND SYMBOLS
 ABBREVIATIONS

 A          — ampere
 AA         — atomic absorption
 AAS        — atomic, absorption spectrometry
 AC         — alternating  current
 Ag-DDC     — silver diethyldithiocarbamate
 cone..      — concentrated
 DC         — direct current
 DMAA       — dimethylarsiriic  acid
 DPP        — differential pulse polarography (polarographic)
 g     •     — gram.'  '  • '   .-• •  • •• •    '"  .  "' •.. ••' -••••:•      ' ' -  •-.-••   • ••  •-•-•••'
 GFAA       — graphite  furnace-atomic  absorption spectrometry    •
               (spectrometer, spectrometric)
 HG         — hydride generation            ;  •
 HPLC-GFAA  — high pressure:liquid chromatography-graphite furnace atomic
               absorption spectrometer  system
 ID      •   — inner diameter   ..'."•  ".-•' • '\-'''•'•.'. :'':.;;.  •'  •   7   •-•••.•
 ICP        —inductively  coupled  argon plasma emission spectrometry
               (spectrometer, spectrometric),  •. v-•      .        .
 L   •  .  .. . ,'T-..liter •','  ._ •-;   '   -• ••     .-.•••;  -;  '••-... ..  .•'.-/:.'•." -';-;••-'•".'
 M.        ,  —molar '                   .       •       .  .  '   '     '
 mA         — milliampere                 T
'MAA        —methylarsonic acid.,,,.    ';;.)::-•,:•,,   :  :•   "" ."  ;
 MeOH       —methanol
 mg         — milligram   .     •..         -^     , ,
 min        — minute(s)
 mL         —milli-liter       ..•••••''"•   ;
 nm         ~ millimeter  (lxlO~3 meter)             ,         ..      ....
 y          —micron (IxKT*5' meter), micro (xlO~6)
 yA         — microampere
 yC         — microcurie
 yF         —microfarad
 yg         — microgram
 yL         — microliter         .-...."        •    .,.,
 nm         — nanometer.(lxlO~9 meter)
 OD         — outer diameter                      '
 ppb        — part per  billion;  the concentrations  of arsenic  compounds
               are expressed in .terms of arsenic, e.g.,  1  ppb As(arsenate)
               means that  the solution  contains 1 ppb of arsenic  in form of
               arsenate  _
 pH         —negative  logarithm of hydrogen ion concentration
                                       xiii

-------
pKa        — negative logarithm of the acid dissociation constant
ppm        — part per million (see ppb for note)
psi        — pounds per square inch
RI         — refractive index (detector)
TBAP       — tetrabutylammonium phosphate
THAN       — tetraheptylammonium nitrate
UV         — ultraviolet
V         ' —.volt
W          -- watt
                                     xiv

-------
                               ACKNOWLEDGMENTS
      The assistance of the following' individuals  in collecting water samples
 is  gratefully acknowledged:   Mr.  E.  Western,  Utah State Department of Health,
 Division of  Environmental Health  (samples- from Hinckley and Delta, Utah);  Dr.
 J.  M.  Borgono,  Servicio Nacional  de  Salud,  Ministerio.de Salud Publica,
 Santiago,  Chile;  Mrs.  Alicia Araya Fuentes, Acting Director, Laboratorio
 Bromatologico,. Antofagasta,  Chile; Mrs.  Stefania  Razmilio Valdes,  Antofagasta,
 Chile (samples  from the Antofagasta  drinking  water supply); Ms. Jane L.
 Valentine, Associate Professor of Public Health,  University of California,
 Los Angeles  (Antofagasta water samples);  Mr.  Pete McGee, Alaska Department
 of  Environmental  Conservation,  Fairbanks (Alaska  water samples); Dr. W.
 Tseng,  Department of Medicine,  National  Taiwan University, Taipei  (Pu-Tai
 and Pei-Men  samples);  Dr. Fung-Jou Lu, Associate  Professor of Biochemistry,
 College of Medicine, Taipei  (Yenshei samples);  and Mr.  C. E. Tupper,
 Administrator,  Environmental. Health,. Department of Public Health,  Nova Scotia
 (Nova Scotia, samples).

      The following postdoctoral associates  and graduate students participated
.in  the .analytical work:  Dr.  R.. Bess,. Mr. •• R.  Nichols (differential pulse
 polarography) ; .Dr. M..  West (graphite furnace  AA) ;  Mr.. D. Aylmer (adsorption
 of  arsenic compounds by .container walls);; Mr. R.  Stockton (HPLC^GFAA system,
 ICP, arsenocholine/arsenobetaine  separation) ;. Dr.. P. Clark and Ms^__P. Micks
 (hydride generation);  Mr. K.. Ehrhardt (mass spectrometry, arsehite/arsenate/
 methylarsonic acid/dimethylarsinic acid  separation)  and Dr. D. Chakraborti
 (graphite furnace AA) .                                ••.....'

      Special, thanks: are owed to Mr.  D. Heinrich,  Radian Corporation, Austin,
 and Ms..Denise  Painchaud, Applied Research  Laboratories, Sunland,  California
 for performing  ICP analyses  on the- water samples.                            •

      Dr. I.  Warner, Assistant Professor,  Department of Chemistry,  Texas  A&M
 University and  Mr. Dennis Shelly  carried out  the  work on the fluorescent
 compounds in the  Taiwan well" water.

      Elsevier Scientific Publishing  Company,  and  Gordon and Breach Science
 Publishers granted permission to  use figures  from references 57 and 58 and
 quote passages  pertinent to  these figures.
                                       xv

-------
                                   SECTION 1

                                 INTRODUCTION.
      Arsenic is an element which possesses a rich chemistry.  Inorganic
 and organic arsenic compounds may contain trivalent or pentavalent arsenic..
 The trivalent arsenic compounds are generally more toxic than the pentavalent
 derivatives (1).  Many inorganic and organic arsenic compounds are linked in a
 cycle with chemical and biologically mediated-reactions changing the compounds
 into each'other.  The input, of arsenic into this-cycle:is suppliedby
 weathering of arseiiic-containing rocks and1human.use and disposal of .various
.arsenic compounds.              ...         -.,'        .  .    ..,,-...    ... .

      Since arsenic is ubiquitous,  man consumes small,amounts1 of arsenic
 compounds with the food he eats and the water he drinks.  Life- developed in
 the presence, of arsenic.  Organisms are, therefore, expected to tolerate a .....
 certain,  not yet clearly defined dosage of arsenic.  Certain geographically
 limited groups of people have taken into their-systems arsenic compounds
 present in their drinking water supplies over extended periods'of time.  The
 most famous- localities where, arsenic-containing water" has been consumed' are
 certain regions in. Taiwan (71^78)  and the city, of Antofagasta in Chile (69).
 Hyperpigmentation, skin cancer, vascular problems and other ailments have
 been attributed to the arsenic present in the drinking water.  Other groups,
 for which the people of Fallen, Nevada, may serve as examples, have been
 exposed to similar arsenic levels  in their drinking water without any ill
 effects (82).  This project was undertaken in'order to determine the arsenic
 compounds and other trace elements present in arsenic-containing water-
 supplies and to check whether these drinking water supplies contain the same
 or different arsenic -compounds.

-------
                                  SECTION 2

                                 CONCLUSIONS .
     As the result of this project, together with previously developed
analytical techniques, reliable methods are now available to determine
arsenite, arsenate, methylarsonic acid, dimethylarsinic acid, trimethylarsine
oxide, methylarsines, arsenocholine and arsenobetaine at the low ppb level.
The high pressure liquid chromatography system employing a Hitachi Zeeman
graphite furnace atomic absorption spectrometer (HPLC-GFAA) as a sensitive,
element-specific detector is specially useful for the detection of yet
unidentified arsenic compounds present in water samples. 'In order to increase
the confidence in and the reliability of the analytical data, total arsenic
concentrations and the concentrations of individual arsenic compounds must be
determined by at least two independent methods.

     The analyses of water samples from Taiwan, Antofagasta, Alaska, Utah
and Nova Scotia produced total arsenic concentrations in the range 18 ppb
to 8 ppm.  The arsenite/arsenate ratios were in the range 0.007 to 3.4.  No
indications of the presence of methylated arsenic compounds, reducible to
methylarsine or dimethylarsine, at levels above 1 ppb have been found.
Experiments with the HPLC-GFAA system showed only the presence of arsenite
and arsenate in the water samples.  Most of the water samples contained
Al, B, Ba, Ca, Cu, Fe, K, Li, Mg, Mn, Na, P, S (probably as sulfate), Si and
Sr (determined by ICP).

     Various physiological effects observed in populations exposed to arsenic-
containing drinking water supplies could have been caused by the presence of
varying amounts of arsenite and arsenate.  It is also conceivable that one
or more of the trace elements present in the water supplies acted in concert
with arsenic to cause the observed effects.  Many more samples need to be
analyzed and the results of these analyses correlated, with epidemiological
studies before ^a definite statement can be made about the interactions''of
trace elements with arsenite.or arsenate.

     The chromatographic work on the fluorescent compounds in the Taiwan
well waters strongly suggests the presence of alkaloids such as ^-lysergic
acid, ergometrine and calciferol.  These findings call for a reexamination
of the presumed cause-effect relationship between arsenic and vascular
diseases in Taiwan.

-------
                                  SECTION 3

                               RECOMMENDATIONS
     The availability'of the hydride generation technique,, differential pulse
polarography, high pressure liquid chromatography coupled to sensitive,
element-specific detectors and colorimetric methods for the determination of
total arsenic concentrations and concentrations of arsenic compounds, and the
availability of simultaneous, inductively coupled argon plasma emission  .:••'••
spectrometers for trace element determinations and of ion chromatography for
anion analyses;make the thorough characterization of water samples a
relatively easy and not too-time-consuming task.  Many more arsenic-
containing water samples must be.analyzed in support of or in preparation for
epidemiological studies.               .       ,/            -

     Arsenic has been declared a priority pollutant.  Many environmental
samples will have to be analyzed for arsenic compounds before final      ,
conclusions can be drawn about the health effects of arsenic compounds.
Experience has shown.that .analysis by one method is not sufficient-to produce
reliable results.  At least two independent methods should;be used for' the : ?:
determination of; arsenic-'-.compounds.    .     :      -              - ,

-------
                                  SECTION 4

                  ARSENIC COMPOUNDS IN NATURAL WATER SAMPLES
     Arsenic is associated with igneous and sedimentary rocks in form  of
inorganic compounds.  The average concentration of arsenic in these rocks
varies between 0.1 ppm and several hundred ppm (1).  Arsenic is,found in
high concentrations in sulfidic ores..  Special conditions created by recent
or geologically ancient volcanic, or hydrothermal activities have also caused
high concentrations of arsenic in localized areas.  Weathering of arsenic-
containing rocks, produces arsenic trioxide, arsenites and arsenates which
might be retained in soils or dissolved in water and transported to the ocean.
In the absence of living matter water most likely will contain only arsenite
and/or arsenate.            •

     Many reactions are known.which produce organic arsenic compounds from
inorganic arsenic compounds (2)..  Such reactions generally do not occur in
aqueous systems unless organisms capable of transforming arsenic compounds
are present.  The conversion.of inorganic to methylated arsenic compounds
mediated by various organisms is well established  (3,4,5).-  Methylarsine,
dimethylarsine and trimethylarsine may thus be formed from arsenite or
arsenate.  The formation of methylarsonic and diinethylarsinic acid was also
observed (6).  Microorganisms can also demethylate methylarsenlc derivatives
(7).  These methyl-transfer reactions, the reduction of arsenate, the oxida-
tion of arsenite,. the'-reduction of methylarsenic acids to arsines and the ox-
idation of these arsines to the arsenic acids set up an arsenic cycle (Fig.
1).  The arsenic cycle in nature ,is: much more complex than the one shown in
Mg. 1.  Arsenic compounds are distributed among the various compartments of
the environment including air,.water, soil, sediments, rocks and plant and
animals (7).  Living matter seems to be a rich source of many organic arsenic
compounds.  Arsenic has been shown to be incorporated into both marine and
freshwater organisms in the form of water and lipid soluble organic arsenic
compounds (8).  The suggestion was made (8) that arsenic is incorporated into
lipids by replacing nitrogen in lecithins.  A phospholipid containing 0.5
percent of arsenic (9), which might have formula^ (R = H), was isolated from
the alga Tetvaselmis chui.  Hydrolysis of the arsenic-containing lipid JL^would
yield water-soluble arsenocholine 2^  (R = H).  Arsenobetaine ^3, an oxidation
product of arsenocholine 2.(R = H), was isolated from the Australian rock
lobster (10).  Chaetpceros aoneauicoimis grown in arsenate-containing, axenic
media is reported (11) to have synthesized 0-phosphatidyltrimethylarsoniolac-
tic acid 2  (R = COOH).  The arsenic-containing sugar ^ was isolated from
Eaklonia radiata (12).  Most of these investigations employed marine
organisms.  It is very likely, that freshwater organisms are capable of
synthesizing the same or similar arsenic compounds.  The isolation and

-------
  CH3AsH2 .^p—±: CH3As03H2
(CH3) 2AsH
                    ) 2As02H
           methylation

           demethylation
                                H3AS04:
                               I
                                                     As-containing sugars;
                                                     other, yet unidentified
                                                     arsenic compounds;
                                                     As-containing lipids
                                I
 (CH3)2As02H.

*''*'
          [(CH3)3AsCH2CH2OH]n
                                                         ) 3AsCH2COO"-*
Figure 1.   The transformation of arsenic compounds in.the environment.
     CH—COR*
            H,. COOH
                                                     !3As-CH2-CH-OH
                                                       - II,  COOH

                                                       .'-   2
     (cH3)3AsCH2COO-


            .3-   ' '
                                  CH3
                               O-AS-CH
                                  CH,
                 CH2CHOHCH2R
                                                              R = OH,  S02H

-------
and identification of organic arsenic, compounds has just begun.  Many more,
yet unknown arsenic compounds will be discovered and their properties deter-
mined as analytical techniques are improved for the speciation of such
derivatives.

     In. natural fresh water not polluted with arsenic compounds through
human activities, arsenite and/or arsenate are expected to be found.  Based
on the current knowledge of the arsenic cycle in nature methylarsenic
compounds, arsenobetaine and arsenocholine might also be present.

     Man-made arsenic compounds are used as pesticides, herbicides, desic-
cants,. wood preservatives, feed additives and drugs (13).  Most of these
compounds are either inorganic or belong to the class of methylated arsenic
derivatives and will be detected and quantitatively determined upon routine
analysis.  Pollution by the. aromatic arsenic compounds employed as feed
additivies and wood preservatives is unlikely unless there is a manufacturing
or processing plant for such compounds near the origin of water supplies.

-------
                                  SECTION 5

           PRESERVATION OF ARSENIC COMPOUNDS .IN AQUEOUS SOLUTIONS
     In contrast to most metal ions, which do not change valence easily and
are not transformed into organic compounds in aqueous solution, arsenic
possesses a complicated chemistry characterized by equilibria, steady state
conditions or situations approaching such conditions between trivalent and
pentavalent inorganic arsenic compounds and various derivatives containing
arsenic-carbon bonds.  A brief raview of the chemistry of arsenic is avail-
able in the literature (14).  The analysis for arsenic compounds in water
samples, for which the term speciation shall be used, will correctly give the
composition of the sample at the time of collection only if no arsenic com- '
pounds were lost, interconversions of arsenic compounds present in the sam-,
pies did not occur and no  new arsenic compounds were formed during the time
between collection and analysis.'  The determination of. total arsenic:concen-
trations in samples requires only that no arsenic shall be lost and that the
analytical techniques to be employed shall be applicable to all the arsenic-
compounds present in the sample.

     Water samples can generally not be analyzed -for trace elements immedi-
ately after collection..  Several hours or more often several days elapse
between collection and analysis.  During this time the chemical nature of a
particular trace element can change, trace elements can be lost by volatili-
zation and/or can be absorbed on cqntainer walls.  The adsorption of many
metal ions (15, 16) and of phosphate ions (17) were studied but arsenic was
rarely included.              '     •    ., • ' :; '." ""."'.   ....   '",     "

     Disagreements exist in the literature as to the extent of loss of ar-
senic from solutions stored in various containers.  These disagreements are
probably caused by differences in composition of the samples and sample con-
tainers, the species of arsenic present and the sensitivity of analytical
methods.

     Acidification of samples is usually considered necessary when trace ele-
ments are to be determined (15, 16).  Several acids (nitric, perchloric,
hydrochloric and acetic acid) were suggested as perservatives.  Guimont and
co-workers (18) mention the use of acetic acid as a preserving agent for
arsenic solutions but give no details.  Ray and Johnson (19) report losses of
arsenic up to 70 percent within one week from seawater acidified with 9 raL
concentrated hydrochloric acid per liter of sample.  Talmi and Bostick (20)
observed no losses of arsenate from distilled-water solutions containing
15 percent nitric acid or 5 percent perchloric acid upon storage for three
weeks in polyethylene or soft glass containers.  Whitnack and Brophy (21)

-------
reported no loss of arsenite from samples of well water kept for one week
in polystyrene vials with snap caps of polyethylene.  Al-Sibaai and Fogg
(22) found that the total arsenic concentrations in dilute standard solutions
containing only arsenite or arsenate (20 yg. mL"1) remained unchanged for at
least 50 days when stored in borosilicate glass, soda glass or polyethylene
in light or in the dark.  Experiments with seawater containing 2 yg L   of
arsenic gave different results (23).  When such filtered seawater samples
equilibrated with 5 yC of arsenic-74 were placed 'in high density polyethylene,
soda glass or Pyrex bottles and stored in the dark over a period of five
weeks, losses of arsenic occurred with all types of bottles (23).  With soda
glass bottles sixteen percent had been adsorbed by the 15th. day when equili-
brium was attained..  The uptake by polyethylene and Pyrex bottles was com-
plete within ten days and amounted in both cases to approximately six percent
(23).

     Storage of water: samples in the frozen state has been recommended (16,
23, 24, 25).  Harrison and co-workers (26) found that frozen solutions (1 ppm
As) did not lose arsenic upon freeze-drying.  The average retention was
96.3±1:8 percent-  Natural water samples stored below -15° or under dry ice
did not lose arsenite (27).  Seawater samples kept in a refrigerator at 4°
showed no measurable change in the arsenite/ar senate ratio within about ten
days (28).  For longer storage the samples had to be frozen quickly with
crushed dry ice within 15 minutes.  They could then be stored frozen.  In
samples frozen in a freezer,, spurious losses of arsenite were observed (29) .
     Sndreae ~(27y~s~tated~tfiat~ cono^/'di^ and" tfimethylarsihe~ were" slfable~in .
aqueous' solutions for a f ew~days~whenT"storie3~in aTF-tightPcontainers .  Me thy 1
arsonic acid and dimethylarsinic acid, were lost measurably after a period of
approximately three days from untreated water samples. .Acidification with
hydrochloric acid to make the sample 0.05 M prevented these losses.

     It is not possible to recommend at this time a method for the preser-
vation of water samples applicable to all the various arsenic compounds which
may occur in water.  Not enough is ;known, for instance, about the adsorption
behavior of inorganic and organic arsenic derivatives in the presence of
other compounds generally found in water.  Even less . knowledge exists about
the conditions under which arsenic compounds are converted to other species.

     Arsenite and arsenate are rather easily changed into each other.  Ar-
senite in natural water samples is slowly oxidized to arsenate.  Acidifica-
tion of the sample increases the oxidation rate (27).  Fifty percent of the
arsenite in an aqueous standard solution (20 yg mL  ) was oxidized to arse-
nate after storage in a glass container for 33 days (22) .  Arsines were
oxidized by traces of air (27) .  An arsenic cycle operates in nature (7)
connecting inorganic with organic arsenic compounds.  The chemical nature of
many of the organic arsenic compounds participating in this cycle are still
unknown .
EXPERIMENTAL

     Conditions which will prevent adsorption are likely to favor transfor-

                                      8

-------
mation of arsenic compounds and vice versa.  A universal preservation agent
which would prevent adsorption and transformation is not' known and: will
probably not be found.  It will be necessary, to develop preservation techni-
ques for each sample dependent, upon the type of arsenic compounds present.
Investigations were carried out to find conditions under which arsenic'
compounds would not be adsorbed on container walls and arsenite would not be
converted to arsenate..          .

     Most of"the water samples, which are the subject of these speciation
studies,.contained arsenic at concentrations of approximately 1 ppm.  To ob-
tain data on the amount of arsenic lost to container walls, the adsorption
of arsenate, arsenite and dimethylarsinic acid from aqueous 1 ppm As solu-
tions on polyethylene and Pyrex was studied.

     New polyethylene Cubitainers (soft polyethylene manufactured by Kimberly)
and Pyrex glass containers..were filled with 7.5 M nitric acid and kept for  .
24 hours.   They were then emptied and rinsed thoroughly with deionized water.
The arsenite and arsenate solutions were prepared using either distilled,
deionized water of pH 6.5 or tap water of pH 8 with a sodium concentration
of 211 ppm.  Neither water contained arsenic.

     Weighed quantities (required for the preparation of stock solutions) of
Na2HAsO^-7H20, NaAsO^ or &S2°5 were irradiated with neutrons at the Texas
A&M University Nuclear Science Center to produce 76As, a gamma-emitter with.
a half life of. 24.6 hours:.  For- the experiments with Pyrex flasks the stock
solution was prepared by dissolving irradiated As2<35 in distilled water.
The stock solutions for the polyethylene experiments were made; from Na2HAsO^-
7H20 or NaAsC>2.  Working solutions of 1 ppm arsenic (As) were prepared by
appropriate dilution.                     .                                  .

     Acid-cleaned 25 mL volumetric .flasks made from Pyrex were filled to the
mark with the 1 ppm As solution (prepared from As2^5)•  One flask each was
emptied after one, two, four,, five,  six and ten days :had elapsed.  The flasks
were rinsed, then filled, to the mark with distilled water and placed into the
gamma-ray counter well equipped with a Li--drifted germanium detector.  The
entire gamma spectrum was recorded..  The areas under the.arsenic peaks were
computed and corrected for the decay of the 76As-isotope.   A calibration
curve was obtained by counting 25 mL-volumetric flasks containing 0.5, 1.0,
1.5 or 2.0 mL of the 1 ppm As solution in 25 mL.

     For the adsorption studies in polyethylene bottles, 1 mL of a solution
containing 100 ppm As of either arsenate or arsenite was pipetted into
100 mL of distilled water or tap water stored in the plastic bottles.  These
bottles were treated and counted as described above for the Pyrex studies.

     Samples of 1 ppm As solutions containing dimethylarsinic acid in poly-
ethylene bottles were prepared similarly.  The arsenic concentrations in
solution as a function of time were determined by flameless atomic absorp-
tion spectrometry after a ten-fold dilution of an appropriate aliquot.

     For the polarographic determination of arsenite in solution, 45 mL were


                '"..'•••           9        -  •                 '  '

-------
withdrawn from the plastic bottles and transferred into the polarography
cell.  After addition of 5 mL of cone. HC1 nitrogen was bubbled through the
solution for five minutes before performing the analysis.  To determine
arsenate, 30 mL of the solution, 10 mL cone. HCIO^ and 3 g pyrogallol were
placed into a 50 mL volumetric flask.  After filling to the mark with dis-
tilled water, the cell was charged with this solution.  Calibration curves
were obtained by standard addition.

     The following systems were investigated:

          Arsenate in distilled water - Pyrex
                   in distilled water - Polyethylene
                   in tap water - Polyethylene

          Arsenite in distilled water - Polyethylene
                   in tap water — Polyethylene

          Dimethylarsinic Acid in distilled water - Polyethylene
                               in tap water - Polyethylene
RESULTS

     When 1 ppm As (arsenate) solutions were kept in 25 mL acid-cleaned Pyrex
volumetric flasks activity corresponding to 0.22 ug of arsenic was found
on the walls after one day.  The amount of arsenic adsorbed did not change
during the ten-day experiment.  The samples could not be counted after more
than ten days had elapsed because of decay of the '^As-isotope.

     In all the experiments in polyethylene bottles no  adsorption of arse-
nate, arsenite or dimethylarsinic acid was detected.  The results for dime-
thylarsinic acid are tentative because the results of atomic absorption
spectrometry are accurate only within ±5 percent (30).  ^As-Dimethylarsinic
acid was not available for these experiments.  After two days all the arse-
nite in the NaAs02 solutions had been converted to arsenate (checked by
polarography).  These studies indicate that polyethylene bottles are the con-
tainers of choice to store aqueous solutions of arsenite and arsenate.

     Among the chemical changes to which arsenic compounds may be subjected
in aqueous solutions, the oxidation of arsenite to arsenate has' been re-
ported repeatedly in the literature (7, 22, 27, 31).  Preliminary analyses of
water samples to be speciated for arsenic compounds indicated the presence
of arsenite and arsenate.  To keep these two ions from converting into each
other a search for an appropriate preservation technique was made.

     The most convenient and experimentally tested procedure for the pre-
servation of arsenite was reported by Feldman (31).  Ascorbic acid at a con-
concentration of 1 mg mL   prevented the oxidation of arsenite to arsenate in
aqueous standard solutions at room temperature.  Arsenate was not reduced
to arsenite under these conditions (31).  However, special care must be
taken, not to expose arsenite solutions preserved by ascorbic acid to
                                      10

-------
elevated temperatures.  Ascorbic acid has been used to reduce arsenate to
arsenite in boiling seawater (23).   Whether ascorbic acid scavenges oxygen
from the solution, combines with arsenite producing a species less reactive
toward oxidation than arsenite or acts both ways in preventing oxidation of
arsenite, is not known.  Natural water samples contain many other metal
ions.  Comprehensive experiments to elucidate the influence of such metal
ions on the ability of ascorbic acid to preserve arsenite have not yet been
carried out.  Metal ions do catalyze the oxidation of ascorbic acid fay
oxygen (32).

     The water samples which were analyzed as part of this project contained
arsenite and arsenate (see Section 7).  Additional water samples were analyzed
during the past six months.  On the basis of the experience gained with these
samples the following procedure is  recommended for the analysis of arsenite
and arsenate in water samples:

     • Collect three samples for each water supply to be analyzed in
       one-quart polyethylene Cubitainers.
     • Preserve-the first sample by addition of 0.1 percent by weight
       of ascorbic acid.
     • Preserve the second sample by addition of 7 mL ultrapure,
       concentrated nitric acid to. one quart of sample.
     ••Leave the third sample unpreserved.
     •'Fill all,the containers to the brim;to avoid air space's. /
     • Analyze the samples as soon as possible after collection.
     • The samples should be kept at or below room temperature during
       shipment and storage.          •

If the concentrations of arsenate and arsenite in the three samples deter-
mined by at least, two independent methods agree, arsenite and arsenate were
not chemically changed during the period between collection and analysis.
Among the fifteen different water samples analyzed by this procedure thus far,
the arsenite and arsenate values obtained from a,set of preserved and un-
preserved samples did not agree in only one :case.

     Since methylated arsenic compound's were not.present in any of these
samples, no statement can be'made about the effect of this procedure on .
organic arsenic derivatives.
                                      11

-------
                    ,             SECTION 6
                                                      X
                DEVELOPMENT OF ANALYTICAL TECHNIQUES'FOR THE
                     DETERMINATION OF ARSENIC COMPOUNDS
     Several methods are available to determine total arsenic concentrations:
colorimetry (silver diethyldithiocarbamate - AsH, complex; arsenomolybdic
acid); neutron activation analysis employing the gamma-active isotope   As;
polarography of arsenite; coulometry; X-ray fluorescence spectroscopy;
atomic emission spectrometry; plasma emission spectrometry; and gas chrom-
atography of volatile arsenic compounds.  Recent reviews of these techniques
are available (33, 34).  The detection limits may be as low as picograms of
arsenic per milliliter.

     Whereas many methods exist for the determination of total arsenic, the
choice of techniques for the determination of individual arsenic compounds
is limited.  The following methods were available for the determination of
arsenic compounds when this project was initiated.  The arsenic compounds,
determined by these methods, are listed as part of the heading.  It is
likely that other arsenic compounds not yet investigated may also be
determinable by these methods.

     Silver diethyldithiocarbamate (Ag-DDCl. / colorimetry for inorganic
arsenic and methylarsonic acid:  Arsine generated by reduction of arsenite
and/or arsenate and methylarsine obtained by reduction of methylarsonic
acid, reacts with Ag-DDC in pyridine to produce colored complexes.  Zinc/
hydrochloric, acid serves as,the reducing agent.  The absorbances of the
complexes are measured at 540 and 440 nm.  Detection limits as low as 20
ppb are achieved  (35).
                                                    *
     Ammonium molybdate for arsenite and arsenate:  Arsenate (but not
arsenite) reacts with ammonium molybdate to form blue arsenomolydate,
the absorbance of which is measured at 865 nm.  Arsenite can be oxidized
to arsenate for the determination of total inorganic arsenic.  Reduction
of arsenate to arsenite with sodium disulfite and sodium thiosulfate in
aqueous sulfuric acid allows the determination of phosphate (36).

     Polarography for arsenite;  Arsenite can be determined polarographically
in acidic aqueous media by single-sweep polarography (21) (sulfuric acid,
detection limit 5 ppb) or by differential pulse polarography (37) (1 M
hydrochloric acid, detection limit 0.3 ppb).  Arsenate is polarographically
inactive under these conditions.  Dimethylarsinic acid, methylarsonic acid
(38), and other aliphatic (39) and aromatic (40) arsinic and arsonic acids
are reducible on  the dropping mercury electrode.  Whether inorganic arsenic


                                     12

-------
 compounds and organylarsenic acids can be determined simultaneously by
 polarography must yet be decided by experiment.
                                      /
      Hydride generation for arsenite,?;. arsenate;  methyl-,  ethyl-,  propyl,
 butyl-,  phenylarsonic acid;  dimethylarsinic acid;  trimethylarsine oxide:
 Arsenite, arsenate,  organylarsonic acids, diorganylarsinic acids  and tri-
 methylarsine oxide are reduced in aqueous medium by sodium borohydride to
 arsines  (27, 31,  41, 42, 43).  Speciation of these compounds is accomplished
 by their pH selective reduction to arsines, transferring  the arsines into
 an organic solvent (43), or adsorbing the arsines  on Chromosorb 101 (43),
 selective volatilization of arsines from a cold  trap (27,  41, 42, 43)  or
 gas chromatographic separation of the arsines (27, 43)  and detection of the
 separated arsines by helium-DC discharge spectrometry (31, 41, 42), argon
 microwave emission spectrometry (43), atomic absorption spectrometry using a
 hydrogen/air flame (27, 44), electron capture or flame  ionization detectors
 (27). Detection  limits in the part per trillion range  have been  reported
 (27). Great care must be exercised when arsenic compounds are reduced with
 sodium borohydride.   Some arsenate may be reduced  to arsine under conditions
.under which only  arsenite should react, and cleavage of methyl groups  from
 methylarsenic compounds has been observed during their, reaction with sodium
 borohydride (31).  Improvements••concerning the reduction,conditions (31) and
 modifications of  the analytical system (31, 45)  made the  hydride  generation
 technique a more  reliable method for the determination  of arsenic: compounds.

      The-disadvantages.of the hydride generation method include the1 require-
 ment that the arsenic compounds must be reducible,.to arsines, and the
 necessity that the .arsines,. possess, sufficient volatility  to allow their trans-
 fer to the detectors.  Many arsenic compounds exist which do not  fulfill
 these requirements.         : ;  .  ,-.;

      Derivatization/gas chromatography. for-methyrarsonic  acid and dimethy1-
 arsinic  acid:  Methods which digest samples to convert  arsenic compounds to
 inorganic, derivatives (46) and subsequently transform these into  organic
 derivatives (20,  47) suitable for gas chromatography cannot be employed
 for speciations.   All information about the chemical nature of the original
 compounds is lost during digestion.  Volatile arsenic compounds,  such  as
 methylarsines, can be determined without prior chemical modification by gas
 chromatography (27,  48)..

      Chemical modification has been successfully employed to convert methyl-
 arsonic  and dimethylarsinic acid to volatile derivatives.   Methylarsonic
 acid combines with ethylene glycol (49), dimethylarsinic  acid reacts with
 hydriodic acid (50), and both acids interact with  allylthiourea (51) to
 form arsenic derivatives which can be determined by chromatography. The
 reactions.of arsenic compounds with trimethylsilyl chloride produce volatile
 arsenic  compounds.  Unfortunately,.the  method is non-selective.   Several
 arsenic  compounds gave the same product (52).

      Gas chromatbgraphy/atomic absorption spectrometry  for arsenite, arsenate,
 methylarsonic acid,  dimethylarsinic acid, trimethylarsine oxide and
                                       13

-------
methylarsines:  Arsenic compounds, which possess adequate volatility or can
be converted to volatile compounds, can be separated by gas chromatography
and quantitated employing graphite furnace (53) or flame (27) atomic absorp-
tion spectrometers as element-specific detectors.  This method has not been
used extensively;           .   '

     Gas chromatography/mass spectrometry for methylarsines and ethyl, propyl-
and butylarsine:  After separation of volatile arsenic compounds by gas
chromatography a high-resolution mass spectrometer may serve as an element-
specific and compound-specific detector which can also provide the elemental
compositions for the compounds (43).

     All the methods available for the speciation of arsenic compounds at
the time this project was initiated had severe limitations.  The methods
were applicable only to the determination of a few arsenic compounds or
required their conversion to volatile derivatives.  Practically, the quanti-
tative determination of arsenic compounds was thus limited to arsenite,
arsenate, methylarsonic acid,  dimethylarsinic acid, trimethylarsine oxide,
ethyl-, propyl- and butylarsonic acid, and the arsines obtainable from these
compounds.  Non-volatile arsenic compounds and arsenic compounds not reducible
to volatile arsines could not be determined with the existing methods.
Therefore, an analytical system wi.th an element-specific detector had to be
developed capable of separating volatile and non-volatile arsenic compounds
from complex matrices.

     The development efforts produced a high pressure liquid chromatography -
Hitachi Zeeman graphite furnace atomic absorption system, a differential
pulse polarographic method for the determination of arsenite and arsenate,
and the elucidation of the mass spectral behavior of organylarsenic
acids, organyliodoarsines,' arsenocholine and arsenobetaine.
HITACHI ZEEMAN GRAPHITE FURNACE ATOMIC ABSORPTION SPECTROMETER AS AN ELEMENT-
SPECIFIC DETECTOR FOR HIGH PRESSURE-LIQUID CHROMATOGRAPHY

     Liquid chromatography and specially high pressure liquid chromatography
(HPLC) with the great resolving power of its micro-particulate.column
materials are potentially the best techniques for the separation and
simultaneous determination of arsenate, arsenite and organic arsenic com-
pounds.  A water sample may contain many substances in addition to arsenic
compounds, which can be separated by HPLC techniques.  The common detectors
(refractive index, UV-visible spectrometer) will not respond specifically
to arsenic compounds.  The identification of arsenic-containing fractions is,
therefore, difficult if not impossible unless element-specific detectors
with high sensitivity are available.

     Flame atomic absorption spectrometers have been interfaced with
chromatographic columns for the element-specific detection of several metal
ions (54).  The detection limits were relatively poor primarily because of
peak spreading in the chromatographic column.  Cantillo and Segar described
a system in which the effluent flow is stopped at fixed time intervals


                                      14

-------
for introduction of a sample into the graphite  furnace  (55).   This  system
gave better sensitivity but the total analysis-time was very  long.   A
graphite furnace atomic absorption spectrometer (GFAA) combines  the advantage
of element-specificity with high sensitivity for many elements.   Two automated
systems combining HPLC with GFAA as an element-specific detector were
simultaneously and independently developed.  The first system uses  a Perkin-
Elmer GFAA with a specially adapted autosampler (56).  The  second system,
developed for this project, employs a Hitachi-Zeeman GFAA with a sample
valve, an injector and associated electronics to control the  analysis sequence
(57).                                 .

Description of the HPLC-Hitachi-Zeeman GFAA System

     In order to determine arsenic quantitatively in the HPLC column
effluents an automated HPLC-GFAA interface had  to be developed capable of
controlling the following functions sequentially and reliably.

     • Lowering .the*injection tube into the cool graphites cup.
     • Turning the-sample valve into the inject position.
     • •Actuating the pump to inject nickel'solution into the  graphite
       cup (if desired) and controlling time of its operation to
       deliver a predetermined volume of this solution.     :
     • Turning on the Start/Stop switch to. initiate the AA  analysis,
       sequence-.     •   '••••   •'  ' '''-  •  .' .  ' :-''• '•'.""; >•"'         ''.'"'••   •
     • Raising: the injectipn tube out of,.the graphite cup and the   -,,:,   ...
   .    furnace. .......  ..  '.'..'               '
     • Returning the sample valve to the-'load: position in preparation  -•.••••••-
       for the next injection..        '     V
     •Allowing the GFAA-to perform the analysis and the graphite
      , furnace to cool.sufficiently to permit injection of.  the next     . '
       sample without splattering. .     . . ;x    •     '           '••••••
     • Actuating the Start/Stop switch to terminate the graphite
       furnace cooling cycle.          .  :
     • Returning to the first function to repeat.the sequence of events.

     Although there are several possibilities to accomplish .the  required
sequencing and timing of events, the design chosen divides  the interface
into the sampling sub-system, the injection sub-system and  the control and
timing circuit.  A block diagram for the HPLC-GFAA system is  given  in Fig. 2
and schematics for the interface are presented  in Fig. 3.   The specifications
and sources for the electronic components are summarized in Table 1.

The Sampling Sub-System:  The sampling sub-system consists  of an Altex,  eight
port, Teflon Slider Injection Valve (Altex catalog #201-14) with two attached
pneumatic actuators and connecting tubings as shown in Fig. 4.   One of the
eight ports (Fig. 4) is closed in one of the two possible slider positions.
To this port the nitrogen line is connected.  When the nitrogen  port is  in   -.
the open position.it will connect to the. "bottom sample loop  port"  (BSLP).
The BSLP is connected by a 0.6 mm ID Teflon tubing to the port directly
above it (top sample loop port, TSLP).  The length of the Teflon tubing  is
adjusted to deliver the volume (40 uL in the case; of the HPLC-GFAA  system


                                      15

-------
SOLVENT
RESERVOIR
1
HPLC
PUMP
I
SAMPLE
INJECTION
PORT




—
Flow
Elec

RECORDER


UV :
DETECTOR

.;"



HPLC COLUMN '
of Liquid
trie Connections


RECORDER

DIFFE
REE
INDEX
' 1
SLID
INJE
VALV

;.
RENTIAL
RAQTIVE
DETECT,
;
ER ••
CTION
E
.
.
Y

X
T—
•
^

NITROGEN
RESERVOIR

Mi SOLUTION
RESERVOIR
AND PUMP



TIMING
AND
CONTROL
CIRCUIT

1 '/
SAMPLE
INJECTION
SYSTEM;
LINEAR
ACTUATOR

FRACTION
COLLECTOR
.u— Jj



HITACHI
ZEEMAN
GEAA


RECORDER
Figure 2. Block diagram for the HPLC-GFAA system.





(From reference 57; redrawn with permission of Gordon and Breach Science Publishers)

-------
  i* -E
Spll
  ^ to I  .-
Boara 1
.(•c'T


j_
T
0
S
CD
in ,
m
•
OJ
o
a
ftnqrfl
X
T
CM
U
-2.
1
7
M
U

1
O

1
in
u
a •
-a. "i
0
a.
                                                                                             Trioaer/Carner Culoll ConlroB
                                                           LOAD. SAMPLE LOOP
                                                    1. DETECTORS  '•'.     .;  '.t
                                                     NITROGEN
                                                     FRAC. COLLEC TOR
                                                    4. A.A.       '     v   ,'• ';
                                                     SAMPLE LOOP      "  4
                                                    6 BVPASS  .LOOP    i
                                          1   P:  Match points
                                          i •  B:  Interconnecting  cables  between boards  and other components
Figure 3.   Schematics for the HPLC-GFAA interface.

-------
TABLE 1.  SPECIFICATIONS AND SOURCES FOR ELECTRONIC COMPONENTS FOR THE
          HPLC-GFAA INTERFACE
          COMPONENTS DESIGNATION*'
                                     SPECIFICATIONS AND SOURCES
          C 1,2
          C 3,7,9,22,36
          C 4,5,15,29,43-49
          C 6

          C 9,10,11,18,19,25,26
            32,33,39,40
          C 12,17,20,24,27,31,34,
            38,41
          C 13,14,16,21,23,28,30,
            35,37,42

          D 1-26
          F 1

          1C 1


          1C 2

          1C 3-7


          R 1,4,5,8,9,12,13,16,
            17,20
          R 2
          R 3,6,7,10,11,14,15
            18,19

          Relay 1-6,9,10

          Relay 7

          Relay 8
          S 1,4,5
          S 2,3

          SO 1-4
           SD 5



           T 1

           VR 1-4



           dpdt
100 vf, 50 V electrolytic capacitors
0.1 uF, 500 V ceramic disc capacitors
0.01 uF, 1000 V ceramic disc capacitors
10 yF, 25 V tantalum electrolytic
  capacitors
0.01 yF, 50 V ceramic disc capacitors

100 uF, .50 V electrolytic capacitors

0.001 uF, 50 V ceramic disc capacitors
600 V PRV, 1.0 A silicon diodes,
  Sylvania model ECG116

1.5 A quick blow fuse

30 A, 70 PRV epoxy bridge rectifleer,
  type VK 148 distributed by Allied
  Electronics
+12 V- voltage regulator, type LM 340-
  12
LM 556 CN dual monostable multi-
  vibrators (timers)

10 kn, 1/4 W resistors

5 Mfl, 1/4 W variable resistors
2 MSJ,. 1/4 W variable resistors
12 V-DC, 60 mA SPOT relays; Calectro
  part #974
SPDT mercury-wetted reed relay, series
  W133MPC, Allied Electronics
5 V-DC, 2.3 mA SPDT relay type R10S-
  E1-Y1J, Allied Electronics

Single pole, single throw switch
Single pole, single throw switch

110 V-AC, 9W solenoid valves, NC model
  #61P18C3, Valcor Engineering Copr.,
  Kenilworth, N.J.

 110 V-AC,  7  W'three-way solenoid  valve
  model  0V55KJ8NGV,  Peter  Paul
  Electronics, New  Britain,  Conn.

 35 V,  1.5  A Centertap transformer

 150 V  RMS  metal  oxide varistors,  model
   #750,  Workman  Electronics,
   Sarasota,  Florida

 Double pole double throw  switch
           *Coraponent designation as used in Figure 3.
                                     18

-------
                                                         sample loop
                                                   TSU>A
to fraction
 collector
                   VWVWWWVWVW
                         .  pressure  relief  loop

    Figure A.  The sampling  sub-system of the HPLC-GFAA  interface.
to GFAA
                                                                             from HPLC column

-------
described here) required by.  the graphite furnace spectrometer.  When  the
nitrogen port is open,  the TSLP is connected to .the port leading to the
injection device.  The  bypass  loop uses the .two ports opposite  the BSLP and
the TSLP.  The volume of the bypass loop should be made as .small as
conveniently possible.  To the port opposite the nitrogen port  the tubing
coming from the HPLC column  is connected.  The port above the nitrogen port
and opposite the ports  leading to .the injection device is the outlet  port
for the HPLC eluent going to a fraction collector or to waste.  Several feet
of a 0.3 mm ID Teflon tubing (pressure relief loop, Fig. 4) connect the
tubing coming from the  column  with the line leading .to the injector.. This
restrictor tubing is required  for relief of any pressure built  up during the
time the valve switches between load and injection postions.  Pressure
build-up may damage the windows in the refractive index detector.

     During the load cycle the nitrogen port, the bypass loop and the
injection port are closed.   The effluent passes from the inlet  line through
the sample loop to the  fraction collector.  The valve is switched from
loading to injection by two  pneumatic actuators, which .are controlled by the
pressure control system shown  in Fig. 3.  ^Nitrogen at 60 psi is introduced
via the inlet to selenoids SD  2 and SD 3, which determine whether nitrogen
is routed to outlet 1 or outlet 2.  Outlets 1 and 2 are connected to  the
actuators on the slider injection valve.  Outlets 3 and 4 are exhaust ports
controlled by selenoids SD 1 and SD 4, respectively.  Selenoids SD 1  and SD
3 operate simultaneously as  do selenoids SD 2 and SD 4 releasing pressure
from one actuator and applying it to the other thereby changing the slider
injection valve from the loading to the inject position.  During the
injection cycle the column effluent travels from the inlet line through the
bypass loop to the fraction  collector.  The sample contained in the sample
loop is pushed .out from the  loop into the injection line by nitrogen  pressure.
The nitrogen at approximately: 5 psi comes from a nitrogen tank.  To deliver
nickel solution to the  graphite cup a peristaltic pump with a capacity of 200
pL per minute is activated for the required amount of time to push the
appropriate volume of nickel solution (10 yL in the.case of the HPLC-GFAA
system described here)  into  the nitrogen line just before the nitrogen
port.  Nitrogen pressure then  transports the nickel solution through  the
_sampleL loop and theinjection  line into_the ..graphite, .cup_.  The  entire cycle
begins \ag-ain after the  slider  injection, valve .is__switched._b.ack_.to__the_load
position. .

The Injection Sub-System:  The injection system consists of an  American
Design Components 2-% inch stroke 24 V-DC linear actuator, a filter
capacitor  (0.2 uF, 200  V), injection tubing and mounting brackets  (Fig. 5).
The positioning bracket is slotted for exact positioning of the injection
probe on the mounting bracket. The delivery tube is a section  of 1/16
inch OD 316 Stainless Steel  tubing inside a section of 1/8 inch OD 316
Stainless Steel tubing  (guide  tube).  The back support plate has a slot into
which the injection bar extends holding it straight and providing stops at
the top and the bottom  for the actuator stroke.  The mounting bracket has a
viewing port through which the injector can be exactly positioned.  The
injection sub-system must be very carefully positioned and adjusted.  The
samples must be deposited into the center of the cup without splattering
and without touching the sides of the graphite cup.  The delivery tube must

                                      20

-------
                                  electrical
                                  connections	
                                 filter-capacitor
                     delivery tube-
           positioning bracket...
                       guide tube-
                                       to
                                       sampling
                                       valve.
                                                           o
                                                           u
                                                           tj
                                                           OJ
              0)
              B
                                           injection bar'
1/8" pipe to
  1/8" tube
  fitting
                                                mounting bracket
      •walls of GFAA furnace and. magnet cavity-
Figure 5. The injection sub-system of the HPLC-GFAA interface.
                                    21

-------
just touch the deposited sample but not the sides of the cup.

The Timing and Control Circuits:  The timing and control circuits are divi-
ded into three printed circuit boards (Fig. 3).  Board 1, the power supply
board, consists of the. transformer T 1, which converts 110 V-AC to 17.5
V-AC, and>the rectifier IC'l, which changes 17.5 V-AC to 24.7 V pulsating
DC.  Ripple is removed by.capacitor C 1.

     Board 2. contains additional filtering and despiking capactitors C 2-5
and relays 1-6.  The despiking capacitors remove line transients, which
might damage the voltage regulator 1C 2 on board 3.  The SPDT relays 1-6
control the GFAA, the injector and the actuators on the slider injection
valve.

     Relay 1 is closed by a signal from: 1C, 3B set at 1 second.  Closing of
relay 1 actuates relay 7 which is connected in parallel with the start/stop
switch on the AA and thus terminates the cooling cycle, the length of which
is controlled by 1C 3A.

     Relay 2 activated by timer--IC 4A switches-on the motor of the linear
actuator lowering the delivery tube into the graphite,cup.  1C 4A is set to
allow the injection bar to reach the bottom stop in the back support plate.

     Relay 3 controlled by timer 1C 4B activates solenoids SD 1 and SD 3
which cause the slider injection valve to change from the load to the inject
position.  At the same time the nickel pump is activated and kept on by 1C
4B long enough for 10 yL nickel solution to flow into the injection system.

     Relay 4 controlled by timer 1C 5B switches on the motor of the linear
actuator to raise the delivery tube out of the graphite cup..  1C 5B is
set to allow the injection bar to just reach,the top stop in the back support
plate.

     Relay 6 controlled by timer 1C 6A set at one second activates solenoids
SD 2 and SD 4 which return the slider injection valve to the load position.
The silicon diodes D 21 - D 26 and the capacitors C 41 - C 49 minimize line
transients propagating into the timing circuit.

     Board 3 contains the timing circuit for the HPLC-GFM interface
consisting of five cascaded LM 556 CN dual, monostable multivibrators (Fig.
3, Table 1) and one 12 V-DC LM 340T 12 voltage regulator, 1C 2.  1C 2
converts the 24.7 V-DC coming from Board 1 to a regulated 12 V-DC providing
power for the timing circuit and the relays 1-6.  Each one of the dual
multivibrator packages (1C iA and 1C iB; i=l-7) controls two events.  The
timing in each multivibrator is controlled by a capacitor and a variable
resistor.  This arrangement makes it possible to time each multivibrator
independently of the other nine multivibrators.  The first multivibrator 1C
3A (Fig. 3) is triggered by relay 9, which is activated through relay 8 by
the timer (1C'3.4-9). in the GFAA determining the length of the atomization
cycle (Trigger/Carrier Cutoff Controls, Fig. 3).  1C 3A remains on until
the graphite cup has cooled.  The output pin of each multivibrator (with
the expection of the last multivibrator 1C 7B) is connected to the input pin

                                     22

-------
 of the following multivibrator through a 0.001 uF capacitor.  When a multi-
 vibrator has timed out,  the output pin returns to ground triggering the next
 multivibrator.   This arrangement assures that the events occur in the intend-
 ed sequence, because a timer cannot start until the preceding one has timed
 out.   The pin configurations and other pertinent specifications of the multi-
 vibrators are described  in the manfacturer's literature (59).   ••

      The timers 1C 6B, 7A and 7B are not utilized presently.  They provide,
 however, the capability  of dual injections of sample into the graphite cup
 to increase the sample size to be analyzed.  The second injection will come
 after the drying cycle for the first injection has been completed.

      The diodes D 1 - D  10 sharpen the triggering pulses to the respective
 timers 1C 3 - 1C 7.  The bypass .capacitors C 10,11,18,19,25,26,32,33,39 and
 40 reduce the susceptibility of the timers 1C 3 - 7 to noise.,  The
 despiking capacitors C 8,15,22,29,36,43 remove line transients.

      The switch S 2 interrupts the timing sequence and resets'all timers.
 The switch S 3 allows the manual starting of the timing sequence.          ,,...

      The Trigger/Carrier Cutoff Controls of the interface are mounted inside
 of the power supply of the GFAA-  Relay 8is connected to 1C 34-9" in the.   '  ;
 GFAA power unit-  When 1C 34-9 initiated, the atomization .cycle relay 8 is.
 activated, which in turn activates relays 9 and 10.  Relay 9 triggers timer
 1C 3A.  Relay 10 turns on solenoid,SD 5 which terminates the carrier gas flow
 through the graphite furnace during the atomization cycle.  Capacitor C 43
 and varistor VR 4 control voltage transients.:   " "        - :

      Relay 7 is mounted  outside the GFAA power unit together with the switch.
 dpdt (double pole-double throw).  This switch isolates the GFAA from the
 interface allowing normal operation of the spectrometer.

      The HPLC-GFAA system has functioned almost flawlessly during the past
 three years..  The major  problems encountered.initially were;caused,by line
 spikes propagated through the building wires' to the electronics of the
_interface, where tphey triggered^relays                         ^.^
 events.   The capacitors installed as shown in the. schematics (Fig. 3) .elim-
 inated this interference.

      The time interval between two analyses can be made as short as 30
 seconds.  In most cases analyses  carried out in one minute intervals pro-	
 duce satisfactory chromatograms.   The1 firing of the graphite furnace Pvpry
_?^e_^aHsld_lhA..PA^^^^	._
 fore, a refrigerated glycol/water mixture;was circulated through the magnet's
 cooling system replacing tap water used under normal operating conditions.
 Because the glycol/water mixture  was pumped at.much lower pressure than the
 tap water, the water pressure switch in the GFAA had to .be deactivated.  The
 instrument is still protected by  an overheat shut down switch.  The time
 between analyses can be further shortened by using cooled gases to bring
 the graphite cup to room temperature after completion of the atomization
 cycle and by shortening the drying cycle.


                   .  ••                23          "••'•''            ' •'•'• '"•

-------
     The Hitachi-Zeeman GFAA Model 170-70 has a detection limit for arsenic
of 10 picograms.  This sensitivity is, of course, retained in the HPLC-GFAA
system for each separate analysis.  Upon migration through the chrbmatographie
column the arsenic compounds are separated and spread out into bands.  Ali-
quots of 40 pL are withdrawn from the effluent.  The 40 \iL aliquots taken
from the center of a band, where the arsenic concentration is highest, must
each contain at least 10 picograms of arsenic.  The detection limit of the
HPLC-GFAA system is therefore strongly dependent on the degree of band
spreading.  Under optimized conditions ten nanograms of an arsenic compound
placed on the column should produce discernible GFAA signals.  The sensitivity
of the HPLC-GFAA system can be increased by optimization of the chromato-
graphic parameters, by using multiple injection - drying cycles before
atomization or by pre-concentration of the arsenic compound on the top of
the chromatographic column.
THE SPECIATION OF ARSENITE, ARSENATE, METHYLARSONIC ACID; DIMETHYLARSINIC
ACID, ARSENOCHOLINE AND ARSENOBETAINE USING" THE. HPLC-HITACHI ZEEMAN GFAA
SYSTEM       .

     Arsenite, arsenate,, methylarsonic acid, dimethylarsinic acid, arseno—
choline and arsenobetaine are arsenic compounds which might be present in
natural water-not polluted by human activities.  Methods were developed for
the separation and determination of these compounds by HPLC employing a
Hitachi Zeeman graphite furhance atomic absorption spectrometer as an
arsenic-specific detector.  The HPLC-GFAA system will not only indicate the
presence or absence of the arsenic compounds for which the separation methods
were designed but will also show the presence of other arsenic compounds.
The possibility that an unknown.arsenic compound has the same retention
time (volume) as one of the arsenic compounds named above does exist.
However, chromatograms using several mobile phases and mass balances should
reveal, the presence of additional compounds.

Separation of inorganic arsenic, arsenocholine and arsenobetaine:

     Arsenocholine, arsenobetaine arid inorganic arsenic  (arsenite and
arsenate) were separated on a micro-particulate (10 y) C-18 reverse phase
column with the sodium salts of heptanesulforiic acid or dodecylbenzenesul-
fonic acid as counter ions for the arsonium salts (57 i 58).

     Arsenocholine, arsenobetaine and inorganic arsenic required mixtures of
acetonitrile, water and counter ions as mobile phases, which are made 1.0 M
with respect to acetic acid to suppress the dissociation of the carboxylic
acid group in arsenobetaine and thus achieve greater retention on the
reverse phase column.  By varying the ratio acetonitrile/water in the eluent
the retention of arsenocholine can be changed from total retention to co-
elution with arsenobetaine (Fig. 6).  Arsenobetaine cannot be totally
retained with heptansulfonic acid as the counter ion.  However, total
retention on the top of the column was achieved with dodecylbenzenesulfonate.
This procedure is useful for the pre-concentration of arsenobetaine from
very dilute solutions.  Arsenite and arsenate were not retained at all with
either of the sulfonates on the C-18 reversed phase column.  These inorganic

                                     24

-------
mobile phase:

0.005 M
water 20 mL
acetonitrile 30 mL
acetic acid 6 mL
Na dodecylbenzenesulfonate
C-18 reversed-phase column
arsenite
arsenate
1 Ug

injection
/
.... 1










arsenobetaine





1 ug



1 - •' '



i i i i 1 i i i i 1 i i









1
flow rate: 0.5 mL/min



arsenocholine

1 Ug
> : * '
mobile phase
changed to
. 100% acetonitrile .





















i 1 I 1 i i 1 t i i I
o  :
"inutes
15
2O
                                   25
30
' I ' 1 1 1 I I 1 j 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
"-^injection
mobile phase: water 95 mL



• -.,
	 ?i.....J!\,
I:
arsenite ji
arsenate j
1 Ug j
i
'*

acetonitrile 5 mL








^

|






. • • acetic. acid 6 mL.
0.005 M heptanesulf onic acid
• '.i:.-.:'. flow rate: 0.5 mL/min




C-18 reversed-phase column .



If
I
arsenobetaine
1 ug








"










arsenocholine
1 ug
1
Figure 6. Separation of inorganic arsenic  (arsenite and  arsenate),  arseno-
          choline and arsenobetaine by HPLC-GFAA on a C-18  reversed-phase
          column.


(From reference 57; redrawn with permission of Gordon and Breach  Science
Publishers)
                                    25

-------
arsenic compounds eluted with the solvent front.  Peaks caused by arsenic-
free impurities in. the solvent or the samples were recorded by the refractive
index (RI) and UV detectors.  These detectors were not sensitive enough to
respond to the arsenic compounds.  However, the GFAA detector has the required
sensitivity to show the presence of the arsenic compounds (Fig. 6).  No
peaks other than those indicating the presence of arsenic compounds will
appear in the GFAA chromatograms facilitating the quantitation and identifi-
cation of arsenic compounds even in complex matrices which would produce a
large number of RI and UV signals.

The separation and determination of arsenite, arsenate, methylarsonic acid and
dimethylarsinic acid:  A C-18 reversed-phase column can be used to separate
ionic species provided the ions are paired with a suitable lipophilic counter
ion to produce a rather hydrophobic ion pair.  Tetrabutylammonium phosphate
(TBAP) at 0.005 M concentration in the appropriate mobile phase has been
recommended as the reagent for such paired-ion chromatography.  Arsenite and
arsenate migrated almost with the solvent front through a C-18 reversed-
phase column with water-methanol (95:5 v/v) at pH 7.3 as the eluent in the
absence of TBAP but were completely separated in the presence of TBAP
(0.005 M) with retention times of 19 min and 11 min for arsenate and
axsenitE^ respectively (Fig. 7).  When a mixture of arsenate, arsenite,
dimethylarsinic acid (DMAA) and methylarsonic acid (MAA) was chromatographed
with 0.005 M TBAP in pure water as eluent, which should increase the
retention time of lipophilic ion pairs, the peaks caused by arsenite and
DMAA overlapped.  The MAA signals were located under these conditions between
those belonging to arsenite and arsenate.  TBAP does not cause sufficient
retention and separation, even with water as the mobile phase and has the
additional disadvantage, of containing phosphate.  Phosphoric acid used to
adjust the pH of the mobile phase and ammonium phosphates added as a source
of counter ions depressed the arsenic signals (Fig. 8)  Such interferences
have been mentioned in the literature (60).  Although phosphates have been
recommended for use in HPLC because of their compatibility with the
stainless steel columns, nitrates or perhaps organic anions should be used
for work with arsenic compounds.

     A much better separation was achieved with tetraheptylammonium nitrate
(THAN).  Utilizing water-methanol (75:25 v/v) of pH 7.6 saturated with THAN
as the mobile phase all four arsenic compounds were satisfactorily separated
(Fig. 9).  Arsenite eluted almost with the solvent front, followed by DMAA,
MAA and arsenate.  The retention times of these compounds increase with
increasing pKa values of the arsenic acids.  Arsenous acid (pKa = 10) remains
undissociated under these conditions, does not pair with the tetraheptyl-
ammonium cation, retains its polar character and is, therefore, not adsorbed
onto the organic, reverse-phase column material.  The weak acid DMAA (pKa6.3)
owes its retention time of 10 min to the two methyl groups rather than to its
presence in solution as the tetraheptylammonium dimethylarsinate ion pair.
The moderately acidic MAA (pKa3.6) is very likely combined with the ammonium
ion and adsorbed by the column material as the tetraheptylammonium methyl-
arsonate resulting in a retention time of 23 min.  Arsenic acid with a pK
value of 2.2, the most acidic compound among the four arsenic derivatives
investigated, is strongly retained on the column head probably as the ion


                                      26

-------

     RI
     mobile phase: water 95 mL   \
                 methanol 5 mL ';

     flow rate:  0.5 mL/mln       \
     C-18 reversed—phase column   \
   injection
                      arsenite 5' ug
                    arsenate  4.5 ug
        without
        tetrabutylamraonium phosphate
                                  minutes
                                             12
             16
2O
 	inj ec tion
    mobile phase: water 95 mL
                 methanol 5 mL

    flow rate:  0.5 mL/min

    C-18 reversed-phase column.
                            arsenite
                            5.0 ug
with .'•'•'•
•0.005 M tetrabutylammonium
        phosphate
             arsenate
            4-5 pg:'-
Figure 7. Separation of  arsenite and arsenate by HPLC-GFAA on a C-18  reversed-
           phase column in the absence and presence of tetrabutylammonium
           phosphate.


 (From reference 58; used with permission of Elsevier Scientific Publishing
Company)                             .
                                         27

-------
           pH 7.3
                mobile phase: water    95 mL
                              methanol  5 mL

                flow rate: 0.5 mL/min
                C-18 reversed-phase column
                0.005 M tetrabutylammonium phosphate
                         pH 5.0
  /I
 1  \
                                pH adj.usted with phosphoric
                                                 acid

                                arsenate  5 ug
1
pH 4.0
h
1 1 (
     19
21
       16       18
retention time in minutes
                                                        16
18
Figure 8.  The influence of phosphate on the GFAA arsenic signals.
pair (R.N)'t' HAsO,.   The mobile phase had to be changed to pure methanol to
elute the arsenate.

     The GFAA signals produced by the arsenic compounds eluting from the
columns can be quantitated either by calculating the area under the curves
obtained by connecting the maxima of the GFAA signals or by summing the
heights of the signals corresponding to a particular compound.  A comparison
of the two methods to convert the GFAA signal clusters corresponding to a
particular arsenic compound to numerical values to be plotted against the
amounts of arsenic, proved that the summation of peak heights produces the
same results as the determination of the areas under the signal clusters
(58).  Examples of calibration curves are given in Figs. 10 and 11.

     The sensitivity and general response of the GFAA instruments is rather
strongly dependent on the characteristics of the graphite cups and other
parts of the furnace system.  The cups, for instance, tend to deteriorate
with use and reduce the sensitivity.  A new cup does not necessarily produce
                                     28

-------


4JD
t
1 3JB










2 2 . ' ' ' •: '

.y
1.,,
' , ' ' \ ..,
20










CH











3/



•










,""""


03H2








Na2HA&04
-





• —

. —
1
1 , 1
30 40
                                   	 RETENTION TIME (MINUTES)
Figure 9.  The separation of-arsenite, arsenate, dimethylarsinic acid and methylarsonic acid by
           HPLC-GFAA on a reversed-phase column in the presence of tetraheptylammonium nitrate.

(From reference 58; used with permission of Elsevier Scientific Publishing Company)

-------
         120
UJ
O
       to  6tQ
         4.0
                             200
           •  (SO®
ng ARSENIC	-4
         Figure 10.  HPLC-GFAA calibration curve for sodium arsenite.

         (From reference 58;  used with permission of Eisevier Scientific Publishing Company)

-------
   12.O
I
        8.0
£   §
^   I—I
    to




    tn
   4.0
   0.0
                                         NaAsOz x   x>—-'
          .^•-"
                                                               [CHj)2AsQzH
                                                           I
                      100
 200              300



	^ng ARSENIC—	j
                                                                               400
500
   Figure  11.  HPLC-GFAA  calibration curves for sodium arsenlte, dimethylarsinic acid, methylarsonic

              acid  and disodium hydrogen arsenate.


   (From reference  58; used with permission of Elsevier Scientfic .Publishing Company)

-------
the same response as one subjected to a number of analysis cycles.  This
behavior necessitates the frequent use of standards.  The Hitachi Zeeman
GFAA with a sensitivity of 10 picograms of arsenic serving as a detector
for the effluent from a C-18 reverse phase column is capable of detecting
several nanograms of arsenic under ideal conditions (Fig. 10) and approxi-
mately 10 nanograms of sodium arsenite in the presence of DMAA.  The
determinations are accurate with ±5-10 percent and are intrinsically
background-corrected.  The causes of the' changes in the slopes of the calibra-
tion curves are not clear yet and need to be .investigated further.

     Various arsenic compounds produce different GFAA responses for the same
amount of arsenic (58^.  As much as a two-fold difference in intrinsic GFAA
sensitivity was: ojjseryedjwith: DMAA and MAA as well as between their sodium
salts.  The less volatile species give a more intense signal than the more
volatile compounds.  Onrthe basis of these results quantitative information
for unknown arsenic compounds obtained'with the'HPLC-GFAA system must be
regarded to be qualitative, until the compounds are identified and calibra-
tion curves constructed with pure samples.

     The HPLC-GFAA system has been shown to be applicable to the qualitative
and quantitative, determination of arsenite, arsenate,- methylarsonic acid and
dimethylarsinic acid in the" ppb range with detection limits approaching 5 ppb
under favorable conditions.  The GFAA detector responds only to a specific
element and is better suited for trace element compound determination than
UV detectors.  Development of column materials will lead to better separa-
tions.  Recently the use of a proprietary special anion-exchange column for
the separation of arsenic compounds (61) and an automated system for-the
analysis of arsenite, arsenate,. methylarsonic acid, 4-aminophenylarsonic
acid and dimethylarsinic acid 'based;on -the separation with a Dionex-anion
exchange column (62) was reported.                ,   .  .        .   :
DIFFERENTIAL PULSE POLARQGRAPHIC DETERMINATION OF ARSENATE AND ARSENITE

     Arsenite is reducible at the dropping mercury electrode.  A very de-
tailed review of the literature about the polarographic behavior of arsenite
appeared in 1969 (63).  In hydrochloric, perchloric, sulfuric or acetic acid
media arsenite is reduced polarographically to elemental arsenic and then to
arsine.  The best supporting electrolyte among those thus far investigated
is hydrochloric acid.  In 1 M hydrochloric acid the detection limit is 0.3
ppb with a linear response up to 60 ppm (37).

     Arsenate is polarographically inactive under the conditions used for
the polarographic determination of arsenite.  However, arsenate produced
polarographic waves in '11.5 M hydrochloric acid (64) and in 2 M perchloric
acid in the presence of pyrogallol (65).  Arsenate gave three polarographic
waves in pyrogallol, the heights of which increased linearly with concentra-
tion.  Solutions with concentrations lower than 37 ppm arsenic were not in-
vestigated  (65).  Pyrogallol is difficult to handle, is easily oxidized by
air and may pose health hazards.  The use of a 11.5 M hydrochloric acid
medium for routine analysis is untenable.  Several other polyhydroxy


                                     32

-------
compounds were investigated as reagents to make arsenate polarographically
active and develop a method to determine arsenate and arsenite simultaneously
at ppb levels.

Differential Pulse Polarographic(DDP) Determination of Arsenate in the
Absence of Arsenite

     Pyrocatechol, pyrogallol, 1,2-ethanediol, 1,2-propanediol, 3-chloro-
1,2-propanediol, 1,2-butanediol, 2,3-butanediol, D-ribose, sucrose, fructose
and D-mannitol were investigated..  All. of these polyhydroxy compounds made
arsenate polarographically active..  The peak heights of the DPP curves for
the reduction of 10 ppm As(arsenate) in solution 2.0 Q with respect to
perchloric acid and 0.5 >1 with respect to the polyhydroxy compound are
listed in Table 2.  An example of the DDP curve produced by arsenate in the
presence of mannitol is given in Fig. 12.  Under these conditions D-mannitol


  TABLE 2.   PEAK HEIGHTS OF THE DIFFERENTIAL PULSE POLAROGRAPHIC CURVES
            AT -0.55 V OBTAINED WITH 2.0 M AQUEOUS PERCHLORIC ACID     ,
            SOLUTIONS OF 100 ppm As(ARSENATE) IN THE PRESENCE OF          .
            VARIOUS POLYHYDROXY COMPOUNDS AT 0.5 MCONENCTRATION


            Polyhydroxy Compound          Peak: Height at -0.55 V in mm


            3-Chloro-l,2-propanediol                  45.7

            Pyrocatechol                              68.6

            Pyrogallol   .                     "        71.1

            1,2-Ethanediol                            93.9

            1,2-Propanediol  :   ."'""           ,     132.1

            1,2-Butanediol           .                142.2

            D-Ribose             :   •   :              170.2

            Sucrose                                  172.7

            Fructose                                 182.8
            2,3-Butanediol                           185.4

            D-Mannitol                               203.2
produced the largest signal.

     The heights of the reduction waves are influenced by the concentration
•of mannitol.  As shown in Fig. 13 these heights become independent of the
mannitol concentration after three grams of mannitol had been added to 50 mL
of the sample.  To make certain that, an excess of mannitol is always present

                ,••..•"•'•'•'•'•• 33 •••'"••''"''•  ,. •         '   .  •

-------
  AE
           current: 0.1 mA
           scan rate: 2 mV/sec
           drop time: 2 sec
           ref.  electrode: satd. KC1
           mod.  amplitude: 100 mV
           initial potential: 0.0 V
           scan direction: negativ
           supporting electrolyte: 2 M
           mannitol: 0.5 M
                    1.   2 ppm As(arsenate) '
                    2.   4 ppm As(arsenate)
                    3.   6 ppm As(arsenate)
                    4.   8 ppm As(arsenate)
                    5.  10 p'pm As(arsenat'e)
                             I
I
I-
                           -QA   Potential, V    -Q.6
                                -O.8
                             I
I
I
  Ai
           current:  0.01 mA
           other parameters same as
           given above
                                           500 ppb As(arsenate)
                                              with mannitol     ..••''
                               ..X
                                                           500 ppb As(arsenate)
                                                               without mannitol
Figure 12.  Differential pulse polarograms of  arsenate  solutions two molar
             with respect to  perchloric acid  in the presence of  D_-mannitol.

-------
         2O
       0 1§
         10
          o
                                current: 0.01 mA
                                scan rate: 2 mV/sec
                                drop time: 2 sec
                                ref.. electrode: satd. KC1
                                mod., amplitude: 100 tnV
                                initial potential: 0.0 V
                                scan direction: negative
                                supporting electrolyte:
                                             2. M HC104

                                20 ppm-AsCarsenate)-       —
                                               1
            o
                        2           3           4
                  Grams of D-Mannitol; Added to 50 mL
Figure 13.
The dependence of the height of the differential pulse
polarographic curve of arsenate at -0.55 V on the concen-
tration of D-mannitol.
in the solutions to be polarographed and thus assuring the independence of
the heights from the mannitol concentration, all further experiments were
carried out with solutions prepared by addition of 4.5 g of mannitol to 50 mL
of the acidified sample.  These solutions are approximately 0.5 M with
respect to mannitol.

     Sulfuric, hydrochloric, acetic, phosphoric and perchloric acid were
investigated as supporting electrolytes in the DPP determination of arsenate
in solutions containing mannitol at 0.5 M concentration.  The solutions were
all 2.0 M with respect to the acid to be investigated.  The results of these
                                     35

-------
experiments  are  summarized  in Table  3.  Perchloric  acid produced  the  largest
signal among the acids  tested and was used  in  all further experiments.


   TABLE 3.   HEIGHTS  OF THE FIRST DIFFERENTIAL PULSE POLAROGRAPHIC
              MAXIMUM  IN THE REDUCTION OF  9.5 ppm As(ARSENATE)  IN  AQUEOUS
              ACIDIC MEDIA IN  THE PRESENCE OF D_-MANNITOL AT 0.5 MOLAR
              CONCENTRATION                          	_^

               Acid           Peak Height  in mm     Potential of Maximum
Sulfuric
Hydrochloric
Acetic
Phosphoric
Perchloric
7.6
15.2
17.8
20.3
22.8
-0.45 V
-0.55 V
-0.90 V
-0.45 V
-0.44 V

              *A11  solutions were  2.0 M with  respect  to  the  acid.
     The reduction waves and the current maximum shifted anodically and the
heights.of the shoulders on the anodic and cathodic side of the current
maximum increased with increasing perchloric acid concentration (Fig. 14)..
For example, an increase of the perchloric acid: concentration from 0.94 M
to 5.8 M shifted the first reduction wave from  -0.57 V to ^-0.5CTV lind   ~""~
.doubled the peak height. . Because of the difficulty of adjusTing"water
samples to a perchloric acid concentration of five molar, two molar
perchloric acid was chosen as a compromise providing satisfactory sensitivity
'and relative ease of sample preparation.

     The DPP curve of arsenate in an aqueous perchloric acid/mannitol
solution is characterized by maxima at -0.55 V  and -0.75 V (Fig. 12).  Above
an As(arsenate) concentration of 500 ppb a current maximum appears at
-0.59 V which increases in intensity with increasing concentration.  The
peak at -0.55 V merges at As(arsenate) concentrations of approximately 5 ppm
into the current maximum and becomes a shoulder.  Another rather low
intensity peak between -0.7 V and -0.8 V might  be obscured at low arsenate
concentrations by the solvent breakdown and at  high arsenate concentrations
by the current maximum.  Since arsenite is reduced in 2.0 M perchloric acid
in the presence of mannitol at -0.34 V, it is likely that the first arsenate
reduction wave at -0.55 V is caused by the reduction As(V) -* As(0) followed
by the conversion of As(0) to AsH3 at approximately -0.7 V.  Experiments
which would verify these postulated steps have  not been carried out yet.

     The differential pulse polarographic maximum or shoulder at -0.55 V
generated by arsenate in 2.0 M perchloric acid  in the presence of mannitol
is suitable for the quantitative determination  of arsenate.  The current
maximum could perhaps be used for the determination of arsenate at

                                     36

-------
Ai
AE
                                     molar concentration of HC10.

                                      5.8     2.8  .  2,0   0.94
         current:  0.1 mA
         scan rate:  2 mV/sec
         drop time:  2 sec
         ref.  electrode: satd. KC1
         mod.  amplitude: 100 mV
         initial potential: 0.0 V   .
         scan direction: negative  /
                 A     1. A  /
             - y  >.    ;v  \/
                          10 ppm As(arsenate)

                          0.5 M mannitol
  -Q2
-0.4
         -Q6
Potentially
-08
Figure 14.. Differential pulse polarographic curves for 10^ppm As(arsenate) in solutions
           0.5 M^with respect to D-mannitol at various perchloric acid  concentrations.

-------
concentrations above 1 ppm As(arsenate).

     Calibration curves for the determination of arsenate in the absence of
arsenite were obtained by addition of 20 yL aliquots of an arsenate solution
of appropriate concentration to 50 mL of sample solution in the polarographic
cell.  After each addition the differential pulse polarogram was recorded.
An example of a calibration curve using the -0.55 V maximum (shoulder), is
given in Fig. 15.  By proper adjustment of the current, linear calibration
curves were obtained for As(arsenate) concentrations ranging from 20 ppb to
900 ppb at 2 yA, 1 ppm to 18 ppm at 50 yA and 20 ppm to 160 ppm at 0.5 mA.

Differential Pulse Polarographic Determination of Arsenite in the Absence of
Arsenate

     Arsenite does not require mannitol for its reduction at the dropping
mercury electrode.  In 2.0 molar perchloric acid solution arsenite is
reduced to As(0) at -0.425 V.  As(0) is converted to arsine at approximately
-0.75 V.  A current maximum appears at an As(arsenite) concentration of
approximately 300 ppb.  The addition of mannitol to 2.0 M perchloric acid
solutions of arsenite shifts the As (III) ->• As(0) reduction anodically to
-0.34 V, but does not affect the position of the current maximum and the
wave at -0.75 V  (Fig. 16).  The heights of all the maxima are increased.
The location of  the maxima, is dependent on the perchloric acid concentration.
The As (III) -»• As(0) peak moves from -0.49 V at 0.7 M to -0.31 V at 4.7 M
perchloric acid  concentration.

     An example  of a calibration curve for the determination of arsenite in
the range 0.2 to 2.4 ppm As(arsenite) in the presence of mannitol is given
in Fig. 17.

Differential Pulse Polarographic Determination of Arsenate and Arsenite in
Solutions Containing Both of these Arsenic Compounds

     When arsenate and arsenite are present in the same solution, differen-
tial pulse polarography employing 2.0 M aqueous HClOi^ as supporting
electrolyte, which is 0.5 M with respect to D-mannitol, will produce
polarograms with maxima at -0.34 V [As(III) -»• As(0)],. -0.55 V (arsenate
reduction), -0.59 V  (current maximum) and -0.75 V [As(0) ->• As (-III)]
(Fig. 18).  The  arsenite reduction peak at -0.34 V [As(III) •*• As(0)] is well
separated from the other maxima and can be used to quantitatively determine
arsenite using the standard addition technique.  The maximum at -0.55 V
caused by the reduction of arsenate can be used for the determination of
arsenate with some confidence only when the concentration of As(arsenite)
is between 100 ppb and approximately 500 ppb, the current is not lower than
2 yA and the arsenate concentration is equal to or greater than the arsenite
concentration.   At As(arsenite) and As(arsenate) concentrations lower than
100 ppb (Fig 19) the polarograms must be obtained with a current of 0.5 yA.
Although the current maximum at -0.59 V does not appear at these low
concentrations,  the arsenate reduction peaks occur as shoulders on the anodic
side of the steep solvent breakdown signals.  The shoulders can be used to
ascertain the presence of arsenate and to semi-quantitatively determine
arsenate.  At an As(arsenite) concentration of 100 ppb and As(arsenate)

                                     38

-------
                             5O  ppm As(arsenate)   •JQQ
                        150
          -ill
       15
     :*>
     MS
     T*
     01
i
r
                  current: 0.5 mA
                  scan,race:  2 mV/sec
                  drop time:.2 sec
                  ref. electrode: satd. KC1
                  mod. amplitude: 100 mV
                  initial potential: 0.0 V
                  scan direction: negativ
                  supporting  electrolyte: 2.0 M HC10
                  0.5 M mannitol
                                                          current: 0.05 mA
                                                          other parameters
                                                          as given above
                                  I    I    I    I
        I    I    I   I
                                     I	I
                              5   ppm As (arsenate)   1Q
                        15
                  15
                  10
                           current: 0.002 mA
                           other parameters same
                          . as given above
                                              500
                                         ppb As(arsenate)
                       1000
Figure 15.   Calibration  curves  for the determination of  arsenate by
              differential pulse  polarography  in aqueous 20 M perchloric
              acid  in the  presence of jD-mannitol.
                                           39

-------
       Aj
       AE
                current: 0.01 mA
                scan rate: 2 mV/sec
                drop time: 2 sec
                ref. electrode: satd. KC1
                mod. amplitude:; 100 mV
                initial potential: 0.0 V
                scan direction: negativ
                supporting electrolyte:
                2.0 M HC10,
           __   20 ppm As(arsenite)
                              -O.2
   -0.4

Potential, V
-O.6
-O.8
Figure 16.  Differential pulse  polarograms for arsenite  in 2.0 M aqueous perchloric  acid in  the
             presence and absence of I)-mannitol.

-------
   25
  2O
e
o
•>
sr
 oc
•H-
 0)


1:
   15
   10
        i  i   i  i
                                i  i   i  i     i  i  i   i
i.i  i
             current: 0.01 mA             .
             scan rate:. 2 mV/sec
             drop time: 2 sec
             ref.. electrode: satd. KC1
             mod. amplitude: 100 mV
             scan direction: negative
             supporting electrolyte:. 2.0 M'HCIO,
                     0.5 M mannitol
        I  I  I  I  I   I  I  I   I  I  I  I  I  I
                                           I  I  I
                0.5
                                                    2O
      2.5
                                 10            15
                                   .ppm As(arsenite)
Figure 17. Calibration.curve for the-determination of.' arsenite by differential
    pulse polarography in aqueous .2.0 MHCi04 in the presence of D-manhitol.
                                41

-------
   Ai
   AE
                             .0.5 ppm As(arsenate)
                              1.5 ppm As(arsenite)
                              with mannitol
                                                    •v
                                                     \  i
                                                      \  \.
                                             current: 0.01 mA
                                             scan rate:  2 mV/sec
                                             drop time:  2 sec
                                             ref. electrode: satd. KC1
                                             mod. amplitude: 100 mV
                                             initial potential: 0.0 V
                                             scan direction: negative
                                             supporting  electrolyte: 2.0
                                             0.5 M mannitol	

                                                 ~    /      "V.'
                                                                                     HCIOV
                                                                                          / /
                                                                                           /
       —     /
  11
II
; I
0.5 ppm As(arsenate)
1.0 ppm As(arsenite)
with mannitol
                                                  0.5 ppm
                                                  As(arsenate) with mannitol
                                                                          0.5 ppm As(arsenate)
     -0.2
Figure 18.
                 -0.4
                                                         -0.8
                                        -0.6
                                  Potential, V
Differential pulse polarograms  of mixtures of arsenite  and arsenate in 2.0 M
perchloric  acid in the presence of ID-mannitol.

-------
           M
           AE
     current: 0.5'yA
     scan rate:' 2.mV/sec
     drop time: 2 sec
     ref.. electrode: satd. KC1
     mod., amplitude: 100 mV
     initial potential: 0.0 V
     scan direction:, negative
     supporting-electrolyte:.
        .2.0 M HC10A
     0.5 M mannitol:
                                                           77
                                                               1
                                                                 -i
                                                            '    /
                                        I
                                     '
                           .•'• i
                           
-------
concentrations in the range 200 to 300 ppb the polarogram can be obtained
at 2 yA.  Under these conditions the solvent breakdown curve is shifted
cathodically, a current maximum is not present, and arsenate, therefore,
may be determined by measuring the peak height of the -0.55 V maxima.

     At As(arsenite) concentrations of 500 ppb with a current setting of
10 yA the solvent breakdown curve rises from the baseline beyond -0.8 V and,
therefore, does not interfere with the -0.55 V maxima.  However, at
As(arsenite) concentrations of 500 ppb the current maximum at -0.59 V
appears.  The arsenate reduction peaks at -0.55 V are not resolved from the
current maxima and appear as shoulders which make the quantitative determi-
nation of arsenate difficult.

     At a concentration higher than approximately 0.5 ppm As(arsenate plus
arsenite) the current maximum at -0.59 V, which is produced by arsenate as
well as arsenite,. could perhaps be used to determine total inorganic arsenic.
Additional experiments are needed to clarify this issue.

Oxidation of Arsenite Followed by Polarography of Arsenate

     The difficulties encountered in the direct-determination of arsenate in
the presence of arsenite can be overcome by quantitatively oxidizing
arsenite to arsenate and determining the total inorganic arsenic as arsenate
by differential pulse polarography with I)-mannitol as reagent.  As reported
in the literature, the oxidation of arsenite to'arsenate was achieved by
addition of an approximately 200-fold molar excess of cerium(IV) ammonium
nitrate in two equal proportions (66).  An example of a.polarogram of
arsenate obtained by cerium(IV) oxidation of arsenite is presented in Fig. 20.
The arsenite. polarographic wave (curve 1) disappears completely upon
oxidation of the arsenite'.  The arsenate-containing solution does not
produce a polarogram in the absence pf mannitol (curve 2). .After addition of
marinitol the arsenate reduction peak at -0.55 V can be used to determine the
concentration of arsenate (curve 3).

     The amount of cerium(IV) required for complete oxidation of 2xlO~  moles
of arsenite was ascertained by successive addition of ; 40 uL.ofxa. 0.58 M
cerium(IV) ammonium nitrate solution to the solution in the polarographic
cell.  After each addition the polarogram was recorded and the peak height
at -0.34 V measured.  A plot of the peak height versus the amount of
cerium(IV) added (Fig. 21) shows that at least a 30-fold molar excess is
required to completely oxidize arsenite to arsenate;  To ascertain complete
oxidation of arsenite, a 200-fold excess of cerium(IV) was employed in all
experiments as recommended in the literature (66).

     The excess cerium(IV) must be reduced to cerium(III) to avoid inter-
ferences in the polarographic determination of arsenate.  It has been
reported in the literature that the mercury on the bottom of the cell causes
this reduction (66).  However, erratic polarograms were obtained several
times, which were caused by cerium(IV) still present in the solution.

-------
Ai
AE
current: Oi05 mA
scan rate: 2 mV/sec
drop time: 2 sec    •
ref. electrode: satd.  KC1
mod. amplitude: 100 mV
initial potential: 0.0V
scan direction: negative  •
supporting electrolyte: 2.0 M HC10,
            5 ppm	4™
            As(arsenite) /
            X
                                                   jl '\
                                                   J. . l\
                                                   /! \\
                                                   /i  \
                                                   '

                                                           5  ppm As(arsenite)
                                                           oxidized by Ce(IV)
                                                           plus mannitol       /
                                                                  ''•••..          /
                                                                        /
                                                                                  /
                                                                                  .
                                                                   5 ppm As(arsenite)
                                                                   bxidlzed by Ce(IV)
                                                     i
   -0.2
                          -0.4
                                            ,  V
                                      -06.
-as
Figure 20.  Differential pulse polarograms  of  arsenite^ and of arsenate obtained by
            Ce(IV)  oxidation of arsenlte.

-------
                  current: 0.01 mA
                  scan rate: 2 mV/sec
                  drop time: 2 sec
                  ref. electrode: aatd. KC1
                  mod. amplitude: 100 mV
                  20 ppm As(arsenite), 50 mL
                  1.33 x 10  moles
                  supporting'electrolyte:
                  2.0 M HC104
                   0.5 M mannitol
                                 1.9       2.8
                             Moles  of Cerium(IV) x
              Figure  21.   The dependence  of  the  peak height of  the
                          differential pulse polarographic  maximum
                          at -0.34  V (arsenite reduction) as a
                          function  of the amount of  cerium(IV)
                          ammonium  nitrate added.
     The addition of approximately 0.35 g of hydroxylamine hydrochloride
    lO"3 moles) to a solution containing excess cerium(IV) assures that all
the cerium is reduced to the trivalent state.  Excess hydroxylamine hydro-
chloride does not interfere with the polarographic reduction of arsenate
(Fig. 22).

     This procedure allows the determination of arsenite in the form of
arsenate in the absence of arsenate and the determination of total inorganic
arsenic (the sum of the arsenite and arsenate concentrations).

     Fig. 23 shows calibration curves for the determination of arsenate in
the ranges 20 to 160 ppb and 500 to 1500 ppb As(arsenate).  These curves
were obtained by oxidizing 20 ppb As(arsenite) or 500 ppb As(arsenite) by
cerium(IV), reducing excess cerium(IV) by hydroxylamine hydrochloride and
recording the polarograms after addition of D-mannitol.  Aliquots of
                                     46

-------
    Aj
    AE
.1.  0.5  ppm As(arsenite)  + Ce(IV) +
    + NH2OH-HC1
 2.  0.5  ppm As(arsenite)  + Ce(IV) +
    + NH2OH-HC1  + mannitol
                                                      /8\
                                                          \
               standard  addition  pf  arsenate
               to  solution  2.               !
3.
4.
5.
6.
7.
                 1  ppm  As (arsenate)
                 2  ppm  As(arsenate)
                 3  ppm  As (arsenate)
                 4  ppm  As (arsenate)
                                                        current:  0.01  mA
                                                        scan rate:  2 mV/sec
                                                        drop time:  2 sec
                                                        ref. electrpde:  satd.  KC1
                                                        mod. amplitude:  100  mV
                                                        initial  potential: 0.0 V
                                                        scan direction:  negative
               0.5  ppm As(arsen.ate)
         _ 8.  1.0  ppm As (arsenate)
supporting electrolyte:
   2.0 M HC10,
       -0.2
                  -0.4
                                           -O.6
                                Potential, V
                                                                             -O.8
Figure 22.  Differential pulse polarograms of 500 ppb As(arsenate)  obtained  by  cerium(IV) oxidation
            of arsenite in 20 M perchloric acid,  followed by addition of  hydroxylamine hydrochloride,
            addition of D-mannitol and standard addition of arsenate.

-------
00
                                 5O
                                             ppm As(Arsenate)


                                                  1OO
             ISO
      B
      o
      o
      •s
      •H
      01
      X



                                        IT)



                                        ?
                                                                                                     00

                                                                                                     •H ,

                                                                                                     0)
                                                                                                         21
                                    1.5
    Figure  23.   Calibration  curves  for  the  determination of  arsenate  obtained  by  oxidation of  arsenite

                with  cerium(IV).

-------
appropriate standard arsenate solutions were  then pipetted into the polaro-
graphic cell and the solutions polarographed  after each addition.   The
detection limits for arsenic under these  experimental conditions were
calculated to be 6 ppb at the 95 percent  confidence level and 10.5 ppb at
the 99 percent confidence level.

     To check, the reproducibility of  the  method,, a 200 ppb As(arsenate)
solution was polarographed seven times with 2.0. M: HClOi^ as supporting
electrolyte in the. presence of mannitol.  The peak heights were measured in
inches.  The mean of the peak heights was 0.543 inches with a. standard
deviation of 0.012 and a variance of  0.0013:.

     Based on the experimental results described in the previous sections,
the following procedure is recommended for the  quantitative determination
of arsenate and arsenite in water samples by  differential pulse polarography.

     1..  Pipette 200 mL, of the water  sample into a 250 mL volumetric flask.

     2.  Add 42.5 mL of 70 percent perchloric acid and fill to the mark
         with distilled water.       ;  .,vv:    • -;;!.      ;.,.;.:.       ,;.-•

     3.  Pipette 50 mL of this solution into  the polarographic.. cell and
         deaerate for 15 minutes.

     4.  Determine the arsenite concentration by standard addition
         employing the maximum at -0.425  V.                 -

     5.  To another 50 mL aliquot of  the .sample add 4.5 g D_-manriitol,.
         record: the polafogram1.and check  for  an arsenate reduction
         peak-at -0.55 V..,;..;..:...•.„',:'. .- :; •'•'•;•;• ''-....-•  ''.''.'•' ";.V--'..;..;.  .,;...      .

     6.  Determine the arsenate concentration by standard addition
         employing the maximum at -0.55 V if  the. arsenite and. arsenate
         concentrations permit it.

     7.  Oxidize arsenite to^arsenate with a  .200-fold molar excess of
         cerium(IV) ammonium nitrate by addition of the appropriate
         number, of microliters of a 0.58  M cerium(IV) solution.

     8.  Reduce the excess cerium(IV) by  addition of  0.35 g of
         hydroxylamine hydrochloride.

     9.  Record the polarogram and determine  arsenate employing the
         standard addition technique.  Measure  the peak height at
         -0.55 V.

    10.  Subtract the concentration of As(arsenite) obtained in Step 4
         from the As(arsenate) concentration  determined in Step 9  to
         find the( As(arsenate) concentration  originally present .in
         the water sample.
                                     49

-------
     A flow sheet for the recommended analysis procedures is given in
Fig. 24.  If desired, arsenite.can also be determined in another aliquot
after addition of D_-mannitol using the peak at -0.34 V.

     Lacking a natural water sample containing both arsenite and arsenate,
Texas A&M University tap water was spiked with 100 ppb arsenite and 300 ppb
arsenate solutions.  In 2.0 M perchloric acid 103 ppb As(arsenite) were
found in the absence and in the presence of 4.5 g D-mannitol.  The
As(arsenate) concentration was determined as 300 ppb.  After oxidation of
50 mL of the spiked tap water sample with cerium(IV) and addition of
hydroxylamine hydrochloride 392 ppb total arsenic were detected as arsenate.

Experimental                        .

     Reagents

     The sources of the reagent grade chemicals used in these investigations
are listed below.

     1,2-Butanediol:  Aldrich Chemical Company
     2,3-Butanediol:.  Aldrich Chemical Company
     3-Chloro-l,2-propanediol:  Aldrich Chemical Company
     1,2-Propanediol:  Aldrich Chemical Company
     Vanadium(III) Chloride:.  Alfa Division Ventron Corporation
     Acetic Acid:.Allied Chemical
     Hydroxylamine Hydrochloride:  American Drug and Chemical Company
     Fructose:  Eastman Organic Chemicals
     Ceric Ammonium Nitrate:  Fisher Scientific Company
     Ethylene Glycolr  Fisher Scientific Company
     Hydrochloric Acid:  Fisher Scientific Company
     Nitric Acid:.  Fisher Scientific Company     . ,
     Phosphoric Acid:  Fisher Scientific Company
     Pyrogallol:  Fisher Scientific Company
     Disodium Hydrogen Arsenate Heptahydrate:. Fisher Scientific Company
     Sodium Arsenite (NaAs02):  Fisher Scientific Company
     Sucrose:  Fisher Scientific Company
     Sulfuric Acid:  Fisher Scientific Company
     Pyrocatechol:  J. T. Baker Chemical Company
     Zinc:  Mallinckrodt, Incorporated       .
     D-Ribose:  Pfanstiehl Laboratories, Incorporated
     Mercury:  The Instrument Grade, triple-distilled mercury was obtained
     from D. F. Goldsmith Chemical and Metal Corporation.

     Instrumentation

     All polarograms were recorded employing a PAR (Princeton Applied
Research, Incorporated) Model 174A Polarographic Analyzer, with a PAR Model
172A Drop Timer in the differential pulse mode on a Houston Instrument 1000
x-y Recorder.
                                     50

-------
                            {Sample Containing Arsenite and/or Arsenate |
 Determination of Arsenite
                                   Simultaneous Determination
                                     of Arsenite and Arsenate
Determination of Total
 Arsenic as Arsenate
Acidify Sample to 2.0 M HC10.    Acidify Sample to 2.0 M HC10,      Acidify .Sample to 2.0 M HC10,
                            t          ••'•''•' i ••"•'    —     «t                            —

          .*'                               1                                     I
                                      Add D-Mannitol                 Add 200-Fold Molar Excess  of
                                                                    Ce(IV), then NfH-HCl
            .
      DPP of Arsenite
   Peak Height at -0.43 V
                                                                     Followed by Mannitol
                                  DPP of Arsenite and Arsenate*         DPP of Total  Arsenic as
                                  Peak Height at -0.34 V [As(III)]   Arsenate; Peak Height at -0.55 V
                                  Peak height at -0.55 V [As(V)]

*                            '.'•••'
 Determination of arsenate in the presence of airsenite possible only in a restricted range of arsenite
 and arsenate concentrations and arsenite/arsenate ratios.
Figure 24. Flow chart for the determination of arsenite and arsenate by differential pulse polarography.

-------
     The polarographic cell had a side arm and Luggin probe for the
reference electrode.  The cell was equipped with a medium porosity gas
dispersion tube, a platinum wire counter electrode, and the dropping mercury
electrode.  A Coleman Model. 3-711 Asbestos Wick Saturated Calomel Electrode
was used in the side arm as the reference electrode.  The calomel electrode
must frequently be soaked in distilled water to avoid plugging of the
electrode tip by a potassium perchlorate precipitate.

     Degassing of Solutions

     Nitrogen was passed through two scrubbing towers each containing an
aqueous solution of vanadium(III) chloride over amalgamated zinc.  Before
entering the cell the nitrogen was passed through a presaturation tower
containing the supporting electrolyte.  The cell contents were degassed for
15 minutes by passing the scrubbed nitrogen through the solution.  During
the polarographic determination nitrogen was passed over the surface of the
solution.                                                    .

     General Procedure for Sample Preparation

     The stock solutions of arsenite and arseriate were prepared fresh for
each analysis using NaAsC>2 and Na2HAsO^'7H;>0, respectively.  The concentra-
tions of these stock solutions were chosen in such a manner that the
addition of 20 microliters to the contents of the polarographic cell would
produce the desired concentration.  For example, 20 microliters of a 2500 ppm
As(arsenite) solution pipetted into, a 50 mL solution in the polarographic
cell produced a 1 ppm As(arsenite) solution.  The pipette used was a fixed.
volume Eppendorf micropipette.    •

     Standard Addition

     The method of Standard. Additions was used in the analyses for arsenite
and arsenate.  The additions were made by pipetting microliter aliquots of
the arsenite and arsenate solution into the side arm of the polarographic
cell.  All of the solution in.the side arm was then forced into the main
chamber of the cell by applying air pressure by means of a pipette bulb.

     Preparation of Natural Water Samples for Analyses

	A 2QO-mL aliquot of the natural water sample was pipetted into a
150-mL_volumetric~ flask".  Then~42T5 mlTof"~70% :HCTOiT~wefe" addedT  Distilled","
deionized water was used tofill~the"~flisk"t6~'the mark."'"'

     Supporting Electrolytes

     The following acids were used as supporting electrolytes:  sulfuric,
hydrochloric, acetic, phosphoric, and perchloric.  The concentration of the
supporting electrolytes was made 2.0 M with the single exception of
perchloric which was prepared in the range from 0.7 M to 7 tl.  The supporting
electrolytes were prepared by adding the appropriate amount of acid to
deionized distilled water.  For example, 8.5 mL of 70%  (12 M) perchloric
acid were pipetted into a 50 mL volumetric flask containing distilled,

                                     52

-------
deionized water, the contents mixed, and the flask filled to the mark with
distilled, deionized water.

     Polyhydroxy Compounds

     The following polyhydroxy compounds were used as reagents:  3-chloro-
1,2-propanediol, pyrocatechol, pyrogallol, 1,2-ethanediol, 1,2-propanediol,
1,2-butanediol, D_-ribose,, sucrose, fructose, 2,3-butanediol, and D_-mannitol.
The concentrations of the polyhydroxy compounds were 0.5 M_.  They were
weighed only with one-tenth gram accuracy.  For example, 4.5 g of D_-mannitol
were added to 50 mL of solution in the polarographic cell.  The mixture was
stirred until all the mannitol had dissolved.  The resulting solution was
approximately 0.5 M with respect to mannitol.
MASS SPECTROMETRY OF ORGANYLARSONIC ACIDS, DIORGANYLARSINIC ACIDS,
ORGANYLIODOARSINES, ARSENOCHOLINE AND ARSENOBETAINE

     Mass spectrometry has the advantage of high sensitivity, of requiring
only small samples and, of providing both elemental composition and
structural information.  High resolution mass spectrometry will also
distinguish between substances containing only lighter elements and those
having one or more heavy elements in their-molecules.     ;    '.'.",':.

     Arsinic and. arsonic acids could perhaps be determined by mass spectral
analysis in.the residues obtained by evaporation of water samples.  The mass
spectral behavior of'.five organylarsonic'acids and fourteen diorganylarsinic
acids was studied using a.CEC 21-110B mass spectrometer.

     At probe temperatures between 110° and 250° required to. obtain
satisfactory mass spectra,., the arsinic and arsonic acids form anhydrides and
decompose.  The products of these thermal .reactions are then ionized and
fragmented yielding complicated mass spectra with many peaks at m/e values
higher than those expected for the molecular ions of the arsinic and
arsonic acids (Figs. 25,. 26),         .                 .

     There are several reports in the literature concerning the dehydration
of arsonic acids.  Differential thermal and vacuum thermogravimetric
experiments indicated that arsonic acids,;RAsO^H^, are first converted to
diarsonic acids, H02(R)As-0-As(R)02H, and then via polyarsonic acids to the
polymeric arsonic acid anhydridesj (RAs02)n> which decompose on further
heating to arsenic trioxide (67).  Arsinic acids, R2AsOOH, can only form
the anhydrides, [R2As(0)]20.

     The mass spectra of the arsonic acids do not contain molecular ions
for the acids, the diacids or the anhydrides.  The highest mass peak in
the spectrum of methylarsonic acid corresponds to Cl^As^Og.  In the spectra
of most of the acids As^Og is the highest mass peak.  Characteristic
fragments are formed by loss of arsenic and oxygen from As^Og and by loss
of methyl, oxygen and hydrogen from CHsAsOsH.  In all the spectra peaks of
considerable intensity are observed for Asn (n = 1-4).  The arsenic is


                    .                 53              -     '

-------
TOO


S
-S
i

so



s
8
0
• .

—


—

	
=C
™~ ' o
\
\
- j
1
Cj
_f I
eo \
tT ^ .
i fi ^io !:
i I*!* \






S j \
l!
1

111
1


1








° CHiAsO,H, %
x1" ^>j ^ ----- PEAK HSIGHT jc 0 1 • " **
" ' § § o" • 	 «AK HtlGHT»O.OS —





<


i
J|
1 III i
i'g" =~ J, 0~
. • cT* 5


i t




I

a1


1





s
f
tl Uu

C -C ' ' a"
If. |

iii ,
	 	 1 . , ... i . i . i . 1 .
3 5O 10O 15O 20O 25O

c?
' . 9

j * cS" ^, 	
. . | 1 . - C 3*
0^


-------
1UU





01

c
to
. -o
c
3
J3
«T

50










- OJ
•r-
«O
i

°c

- (CH3)2AsOOH

_ — — — DFAIf MFIfXMT v (11


~~ R^CH3

CO
3:
o


— - . in
3:
CM
CJ
	













— •


r«*
or
co w»
0 <








1
I
I i
,1 , ill 1.

) 50































J i

o
I/I

-------
probably formed by decomposition of arsenic-containing materials on the hot
filaments in the ion source.  The most important ions in the mass spectra
of arsonic acids are summarized in Table 4.

     Diorganylarsinic acids- appear to produce molecular ions (Fig. 26,
Table 5) .  The signals for  these ions decrease with increasing molecular
mass of the arsinic acid.   The ions with m/e values equal to or lower than
those of the molecular ions and the ions containing arsenic and oxygen only
are listed in Table 5.  Molecular ions formed from arsinic acid anhydrides
have not been observed in any of the mass spectra.  However, intense peaks
corresponding to R3AS203 are present in the spectra.  The signals with m/e
values higher than those of the molecular ions of R,AsOOH are generated by
fragments arising through loss of alkyl groups, oxygen, hydroxyl groups and
hydrogen from R3As203,

     Mass spectrometry is of little value for the identification of arsonic
acids, but can be used to establish the presence of diorganylarsinic acids
in the residues obtained from water samples.  Exact mass measurements by
high resolution mass spectrometry might be necessary to distinguish
arsenic-containing from' arsenic-free fragments.

     Organyliodoarsines, RjjAsIj^  (n = 1,2) are much more volatile than
arsinic or arsonic acids, are not capable of forming anhydrides and are
easily prepared by treating arsinic and arsonic. acids with hydriodic
acid  (68).  These reactions should be- useful for the conversion of organyl-
. arsenic, acids, in environmental samples to Organyliodoarsines, which can then
be determined by mass spectrometry.  Arsenite and. arsenate are both
transformed to arsenic triiodide under these conditions.

     Purified samples of Organyliodoarsines were employed to obtain their
mass spectra at probe and source temperatures given in Tables 6 and 7:  All
the iodoarsines gave intense molecular ion peaks .  Fragmentation proceeds
by loss of alkyl groups, iodine and hydrogen abstraction..  Peaks of lower
intensity corresponding  to  AsI3, R2 As I. and Asn  (n = 1-4) were observed in
many of the spectra.  Thermally induced disproportionation of RasI  into
R2AsI and Aslj is a probable mechanism for the formation of these compounds.
Interesting series of fragments represented by the formulas CnH,n+1AsI,
CnH2a+2As anc* cnH2nAs are Part °^  tlle mass spectra of Organyliodoarsines.
It appears that methylene groups are lost generating signals fourteen mass
units apart from n = 0  to n equal  to the number of carbon atoms in the
Organyliodoarsines.  Additional experiments are in progress to verify this
unusual cleavage of methylene.  As examples for the mass spectra of Organyl-
iodoarsines the spectra  of  butyldiiodoarsine and dioctyliodoarsines are
presented in Fig. 27 and Fig. 28,  respectively.  Organyliodoarsines are
well  suited for the mass spectrometric identification of organic arsenic
compounds which can be  converted to their iodo derivatives .

     Arsenocholine chloride [(CH3) 3AsCH2CH20H]+Cl~,  (Fig. 29) and
arsenobetaine chloride,  [(CH3)3AsCH2COOH]+Cl,  (Fig. 30) produce rich mass
spectra, which, however, do not contain molecular ion peaks.  The highest
mass peaks correspond to (CHa^AsCI^CI^O and  (CH3)2AsCH2COOH, which are
                                       56  .

-------
ABLE 4.  RELATIVE ABUNDANCE OF IONS IN THE.ELECTRON IMPACT MASS SPECTRA OF
         ORGANYLARSONIC ACIDS

Ion
R
RAs
RAs02
RAs02H
RAs03
RAs03H
RAs03H2
RAs°3H3
R2As
TJ A oC\
*\fj£\9\s - ,
R2As02
R2As02H
13 AsO "H
As
AsO
As62" ''•••'• :
As03
AS202
As2°3
'^394
As406
Probe Temp PC
Source Temp °C
CH,
0.00
3.60
1.15
100.00
24.04
0.96
0.00
8.30
1.63
7.20
.53.90
24.0.0
4.00
4.40
75.00
5&. 70
—
4.70
0.96
40.38
75.96
200
210
R in RAsO
11.32
0.40
3.40
100.00
0.00
1-21
0.16
8.29
. 0.00
0.58
18.29 .
7.63
3.95
1.84
53.95
0.53
0.89
3.80
0.76
19.00
21.60
165
235
3H
100.00
0.57
0.64
29.90
0.50
0.64
0.00
2.36
0.00 ,.-
0.00
5.36
1.57
-1.64
2.86
54.29
3.29
4.79
3.43
0.57
9.29
8. 57
180
210
C9H19
0.38
7.30
1.03
0.23
0.00
o.oo
0.00
0.00
2.40
0.34
0.00
0.00
0.00
0.34 ~"
0.91
0.11
1.14
0.11.
5.14
0.80
2.40
210
230
C6H5- '
31.43
21.43
0.00
0.00
0.00
o.oo,
0.00
0.00
19.29
, 3.29
,'11.43
3.00
1.71
4.71
80.71
0.00
0.43
9.29. *
0.00
62.86
92.86
255
210
                                 57

-------
     TABLE 5.   RELATIVE ABUNDANCES OF IONS IN THE ELECTRON IMPACT MASS SPECTRA OF DIORGANYLARSINIC ACIDS
01
00

Ion
R
RAs
RAsO
RAs02
RAs02H
RAs03H
RAs03H3
R2As
R2AsO
R2As02
R2As02H
R2As02H2
As
AsO
As02
As03
As202
As 2 03
As30^
As4°6
Probe Temp °C
Source Temp °C
™3
5.56
7.22
4.56
8,33
6.67
0.56
0.77
40.56
51.11
0.67
15.56
8.89
2.33
36.67
26.67
6.67
4.56
3.33
0.00
0.00
184
210
C2H5
o.oo
0.96
0.93
0.00
0.37
0.00
0.00
1.00
2.11
1.15
7.41
0.44
0.33
10.37
0.11
0.00
20.37
0.33
2.22
1.89
110
200
S*7
3.48
1.96
0.43
0.30
0.30
0.07
0.20
0.50
0.39
1.17
0.20
5.87
0.02
11.02
1.24
0.04
2.61
0.28
0.91
1.30
120
200
R in R_
C4H9
5.00
3.11
2.30
0.22
0.32
0.00
11.89
0.22
1.62
0.00
0.11
0.49
0.03
2.30
1.11
0.00
2.30
0.24
1.62
0.92
130
220
AsOOH
C5H11
4.00
5.00
3.17
0.67
0.50
0.00
0.00
0.33
1.98
0.00
0.00
0.00
0.25
3.00
1.42
0.00
2.08
0.25
1.42
10.00
165
210
'"tfu
100.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
9.77
4.65
4.65
23.26
1.16
1.63
1.86
0.00
1.16
0.00
0.93
0.70
150
210
C8H17
1.19
17.84
4.83
12.27
4.46
0.52
1.86
1.64
1.78
0.74
6.15
4.01
0.37
5.58
3.94
0.52
0.89
0.97
2.68
8.18
202
210
C6H5
18.80
25.56
2.33
0.33
0.00
0.00
0.00
46.94
23.89
14.44
3.33
5.28
1.28
10.56
0.00
0.00
3.89
0.00
1.83
6.33
220
220
C6WH2
100.00
24.27
0.78
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
•;"••' o.oo
0.58
9.90
0.78
0.00
3.50
0.00
12.62
18.45
„ 220
/** 220

-------
TABLE 6.  RELATIVE ABUNDANCES OF IONS IN THE ELECTRON IMPACT MASS SPECTRA OF ORGANYLDIIODOARSINES
Ion
R2AsI
RAsI2
RAsI
AsI2
Asl
R2As
RAs
AsI3
I
As
R
I2
Probe Temp QC
Source Temp °C
CH3
0.19
100.00
51.58
11.84
1.42
0.08
10.53
0.10
13.95
3.42
0.47
2.92
40
220
c2
0
100
48
24
31
0
3
d
16
7
9
1


V
.15
.00
.75
.88
.86
.06
.25
.31
.86
.81
•69 i
.31
45
190
C3H?
0,14
100.00
23.88
32.24
41.63
0.05
2.45
0.86
16.73
4.29
61.22
1.69
30
190
R
0
57
13
19
20
- -
4
0
9
3
100
0


in RAsI2
.24
.58
.33
.39
.61
r— '_ .
.24
.18
.09
.21
.00 ^
.72
40
210
0.30
65.57
21.52
24,78
20.87
0.04
3.48
0.83
7.83
2. 83
63.04
0.69
35
210
C7H15
21.05
7;02
7.02
6.14
7.50
2.10
37.74
	
4.34
0.87
1.66
0.46
60
190
C14H29
3.23
35.48
100.00
85.48
30.65
0.09
6.77
	
48.39
5.00
14.35
0.32
85
250
C6»5
0.03
31.03
100.00
2.53
2.64
0.14
30.34
2.76
4.60
0.92
0.60
1.84
50
200
Cx-Hg-CHnCHn
0.26
2.30
57.45
2.77
3.19
0.18
31.06
0.91
5.74
0.98
100.00
1.96
55
: 20°

-------
 TABLE 7.   RELATIVE ABUNDANCES OF IONS IN THE ELECTRON IMPACT MASS SPECTRA.OF
           DIORGANYLIODOARSINES

Ion
R^I
RAsI2
RAsI
AsI2
Asl
R2As
RAs
As
R
I
*2
CH3
100.00
0.41
63.84
0.32
16.07
1.09
61.61
8.04
3.57
15.18
0.15
C-C6H11
25.24
0.10
12.38
0.14
1.31.
1.21
3.33
0.21
100.00
2.00
0.29
R in R2AsI
C8H17 C10H22
35.96
1.05
3.93
2.13
7.30
3.47
100.00
1.01
4.83
5.39
0.20
44.96
3.88
8.37
4.42
6.36
5.81
100.00
0.62
4.19
4.57
0.23
C6H5CH2CH2
6.82
1.36
0.07
1.23
97.73
97.73
0.93
100.00
2.05
0.07

formed by thermal cleavage of HC1 and CHsCl from the arsonium salts.
Fragmentation proceeds by loss of methyl groups, cleavage of hydroxyl groups
and loss of ethylene..  Fragments containing arsenic-oxygen or arsenic-
chlorine bonds were either formed by thermal decomposition of the arsonium
salts or through intramolecular rearrangements.  In spite of the absence of
molecular ions mass spectrometry can provide an indication of the presence
of arsenocholine or arsenobetaine in a sample.
                                      60

-------
1UU



O)
u
03
c
a
•f
so






01
.^
fO

O)
a:





















co-








r* <
2: z
m
•



in
CVJ
o









j






J>

a*



^ •
QK







1 A 1
bAsh
PEAK











HEIGHT x 0.1











•

z
(fl
eC
CT»-

U







f

"^n 
Kj <^

Z
ro
o
•




I

1 ,I.I,L











ll
K.J..-I-..I 	 ....1.
100








i
«




i 1 i

"LO "%x "^n

c*j « «•
00 0



fl
c






























15O 2OO
• .,-•• ... .... -, m/e ,. .., .....'
  100
   so
             <—•      VI    (/I      l/l
             J2     <    <     . <
                   00      O
           11.1
                                                                                         <:   vi

                                                                                         3?!
                                                                                         O


                                                                                         i
                                                                                         i
                                J....1.-.L..I....I	I....I....I.	I	I	L..I....I....I....I....	l....i....l....i....l.-.i....l-.i..
20O
                          25O
 30O
m/e
35O
4OO
Figure  27.   The mass  spectrum of  butyldiiodoarsine.
                                                61

-------
                                                                       	PEAK HE KMT ix 0.1
                                   150
                                               200  m/e    2SO
                                                                     300
                                                                                350
                                                                                            400
Figure 28.  The mass  spectrum of dioctyliodoarsine.   The peaks marked by  *  and ** belong to the
            series  CnH2n+1AsI and cnH2n+2As, respectively.

-------
 1OO
  OJ
  0
  CO
  TJ
  -Q
  SO
  OJ
  >
((CH3)3AsCH2CH2OH)c

" CO
DC


,_ —




>M_




^"T


^ 	 	




•HW



"•^T"





_
.11




























h
JL








-r-
O
cv
3C
(_J

































*•
3
™

H
• c_

' 3
c_










CJ







1
1

	 J....I....1 	 I....


31
i

•






|o


i





i ,
i i
'*.


'



5
NJ

5
CM

3


















.
U L

50



'f—.
o
CO
3C
O


















5 ' 	 PEAK HEBGHTx 0.1
o .......
CO CO :—
. =t =t O
: CM co to









to
^^
CO
"'.'.- DC
O
\







10
•a
";






...1
...J....I....I....



;

\


1'
.
.1










i
i
U
M
n
ii
^•^ ^C "3T T"
CO CO CO CM CD
ac ac — » oc • ac CM —
CJ CJ CO O O '3C
	 • / 3= CM CM 0
U 3Z 3: CM
«— • • • O O 3C
























1
1
1 i 1 i
CO CM O 	
i - .- • ~f— •
VI ^ " ^t—

O

\

_,


..-— ^
3C
















(

,t- 	 nc A i/ LJCI/^
PEAK HEJu


HTxO.05
HTxttl

01
^C
C^
nz
t_5





»~~
CJ



Nl
O
^



1 \

o
— o

|
L
,...,...,1 	 .,.,( 	 i....
j




1




i

I
CO
oc




CO
•*

i
i • '
• •
: I 1 •
Li) ilJiLjiliL



















li

CO
CO
*~^
CO CO r-
< Z C
CM O •— C
-— ~ 	 <_> <
CO : CO
n: i «a: ^~
CJ • CO
^__^«















JihllllilLJ
...i 	 ...i, 	 .,i,,,i...,i....l....i....l.........i....:....l.......,,
or n
o c.

; •
/ ;


;







g
g
I
Jl









....I....I...J,...

















j
0
C
CM
^
CO

5
_^








_











3C
O
O
C_)
CM

t_>
CO
=t
CO
:n
O






,






l_5 _
"-•-. co
CO
CO 	
a:
o




T*
0
o
t_3 — —
CM

CJ
CO __ _

CM
CO
^C <~w«
o
.

• • — ,
/ -'••''
•
, 1
IidiQtQiln
    O
                           5O
1OO
200
Figure 30.  The mass spectrum of arsenobetaine chloride.

-------
                                 SECTION 7

                         ANALYSIS OF WATER SAMPLES
     Samples of arsenic-containing drinking water supplies selected.by the
project officer were collected and shipped to College Station as quickly as
possible.  Total, arsenic concentrations and the concentrations of arsenite
and arsenate' were, determined by several methods.  Each water- sample was
checked for the presence- of. methylated arsenic compounds and. other organic
arsenic derivatives.  Water, samples from Hinckley, Utah; Delta, Utah.;
Barefoot Site, Alaska;. Mauer Site, Alaska; Antofagasta, Chile; Yenshei,
Taiwan; Hartlin Site,. Nova. Scotia; and Sullivan Site, Nova Scotia were
investigated.

     The following methods were: employed for the determination of total
arsenic and arsenic compounds.

        ® Graphite furnace atomic absorption spectrometry:  A
          Hitachi Zeeman graphite furnace atomic absorption
          spectrometer with 'graphite cups was employed for total    •
          arsenic determinations. . Forty microliters were injected
          in most' cases.. - Since GFAA is subject' to many inter-
          ferences, arsenic was extracted from water samples.
          The extract was then injected into-the graphite cup.

               A suitable aliquot of the water sample is made
          0.1 M with respect to hydrochloric acid.  Arsenate is
          reduced to arsenite by addition of 5 mL of a solution
          containing NaHSOs, NaaSgOs and HC1.  The solution is
          extracted with 5 mL of 0.01 M (CifH90)2PS2H in hexane..
          The extraction is repeated four times with 2 mL each
          of the reagent solution.  The combined organic layer
          is washed first with 4.5 M HC1 and then with water.
          The organic layer is extracted twice with 2 mL of
          bromine water.  The aqueous phases are transferred
          to'a.volumetric.flask, adjusted to 0.1" M acid with
          nitric acid, brought to 200 ppm with respect to
          nickel by addition of nickel nitrate and treated with
          50 vL of 30% hydrogen peroxide.  After filling to the
          mark with distilled water, 20 yL are injected into the
          graphite cup for analysis.  The standards and blanks
          are prepared in the same manner.
                                     65

-------
        © Differential pulse polarography:   Arsenite, arsenate and
          total arsenic were determined by DDP as described in
          Section 6.

          High pressure liquid chromatography/GFAA as element-
        ® specific detector:  Concentrated water samples were
          analyzed by the HPLC-GFAA method to check for the
          presence of organic arsenic compounds as described in
          Section 6.

        @ Hydride generation technique:  Arsenite, arsenate and
          methylated arsenic compounds are reduced by sodium
          borohydride at pH 1..  At pH 6.5 only arsenite is reduced.
          The reductions are carried out in a chamber capable of
          holding 10 mL of sample.  After the addition of the
          appropriate buffer'solution a measured volume of sodium
          borohydride solution is pumped into the chamber.  A
          stream of helium transports the arsines through a U-trap
          cooled in an isopropanol/dry ice bath to remove water.
          The arsines are collected in a U-trap cooled by liquid
          nitrogen.  One leg of this trap is packed with silanized
          glass beads.  After the arsines had been transferred to
          the.U-trap, they are rapidly volatilized by electrically
          heating the trap.  A stream of helium transports the
          arsines through a short tube packed with sodium hydroxide
          pellets to remove carbon dioxide into the helium discharge
          created, by passing direct current between two helium-arc
          welding electrodes.  The radiation from the discharge
          passes through the quartz window of.the emission chamber
          and is focused on the. entrance slit of the monochromator
          set at the 228.8 nm arsenic emission line.  A gas
          chromatograph can be attached to the exit of the arsine
          trap to improve separation of AsH3 from methylarsines.
          The hydride generation technique has detection limits
          of 0.2 ppb for arsine, 0.3 ppb for methylarsine and
          1 ppb for dimethylarsine.

        ® Inductively coupled argon plasma emission spectrometry:
          ICP is the most convenient method to determine the trace
          element concentrations including arsenic concentrations in
          water samples.  An ICP spectrometer became available at
          Texas A&M University during the last months of the
          project.  The recently collected samples were analyzed
          for 48 elements by the simultaneous ICP Model 34000
          manufactured by ARL.  Prior to this time ICP analyses
          were carried out with ARL instruments at Radian
          Corporation in Austin, Texas or at the ARL plant in
          Sunland, California.

As an illustrative example for the sampling procedures the sampling
instructions for the Taiwan well waters are cited.  Sampling instructions
                                      66

-------
were pasted to each bottle.  The correct amounts of preservations in vials
were also attached to the bottles (with exceptions of nitric acid).

Rinsing of all bottles before filling:

     1.  Fill bottle half full with well water, close bottle with
         cap and shake thoroughly.  Empty bottle.

     2.  Repeat procedure of item 1 three times.

     3..  Fill the bottle as described below.

     4.  Tightly close the cap, wipe it dry and tape securely with
         the tape provided.

Glass container for unpreserved sample (Well I):

Fill the rinsed bottle with water to the brim, close the. cap tightly and
secure by taping the cap with the tape provided.  Provide the information
requested on the label.

Glass container for sample preserved with HgCl^ (Well I):

Fill the rinsed bottle with water almost to the brim.  Add ,all of the
mercuric chloride in the vial to the sample.  Fill to the brim, close the
cap tightly and secure by taping the cap.  Shake the closed bottle to
dissolve the mercury salt..  Provide the information requested on the label.

Plastic containers for unpreserved sample (2 containers for Well I;
2 containers for Well II):     .

Fill the rinsed bottles with water to the brim.  Make sure no air space
remains in the bottle.  Close the cap tightly and tape.  Provide the
information requested on the label.

Plastic containers for sample preserved with 0.1 weight % ascorbic acid
(2 containers for Well I; 2 containers for Well II):

Fill the rinsed bottles with water almost to the brim.  Add all the
ascorbic acid from the vial to the sample.  Fill to the brim, close the
cap tightly and secure by-taping.  Turn the bottles several times to aid
the dissolution of ascorbic acid.  Provide the information requested on the
label.

Plastic containers 'for samples preserved with nitric acid
(2 containers for Well I; 2 containers for Well II):

Fill the container almost to the brim.  Add 7 mL of ultrapure nitric acid.
Fill to the brim, close the cap and secure by taping.  Shake the bottle
several times to make the sample homogeneous.  Provide the information
requested on the label.


                                      67

-------
Nitric Acid Control Solution;

Measure 7.0 mL concentrated nitric acid from the nitric acid bottle used
for the preservation of the water sample into a clean 1.0 L volumetric flask.
Fill to the mark with distilled water and shake to homogenize the solution.
Fill the two 125 mL-plastic bottles with this solution, tighten the cap
securely and tape the cap.
                                         X
The results of the analyses of the water samples are presented in the
following sections..

     The samples were shipped in sturdy cardboard or wooden boxes.  None of
the samples were shipped frozen.
     Water samples from Delta and Hinckley, Utah were collected by
Mr. E. Western, Utah State Department of Health, Division of Environmental
Health on July 16, 1979.  Pyrex glass bottles and Cubitainers were employed
as containers.  Some of the samples were preserved by addition of 0.1 weight
percent of ascorbic acid.  The samples were shipped to College Station as
air freight and received July 17, 1979.  The analyses were performed during
the period July 17 to 25, 1979.

     The following analytical methods were employed:  graphite furnace
atomic absorption spectrometry (GFAA) using a Hitachi-Zeeman GFAA Model
170-70; differential pulse polarography (Princeton Applied Research Model
174A); hydride generation with atomic emission spectrometric. detector
(Braman method); inductively coupled argon plasma emission spectrometry
(ARL 34,000).               ~~

Hinckley Water Samples

     The results of the total arsenic analyses of the Hinckley water samples
are listed in Table 8.  The average arsenic concentration in the Hinckley
water samples obtained from 36 determinations (twelve each by graphite
furnace atomic absorption spectrometry, differential pulse polarography and
hydride generation) is 185±8 ppb.

     The analyses of the water samples by ICP were carried out several
weeks after the collection, because the instrument had not been available
earlier.  The samples stored in plastic containers gave ICP arsenic
concentrations of 170 ppb (without ascorbic acid) and 160 ppb (with ascorbic
acid), which are in fair agreement with the average value from the other
methods.  The ICP values in glass containers (140 ppb and 120 ppb) were
considerably lower than the values obtained from samples stored in plastic
containers.  Glass containers in contrast to polyethylene containers are
known to adsorb arsenic on prolonged storage (see section 5).  Samples
stored in glass containers are, therefore, expected to have less arsenic in
solution than fresh samples or samples stored in plastic Cubitainers.
                                      68

-------
TABLE 8.  TOTAL ARSENIC CONCENTRATIONS IN ppb IN THE HINCKLEY AND DELTA WATER SAMPLES

)H H
y <0 (U
•H O Ci C!
' •? "1 'd * "3
M 4J n) X id
O TJ 0) 4-1 
-------
     The determinations of arsenic compounds were performed employing the
hydride generation (HG) technique and differential pulse polarogrphy (DPP).
The results are summarized in Table 9.  Only arsenite and arsenate were
found.  The average concentration of arsenate was 172±7 ppb (by HG) and
187±8 ppb (DPP).  The ars'eni.te concentrations averaged 11 ppb.  The arsenite
concentrations  (by HG) in the samples preserved with ascorbic acid were
approximately four times higher than in the unpreserved samples.  Ascorbic
acid might have reduced some arsenate (see section 5) at these low arsenic
concentrations.  The sum of the concentrations of arsenite and arsenate agree
well with the values of the total arsenic concentrations which were deter-
mined separately.

     Organic arsenic compounds, if present at all, cannot occur at concen-
trations larger than a few part per billion.  The search for methylated
arsenic compounds by the hydride generation method, which has detection
limits for- such compounds of approximately 1 ppb did not produce any
indication of their presence.

     The results of the ICP analyses of the Hinckley water samples are
summarized in Table 10.  With the exception of the concentrations of As, B,
Ba and Si the listed values are averages obtained from the concentrations of
the following four samples:  Pyrex container with 0.1% ascorbic acid;
Pyrex- container without ascorbic acid; plastic Cubita'iner with 0.1% ascorbic
acid; plastic Cubitainer without ascorbic acid.  The values for As, B, Ba
and Si are averages of the samples stored in Cubitainers.  The elements
present in concentrations larger than the detection limits (Table 10) are
As, B, Ba, Ca,  Cu, Mg,. Na, Si and Sr.  These average concentrations have
average deviations of ±5 units in the last digit of the listed numbers.
The boron, barium and silicon concentrations in the samples stored in
Pyrex containers were 2.25 ppm (B), 0.13 ppm (Ba) and 15.5 ppm  (Si),
significantly higher than in the samples stored in Cubitainers.  According
to information  supplied by the Wheaton Glass Co., the manufacturer of
the Pyrex containers, a- soda ash-barium-borosilicate melt was; .used in the
manufacture of  the containers.  The sodium concentration averaging 233±3 ppm
was the same in all four samples.

     No significant changes, in arsenite, arsenate and total arsenic
concentrations  were observed during the one week period, during which the
initial analyses were performed.

Delta Water Samples

     The results for the Delta water samples are summarized in Table 8
(total arsenic  concentrations), Table 9 (arsenite and arsenate concentra-
tions) and Table 11  (trace elements by ICP).  The total arsenic concentra-
tion is approximately 21 ppb.  Arsenite and arsenate concentrations are
comparable in magnitude and close to the detection limits of the methods
employed.  According to ICP results  (Table 11) Ba, B, Ca, Cu, Fe, Mg, Na,
Pb, Si and Sr are present in the water samples.  The Ba and B concentrations
in the samples  stored in glass containers are again higher than those of
samples kept in plastic containers.  The Delta water samples are harder
                                      70

-------
!TABLE 9.   CONCENTRATIONS OF ARSENITE AND ARSENATE IN ppb IN THE HINCKLEY AND DELTA WATER SAMPLES

U M '
0 a) Q)
! -H o d rt . • ,
•° 'H -rl °d TT/mat
, • l-i 4-1 ctt • X ni . : HG**
O TJ • COW CU-W
o -H •• a a ^ o ;
, Location 
-------
TABLE  10.  TRACE ELEMENT CONCENTRATIONS IN THE HINCKLEY WATER SAMPLES DETERMINED BY ICPf
H
•°ff
5.5
Na
233
42
K
Rb
Cs
Fr

0.46
Be
34
Mg
1.83
45
Ca
3.24
0.44
Sr
0.10
0.58
Ba
).026
Ra
Concentration of the
Element in ppra Found
in the Sample*
*The absence of a yyyy
concentration is at 01
8
Sc
2
Y
La
Ac
16
Ti
4
Zr
Hf
8
V
Nb
Ta
8
Cr
8
Mo
18
w
E
•yyyy

	 JJCLCI.I-J.IJII IJ-LUU.U in yy
figure indicates that the
r below the detection limit.
Mn
Tc
Re
8
Fe
Ru
Os
6
Co
Rh
Ir
10
Ni
Pd
25
Pt
4
Cu
0.09
2
Ag
48
Au
10
Zii
6
Cd
32
Hg
u
9
B
1.2
50
A!
Ga
55
In
91
Tl
C
16
Si
13.4
Ge
120
Sn
84
Pb
•H
150
P
57
As
0.165
32
Sb
49
Bi
O
S
84
Se
103
Te
Po
H
F
a
Br
1
At
He
Ne
Ar
Kr
Xe
Rn

                 Ce
Pr
Nd
Pm
Sm
Eu
Gd
Tb
Er
Th
Pa
64
U
Np
Pu
Am
Cm
Bk
Cf
Es
Fm
Md
No
Lr
          ''"The average deviation of  the listed concentrations is ±5 units of the last digit.

-------
T^.BLE 11.  TRACE ELEMENT CONCENTRATIONS IN THE DELTA WATER SAMPLES DETERMINED BY ICP1
H
?ff.
5.5
Na
65
42
K
Rb
Cs
Fr
?'

0.46
Be
34
Mg
6.63
45
Ca
1.5.4
0.44
Sr
0.52
0.58
Ba
0.045
Ra
Concentration of the
Element in ppm Found
in the Sample*
*The absence of a yyyy
concentration is at o
8
Sc
2
Y
La
Ac
16
Tf
4
Zr
Hf
8
V
Nb
Ta
8
Cr
8
Mo
18
w
E
•yyyy


figure indicates that the
r below the detection limit.
Mn
TC
Re
8
Fe
0.007
Ru
Os
6
Co
Rh
Ir
10
Ni
Pd
25
P*
4
Cu
0.26
2
Ag
48
Au
10
Zn
6
Cd
32
Hg

9
B
0.01
50
AS
Ga
55
in
91
Tl
C
16
^
Ge
120
Sn
84
Pb
1.17
N
150
P
57
As
32
Sb
49
Bi
0
S
84
Se
103
Te
Po
H
F
Ci
Br
1
At
He
Ne
Ar
Kr
Xe
Rn

Ce
Pr
Nd
Pm
Sm
Eu
Gd
Tb
Dy
Ho
Er
Tm
Yb
Lu
Th
Pa
64
u
Np
Pu
Am
Cm
Bk
Cf
Es
Fm
Md
No
Lr
      average deviation or the listed concentrations is ±5 units of the last digit.

-------
 (15 ppm Ca, 6.6 ppm Mg)  than  the Hinckley samples.  The sodium concentra-
 tion is 65 ppm.  The relatively high Pb  (1.3 ppm) and Sr  (0.5 ppm)
 concentrations are noteworthy.

     Organic arsenic compounds were not  detected in the Delta water samples.
WATER SAMPLES FROM ANTOFAGASTA, CHILE

     Water samples from  the drinking water supply for the city of
Antofagasta, Chile were  collected by Dr. K. J. Irgolic at the water treat-
ment plant on October  21, 1980.  Three samples were taken from the untreated
water just before it entered:  the treatment plant.  The faucet attached to
the large delivery pipe  was opened and water was allowed to run for five
minutes.  Then  the one-quart  plastic Cubitainers were rinsed with the water,
filled to the brim and closed tightly.  No air was left in the container.
One of these samples was preserved with 0.1% by weight of ascorbic acid.
Two samples were similarly taken from the pipe leaving the water treatment
plant.  All samples were collected in plastic Cubitainers.  The analyses
were performed  during  the period October 25 to November 10, 1980.

     The following.analytical methods were employed:  graphite furnace
atomic absorption spectrometry  (GFAA), differential pulse polarography;
hydride generation and inductively coupled argon plasma emission spectron-
etry.

     The results of the  total arsenic analyses on the five water samples
from Antofagasta are summarized in Table 12.  The average arsenic concen-
tration  (average deviation) obtained from all values for the untreated
samples listed  in Table  12 is 0.910.1 ppm.  This, value is close to the
concentration of 0.8 ppm reported earlier for this water (69).  The results
for the untreated samples with and without ascorbic acid as preservative are
in good agreement when the values obtained with a particular method or
instrument are  considered.  The ICP analyses were carried out on three
different instruments  and with three different standards.  The treated
water samples gave arsenic concentrations in the range of 0.25 ppm to
0.51 ppm  (Table 12).   After passing through the alum treatment for arsenic
removal the treated water has generally an arsenic concentration of
approximately 80 ppb  (70).  The high arsenic values in the treated water
samples collected at Antofagasta on October 21, 1980 could have been caused
"by a malfunction of the  water treatment plant.  The removal of arsenic by
alum coagulation is highly pH dependent.  The percent arsenic removed from
arsenite-containing, chlorinated well water and from arsenate-containing
tap water decreased from 90 percent at pH 7 to approximately 20 percent at
pH 8.5  (70a).   The treated water sample could have had a pH above 7.  The
pH values of the treated and  untreated samples after nine months of storage
in Cubitainers  were 7.8  and 8.2, respectively.  Another cause for the high
arsenic values  of  the  treated water samples could have been a local region
of trapped arsenic-rich  water or soluble deposit in the vicinity of the
faucet from which  the  samples were taken.
                                      74

-------
TABLE 12.  TOTAL ARSENIC CONCENTRATIONS IN ppm IN THE ANTOFAGASTA WATER SAMPLES

Sample
Untreated I
Untreated II

Untreated
Treated I
Treated II
Ascorbic Acid HG*t GFAA*t DPP*t ICP*t
No 0.71±0.02 0.75+0.02 0.78*0.03 0.90, 0.78, 1.06
No .'••!- 0.73+0.02 0,78*0.03 0.91, 0.81, 1.02
i
Yes 0.72±0.06 0.70±0:.02 - 0.91, 0.82, 1.05
No O.;43±0,02 0.40+0.01 0.41±0.01 0.51, 0.42, 0.33
No - 0.36+0.01 ' - 0.32, 0.33, 0.25
  HG:  Hydride generation.
 DPP:  Differential pulse polarography.
 ICP:  Inductively coupled argon plasma emission spectrometry.
GFAA:  Graphite furnace atomic absorption spectrometry.

 Average + average deviation of at least three determinations for each sample.

-------
     The arsenic compounds present in the Antofagasta water samples were
determined by the hydride generation technique and by differential pulse
polarography.  Most of the arsenic is in the form of arsenate (Table 13).
The arsenite concentrations in the preserved and unpreserved, untreated
water samples are at most 16 ppb.  The treated.water sample contained 3 ppb
arsenite.  No-methylated or other arsenic compounds were detected.

 TABLE 13.   CONCENTRATIONS OF  ARSENITE AND ARSENATE IN ppm IN. THE ANTOFAGASTA
            WATER' SAMPLES                    .
 Sample               Ascorbic Acid           Arsenate**           Arsenite**
Untreated
Untreated
Treated
No 0.70±0.02*
0.78±0.03t
Yes 0.7010.02*
0.78±0.03f
No 0.43±0.02*
0.41+0. Or1"
0.016±0.001*
0.016±0.001*
0.003+0.001*
 * Hydride generation method  1" Differential pulse polarography
 **Average + average deviation from at least three determinations given.


     Trace elements  in the Antofagasta water samples were determined by
 inductively coupled  argon emission spectrometry employing three different
 instruments.  Among  the 41 elements on the  analytical arrays of these
 instruments Li, Na,  K, Mg, Ca,  Sr, Ba, Fe,  B, Al,  Si, P, and As were present
 in the untreated  (Table 14) as  well as the  treated  (Table 15) water samples
 above the detection  limits, which are also  listed  in these Tables.  The
 values obtained for  Mn and Cu are very close to the detection limits.
 Selenium was not  found in the samples (detection limit 84 ppb).  The
 beryllium content (7.4 ppm) of  the untreated samples is noteworthy.
 Beryllium is a proven carcinogen  (70b, c) ,  and beryllium sulfate has been
 shown to cause enhancement of SA7 transformation of fetal cells of the
 Syrian hamster (70d).  Additional analyses  for beryllium in the untreated
 water samples should be carried out.  The treated water samples did not
 contain beryllium.

     The  trace element concentrations obtained for  the untreated, not
 preserved, samples did not differ from those found  in the untreated sample
 preserved with 0.1%  by weight of ascorbic acid with the exception of the
 values for Fe and Mn.  Therefore  the nine values available for each of  the
 other elements were  averaged.   These average values with their average
 deviations are listed in Table  14.  The  Fe  and Mn  concentrations in the

                                      76

-------
TABLE 14.  TRACE ELEMENT CONCENTRATIONS IN THE UNTREATED ANTOFAGASTA WATER  SAMPLES  DETERMINED BY ICP"1"
H
°tf
.62/2
5.5
Na
102/2
42
K
L3.3/7
Rb
Cs
Fr

Be
7.4/2
34
Ma
7.4/2
45
Ca
20JD/2
0.44
Sr
.28/2
0.58
Ba
008/1
Ra
Concentration of the-—
Element in ppm Found
in the Sample* :
*The absence of a yyyy
concentration is at 61
8
Sc
2
Y
La
Ac
16
Ti
4
Zr
Hf
8
V
Nb
Ta
8
Cr
8
Mo
18
w
E
•yyyy


figure indicates that the
r below the detection limit.
Mn
.002
Tc
Fie
8
;f&
Ru
Os
6
Co
Rh
Ir
10
Ni
Pd
25
Pt
5M
2
Ag
48
Au
10
Zn
6
Cd
Hg

9
B
2.8/2
50
.0873
Ga
55
In
91
T!
C
16_
3.8T/6,
Ge
120
Sn
84
Pb
.N
150
P-
.'34/3
57
As
.91/8
32
Sb
49
Bi
O
S
84
Se
103
Te
Po
H
. F-
Cl'
Bf
I
At
He
Ne
• Ar
Kr
Xe
Rn
'- ' ' ' • ' • "
Ce
Pr
Nd
Pm
Sm
Eu
Gd
Tb
Dy
Ho
Er
Tm
Yb
Lu
Th
Pa
64
u
Np
Pu
Am
Cm
Bk
Cf
Es
Fm
Md
No
Lr
'''Average concentrations/average deviations are given in the form 20.0/2 to be read as 20.010.2.

-------
TABLE 15.   TRACE ELEMENT CONCENTRATIONS IN THE TREATED ANTOFAGASTA WATER SAMPLES DETERMINED BY ICP
H
OA3
• Li
.64/1
5.5
Ha
103/1
42
K
112/8
Rb
Cs
Fr

0.46
Be
34
Mg
7.572
45
Ca
20.3/;
0.44
Sr
.29/2
0.58
Ba
008/1
Ra
Concentration of the ••••
Element in ppm Found
in the Sample*
*The absence of a yyyy
concentration is at 01
8
Sc
2
Y
La
Ac
16
Ti
4
Zr
HI
8
V
Nb
Ta
8
Cr
8
Mo
18
w
xxxx-
E
•yyyy

	 Detection Limit in pp
figure indicates that the
r below the detection limit.
Mn
.006
Tc
Re
8
Fe
.3/1
Ru
0s
6
Co
Rh
8r
10
Ni
Pd
25
Pt
4
Cu
.007
2
Ag
48
Au
10
Zn
6
Cd
32
Hg
b
9
B
2.8/1
50
A\*
Ga
55
In
91
TI
C
16
Si
363/9
Ge
120
Sn
84
Pb
N
150
P
.26/5
57
As
.36/7
32
Sh
o
49
BJ
O
s
84
S@
103
Te
Po
H
F
Cl
Br
,|.=
At
He
HQ
Ar
Kr
Xe
Ro

                                                    Eu
Gd
                                     64
t
                                                                                              LIT
 Average concentrations/average deviations are given in the  form  20.0/2  to be read 20.0+0.2.

-------
untreated samples with ascorbic acid, were  five- and  twenty-times higher,
respectively,  than  the values in the untreated samples without ascorbic acid.
It could be  that Fe and Mn were introduced by the ascorbic  acid.

     Another significant' difference between  the trace element, concentrations.
in the untreated and treated water samples was noted for aluminum and
arsenic.  The  untreated water had 0.08 ppm Al and 0.91 ppm  As, whereas the
treated water  gave  3.1 ppm Al and 0.36 ppm As.  The;  water treatment process
uses potassium aluminum sulfate to produce aluminum  hydroxide.  The higher
aluminum concentration in the treated water  is, therefore,  not unexpected.

     Another set of samples from Antofagasta had been received by mail from
the EPA Laboratories in Cincinnati during  October 1977.  The six samples
were provided  by Dr.  Jane L". .Valentine,;Assistant Professor of;Public -
Health, University  of California:, Los Angeles.  The  samples had been
collected sometime  before August 30, 1977.   The exact collection date was
not stated in  the transmittal letter.  The samples were received in
thin-walled  polyethylene containers (Cubitainers) which were closed by
polyethylene caps with paper liners.  All  samples, were colorless and
contained little if any suspended matter.  All samples were filtered through
0.45 um Millipore filters.  The filtrates  were stored in 250 mL acid-cleaned
polypropylene  bottles in a refrigerator after acidification with the acids
listed in Table 16.   The analyses of the samples began on.October 14, 1977.

     The results of total element analysis for As, Na, Fe,  Mn, Zn and Cu by
.atomic absorption spectrometry are summarized in Table 16.
 TABLE 16.  RESULTS OF TOTAL ELEMENT ANALYSIS FOR WATER SAMPLES FROM
            ANTOFAGASTA,. CHILE COLLECTED AND ANALYZED IN 1977

Sample
Untreated


»
Treated

2 mL Cone.
Acid Added
H2S04
HN03
none
HC104
HN03
H2S04
As*
ppm
0.52
0.89
0.91
0.91
0.088
0.081
Na**
ppm
54
106
106
106
107
107
Fe**
ppm
0.18
0.18
0.15
0.55
t
t
Mn**
ppm
J-
I
t
t
t
t
t
Zn**
ppm
0.042
0.10
0.047
0.058
0.10
0.10
Cu**
ppm
t
t
t
t
t
t
 *  Graphite furnace atomic absorption spectrometry.
 ** Flame atomic absorption spectrometry.
 t  Less than the detection limit  (Fe  0.075  ppm,  Mn 0.075  ppm,  Cu 0.10 ppm).
                                      79

-------
      The total arsenic concentrations in the untreated Antofagasta water
 samples ranged from 0.5 to 0.9 ppm as determined by flameless atomic
 absorption spectrometry, in good agreement with the values found in the
 samples collected in 1980.  The two samples which had passed through the
 water treatment plant had an arsenic concentration of 0.08 ppm in agreement
 with the values obtained in Antofagasta (70).

      Arsenite and-arsenate were determined only in the sample-which had not
 been acidified.  Only arsenite (0.12. ppm) and arsenate (0.80 ppm) were
 detected~~with no evidence for organic arsenic compounds.   These arsenate and
 arsenite determinations were carried out by differential  pulse polarography
 employing pyrogallol as the reagent to make arsenate electroactive.  These
 values are considered less reliable than the ones obtained with the hydride
 generation technique (see Table 13).
 WATER SAMPLES FROM ALASKA

      Water samples from Alaska with arsenic concentrations in the ppm range
 were collected and shipped to College Station by air freight on
 October 17, 1979 and July 29, 1980 by the Alaska Department of Environmental
 Conservation in Fairbanks, Alaska.  The samples came from the Barefoot and
 Mauer sites.  The samples were collected in Pyrex. contaJLners_and plastic
~Cu6ftainers.  Ascorb~ic "acid or "nitric aci"d7wer'e~"ad'de
-------
     TABLE 17.  RESULTS OF TOTAL ARSENIC ANALYSES AND ARSENITE AND ARSENATE DETERMINATIONS IN ALASKA
                WATER SAMPLES **               \
oo
•o
3 , . « .
•^j ' ^5
o .y • • H
.^ . »-*
.•' .'J3w.il
^ i M 4J Tj S -
Sample to to 8 *s
fO.fi O H i-H
O A4 
-------
     TABLE 18.   TRACE ELEMENT CONCENTRATIONS DETERMINED BY ICP IN THE BAREFOOT WATER SAMPLES FROM Af

                (COLLECTED IN 1979)
H
°ff
5.5
Ma
47/2
42
K
Rb
Cs
Fi-

le6
34
Ma
L13A
45
Ca
291/8
0.44
Sr
.47/3
0.58
Ba
.23/1
Ra
Concentration of the — •
Element in ppm Found
in the Sample*
*The absence of a yyyy
concentration is at 01
8
Sc
2
Y
La
Ac
16
Ts
flOS/2
4
Zr
HI
8
¥
.15/1
Nb
Ta
8
Cr
.033/y
0
mo
18
w-
E;
•yyyy


figure indicates that the
r below the detection limit.
Wn
.44/4
Tc
Re
8
Fe
53/f
Ru
Os
6
Co
Rh
IT
10
Ni
Pd
25
Pt
4
Cu
.010? f
2
Ag
48
Au
10
fa
6
Cd
.010/f
32
Hg

9
B
50
Al
<0.06
Ga
55
In
91
• Tl
C
16
Si
11/1
Ge
120
Sn
84
Pb
N
150
P
57
2^3
32
Sb
49
Bi
O
S
84
Se
.103
Te
Po
H
F
CI
Br
1
At
He
Ne
Ar
Kr
Xe
Rn

oo
N>
                                                       Ey
Gd
                                        64
     ^Average concentrations/average deviations are given in the form 20.0/2 to be read as 20.010.2.

-------
       TABLE  19.
TRACE ELEMENT CONCENTRATIONS DETERMINED BY ICP IN THE BAREFOOT WATER SAMPLES FROM ALASKA1

(COLLECTED IN 1980)
H
°tf
5.5
Na
37/1
42
12/1
Rb
Cs
Fr

0.46
Be
34
Mq
52/1
45
180/9
0.44
Sr
ND
0.58
Ba
.17/2
Ra
' , • '- • E •
Concentration of the-— -yyyy

Element in ppm Found .
iii the Sample* :
*The absence of a yyyy figure indicates that the
concentration is at or below the detection limit.
8
Sc
2
Y
La
Ac
16
Ti
4
Zr
Hf
8
V
Nb
Ta
8
Cr
8 '
Mo
18
w
Mn
.40/1
Tc
Re
SR
Ru
Os
6
Co
Rh
Ir
10
Ni
Pd
25
Pt
4
Cu
2
Ag
48
Au
10
Zn
6
Cd
32
Hg
9
B
50
Ai
.50/3
Ga
55
In
91
Tl
C
16
Si
ND
Ge
120
Sn
84
Pb
N
150
P
.27/9
57
As
3.4/1
32
Sb
49
Bi
O
s
28/1
84
Se
103
Te
Po
H
F
Cl
Br
1
At
He
Ne
Ar
Kr
Xe
Rn

00
OJ :
Ce
Pr
Nd
Pm
Sm
Eu
Gd
Tb
Dy
Ho
Er
Tm
Yb
Lu
Tn
Pa
64
u
Np
Pu
Am
Cm
Bk
Cf
Es
Fm
Md
No
Lr
       ^Average concentrations/average deviations are given in the form 20.0/2 to be read 20.0±0.2.

        ND = Not determinedi   •

-------
Zn, Cd, Al, Si, P and As.  Comparison of Table 18 (results for samples
collected in 197-9) with Table 19 (results for samples collected in 1980)
shows considerable differences between the concentrations of Na, Mg, Ca, Fe
and Al in the two samples.  Since some concentrations in the 1979 samples
are higher and others lower than in the 1980 samples a systematic instrument-
related error is unlikely.  Errors in preparing standards should not be so
frequent.  The possibility that the composition of the water supply changed
must be considered and should be checked on additional samples.

     Three of the Barefoot samples collected in 1980 (one preserved with
ascorbic acid, two preserved by acidification to 0.1 M HNC^) were analyzed
again by ICP eight/months after collection.  The analyses of samples
preserved with nitric acid agreed fairly well with the results obtained
shortly after the samples had been collected.  The sample preserved with
ascorbic acid gave considerable lower values'for As, Ca, P and Fe.

     The Mauer samples contain 5.1±0.7 ppm of arsenic when all the values in
Table 17 are averaged.  Arsenite is the predominant form.  The arsenite
concentration of one of these samples collected in a Cubitainer and
preserved by ascorbic acid (bottom line in Table 17) found by the hydride
generation technique was much lower than for the other samples.  Repetition
of. these analyses produced the same value.

     Trace element analyses by ICP (Tables,20 and 21) showed that Na, K, Mg,
Ca, Sr, Ba, Ti, V, Cr, Mn,, Fe, Zn, Al, Si and As were present.  Differences
similar to the ones pointed out for the Barefoot samples were noted between
the trace element concentrations of the Mauer samples collected and analyzed
one year apart.  Analyses of the eight months old nitric acid-preserved
samples gave results in general agreement with those reported in Table 21.
The arsenic concentration in the ascorbic acid-preserved samples was 0.7 ppm,
one-fourth of the value found in the samples acidified with nitric acid.

     Samples collected in 1979, which had been preserved with ascorbic acid
and had not been opened since the collection, had formed a black precipitate
after two to three weeks.  Similar samples in containers, which had
frequently been opened for removal of aliquots for analyses, did not form
this black precipitate.  More than 90 percent of the arsenic originally in
solution was present in the black precipitate.  It is possible that these
samples contained microorganisms which grew in the ascorbic acid medium
reducing sulfur compounds to sulfide.  Sulfide.then would react with iron to
produce black iron sulfides.  Arsenic sulfides would coprecipitate with the
iron compound.  Approximately 20 ppm sulfur were found in these water
samples by an ICP analysis.
WATER SAMPLES FROM TAIWAN

     Two sets of water samples from Taiwan were analyzed.  One set
collected by Dr. Tseng, Department of Medicine, National Taiwan University,
Taipei, arrived in College Station toward the end of May 1977.  These
samples came from wells at Pu-Tai and Pei-Men.  The second set was collected


                                     84

-------
    TABLE 20.  TRACE ELEMENT  CONCENTRATIONS  DETERMINED BY ICP IN THE MAUER WATER SAMPLES  FROM ALASKA

               (COLLECTED  IN  1979)
H
•°ff
5.5
Na
50/2
42
K
118/1!
Rb
Cs
FT

0.46
Be
34
Mg
13979
45
Ca
309/
0.44
Sr
.60/3
0.58
Ba
.35/2
Ra
Concentration of the
Element in ppm Found
in the Sample*
*The absence of a yyyy
concentration is at 61
8
Sc
5
2
Y
La
Ac
16
Ti
009/2
4
Zr
Hf
8
V
.17/2
Nb
Ta
8
Cr
.03
8
Mo
18 '
w
xxxx-
E
yyyy

	 Detection Limit in pp
figure indicates that the
r below the detection limit.
Mn
.76/1C
Tc
Re
8
Fe
.63/3
Ru
Os
6 -
Co
Rh
Ir
10
Ni
Pd
25 •
PI
4
Cu
2
Ag
48;
Ail
10
Zrs
.3/1
6
Cd
32
Hg
b
9
B
50
.#
Ga
55
In
91
Tl
C
16.
10376
Ge
120
Sn
84
Pb
N
150
P
57
As
4.5/1
32
Sb
49
Bi
O
S
84
Se
103
Te
Po
H
F
CI
Br
'J."
At
He
Ne
Ar
Kr
Xe
Rn

00
Ol
Ce
Pr
Nd
Pm
Sm
Eu
Gd
Tb
Dy
Ho
Er
Tm
Yb
Lu
Th
Pa
'64
u
Np
Pu
Am
Cm
Bk
Cf
Es
Fm
Md
No
Lr
        Average concentrations/average deviations are given in the form 20.0/2 to be read 20.0+0.2.

-------
           TABLE 21.   TRACE ELEMENT CONCENTRATIONS DETERMINED BY ICP IN THE MAUER WATER SAMPLES  FROM ALASKA1"

                      (COLLECTED IN 1980)
H
°ff
.007/1
5.5
Ma
35/1
42
K
Rb
Cs
Fr

Go
34
MQ
ssri.
45
Ca
200/9
0.44
Sr
0.58
Ba
29/1
Ra
Concentration of the -—
Element In ppm Found
in the Sample*
*The absence of a yyyy
concentration is at 01
8
Sc
2
Y
La
Ac
16
Ti
4
Zr
Hf
8
V
Nb
Ta
8
Cr
8
Mo
18
w
XXXX--
E
•yyyy

	 Detection Limit in pp
figure indicates that the
r below the detection limit.
Mn
.56
Tc
Re
8
Fe
25/1
Ru
Os
6
Co
Rh
Ir
10
Ni
Pel
25
PI
4
Cu
2
Ag
48
Au
10
Zn
6
Cd
32
Hg
b
9
B
50
Al
.54/1
Ga
55
In
91
Tl
C
16
Si
Ge
120
Sn
84
Pb
N
150
P
57
As
6.0/1
32
Sb
49
Bi
O
17.5/3
84
Be
103
Te
Po
H
F
C3
Br
I
At
He
Ne
Ar
Kr
Xe
Rn

oo
cr>
                                                            Gd
                                         64
                                         u
Cm
id
           'Average concentrations/average deviations are given in the form 20.0/2  to be read 20.0±0.2.

-------
by Dr. Fung-Jou Lu, Associate Professor of Biochemistry,  College of
Medicine, Taipei in the Yenshei-Provlnce  from  two wells and  shipped by air
mail to College Station on July 9, 1980.

Pei-Men and Pu-Tai Samples                                         .'.'•:

     The samples from Pei-Men and Pu-Tai  were  received toward  the end  of
May 1977 in Chlorox containers.  Later, a second set  of Taiwan samples was
received in one gallon Cubitainers with the date of collection unknown.   The
samples had a yellowish hue with some flocculated matter  present.   Analysis
of the Taiwan waters was initiated on June 24, 1977.  All analyses were
performed on the original, unfiltered, unpreserved water  as  received except
that in some cases dilution with water from a  sub-boiling still, or with
0.1 £ Ultrex HNC>3 was necessary to bring  the analyte-  concentration into the
linear region of the analytical working curve.  Data  relating  to the
speciation of arsenic, total arsenic, sodium,  copper, manganese,  zinc,  and
iron concentrations are given in Table 22.; Tseng £t_  al.  (71)  reported that
the total arsenic concentrations for drinking  water .taken from artesian
wells in 37 villages located in the southwest ••-'.of. Taiwan fell within the range
0.017 to 1.097 ppm.  The total arsenic concentrations,, determined by
flameless atomic.absorption, spectrometry  and-, neutron  activatipn analysis  for
the Pei-Men and Pu-Tai samples, fell within this range.   Only  arsenite

 TABLE 22.  ANALYSIS RESULTS FOR THE PEI-MEN AND PU-TAI WATER  SAMPLES
         Element                            Pei-Men              Pu-Tai
 Arsenite,  ppm As   (Hydride  technique)       0.05                 0.09
 Arsenate,  ppm As   (Hydride  technique)       0.52                 0.63

 Arsenite and Arsenate               . .        0.57                 0.72
 total As, ppm   (GFAA)                        0.72                 0.76
 total As, ppm   (Neutron activation)          0.76            ...,.

 Sodium, ppm     (Flame AAS)                 282                  223
 Copper, ppm     (Flame AAS)
 Manganese, ppm  (Flame AAS)
 Zinc, ppm       (Flame AAS)
 Iron, ppm       (Flame AAS)
(^0.07 ppm) and arsenate (^0.57 ppm) could be detected.  Since some of the
well waters are known to be anaerobic, one can expect substantial chemical
changes to have occurred in the samples since, the time of collection.  The
sum of the arsenite and arsenate concentrations is somewhat less than the
total arsenic level found by flameless AAS.  The sodium concentration is

            •   .  '    .                87            •.-••.•••.-,...•

-------
rather high  (200-300 ppm).

Trace Elements in the Yenshei-Province Samples

     The water samples from two wells  (Well I and II) in the Yenshei-Province
of Taiwan were collected  in plastic Cubitainers.  Two samples  each  of
unpreserved,water, preserved by addition of 0.1 weight percent of ascorbic
acid, and preserved by acidification to 0.1 M HNOs arrived by air mail  at
College Station on July 13, 1980 and were analyzed during the  two week
period following arrival.

     The total arsenic concentrations and the concentrations of arsenite  and
arsenate in  these samples are given in Table 23.  The total arsenic concen-
trations determined by the hydride generation technique are in the  range  of
0.84 ppm to  1.1 ppm.  Arsenate is the predominant arsenic compound  in  these
samples.  Unpreserved samples and samples preserved by addition of  ascorbic
acid had the same arsenite concentrations indicating that the  chemical forms
of arsenic in the samples did not change during the time period between
collection and' analysis.

     The trace element concentrations' in these water samples were determined
by ICP shortly after: arrival.  The results,'however, were unreliable.   The
arsenic concentrations in most of,the samples were found to be <60  ppb by
ICP..  When our-;own ICP became operational the water samples were analyzed
again.  The  arsenic concentrations in the .eight month old samples were in
fair agreement with those found by hydride generation in the freshly
collected, samples. 'The Yenshei:. water samples contained Li, Na, K,  Mg,. Ca,
Sr,. Fe, B, Si, P, As and  S.  The results of these analyses are summarized in
Tables 24 and 25 -for-Well, I and Well II, respectively.  The chemical form of
the sulfur in these water samples (0.5 and 1.5 ppm) was not determined.
Sulfate was  found in. several well water samples from the'blackfdot  disease
area in Taiwan (7la).  Additional experiments are needed to determine  the
chemical form of .sulfur."  'A-special search should be. made for  sulfite •,
which caused decreased DNA synthesis in human lymphocytes and  is mutagenic
in microbial test systems (71b) -.•'••   '

Fluorescent  Compounds in  the Yenshei Water Samples

     Blackfoot disease is an affliction peculiar to the south-west  coast  of
Taiwan.  The disease is caused by blockage of the peripheral vascular  system
and  extreme  cases are identified by ulceration and loss of the limbs (72, 73).
Fluorescent  pollutants in artesian well waters have been cited as possible
causes of this disease.   As a result, several studies have been undertaken
to identify  and characterize these fluorescent compounds.  Lu  et_ al_. (73)
have found that the water produces an intense green fluorescence under
.ultraviolet  light.  Chloroform extracts of the water were separated into
nine components by thin-layer chromatography.  The pH of these waters
gradually increased  to 9.0 upon standing in air at room temperature or upon
boiling.  Attempts were made to isolate and identify the fluorescent
compounds in the well waters  (74).  The compounds were adsorbed on  an  anion
exchange resin and eluted with a 5% ammonium chloride solution.  The eluate


                                      88

-------
   STABLE 23.  RESULTS  OF TOTAL ARSENIC ANALYSES AND ARSENITE AND ARSENATE DETERMINATIONS IN THE YENSHEI
   '"=           WATER  SAMPLES*                   '          ,
00
\o

K '3%
Sample "2 °
o-o
•••' : 33
*^C ^1
^Well I -
+
""•
Well II
+

*Average + average
XI •";'••:'".''.' '•.-'•.'
o° :> TOTAL ARSENIC, ppm
w „ Arsenite, ppm Ar senate, ppm
S| 2 .I,' HG ; ICP ;, by HG by HG
i-H • Cu • • • ;.••---•
. H • ••••'•-.; ,V
+ - 0.84+0.02
0.88+0.01
+ ; 0.85+0.01
+ - 1.0510.08:
- 1.16+0.03;
+ 1.1110.02 ;
6.73+0.01 - -
0.72+0.01 0.024+0.001 0.85+0.01
0.73+0.01 0.022+0.001 0.8310.08
0.95+0.01 -
0.96+0.01 - -
Q.98±0.01 0.02410.01 1.08+0.02
deviation of .at least three determinations per sample :
' • . . .' '••.'•;.. '•':'•.: •':-. . f ' ' '.

-------
TABLE 24.  TRACE ELEMENT CONCENTRATIONS IN THE YENSHEI WATER SAMPLES,  WELL I, DETERMINED BY ICP
H
V-3
• La
0.007
5.5
Ma
280
42
K
1.0
Rb
Cs
Fr

0.46
Be
34
Mg
8.6
45
Ca
4.3
0.44
Sr
0.08
0.58
Sa
Ra
Concentration of the — •
Element in ppm Found
in the Sample*
*The absence of a yyyy
concentration is at 01
8
Sc
2
Y
La
Ac
16
Ti
-. 4
Zr
Hf
8
V
Nb
Ta
8
Cr
8
Mo
18
w
E
•yyyy


figure indicates that the
r below the detection limit.
Mn
Tc
Re
8
Fe
0.46
Ru
Os
6
Co
Rh
Ir
10
Ni
Pd
25
PI
4
Cu
2
Ag
48
Ail
10
Zn
6
Cd
Hg

9
B
50
A!
Ga
55
In
91
Tl
C
16
Si
3.4
Ge
120
Sn
84
Pb
H
150
P
4.9
57
As
0.73
32
Sb
49
Bi
O
S
0.5
. 84
Se
.103
Te
Po
H
F
Cl
Br
1
At
He
Ne
Ar
Kr
Xe
Rn

                                    Srrs
                   Eu
         Gd
                 Th
Pa
                           64
Pu
Cm
Cf

-------
TABLE 25.  TRACE ELEMENT CONCENTRATIONS IN THE YENSHEI WATER SAMPLES,  WELL II,  DETERMINED BY ICP
H
0 4.3
Ll
0.01
5.5
Na
196
42
K
13.6
Rb
Cs
Fr

0.46
Be
34
Mg
23
45
Ca
17.5
0.44
Sr
0.20
0.58
Ba
Ra
Concentration of the--
Element. in ppm Found
in the Sample*
*The absence of a yyyy
concentration Is at o.i
8
Sc
2
Y
La
Ac
16
Ti
4
Zr
HI
8
V
Nb
Ta
8
Cr
;8
Mo
18
m
xxxx-
E
yyyy

	 Detection Limit in pp
figure indicates that the
rNbelow the detection limit.
Mn
Tc
Re
8
Fe
0.50
Ru
OS
6
Co
Rh
Ir
10
Ni
Pd
25
Pt
4
CIS
2
Ag
48
Au
10
Zn
6
Cd
32
Hg
b
9
B
0.87
50
Al
Ga
55
In
91
Tl
C
16
Si
3.4
Ge
120
Sn
84
Pb
N
150
P
2.3
57
As
0.97
32
Sb
49
Bi
0
S
1.5
84
Se
103
Te
Po
H
•f1
cs
Br
I
At
He
Ne
Ar
Kr
Xe
Rn

Ce
Pr
Nd
Pm
Sm
Eu
Gd
Tb
Dy
Ho
Er
Tm
Yb
Lu
Th
Pa
64
u
Np
Pu
Am
Cm
Bk
Cf
Es
Fm
Md
No
Lr

-------
 was evaporated to dryness on a rotary evaporator at 50°.  The residue was
 extracted with methanol and acetone.  The extract produced six blue-green
_fluqrescent bands on aJTLC plate^	One: of^ the compounds was tentatively
Tdeh'tified" as either"TysergTc~acid7 dihydr'olyserg'ic~'a'cid~'o'r'^' related'
 compound.  Another study (75) has identified ergocalciferol as~i~contaminant
 in the drinking water. .In addition to the chemical studies, a number of
 publications discuss the toxic (76, 77) and cytotoxic (78, 79) 'effects of
 the fluorescent compounds in the drinking water.
                   f  • '      i                .
      Since ergot alkaloids were suggested to cause the fluorescence in the
 artesian drinking water, ion-pair, reversed phase, high performance liquid
 chromatography was used in an attempt to identify the compounds.  This
 method has been recently suggested for ergot alkaloids.

 Experimental      •
  •   .      "•.•/.         ~                     •
      Liquid chromafeography was performed on an Altex/Beckman Model 312 MP
 High Performance Liquid Chromatograph.  A Farrand Model 801 Spectrofluoro-
 meter equipped with a 10 yL quartz flow cell was used for fluorescence
 detection;  For preparative liquid chromatography a 15 x 250 mm glass
 column was used.  A Fisher Model 80 circulating water :bath was employed to
 provide elevated'temperature chromatography.  Samples were concentrated on a
 Brinkman Rotavapor -rotary evaporator.

      Amberlite IRA-400C strong anion exchange resins, Partisil 10-SAX
 (strong anion exchange column), a Partisil 10-SCX (cation exchange column),
 reversed-phase columns (Altex), a 25 cm Spherisorb-C18 and a 15 cm
 Ultrasphere-ODS column were used for chromatography.  Ergometrine, ergotamine
 tartrate, ergocryptine, ergocalciferol and I)-lysergic acid were obtained
 from Sigma Chemical Co.
    ''       •'.'''   -t-     "                  •      "
 Results          .:-. *r               ..•-..••        ....  ,       '•;
 	     ' '      ' *
      First, the sample preparation procedure of Lu  (74) was•evaluated.  The
 procedure requires the adsorption of the fluorescent compounds on an
 Amberlite anion exchange resin followed by extraction with methanol and then
 with acetone.  One liter of unpreserved Yenshei Well I water, collected in a
 Pyrex bottle, was pumped through a preparative LG•column packed with
 Amberlite IRA-400C using a solvent metering HPLC pump with a 2 y fritted
 inlet filter.  The Farrand fluorometer was used to monitor the column
 effluent for the appearance of fluorescent components.  After about eight
 hours of pumping attempts were made-unsuccessfully toelute the       ,
 fluorescent components with aqueous 5% NH^Cl, aqueous 10% NHi+Cl and aqueous
 10% NHijCl/NHitOH at pH 9.  The dark brown band that had formed at the top of
 the column did not move.  Ethanol, 1 M with respect to HC1, eluted most of
 the adsorbed components in the brown band as well as highly fluorescent
 resin decomposition^products.  The eluate was evaporated to dryness.  Dark
 brown crystals were noted in the residue.  During the extraction of the
 residue with MeOH tlje crystals dispersed but did not dissolve.  Attempts at
 dissolving the brown crystals in water, acetone and cyclohexane failed.
                                       92

-------
      A much  simpler preparation  procedure similar to that of Lu et al.  (75),
 in which the filtered  water is concentrated under vacuum at 50°,  was  adopted.
 The water was filtered through a cellulose filter disk with an average  pore
 size  of  0.5  u.   The filtrate was concentrated  under vacuum at approximately
 50° in a rotary evaporator.  Typically 200 mL  of well water were concen-
 trated to 10.0 mL.

 Chromatographic Analyses  of  Well Water Concentrates

      High performance  anion  exchange  chromatography employing a Partisil
 10-SAX column and aqueous 0.15 tl NH^Cl at pH 3 or 0.1 M aqueous trisodium
 citrate  at pH 6 as  the mobile phases  did  not separate the fluorescent
 components.   Washing the  column  with  water and then methanol revealed that
 most  of  the  injected sample  had  been  tightly adsorbed to the column packing.

      High performance  cation exchange liquid chromatography using a Whatman
 Partisil 10-SCX column gave  better  results.  Using gradient elution
 (0.001 M -»•  .05 M sodium citrate,  at  pH 6)  two highly absorbing and fluorescent
 components and an additional, nonfluorescent UV absorbing component were
 separated (Fig. 31A).   Changing  the ionic strength and/or the pH of the
 -mobile .phases ;did not  improve the, separation.   .The frequent appearance  of
 a  shoulder on the strongest  fluorescence  peak  suggested the presence  of an
 additional compound..  Fractions  up  to 2.6 mi elution volume (Fig.. 31B)  and
 fractions with an elution volume >2.6 mL  (Fig.  31C) were collected from
 four  injections,  concentrated and rechromatographed.  A third fluorescent
 component was observed at an elution  volume of 2.4 mL (Fig. 31C).

      Next the- feasibility of a reversed phase  separation of the fluorescent
 compounds using ion-pair  formation  was investigated.  Several experiments
 with  hexanesulfonic acid  and heptanesulfonic acid at pH 3 were conducted.
 Heptanesulfbnic acid gave the best  results.  Figures 32 and 33 show
 chromatograms of the concentrated water samples and five alkaloid standards
 employing a  UV-detector at 254 nm and a- fluorescence detector, respectively.
 D-Lysergic Acid and ergometrine  were  not  resolved under these conditions.
 Calciferol,  a rather- insoluble compound,  does  not: fluoresce and,  therefore,
 does  not produce a signal when a,fluorescence  detector is employed (Fig.  33).
 The chromatograms contain also peaks,arising" from imparities in the aqueous
 solvent  which fluoresce and  are  UV-detectable.   Comparing the chromatograms
 of the alkaloid standards with the  chromatograms of the water concentrates
 D-lysergic acid and/or ergometrine  and ergotamine could be present in the
 water samples.   The presence of  calciferol cannot, be ruled out.   Ergotamine
 and a solvent impurity have  the  same  elution volume with the impurity
 contributing approximately one third  to the total area of the peak.   The
 chromatograms of the water samples  contain several peaks, which do not
 correspond to any of the  five alkaloid standards.

      The water samples,, which had'been concentrated to one twentieth  of
.their original volume, deposited a  white  precipitate, after approximately
 36 hours. Such a sample  was stored and analyzed after eight weeks to see
 whether  any  changes had taken place.   The absorption and fluorescence
 chromatograms of the aged water  sample are shown in Fig. 34.   The
                                      93

-------
                                           uv
Chromatogram of  Fraction
with Elution Volume <2;6 ml
Fl uorescence Detector
.Column: Partisil  10-SCX  4.6 mm x 250 mm
 Mobile Phase: 0.001  M Trisodium Citrate
        at pH 6 (Solvent A)
              0.05 M  Trisodium Citrate
        at pH 6 (Solvent B)
 Sample Size: 100 Microliters
 Flow  Rate: 1 mL/min
 Elution Profile:  Linear  Gradient from 0%
      .  to 5% Solvent B  from 2.5 to 3.0 min.
                                                            UV at 254 nm
                                                            Fluorescence excitation 358 nm
                                                                         emission   438 nm
                                                                     Chromatogram of Fraction
                                                                     with Elution Volume >2.6 mL
                                                                     Fluorescence  Detector
                             I	I
 0        2             5
          Elution  Volume, mL
                                          2            5
                                          Elution Volume,  mL
Figure  31.   Cation exchange  liquid  chromatogram of  the concentrated Yenshei
               water  sample.
                                             94

-------
            ALKALOID

            STANDARDS
            Columns:-Spher1sorb C-18 3.2 nut x 250 ma
                  connected to U1trasphere-ODS 4.6 ran x
                  150 mm
            Mobile Phase: 0.05 H Heptanesulfonlc Acid. 0.005 M
                  KH2P04 in Water at pH 3.0 (Solvent A)
                  Aceton1tr1le (Solvent B)
            Sample Size: 100 MicrolUers
            Flovt Rite: 1 nt/mln
            Elutlon Profile: Linear Gradient from 01 to
                  1001 Solvent B over 17 Minutes
                 UV-Oetector at 254 nm
                 0.64 Absorption Units Full Scale (AUFS)
                     O —


                    'II-
                               ELUTION
10  VOLUME
                                                     15
             Same Conditions as Above
                               y
                                 WATER
                                 CONCENTRATE
Figure  32.   Ion-pair reversed phase  chromatogram of  the concentrated Yenshei
             water sample  and alkaloid standards employing an UV-detector.
                                         95

-------
                  ALKALOID STANDARDS
                    Sane Conditions as Below
                    Sample Size:  100 Mcrollters Containing

                          0.035 mg  0-Lyserg1c Add
                          0.031 mg  Fr-gome trine
                          0.014 mg  Ergotamlne Tartrate
                          0.031 mg  a-Ergocrypt1ne
                          0.010 mg  Calciferol

                    Fluorescence  Detector Attenuation 0.03


1 t 1
' 1
5
i i

ELUTON •
1 1 1

10
1 1

VOLU« .
i i i

15
i i i


i i i
                    Columns: SpheHsorfc C-18 3.2 nui x 250 run connected to
                           Ultrasphere-OOS 4.6 ran x 150 m

                    Mobile Phase: 0.05 M heptanesulfonlc Acid. 0.005 H
                           KH,PO. in ffater at pH 3.0 (Solvent A)
                           AcitoAttrlle (Solvent 8)
                    Sample Size: 100 Micro]Hers

                    Flow Rate: 1 mL/m
-------
                      Sax Conditions as In Figure 33

                      Fluorescence Detector •
                        Attenuation. 0.01
                                                VOLIW. it
                      Sane Conditions as 1n-Figure 32
                      UV-Oetector-.
                                    ELUTIOJI  -10  VOLUK. rt,
20
Figure 34.   Ion-pair reversed phase  chromatograms of  the  aged,  concentrated
              Yenshei  water sample employing a UV-detector  at 254 nm and a
              fluorescence  detector.
                                            97

-------
chromatograms of the aged sample are markedly different in the elution
volume region from 5 to 7 mL and from 15 to 17 mL when compared to the
chromatograms of the fresh samples (Fig. 32, 33).

     Unconcentrated artesian well water was also analyzed by rapid scanning
f luorometry as described previously (81) .  The emission-excitation matrix
for this samples is shown in Figure 35.  The intensity values represent the
sum of four single frame scans electronically added to comprise a single
integrated spectrum.  The data indicate a blue-green fluorescence with the
emission spread over more than 100 mn from ^350 nm to 500 nm.  Rapid
scanning fluorometry could be used for routine and quick characterization of
well waters containing fluorescent compounds .
                                                                    ,.11753
  276
                            Scattered Light Spectrum
     Figure 35.  Rapid scanning fluorescence analysis of unconcentrated
     '  ""••'•       Yenshei well water.
     Chromatography and analysis of the fluorescence of the Yenshei water
sample showed that several alkaloids and other unknown fluorescent compounds
were present.  A sufficient amount of sample was not available to separate
these compounds chromatographically and analyze them by mass spectrometry.
Additional work in this area is urgently needed.
WATER SAMPLES FROM NOVA SCOTIA

     The initial set of Nova Scotia samples was received in linear-
polyethylene bottles.  The second set was collected in one gallon Cubitainers
at  the same time (April 21, 1977).  Sample 1 was taken from a dug well
suspected of causing chronic arsenic poisoning and anemia.  It contained a
                                     98

-------
TABLE 26.  ANALYSIS RESULTS FOR THE NOVA SCOTIA WATER SAMPLES


Element
Arsenite, ppm As
6 weeks after collection*
at time of sample collection^"*"
24 hrs. after collection''"''
Ar senate, ppm As
6 weeks after collection*
at time of sample collection ' '
24 hrs. after collection ^
Total Inorganic Arsenic (Arsenite and Arsenate)
6 weeks after collection*
at time of sample collection''"''
24 hrs. after- collection"*^
Total Arsenic (6 weeks after collection)
AAS**
. . .';. '---• ••..,.••• -.. NAAtt .. •
Sodium, ppm .
Copper, ppm
Manganese, .ppm
Zinc, ppm - '
Iron, ppm
Ammonia, ppm'"''
Oxygen, ppmft
pHtt. .... . . ....
Sample 1
John Hartlin

0.68
4.5 ,
. -

4.48
3.5
"

5.16
8.0


6.8
8.2
60
0.08
2.4
: «0.1 •'""-'
«0.1
15
<1.0
6.6
Sample 2
H. Sullivan

0.05
0.31
0.21

0.23
0.32
0.18

-0.28
0.63
0.39

0.52
1.1
50
0.21
<0.1
•0.9
«0.1
. -•• •
3.2
8.0

 * Determined by Hydride Generation.
   Neutron Activation Analysis.
** Graphite Furnace Atomic Absorption Spectrometry.
   Data provided by C. E. Tupper, Nova Scotia Department of Public Health.
                                     99

-------
substantial quantity of large particulate matter.  Sample 2 was collected
from a drilled well with no known harmful physiological effects.

     The results of the arsenic speciation studies performed during June 1977
and total arsenic, sodium, copper, manganese, zinc, iron, ammonia, dissolved
oxygen, and pH determinations are shown, in Table 26.  The total arsenic
concentration in sample 1 is approximately 6 ppm.  The arsenite and arsenate
analyses carried out in the field by the.Nova Scotia Department of Public
Health indicated comparable concentrations for arsenite (4.5 ppm) and
arsenate (3.5 ppm> in.sample 1.  The dissolved oxygen content, for sample 1
was very low at less than 1 ppm'.  Six weeks, after the collection the sample
contained 0.68 ppm arsenite and 4.48 ppm arsenate.  Since a considerable
period of time had elapsed between sampling and analysis, aerial oxidation
of the arsenite to arsenate could have occurred.  The sum of the arsenite
and arsenate levels is less than the total arsenic concentration determined
by GFAA.  The differences for samples 1 and 2 are 1.6 and 0.24 ppm,
respectively.  It is not certain whether these differences reflect changes
in the total arsenic content as a function of time (decreasing with
increasing time),, are indicative of undetermined arsenic species or were
caused by interferences ..specific for GFAA.  The dissolved arsenic concen-
tration in sample 2 dropped sharply after 24 hours from 0.63 ppm to 0.39 ppm
with the remainder present as particulate arsenic (0.17 ppm).

     An unusually high manganese content (2.4 ppm) was found for sample 1.
Copper in the range 0.1 to 0.2 was also detected in both Nova Scotia
samples.  Zinc was found in sample 2 at the 0.1 ppm level.  The sodium
levels were quite moderate at ^50 ppm.
DISCUSSION

     The water samples analyzed as part of this project had total arsenic
concentrations in the range 18 ppb to 8 ppm.  The arsenite/arsenate ratios
are in the range 0.007 to 3.4 (Table 27).  No indications of the presence of
methylated arsenic compounds,, which are reducible to methylarsine or
dimethylarsine, have been found.  Experiments with the high pressure liquid
chromatograph/graphite furnace atomic absorption spectrometer system, which
would provide information about the presence of organic arsenic compounds
not reducible to methylarsines, detected only arsenite and arsenate.
Comparison of total arsenic concentrations with the sum of the arsenite and
arsenate concentrations place an upper' limit on the concentrations of any
other arsenic compounds which might be present. .These, upper limits are in
most cases in the low ppb range.

     The other trace elements found in these water samples by ICP are listed
in Table 27.  There are no particularly offensive elements in these water
supplies with exception of beryllium (7.4 ppm) in the untreated Antofagasta
samples.

     The various physiological effects observed in populations exposed to
these arsenic-containing drinking water supplies (Table 28) could have been
caused by the presence of varying amounts of arsenite and arsenate

                                    100

-------
TABLE 27.  SUMMARY OF TOTAL ARSENIC, ARSENITE, ARSENATE AND TRACE ELEMENT CONCENTRATIONS IN DRINKING WATER  SAMPLES*


Total As
Arsenite
Arsenate
Arsenite/
Arsenate Ratio
Inckley
PQ
0.18
0.010
0.18

0.06
CO
4J
01
0
0.02
0.010
0.010

VI. 0
. 8
MH
0)
H
m
3.1
2.4
0.7

3.4
Mtt
01
1
.- '" »
• 4.5-6.0
0.35-4.6
0.1-4.3


ntofagasta
ritreated
< P
0.75
0.016
0.74

0.02
ntofagasta
reated
< H
0.41**
0.003
0.41
•
0.007
eiishei I
w ., '
0.85
0.023
6.84

'0.03
M
M
•H .
01
1
01 .

1.1
0.024
1.08
i
j 0.02
ova Sco'tia 1
2
8.0
4.5
3.5

1-3
ova Scotia 2||
a .
0.63
0.31
0.32

1.0
Al
B
Ba
Be
Ca
Cu
Fe
K
LI
Mg
Mn
Na
P
Pb
S
SI
Sr
Ti
V
Zn
. _ _
1.2
0.026
.
3.24
0.08
,
-
-
1.83

233
-
.
*
13.4
0.10
_
.
•
_
0.01
0.045
-
15.4
0.26
0.007
-
-
6.65
-
65
-
1.17
*
15.3
0.52
-
_
-
0.06-0.50
"
0.17
. -
180-291
- ; •
29-53
12
-
52-113
0.42
37-47
0.27
_
28
11.0
0.47
0.008
0.15
0.14
.0.06-0.54
•'
0.29
- •
200-309
• "• ..'-''••
25
13-118
: 0.007
55-139
0.56-0.76
35-50
• 'f-
.
17.5
11.0
0.60
0.009
0.17
0,30
0.08
2.8
0.008
7.4
20.0
0.003
0.11
1313
0.62
7.4
0.002
102
0.34
-
*
38.7
0.28
..
-
0.10.
3.1
2.8
0.008
-
20.3
0.007
0.30
13.2
0.64
7.5
0.006
103
0.26
-
*
36.3
0.29
- " ,:.
_
0.10
> ' :
,T
'- ' '.
• - ':'
• ^3

6.46
1.0
0.007
: !3.6 ;
•: ^'- ' "
280
4.9
'.
0.5 ,
3.4 •
•- 0;08
'-.-.' :-
•;
*' '/-••
_
0.87
"-
-
17.5
- '
0.50
13.6
0.01
23
* '-
196
2.3

1.5
3.4
0.20
-
- '
•
*
*
*
*
*
0.008
<0.1
*
*
*
2.4
60
* •
*
*
*
*
<0.1
* '
*
*
*
*
*
*
0.21
<0.1
*
*
*
<0.1
50
*
*
*
*
*
<0.9
*
*
 * Not determined.  ** As In treated water Is normally less than 100 ppb.  See text for discussion.
 t The concentrations are given In ppm.   ft  Results from two different samples collected one year apart.

-------
TABLE 28.  ARSENIC-CONTAINING WATER SUPPLIES AND THEIR PHYSIOLOGICAL
          . MANIFESTATIONS IN MAN
Sampling Location
Type of Water and Range
of the Total Arsenic
Concentration....
 Symptoms Observed in
   the Population
Taiwan (37 villages)
Chile (Antofagasta)
Bakersfield, California


Fallen, Nevada,
artesian well waters used
for 45 years; arsenic
leached from geologic
deposits;- 0.017-1.097 ppm;
median 0.5 ppm. .

drinking water supply
since 1958; 0.8 ppm before
water treatment,0.1
after-water treatment.
community drinking water
supply; 0.3-0.7 ppm.

drinking water, 0.1 ppm.
 melanois, keratosis,
 skin cancer; 15% pre-
 valence among: males
 over age 60, normal
 incidence 2-3%.

 melanosis, hyperkera-
 tosis; vascular
 manifestations:
 myocardialischemia,
 hemipleglia with
 occlusion of the
 carotid artery,
. mesenteric arterial
 thrbmbos is, pneumonia.

 no adverse effects
 reported.*

 no known adverse
 physiological effects*
*In a study  (82) of the arsenic exposure of populations in Bakersfield,
 California, and Fallen, Nevada, by a questionnaire .designed to elicit
 information about arsenic related, symptoms and diseases very few symptoms
 were reported.  The incidence of  these symptoms was not significantly
 different from control populations not: exposed to arsenic.


(Table 27).  It is also conceivable that one or more of the trace elements
present in the water supplies act  in concert with arsenic to cause the
observed effects.  Many more samples need to be analyzed and the results
of these analyses correlated with  epidemiological studies before a definite
statement can be made about the interactions of trace elements with arsenite
or arsenate.

     The chromatographic work on the fluorescent compounds in the Taiwan well
waters strongly suggests the presence of alkaloids such as D_-lysergic acid,
ergometrine and calciferol.  Additional experiments (preparative
chromatography, mass spectrometry) could not be carried out because of
insufficient sample.
                                     102

-------
                                  REFERENCES


 1.  National Academy of Sciences,. Committee on Medical and Biological Effects
     of Environmental Pollutants.  Arsenic.  Washington, D.C., 1977.

 2.  Doak,  G.  0.,  and L. D. Freedman.  Organometallic  Compounds of Arsenic,
     Antimony and Bismuth.  Wiley Interscience, New York,  1970.

 3.  Reference 1,  pp. 11, 80; and references cited there.

 4.  Challenger,  F.,  ACS Symp. Series, 812.  (1978) 1.

 5.  Cox,  D.  P.,  ACS  Symp. Series, 7  (1975) 81.

 6.  Braman,  R.  S., ACS Symp. Series, _7 (1975) 108.
..'• .    •    -.   "     ••;..-   •  • •  •  •  •••-'','••• '.>•'  •••• ••/."•
 7.  Woolson,  E.  A.,  Environ. Health Perspectives, i9.  (1977)  73.

 8.  Irgolic,  K.  J..,  E. A. Woolson, R. A. Stockton, R.  D.  Newman,
     N.  R.  Bottino, R. A. Zingarp, P. C. Kearney, R. A. Pyles,  S.  Maeda,
     W.  J.  McShane, and E* R. Cox, Environ. Health Perspectives, j.9_ (1977)
     61, and  references cited there.

 9.  Bottino,  N.  R.,  E. R. Cox, K. J. Irgolic, S. Maeda, W. J.  McShane,
     R.  A.  Stockton,  and R. A. Zingaro, ACS Symp. Series,  J52_  (1978) 116.

 10.  Edmonds,  J.  S.,  K. A. Francesconi, J. R.  Cannon,  C. L. Raston,
     B.  W.  Skeltpn, and A. H. White, Tetrahedron Lett.,  (1977)  1543.

 11.  Cooney,  R.  V., R. 0. Mumma, and A. A. Benson, Proc. Natl.  Acad. Sci. USA,
     _75  (1978) 4262.

 12.  Francesconi.  K.  A.. Nature. 289 (1981) 602.

 13.  Reference 1,  p.  33.

 14.  Reference 1,  pp. 4-15.

 15.  Zief,  M., and J. W. Mitchell.  Contamination Control  in  Trace Element
     Analysis.  John  Wiley & Sons, New York, 1976, pp.  27-33.

 16.  Robertson,  D.  E., Ana1. Chem. Acta.42 (1968) 533.

 17.  Ryden, J. C.,  J. K. Syers* and R. F. Harris, Analyst, 97 (1972) 903.
                                       103

-------
 18.   Guimont,  J. ,  M.  Bichette, and N. Rheaume, At. Absorption  Newsletter,
      16 (1977) 53.

 19.   Ray,  B.  J. ,  and D. L. Johnson, Anal. Chim. Acta,. 62. (1972)  196.

 20.   Talmi,  Y.,  and D. T. Bostick, J. Chromatogr. Sci.,  13  (19?5)  231.

 21.   Whitnack, G. ~C., and R. G. Brophy, Anal. Chim. Acta, 48  (1969)  123.

 22.   Al-Sibaai,  A. -A., and A. G. Fogg, Analyst. 9£ (1973) 732.

 23.   Portmann, J.  E. , and J. P. Rlley, Anal.; Chim. Acta,. 31 (1964)  509.

 24.   Wagner,  R. ,  Z. Anal. Chem. . 282  (1976)  315.

 25.   Reference 15, p.. 27.

 26.   Harrison, S.  H. ; P. D. LaFleur, and W.  H. Zoller, Anal. Chem. ,  4_7
      (1975)  1685.

 27.   Andreae,  M.  0. ,  Anal. Chem., 49  (1977)  820.             .

 28.;  Andreae,  M.  0. ,  Deep-Sea Res. , 25_ (1978) 391.

 29.   Andreae,  M.  0.,  Limnol. Oceanogr., ^4  (1979) 440.

 30.   Owens,  J. W. , and E. S. Gladney, At. Absorption .Newsletter,  15_ (1976) 47.

 31.   Feldman,  C. ,  Anal. Chem., 51 (1979) 664.

 32.   Taqui Khan,  M. M. , and A. E. Kartell,  J. Am. Chem.  Soc. ,  89_ (1967)  4176.

 33.   Reference 1,  p .  255 .     .

 34.   Talmi,  Y. ,  and C. Feldman, ACS Symp. Series, 7  (1975)  13.

 35.   Peoples,  S.  A. ,  J> Lakso, and T. Lais,  Proc. West.  Pharmacol.  Soc. , L4
v .'. '  . (1971)       "    '        ' •  ''''                              -''
 36.   Johnson, D. L. , and M. E. Q. Pilson, Anal.  Chim. Acta, _58_ (1972) 289.

 37.   Myers, D. J., and J. Osteryoung, Anal.  Chem.,  45  (1973)  267.

 38.   Elton, R. , and W. E. Geiger, Jr., Anal. Lett., £  (1976)  665.

 39.   Bess, R. C., K. J. Ir'golic, J. E. Flannery,  and T.  H.  Ridgway,
      Anal. Lett., J9 (1976) 1091.

 40.   Bess, R. C. , K. J. Irgolic, J. E. Flannery,  and T.  H.  Ridgway,
      Anal. Lett., 10 (1977) 415.
                                       104

-------
 41.   Braman, R. S., and  C.  C.  Foreback, Science.  182 (1973) 1247.

 42.   Braman, R. S., D. L.  Johnson,. C. C. Foreback,  J.  M. Ammons, and
      J. L. Bricker, Anal.  Chem., 49^  (1977) 621.

 43.   Talmi, Y., and D. T.  Bostick, Anal. Chem.. 41 (1975) 2145.

 44.   Edmonds,  J.  S.,  and K.  A.  Francesconi, Anal.  Chem., 48 (1976) 2019.

 45.   Crecelius, E. A., Anal. Chem., 5£  (1978)  826.

 46.   Stringer, C.  E., and M. Attrep, Jr., Anal. Chem., 5_1 (1979) 731.

 47.   Talmi, Y., and V. E.  Norvell, Anal. Chem., 47 (1975) 1510.

 48.   Cox, D. P.,  and  M.  Alexander, Bull. Environ.  Contain. Toxicol., 9. (1973)
••....,.  84.,       '.:.-,•   .        '   .     . ...              ;,., ...  .

 49.   Salmi, E- J., K.. Merivuori, and E. Laaksonen,  Suomen Kemistilehti,  19B
      (1946) 102.      '••'••""•   '   •'• •• , /^"^  '  •"':"""-  •'-'  ; '    """';-  •"•'".'.'.,.-"

 50.   Soderquist,  C. J.,  D.  G.  Crosby, and J. B. Bowers, Anal. Chem.,  46
:     ,.(1974)155,.-  .   -......,.,,,.,,/;.   	:,.,'-.,.:.:-..       ...,.-•...  ,>.."     ':,-'.•••

 51.   Lodmell,  Ji  P..  Piss.  Abstr.  Int. B,  34  (1974) 5359.

 52.   Johnson,  L.  D.,,  K.  L.^ Gerhardt, and W. A.. Aiie, Sci. Total Environ.,  I
      (1972) 108.

 53.   Parris, G.. E.; W. R.  Blair, and F.. E. Brinckman,  Anal. Chem., 49_ (1977)
      378.   '  '""  " __..  •.    '.-'':'.'  •"'•'"' ;,/  .''.':' "•'  . '    :•     • • '        '  ' •

 54.   Fernandez, F. J.', At.  Absorption Newsletter,  16 (1977) 33.

 55.   Cantillo, A..  Y., and D. A.  Segar, Proc. Int.  Conf. Heavy Metals  in
      Environment,  Toronto,  Canada, 1976, p.

 56.   Brinckman, F. E., W.  R. Blair, K. L. Jewett,  and W. P. Iverson,
      J. Chromatogr. Sci. ,V 15 (1977) 493.     •  : ,.  •   '               '

 57.   Stockton, R.  A., and K. J.  Irgolic, Intern.  J. Environ. Anal. Chem.,  6^
 . . '   (1979) 313. ''••"... "  ' .    • •  , :;;;-   ."' ' :: " "   ': .   •  '   :'-   .;-  "."  -

 58.   Brinckman, F. E., K.  L. Jewett, W. P. Iverson, K. J. Irgolic,
      K. C. Ehrhardt,  and R. A.  Stockton, J. Chromatogr., 191  (1980) 31.

 59.   Jung,. W.  G.   1C  Timer Cookbook.  Howard W. Sams and Co., Inc.,
      4300 W. 62nd Street, Indianapolis, Indiana,  46268; pp. 244-247.
                                       105

-------
60.   Pierce, F. D., and H. R.. Brown, Anal. Chem., 49  (1977) 1417.

61.   Woolson, E. A., and N. Aharonson, J. Assoc. Off. Anal. Chem.,  6i3_ (1980)
      523.

62.   Ricci, G. R., L. S. Shepard, G. Colovos, and N. E. Hester, Anal.  Chem.,
      53. (1981) 610.

63.   Arnold, J. P., and R. M. Johnson, Talanta,. 16  (1969) 1191.

64.   Meites, L., J. Am. Chem. Soc.. 7$. (1954) 5927.

65.   White, M. C., and A. J. Bard, Anal. Chem., _38  (1966) 62.

66.   Princeton Applied Research, Application Brief A-117.

67.   Smith, M. R., K;. J. Irgolic, and R. A. Zingaro, Thermochim. Acta,  4_
      (1972) 1.:

68.   Irgolic, K. J., R. A. Zingaro, and L. J. Edmonsqn, Jr., Phosphorus,  5_
      (1975) 183. ;    " .  • .  :      .".''/..  .

69.   Borgono, J. M.., P. Vicent, H. Venturing, and A.  Infante,  Environ.
      Health Perspectives, 19_ (1977) 103.

70.   Communication from the  personnel at the Antofagasta water treatment
      plant.                •'.-'.

70a.  Song, T. J., and G.. S.  Logsdon, J. Am. Water Works Assoc., 70  (1978)
      379.

70b.  Groth, D. H., L. Stettler, and G. Mackay, in "Effects  and Dose-
      Response Relationships  of Toxic Metals," G. F. Nordberg,  Ed.,
    ~~ETsevie"r~ Scientific Publishing Company, Amsterdam, 1976,  p. 527.

70c.  Sunderman, F. W., Jr.,  Biol. Trace Element Research, 1 (1979)  63.

70d.  Casto, B. C., W. J. Pieczynski, R. L. Nelson,  and J. A. DiPaolo,
      Proc. Amer. Assoc. Cancer Res.. 1_7_ (1976) 12;  and Casto,  B. C.,
      J.Meyers, and J. A. DiPaolo. ibid, 18  (1977)  155.
71.  .Tseng, W. F.,~¥7~K'. Chu, S. H. How,~T."~M7~Fong ,  C.  S.  Lin,  and S".  Yeh,~
     :' J .
71a.  Shen, Y. S., and C. S. Chen, Adv. Water Poll. Res..  1  (1964)  173.

71b.  U.S. Environmental Protection Agency,  "Health Effects  of  Sulfur
      Oxides" External Review Draft,  1980.

72.   Chen, K. P., and H. Y. Wu, J. Formosan Med. Assoc..  61 (1962)  611.
                                      106

-------
 73.    Lu,  F.-J., C.-K. Yang,  and K.-H. Ling, J.  Formosan Med. Assoc.,  74
       (1975) 596.

 74.    Lu,  F.-J., C.-K. Yang,  and K.-H.. Ling, J.  Formosan Med. Assoc.,  76
       (1977) 209.

 75.    Luv F.-J., M.-H. Tsai,  and K.-H. Ling, J.  Formosan Med. Assoc..  77_
       (1976) 68.

 76.    Lu,  F.-J.,, M.-H- Tsai,  and K.-H. Ling, J.  Formosan Med. Assoc.,  76
       (1977) 58.

 77.    Lu,  F.-J., and K.-H.  Ling. J. Formosan Med. Assoc.. 79 (1979) 314.

 78.    Chen, J.-Y., F.-J. Lu,  and. K.-H Lin, Chinese  J.  of Microbiology, 10_
       (1977) 115.                •;••••.:   ,-••:•"••

 79.    Jam, I.-R., J~.-Y. Chen, F.^-J.Lu, and K.-H. Ling,  J.  Formosan Med.
 ••.-'•    Assoe., 77 (1978) 361.  >,. '•:,.r-.--  ••••••:...•:.:••  •...••••'.   ••-••••'•• ''•-'

•80.    O'sborne,. B.: G.,:;Altex CHfomatogram:,::3  (1980)  5. .      ;               :

 81.    Warner, I. M. ,.M. P..  Fogarty, and D.. C. Shelly,. Anal. Chim. Acta. 109
""  "• '-  -(1979) 361.  -  - .•.••:-:'"  '' "":''" ' '  -•:;>••"- •'-'••"•'•.-•••^•-. • "       •    :  •

 82.    Valentine, J. L., School of Public Health,, University of California,
       Los Angeles, private  communication.         :
                                       107

-------