Perchlorate Chemistry:
 Implications  for Analysis and  Remediation
 Edward T.  Urbansky
 U.S. Environmental  Protection Agency,  National Risk Management Research Laboratory, Water
 Supply and Water Resources Division, Treatment Technology Evaluation Branch, Cincinnati,  Ohio
 45268


      Abstract: Since the discovery of perchlorate in the ground and surface waters of several western states, there has
      been increasing interest in the health effects resulting from chronic exposure to low (parts per billion [ppb]) levels.
      With this concern has come a need to investigate technologies that might be used to remediate contaminated sites
      or to treat contaminated water to make it safe for drinking. Possible technologies include physical separation
      (precipitation, anion exchange, reverse osmosis, and electrodialysis), chemical and electrochemical reduction, and
      biological or biochemical reduction. A fairly unique combination of chemical and physical properties of perchlo-
      rate poses challenges to its analysis and reduction in the environment or in drinking water. The implications of
      these properties are discussed in terms of remediative or treatment strategies. Recent developments are also
      covered.


      Keywords: perchlorate, bioremediation, reductase, thyroid, anion exchange, electrochemical reduction, kinetic barrier, oxidant,
      reverse osmosis, electrodialysis, ion chromatography, capillary electrophoresis, analysis.
Introduction

Several factors have brought about the current interest
in perchlorate (ClOj), which, because of its chemical
and physical nature, presents challenges for analysis
and remediation. Perchlorate has been found in ground-
water and in surface waters in several western states,
including the Colorado River. Concentrations ranging
from 8 ng mL"1 to 3.7 mg mlr1 have been measured.
The extensive use of Colorado River water in this
region and the proximity of some of these sites to the
river have heightened the concern. Other local water
supplies also are affected. Perchlorate targets the thy-
roid, bone marrow, and muscle tissue at sufficiently
high concentrations; however, it is unknown what ef-
fects, if any, occur at the levels currently encountered
in the  contaminated  water sources. Although addi-
tional toxicological studies are  ongoing (Table 1), an
action level of 18 ng mlr1 has been adopted by Cali-
fornia and informally by other affected states.
    The fundamental physical and chemical nature of
perchlorate make it difficult to uniquely analyze for
and to remediate, especially at the low concentrations
typically encountered (i.e., <500 u.g mL"1). Although
ion chromatography is capable of determining very
low levels  (e.g., 5  ng mL"1), retention  time is  not
considered  a unique identifier, and known confirma-
tory tests have much higher detection limits. Perchlo-
rate ion is  unreactive as a ligand and its salts  are
extremely soluble, even in organic solvents. Despite
its strength as  an  oxidizing agent, perchlorate is
nonlabile, that is, very slow to react. This kinetic bar-
rier is well known and widely made use of in chemical
studies. Common reducing agents do not reduce per-
chlorate, and common cations do not precipitate it.
Consequently, standard practices of water treatment
1058-8337/98/$.50
© 1998 by Battelle Memorial Institute
Bioremediation Journal 2(2):81-95 (1998)
                                                                                                   81

-------
 Table 1.  Perchlorate facts at a glance.
 Date of discovery         Discovered in 1997 in Western U.S. ground and surface waters
 Physiological implications   Known to target the thyroid, bone marrow, and muscle at high concentrations

 Medical use              Once used medicinally to treat Graves' disease (hyperthyroidism)
 Remediation implications   Difficult to quantitate and remediate because of its high solubility and low kinetic lability towards
                        reduction
will neither remove it physically nor destroy it chemi-
cally.

Sources and Nature of the Contamination

Ammonium perchlorate (NH4C1O4) has been used as
an energetics booster in rocket fuels, and it appears
most perchlorate contamination is the result of dis-
charge from rocket fuel manufacturing plants or from
the demilitarization of weaponry (missiles). It is im-
portant to emphasize that ammonium perchlorate is
not rocket fuel; it is an additive. Potassium perchlorate
(KC1O4)  can be used as a solid oxidant for rocket
propulsion, and it was the original source for a fraction
of the contamination. However, most of the contami-
nation appears to have come from the legal discharge
decades ago of then unregulated waste effluents con-
taining high levels of ammonium perchlorate. Although
ammonium perchlorate was released initially, the salt
is highly soluble and dissociates completely to ammo-
nium and perchlorate ions upon dissolving in water
(equation 1):
 NH4C104(s)
(1)
It is likely that most of the ammonium has been biode-
graded and the cation is now best viewed as mostly
sodium (Na+) or possibly hydrogen (H+), especially
where levels are below  100 |ig mL"1; nevertheless,
those regions with high concentrations of perchlorate
ion probably retain at least some ammonium ion. At
those sites where contamination dates back decades,
very little (if any) ammonium ion has been found. To
date, there has  been no quantitative determination of
the cations responsible for the charge balance.
    Three states are known to be substantially af-
fected:  Utah, California,  and  Nevada. Arizona also
may be affected since it too  draws water from the
Colorado River. Perchlorate concentrations in Utah
range from 4 to 200 ng mL"1 in groundwater wells on
the property of rocket motor manufacturer Alliant
Techsystems. A level of 13 ng mL"1 at the Kennecott
Utah Copper mines in Magna, Utah has led the com-
pany to supply its miners with bottled water for drink-
ing.
    In Henderson, Nevada, water samples taken 1000
ft (300 m) from the site of the former Pacific Engineer-
ing & Production Company of Nevada (PEPCON)
rocket fuel plant, which exploded in 1988, contain as
much as 630 )J.g mL"1. Wells near the site show con-
centrations ranging from 51.4 to 630 |J.g mL'1. Samples
drawn from 50 wells near  ammonium perchlorate
manufacturer Kerr-McGee Chemical Corporation, lo-
cated about 1  mile  (1.6 km) from  the abandoned
PEPCON site,  also  showed significant perchlorate
contamination. The Kerr-McGee samples showed per-
chlorate levels as high as 3.7 mg mL"1 in the ground-
water. Surface  water samples taken in August 1997
from the Las Vegas Wash,  which feeds into Lake
Mead, had perchlorate  concentrations between 1.50
and 1.68 (J.g mL"1.
    Lake Mead is formed by the Hoover Dam on the
Colorado River and thus affects' the water supply of
southern California,  including Los Angeles. Testing
by the Los Angeles Metropolitan Water District found
8 ng mL"1 perchlorate at its intake anc$VLake Mead
at the Hoover Dam. Lake Mead lies at the southern tip
of Nevada, straddling the Arizona border.  This
prompted the  Southern Nevada Water Authority
(SNWA) to begin testing its water; the SNWA found
11 ng mL"1 perchlorate in tap water.
    The California Department of Health Services
(DHS) began testing wells  in  early 1997, and has
closed more than 20 wells for  exceeding the action
level of 18 ng mL"1. Some wells could not be closed
because of the high water demand of the regions served
by the utility companies. Because the utilities often
blend water from several sources,  it is possible  to
dilute water from some of these wells. The DHS cer-
tifies commercial testing laboratories to perform per-
chlorate determinations, and has therefore stopped its
own testing program (Cal DHS, 1997a). DHS reports
that some regions of Lake Mead showed levels up to
 165 ng mL"1.
 82
                                                                                             Urbansky

-------
Physiological and Health Effects

In 1992, the U.S. Environmental Protection Agency
(EPA) reviewed and assessed the health effects of
perchlorate administered chemotherapeutically to pa-
tients with hyperthyroidism (Dollarhide, 1992, 1995;
Stanbury, 1952). This study showed a no observable
adverse effects level (NOAEL) of 0.14 mg kg'1 day"1.
Doses of 6 mg kg"1 day"1 or more for periods of at least
2 months led to fatal bone marrow changes. The EPA
study recommended the following safety/error factors:
10 (nonchronic study), 10 (sensitive persons), 10 or 3
(database error margin) and allowed for two possible
uncertainty factors, 1000 and 300.
    Using the somewhat arbitrary, but relatively ac-
cepted, uncertainty factor of 300, the California DHS
established 1.8 ng mL"1 as the action level for initiating
remediation and  stopping  water  usage (Cal DHS,
1997b). This cut-off assumes a 70-kg person consum-
ing 0.5  |4.g perchlorate for each kilogram body mass
who drinks 2 L of water daily (18 ng mL"1 = 0.016 mg
mL"1 = 0.14 mg kg"1 day"1 x 70 kg x 1 day/2 L •*• 300).
The 0.5 [J.g  number introduces a rounding error that
was carried through (Cal DHS, 1997b). This 18 ng
mL"1 action level has been adopted informally by other
governmental agencies in the region as well. Using the
same assumptions, we would  calculate that harmful
thyroid  effects begin  to  occur at 49  jig mL"1, and
fatalities occur at 210 to 490 (ig mL"1. Meanwhile, the
European Communities (1982) set a maximum admis-
sible guide level of 20 jig NaClO4 mL"1 for drinking
water. This corresponds to 16 |0.g C1OJ mL"1.
    Perchlorate exerts its most commonly observed
physiological effects on or through the thyroid gland.
The primary effect is a decrease in thyroid hormone
output. The thyroid gland takes up iodide ion from the
bloodstream and converts it to organic iodide in the
form of hormones that regulate metabolism. The mecha-
nism responsible for this  process,  the cellular iodide
pump, preferentially selects for anions on the basis of
ionic volume: I"=SCN" < CIO;, TcO; (Chiovato et al.,
1997; Cooper, 1991;  Foye, 1989; Orgiassi,  1990).
Consequently, the presence of any large anion in the
serum reduces thyroid hormone production.
    This phenomenon was once  used pharmaceuti-
cally  to treat  hyperthyroidism, which is known as
Graves' disease (Foye, 1989;  Chiovato et al., 1997;
Cooper, 1991; Orgiassi, 1990). Chemotherapeutic use
of perchlorate was reduced substantially in the United
States after  several instances of aplastic anemia and
renal  damage  were observed (Foye, 1989;  Hobson,
1961). Domestic perchlorate use now is restricted al-
most  exclusively  to use as a diagnostic tool for the
evaluation of thyroid hormone production. As a diag-
nostic tool, perchlorate is still the standard for evalu-
ating thyroid activity; the protocol at the University of
California, Los Angeles (UCLA) requires a dose of 0.6
g (pediatric) or 1 g (adult) (UCLA, 1997). Following
the administration of radiolabeled iodide, perchlorate
is used to displace iodide anion in the iodide pump.
When thyroid function is low, most of the radioiodine
remains as inorganic iodide (rather than being con-
verted to an organic iodide) and is lost; therefore, very
little intrathyroid iodine  shows the radiolabel.
    Although perchlorate has been used as a treatment
for hyperthyroidism,  under the right circumstances it
also can act as goitrogen in rodents and prevent thyroid
hormone formation by interfering with iodide uptake
(Capen and Martin, 1989). The low level of hormones
is recognized by the pituitary gland which then stimu-
lates the thyroid gland to  work harder, eventually lead-
ing to goiter. A recent study of thyroid hormone levels
in the Sprague-Dawley rat supported the EPA refer-
ence dose  of 0.14 mg kg"1 day"1. Male rats exhibited a
thyroid NOAEL of 0.44 mg kg"1 day"1, but females
exhibited  a thyroid NOAEL of only 0.124 mg kg"1
day"1 (King, 1995). Potassium perchlorate has been
used to treat thyrotoxicosis without toxicity at doses
ranging from 40 to 120 mg day"1 (Cooper, 1996). If we
assume  a  daily intake of 3  L of water, this would
correspond to 13 to 40 (ig KC1O4 mL"1, or about 9 to
12 fig CIO; mL"1. This is a factor of about 1000 times
the California DHS action level, but close in line with
the European Communities level. It is unknown whether
secondary effects resulting from decreased thyroid
function, indirectly caused by perchlorate, will be con-
sequential. No studies link perchlorate to any second-
ary adverse health effects at this time.
    Perchlorate can directly affect organs and tissues
in addition to the thyroid gland. The mouse mammary
gland has  a mechanism  similar to the thyroid iodide
pump that is inhibited by perchlorate (Rillema and
Rowady, 1997); however, it is unclear whether this has
any significance for human health. At high (millimo-
lar) concentrations, perchlorate is known to potentiate
excitation-contraction (E-C) coupling and charge move-
ment in muscle cells (Bruton et al., 1995; Gonzalez
and Rios,  1993; Jong et al., 1997;  Khammari et al.,
1996; Ma et al., 1993; Pereon et al., 1996). At this
level, part of the E-C effect is due to activation of
calcium ion release from the sarcoplasmic reticulum
(Fruen et al., 1994; Percival et al., 1994; Yano et al.,
1995). In  fact,  this property of perchlorate  often is
exploited to study muscle physiology in animals. Much
of what is  known about perchlorate's effects on living
organisms is derived from studies of acute toxicity
over relatively short periods of time rather than chronic
exposure to very low concentrations over a lifetime.
Perchlorate Chemistry: Implications for Analysis and Remediation
                                              83

-------
    The U.S. Air Force Research Laboratories (AFRL)
are conducting animal toxicology studies under the
guidance of the EPA's National Center for Environ-
mental Assessment (NCEA) and in consultation with
state agencies. These studies are intended to refine the
NOAEL and to set a standard to replace the current
action level; preliminary results are scheduled for re-
lease in September 1998. EPA's Office of Water has
added perchlorate  to the candidate  contaminant list
(CCL), but it is unclear whether this will eventually
culminate in the establishment of a maximum contami-
nant level (MCL) for perchlorate in drinking water.


Chemical and Physical  Properties

Perchlorate as a Noncomplexing Anion
Perchlorate is widely known to be a very  poor
complexing agent  (Cotton et al.,  1987) and is used
extensively as  a counter ion in studies of metal cation
chemistry, especially in nonaqueous solution. In this
use, it is comparable with other noncomplexing or
weakly ligating anions, e.g., trifluoromethanesulfonate
(triflate, CF3SOJ),  tetrafluoroborate  (BFj), and to a
lesser extent nitrate (NOj). Some exceptions are known,
but these are rare.  All of these anions have  a highly
delocalized (NOJ, ClOj, CF3SOj) or sterically blocked
(BFJ) monovalent  anionic charge and large  volume;
the low charge density reduces their affinity  for cat-
ions and their  extent of aquation (see Table 2). This
low association with cations is  responsible  for the
extremely high solubilities of perchlorate salts in aque-
ous and nonaqueous media. It is important to point out
that the solubility is not due to association with the
solvent While perchlorate is often described as strongly
retained on anion exchange resins, the truth is that the

        Table 2.  Gibbs free energies of
        formation  for selected  anions in
        aqueous solution.3
        Anion
AG,°, kJ moM
        BF<-
        so|-
        HCOj
        OH-
        ci-
        NOj
        Br
        cio<-
        ClOj
   -1490"
   -1019
    -744
    -587
    -157
    -131
    -109
    -104
     -8.5
     -8.0
          From Barrow (1988), except BF;.
          From Dean (1985).
                        ion in the starting form of the resin (e.g., chloride or
                        hydroxide) is much more hydrophilic than perchlorate
                        (Table 2).

                        Perchlorate Salts as Supporting Electrolytes or
                        Ionic Strength Adjustors

                        In addition to its use in synthesizing transition metal
                        compounds where competition for coordination is un-
                        desirable, sodium perchlorate is used extensively as a
                        means of adjusting ionic strength for equilibrium, ki-
                        netics, and electrochemical studies where potassium
                        nitrate cannot be used. For example, many halogen
                        species undergo multiple simultaneous equilibria in
                       . which a central halogen atom expands its octet and
                        forms a hypervalent species; perchlorate does not act
                        as a ligand in these situations (Urbansky et al., 1997).
                        Inorganic perchlorate  salts  are generally extremely
                        soluble, with potassium perchlorate the notable excep-
                        tion (-17 g L"1 = 0.12 M). The solubility of sodium
                        perchlorate in water is extremely high, just under 8 M;
                        only the mineral acids and the alkali metal hydroxides
                        surpass it in solubility.

                        Kinetics and Thermodynamics of Perchlorate
                        Reduction

                        Besides its weak ligating ability, perchlorate often is
                        used to adjust ionic strength because of its low reactiv-
                        ity as an oxidant. At first glance, this seems surprising
                        given that it contains a highly oxidized central halogen
                        atom,  chlorine(VII). However, the low reactivity is a
                        matter of kinetic lability rather than thermodynamic
                        stability. The standard reduction potentials (Emsley,
                        1989) for the half-reactions (equations 2 and 3) clearly
                        indicate that reductions to chloride or chlorate are very
                        favorable processes from a thermodynamic standpoint:
       8H+ + 8e^Cl- + 4H2O   E° = 1.287V

                                             (2)

C1O4- + 2 H+ + 2 e ;= C1O3- + H2O  E° = 1.201 V

                                             (3)

Given thermodynamics alone, we would expect per-
chloric acid to oxidize water to oxygen because the
water-oxygen couple has an oxidation potential of
-1.229 V (Emsley, 1989).
    2 H20 ,=* 4 H+ + O2  -E° = -1.229 V
(4)
                        Consequently, we can confidently state that the ob-
                        served behavior of perchlorate is dominated largely by
84
                                                               Urbansky

-------
 its kinetics. In fact, perchloric acid is quite unreactive
 toward most reducing agents when cold and dilute. For
 example, digestion of organic material (wet ashing)
 with perchloric acid requires heating the material with
 the concentrated acid (Harris, 1991; Schilt, 1979). The
 kinetic barrier to perchlorate reduction is very useful
 in many oxidation-reduction investigations because it
 allows control  of the ionic strength  with a non-
 complexing anion, even at moderately  acidic condi-
 tions (e.g.,  1 M) where nitrate would be reduced.
     It is fortunate, for the sake of remediation, that the
 behavior of perchlorate is due to a kinetic activation
 barrier rather than a thermodynamic barrier, because
 all kinetic problems can be reduced. It comes down to
 the matter of "where there's a will, there's a way."
 Also fortunate for us, the way is fairly well defined for
 perchlorate. Depending on the reductant, perchlorate
 may be reduced to either chlorate or chloride (Cotton
 and Wilkinson, 1988). Ruthenium(H) reduces perchlo-
 rate to chlorate, whereas vanadium(II), vanadium(ni),
 molybdenum(in), dimolybdenum(in), chromium(n),
 and titanium(in) all reduce perchlorate to chloride.
    The first work to show reduction of perchlorate by
 a metal cation was done by Rothmund (1909). He
 showed that Ti(lH), V(H),  and Cr(H) all reduce per-
 chlorate to chloride in acidic aqueous solution at am-
 bient temperature. Bredig and Michel (1922) refined
 Rothmund's Ti(Tfl) work, and they showed that Mo(ffl)
 also reduces perchlorate to  chloride.
    With the right catalyst, other reductants will react
 with perchlorate. In the presence of ruthenium(ni,IV)
 (Crowell et  al., 1929) or osmium(TV) (Crowell et al.,
 1940), bromide  will reduce perchlorate. Tin(II) will
 reduce perchlorate in the presence of molybdate (Haight
 and Sager, 1952). While these studies were significant
 and substantial at the time of publication, the treatment
 of the data was insufficiently rigorous to apply it here
 directly. Nevertheless,  these papers laid the  ground-
 work for many  of  the later investigations and still
 supply directions for future study.
    King and Garner (1954) published the results of
 the first thorough kinetic investigation of the reaction
 of vanadium(II)  and vanadium(in) with perchlorate.
 Reactions 5 through 7 summarize the behavior they
 observed. The oxidation-reduction reactions of per-
 chlorate with V(II) or V(III) occur on comparable time
 scales (equations 5 and 6). The comproportionation in
 equation 7 is much faster, essentially instantaneous.

   8 V2+ + C1OJ + 8 H+ -» 8 V3+ + Cl- + 4 H2O  (5)

 8 V3+ + C1O4- + 4 H2O ->  8 VO2+ + Cl~ + 8 H* (6)
             Kallen and Barley (1971) published a detailed
         investigation on  the reaction  between hexaaquo-
         ruthenium(n) and perchlorate (equation 8). They also
         discussed the factors that control reaction rates of per-
         chlorate reduction by metal cations.

         2 Ru + C1OJ + 2 H+ -> 2 Ru3+ + ClOj + H2O    (8)

             Duke and Quinney (1954) published the first rig-
         orous study on the reaction of titanium(IH)  and per-
         chlorate (equation  9). They found that the reaction
         proceeds through an initial complexation, after which
         Ti(IH) is oxidized to a titanyl ion, TiO2+(equation 10).

         8 Ti3+ + C104- + 8 H+ -H> 8 Ti(IV) + CT + 4  H2O  (9)

           Ti3+ + ClOj ^= TiO2+ + C1O3°               (10)

         Their postulation of the radical chlorine trioxide as the
         first product is still accepted today.
             Cope et al. (1967) studied the Tim-ClO4-reaction
         in the absence of chloride. They obtained the differen-
         tial rate expression given by equation 11:


         rate = -d[Tini]/df = (ifc + JfcfH+f^TPjaO;]  (11)


         where* = 1.9 x  1Q-4 M'1 s-1 and kf = 1.25 x W4 M~2
            Possibly the most significant paper with regard to
        chemical reduction deals with the redox reaction of
        perchlorate with Af-(hydroxyethyl)ethylenediamine-
        NJf ,Ar-triacetatopentaaquotitanium(in) ion (Liu et al.,
        1984). The net reaction is shown in equation 12.


        8 Ti(Hedta) + CIO; + 8 H+ -> 8 Ti(IV) + CT + Hedta3~

          + 4H20                                   (12)

        This Ti(in) chelate is  reasonably stable in air. The
        Ti(IV) produced begins to form hydrous oxides over a
        matter of an hour or so. Over the course of hours to
        days, fine suspended crystallites of TiO2 develop. They
        found the rate expressible as
rate = -d[lim J/dr =
                                 + £
                                            j [Ti(Hedta)]
                                                     (13)
    VO2+ + V2+ + 2 H* -» 2 V3* + H2O
(7)
where k = 2 x lO"3 M~2 s~' and kf = 2 x lO'8 s~l.
    Based on this study by Liu et al. (1984), we might
propose to treat perchlorate-contaminated waters with
Ti(m) chelates under anaerobic conditions. The chlo-
Perchlorate Chemistry: Implications for Analysis and Remediation
                                                      85

-------
ride produced is harmless, and the TiO2 may be re-
moved by agglutination and sedimentation or by filtra-
tion. TiO2  is very insoluble and quite nontoxic. A
number of stable titanium(H[) chelates and complexes
have been prepared; however, in general, precautions
have been  taken to exclude oxygen (Chaudhuri and
Diebler,  1986). It is unclear if any fairly air-stable
titaniumCEH) chelate may be synthesized that will still
react quickly with perchlorate. Even if other factors
can be overcome, this reaction is still too slow to be
useful. If we could lower the pH to 4, for 1 mM ClO^
and 800 mM Ti(Hedta), the rate of the perchlorate
oxidation of Ti(J3I) would be 1.6 x 10~10 M s'1,  while
the rate of the decomposition would be 1.6 x 10"8 M
s'1, i.e.,  100 times faster! Even  if we stopped the
decomposition, the half-life for this reaction would be
50 days, far too long to be practicable in  a  water
treatment plant.
    Hills et al. (1986) demonstrated that molyb-
denum(ni) and dimolybdenum(m) are capable of re-
ducing perchlorate in acidic solution. This is notewor-
thy in light of the fact that molybdenum is not known
to have any stable "-yl" ions in aqueous solution, the
significance of which is described below.
    Taube (1982) has speculated that the relative sta-
bility of the resulting "-yl," i.e., M=O"+, ions is largely
responsible for the behavior observed for  titanium,
vanadium, and other metals, whereas others (Kallen
and Barley, 1971; Liu et al., 1984) have put forth an
alternative explanation that they feel is superior relat-
ing to the polarizability of the metal d-orbitals and the
interaction of these orbitals with the lowest unoccu-
pied  molecular orbital  (LUMO) of the perchlorate.
Although Taube's explanation is more well known,
the Kallen-Earley postulate is reasonable and fits bet-
ter with some of the experimental evidence.
     Very recently, methylrhenium dioxide (CH3ReC>2)
has been shown to abstract oxygen atoms from halates
 and perhalates (Cl, Br, and I) (Abu-Omar and Espenson,
 1995; Abu-Omar et  al., 1996). The net reaction is
 given by equation 14:

     4 CH3Re02 + CIO; -H> 4 CH3ReO3 + Cl~   (14)

 The  first step in the  process is  a reduction to
 chlorate:

        CH3ReO2 + C1OJ -> CH3ReO3 + C1OJ  (15)

 with a reaction rate given by

   rate = -d[ CH3ReO2 ]/dt = 4fc[CH3ReO2 ][C1O;]

                                             (16)
where 4 it = 29 M"1 s"1; the subsequent chlorate reduc-
tion is about 1000 times faster. The reaction rate for
equation  15 is not dependent on acid concentration,
which is  unusual. The rate is unusually fast, several
orders of magnitude faster than for any other transition
metal compound. Abu-Omar and Espenson  (1995)
provide a very convenient and useful table of rate
constants for the reduction of perchlorate and chlorate
by transition metal complexes.
    There are two serious problems with chemical
reductants: (1) they tend to suffer from oxidation by
atmospheric oxygen, and (2) they are too slow under
normal conditions (i.e., pH and concentration). Conse-
quently, anaerobic conditions would be required for
their delivery as well as for the  duration of their reac-
tion time.
    The electrochemical reduction of perchlorate ion
has been reported for a wide variety of cathodes, in-
cluding platinum (Hordnyi and Vertes, 1975b; Vasina
and Petrii, 1969), tungsten carbide (Hordnyi and Vertes,
1974, 1975a; Vertes and Horanyi, 1974), ruthenium
(Gonzales Tejera and Colom Polo, 1984), carbon doped
with chromium(III) oxide or aluminum  oxide
(Mouhandess et al., 1980), aluminum (Kiss et al., 1973),
and titanium  (Mathieu and Landolt, 1978; Mathieu et
al., 1978; Tsinman et al., 1975). Brown (1986) showed
that chlorate and perchlorate may be reduced success-
fully to chloride by active titanium, and he discusses
the possibility of passivation of the titanium electrode
as titanium(IV) forms, presumably from the deposition
of titanium dioxide.
     Given the exceptionally fast reaction between
methylrhenium dioxide and perchlorate, it seems pos-
sible that alkyl rhenium oxides might be used catalyti-
cally if not chemically. Moreover,  this suggests  an
entire area of research into the use of organometallic
compounds as chemical reductants for perchlorate.
Besides  the possible application to remediation, this
also suggests possibilities for use in kinetic methods of
 analysis since anyone with an ultraviolet-visible spec-
 trophotometer could follow this reaction in as low as
 submillimolar levels of perchlorate based on the molar
 absorptivity of CH3ReO3.


 Quantitative Analytical Chemistry

 Gravimetry
 The strongly dissociative property of perchlorate, which
 makes it so highly desirable  in the uses described,
 makes it correspondingly difficult  to quantitate by
 gravimetry or to remove by precipitation. However,
 some insoluble or sparingly soluble perchlorate com-
 pounds  are  known. The first methods reported  for
  86
                                                                                              Urbansky

-------
 quantitation of perchlorate were based on gravimetry
 with nitron (DOC, 1996; Harris, 1991; Welcher, 1947),
 C20HI6N4, Mr = 312.36 g moH. Other names for nitron
 include 4,5-dihydro-2,4-diphenyl-5-(phenyl-imino)-
 lH-l,2,4-triazolium hydroxide, inner salt;  1,4-diphe-
 nyl-3-(phenylimino)-1,2,4-triazolidine, mesoionic
 didehydroderivative;  1,4-diphenyl-3,5-endanilodihy-
 drotriazol; and 3,5,6-triphenyl-2,3,5,6-tetraaza-
 bicyclo[2.2.1]hex-l-ene.
     As might be expected, nitron also quantitatively
 precipitates BFj, WO^", ReOj, and NOj as well as a
 few other anions. In fact, nitron takes its name from its
 ability to precipitate nitrate anion. It is important to
 note that the precipitates contain the acidic hydrogen
 ion, which protonates the positive nitrogen of the inner
 ring; however, there is no significant association be-
 tween the proton and the inorganic anion. The anion is
 simply a large counterion to balance the charge. Table
 3 gives the solubilities reported for some nitron salts.
 In addition to the anions in Table 3, other large anions
 should be expected to interfere; interference is docu-
 mented for ferrocyanide,  ferricyanide, picrate, and
 oxalate (Welcher, 1947).
     Perchlorate may be assayed gravimetrically using
 tetraphenylarsonium chloride, AsPh4Cl (Ph = C6H5)
 (Harris, 1991). While this material may be very useful
 in the analysis of perchlorate, it is doubtful that it will
 establish itself in any way for remediation because of
 the arsenic.
     The low solubilities of Nit • HC1O4 (Nit = nitron)
 and AsPh4ClO4 immediately suggest several opportu-
 nities for both analysis and remediation. Besides tradi-
 tional gravimetric analyses, this property may be ex-
ploited  for electrochemical analyses. As  might be
 expected, potentiometric titrimetry with nitron has been
 used for perchlorate and other anions (Selig, 1988). An
indirect method  has been used to determine  as little
perchlorate as 0.05 to 0.15 nmol using nitron followed
      Table 3.
      species.3
Solubilities of some nitron • HX
      HX
                            Solubility, g L~1
HCI04
HN03
HI
HSCN
HCrO4H
HCIO3
MONO
HBr
0.08
0.099
0.17
0.4
0.6
1.2
1.9
6.1
         From Welcher, 1947.
 by an iodimetric back-titration (Shahine and Ismael,
 1976).


 Potentiometry and Ion-Selective Electrodes
 (ISEs)

 One technique that holds excellent promise as a rou-
 tine monitoring device is potentiometric measurement
 via an ion-selective electrode. The perchlorate ion-
 selective electrode (ISE) has an extensive history and
 has been under development for about 20 years. Many
 designs and components abound, some with outstand-
 ing response. A liquid-membrane ISE has been shown
 to have nearly Nernstian response to perchlorate in the
 range of 10~5 to 10"2 M with a lower limit of detection
 of -10-6 M (Hassan and Elsaied, 1986). A polyvinyl
 chloride (PVC) membrane impregnated with HNitSCN
 has been used to determine perchlorate down to 2.5 x
 10"5 M, and one impregnated with As(C6H5)4SCN has
 been constructed, but it was not tested for response to
 C1O4  (Elmosalamy et al.,  1987). Perchlorate, along
 with several other anions, has been determined using
 flow  injection analysis with a carbon electrode and
 &M(diphenylphosphino)propane-copper complex  as an
 ion exchanger (Wang and Kamata, 1992). Making use
 of the low solubility of potassium perchlorate, a potas-
 sium  cation ISE was  used  to study the migration of
 perchlorate into a PVC membrane (Verpoorte and
 Harrison, 1992). Perchlorate ISEs based on a barium
 complex with a macrocyclic Schiff base have  been
 developed (Masuda et al., 1991). The newer perchlo-
 rate ISEs are based on large, inert, metal-ligand com-
 plexes that do not undergo  complexation with small,
 hard Lewis bases and have no open coordination sites.
    The perchlorate ISE has already established itself
 as a research tool, and it has  been used to  monitor
perchlorate concentration in a variety of investigations
 (Alegret et al., 1986; Baczuk and Dubois, 1968; Cakrt
 et al., 1976; Ciavatta et al.,  1989;  Efstathiou and
 Hadjiioannou, 1977a,  1977b; Fogg et al., 1977; Hiiro
 et al.,  1979; Hopirtean and Stefaniga, 1978; Hopirtean
et al.,  1976, 1977; Hseu and Rechnitz, 1968; Lnato et
 al., 1980; Ishibashi and Kohara, 1971; Ishibashi et al.,
 1973; Jain et al., 1987; James et al., 1972; Jasim, 1979;
Jyo et al., 1977, 1983; Kataoka and Kambara, 1976;
Manahan et al., 1970; Matei et al., 1986; Nikolskii et
al., 1977; Pathan and Fogg, 1974; Rohn and Guilbault,
 1974; Sharp, 1972; Silber and Zhang, 1991; Sykut et
al., 1979; Tateda et al., 1970; Vosta and Havel, 1973a,
 1973b; Wilson, 1979; Wilson and Pool,  1976). Based
on the significant advances in perchlorate ISEs, the
time is ripe for exploration of perchlorate ISEs  as a
technique both for first-line assay in the field and for
Perchlorate Chemistry: Implications for Analysis and Remediation
                                                                                    87

-------
continuous monitoring within a treatment facility, es-
pecially in regions with known contamination.

Ion Chromatography and Capillary
Electrophoresis
Perchlorate ion is commonly described as strongly
retained on anion exchange resins. What this really
means is that other ions are more strongly attracted to
the aqueous mobile phase. It does not mean that the
resin has a high affinity for perchlorate. Chloride and
hydroxide have  much higher charge  densities than
perchlorate and therefore associate more strongly with
water than they  do with a fairly diffuse quaternary
ammonium site. If eluent components are not chosen
wisely, perchlorate elution times can run over an hour,
allowing substantial diffusion and thus peak broaden-
ing.
    The California Department of Health Services has
established an ion chromatography (1C) method that
usesp-cyanophenoxide to displace the perchlorate from
the resin (Cal DHS, 1997b). Recently, Maurino and
Minero (1997) showed that hydrogen cyanurate ions,
HzA" and HA2" (cyanuric acid = H3A), can be used to
effect excellent separation and peak shape for perchlo-
rate. Biesaga et al. (1997) showed that phthalate can be
used, but retention times are long (40 min) and some
degradation  of peak shape is observed. Jandik et al.
(1990) used acetonitrile to modify the eluent dielectric
constant and a solid-phase reagent to obtain retention
times under 20 min. In a study of retention time and
polarizability, Daignault  et al. (1990) used 1.7 mM
NaHCO3-1.8mMNa2CO3 eluent for perchlorate; how-
ever, they used analyte concentrations of 2 mg mL"1.
Buchberger  and  Haider (1997) used 1C with particle
beam mass spectrometry to detect perchlorate, provid-
ing a definitive  identification of the  ion. An earlier
study used micropacked alumina columns to separate
anions and cations simultaneously, making use of the
amphoteric nature of aluminum oxide (Takeuchi et al.,
1988). Wirt et al. (1998) reported a new 1C method
using only NaOH as the eluent with a retention time of
<10 min.
    One of the advantages of capillary electrophoresis
(CE)  over 1C is readily apparent when considering
strongly retained ions, such as perchlorate. In CE sepa-
ration, the electrophoretic (ionic) mobility is the most
important factor, unlike 1C where interactions with the
stationary and mobile phase are important. In CE,
associations with the wall or a modifier are generally
unintended, undesirable, and, most importantly, avoid-
able. Furthermore, a slow eluter when using 1C may be
a very fast one when using CE. Avdalovic et al. (1993)
showed that perchlorate can be eluted by CE at 10 min;
meanwhile, bromide, chloride, and iodide all take more
than 15 min by their method. Hauser et al. (1995) used
CE with a micro-ISE to quantitate as little as  10 |OM
perchlorate. Corr and Anacleto (1996) used CE coupled
with mass spectrometry with ion spray introduction on
a wide variety of cations and anions. Although it is
common practice to use quaternary ammonium salts,
such as cetyltrimethylammonium chloride, to reverse
the electroosmotic flow, Krokhin et al.  (1997) used
water-soluble polymers with quaternary ammonium
moieties to promote the elution of perchlorate  by CE.
    AFRL and EPA's National Exposure Research
Laboratory  (NERL)  have begun the process of
interlaboratory validation for the California DHS 1C
method (Tsui and Pia, 1998). There currently is no
EPA-approved method for the quantitation of perchlo-
rate in drinking water; however, Cal DHS has estab-
lished its own approval process for laboratories that
seek to analyze for perchlorate. AFRL continues to
work on methods for determining perchlorate in other
media, e.g., soils and plant tissues.


Other Techniques

In addition to gravimetry and electrochemistry, other
techniques also offer promise when coupled with these
reagents, particularly spectrophotometry. Methylene
blue forms an insoluble complex with perchlorate;
the loss of methylene blue from the supernatant is
determined spectrophotometrically (Nabar and
Ramachandran, 1959). The neocuproine-cuprous ion
complex also has been used to extract perchlorate into
ethyl acetate; the fcz'.s(neocuproine)cuprous perchlorate
species has a visible absorption spectrum (Xmax = 456
nm) or, alternatively, the copper content may be deter-
mined by atomic absorption photometry (A, = 324.7
nm) (Collinson and Boltz, 1968). A similar method
designed for biological fluid samples makes use of an
anion  exchange  resin (Amberlite  IR-45) to pre-
concentrate the perchlorate and thereby lower (im-
prove) the quantitation  limit (Weiss and Stanbury,
1972).
    Both tetraphenylarsonium chloride and nitron can
be used to determine perchlorate concentration spec-
trophotometrically by difference in samples contain-
ing nitrate (Shahine and Khamis, 1979). An aliquot of
nitron  solution is added to the sample, and the excess
nitron is determined photometrically  as (HNit)2
[Co(NCS)4] at 625 nm;  perchlorate alone is deter-
mined by precipitation with AsPh^, and the excess is
determined as (AsPh4)2[Co(NCS)4]  at 620 nm.  Al-
though not reported, it seems reasonable that laser-
 88
                                        Urbansky

-------
  induced fluorescence would be be a sound technique,
  especially if the HNitClO4 or AsPh4ClO4 were ex-
  tracted into an organic solvent.
      Reverse-phase high-performance liquid chroma-
  tography  also could be used  on such an extract to
  separate perchlorate from interferents such as nitrate
  or bromide with detection by ultraviolet (UV) absor-
  bance.  The use  of a photodiode array detector may
  provide a UV absorption spectrum sufficiently distinct
  to ensure definitive identification of perchlorate, elimi-
  nating the retention time  problem. It  is known that
  AsPh4ClO4 has a unique infrared absorption spectrum
  that permits  it  to be distinguished  from other
  tetraphenylarsonium salts. Numerous methods using
  wet chemical or instrumental techniques or a combina-
  tion thereof are known, and these have been reviewed
  extensively for the literature prior to  1979 (Schilt,
  1979). Although mulls in  Nujol or perfluorosilicone
 grease may be safe, compressing AsPh4ClO4 in a KBr
 pellet seems to be flirting with disaster. Even if KBr
 pellets can be made safely, there is no guarantee that
 some decomposition of the perchlorate will not occur
 or that a reaction  with bromide will not take place
 during pressurization.
     Walter Selig at Lawrence Livermore Laboratory
 has investigated a number of potentiometric precipita-
 tion titrations to  determine perchlorate using quater-
 nary  ammonium  compounds as titrants (Selig, 1977,
 1979, 1980a, 1980b).
    A carbon paste electrode that uses thallium for
 catalysis has been used to voltammetrically quantitate
 perchlorate in drinking water samples down to 50 ng
 mL~!; however, it suffers from a number of significant
 direct and indirect interferences (Neuhold et al., 1996).
 Of these, bromide, chlorate, and nitrate are most likely
 to be found in a drinking water matrix.


 Remediation and Treatment
 Overview

 It is helpful to keep in mind the following criteria for
 any drinking water treatment technology. The treat-
 ment  must  not (1) adversely affect other treatment
 technologies for regulatory compliance, (2) produce
 water that corrodes the  distribution system, (3) pro-
 duce water that is unpalatable, (4) suffer from degra-
 dation by other components in  the water, (5) fail to
perform reliably, (6) produce excessive  waste, or (7)
fail to meet time and expense constraints. The best
choice for any situation will require a careful evalua-
tion of options and probably some combination of
techniques. We must remember that the potential for
success of any technology is dependent on two factors:
  the establishment of a safe level of perchlorate for
  drinking water and a quantitative chemical analysis
  that ensures this safe level is in fact achieved.

  Remediation and Treatment by Physical
  Processes

  Membrane-Based Techniques.  Membrane-based
  techniques can be effective, but they suffer from sev-
  eral drawbacks.  While reverse osmosis (RO) would
  effect sufficient remediation, it can be impractical for
  a municipal treatment system because of the fouling of
  membranes and the associated cost. RO-treated water
  has to be remineralized with sodium chloride, sodium
  bicarbonate, and other innocuous salts to prevent deg-
  radation of the distribution system and to make the
  water palatable,  since deionized water  generally is
  considered to have an unpleasant taste. Therefore, as
  long as sufficient salts are taken in from food and other
  sources, consumption of deionized water is not likely
  to pose a threat to the normal electrolyte balance. As
  with RO,  electrodialysis also might be used in this
 fashion. These two techniques are probably best suited
 for point-of-use or small systems.

 Anion Exchange.  Although perchlorate ion is
 strongly retained by quaternary ammonium resins, the
 crux of the matter is its initially low concentration in
 most cases. For  example,  it might be necessary to
 reduce perchlorate concentration from 1 jig mL-1 to 20
 ng mL-1. However, consider that bicarbonate, carbon-
 ate,  chloride, and a host of other anions are all likely
 to be present at much higher concentrations. Assuming
 that a chloride-form resin  is used, the presence of
 phosphates, carbonates, and sulfate remains an issue.
 Although it may be possible to produce a resin salt that
 matches the proportions of the major anions in the
 influent water, to do so would be extremely inconve-
 nient. In addition, the low concentration of perchlorate
 in the raw water substantially reduces the driving force
 for its removal. In other words, to adequately remove
 the perchlorate may require  essentially demineralizing
 and remineralizing the water,  depending on its anion
 content.
    It is possible to modify resins  so as  to improve
 their selectivity for particular anions. Kawasaki et al.
 (1993) have used Dowex 1X-8 to selectively precon-
 centrate perchlorate; the  selectivity of the resin for
perchlorate is about 100 times that for chloride and 10
times that for nitrate. In  addition to selectivity  in a
thermodynamic sense, there is the matter of rapid equili-
bration and anion exchange. If the rate of exchange is
too slow, a resin will not be usable no matter how high
its selectivity. The U.S. Department  of Energy has
Perchlorate Chemistry: Implications for Analysis and Remediation
                                                                                                   89

-------
developed a mixed triethylammonium-trihexylammo-
nium resin that is capable of removing pertechnetate
down to the parts-per-trillion level (Brown, 1997).
    Tethered triphenylarsonium or phosphonium moi-
eties or a tethered (through a phenyl  group) nitron
might work in an anion-exchange resin to selectively
preconcentrate  perchlorate as a step in remediation.
The disadvantage of the tethered triphenylarsonium
group is that normal degradation of the resin would
lead to the release of arsenic into the  treated water.
Although the health effects of nitron are unknown, it
would be expected to undergo biodegradation; further-
more, it would  be destroyed readily by UV irradiation
(A,£ 185 nm),  whereas arsenic would remain as an
inorganic oxyanion even if the organic portion of the
species were destroyed.

Precipitation.   The low solubility of the HNitClO4
ion pair reveals a strong association between the pro-
tonated nitron  cation and the perchlorate anion. All
insoluble ion pairs and complexes exist at some level
in solution. It may be possible to exploit this pairing
for purposes of remediation. If the addition of nitron to
perchlorate-containing waters results in formation of
the soluble ion pair, it may be possible to subsequently
induce an intramolecular reaction in which both the
perchlorate and the nitron are destroyed. Photoactivation
 of the perchlorate by  UV or laser irradiation  may
promote an intramolecular redox reaction (probably
 by oxygen atom transfer). The proximity of the HNif-
 ClOj ion pair within a solvent (water) cage means that
 it is not necessary to form an encounter complex. In
 addition, the local concentration of the two species is
 very high within the solvent cage. This should help to
 reduce the effects of the perchlorate kinetic barrier
 (discussed below). Of course, irradiation with UV light
 also will promote destruction of the nitron by hydroxyl
 radical formation. Ideally, the largest  possible wave-
 length (lowest frequency and energy) light would be
 used to reduce side reactions that would destroy the
 nitron.
     Unfortunately, nitron has potential to remediate
 only those sites with very high perchlorate concentra-
 tions unless it can be  synthesized more cheaply. At
 present, nitron is about 52 times more expensive than
 an equal mass of reagent-grade sodium chloride. How-
 ever, should a method involving nitron prove effec-
 tive, bulk synthesis of the material would likely drop
 the cost by 40 to 50%, and use of a less refined tech-
 nical (rather than reagent) grade would probably re-
 duce it by another 10 to 25%.
      At some of the sites where the perchorate concen-
  tration is 0.037 M, nitron could readily be used as a
 precipitant since the nitron-hydrogen perchlorate salt
has a solubility of only 0.19 mM. Although the action
level of 18 ng mL"1 corresponds to 0.18 M, a level of
0.19 mM is certainly preferable to 37 mM. Of course,
one drawback is that a source  of acid (usually  5%
acetic acid) must be present. On the other hand, vin-
egar is probably preferable to 0.037 M ClOj. More-
over, such post-remediation acetic acid and acetate
would be biodegradable.
     In addition to cost, all physical separation pro-
cesses have one major problem: waste disposal. Pre-
sumably, the regenerant from ion exchange and the
concentrate from RO or electrodialysis would contain
perchlorate at concentrations too high to be released
into a sewage system. This waste presents a problem in
terms of cost and post-treatment needs. Although these
techniques take the perchlorate out, they concentrate it
somewhere else where it must be dealt with later.

Remediation and Treatment by Chemical
Processes
Chemical and Electrochemical Reduction.  Here we
refer to reduction specifically in the redox sense of
adding electrons. From the description of the oxida-
tion-reduction reactions of perchlorate above, it is clear
that chemical reduction will play no role in drinking
water treatment in the near future. Chemical reduction
is simply too. slow. Unless safe new catalysts become
 available, this appears unlikely to change. Common-
 place reductants (e.g., iron metal; thiosulfate, sulfite,
 iodide, and ferrous ions) do not react at any observable
 rate, and the more reactive species are too toxic (and
 still too sluggish). In addition, any reductant will nec-
 essarily have oxidized by-products. The toxicity of the
 by-products must be considered.
     There is more hope for electrochemical reduction.
 A decided advantage  of electrochemical reduction is
 the large amount of control over kinetics that results
 from control of the operating potential. Electrode re-
 duction kinetics  reasonably can be viewed as being
 limited by three factors: (1) diffusion of the ions to the
 electrode surface, (2) association  with the electrode
 surface, and (3)  activation past the overpotential re-
 quired to reach  the transition  state.  Although
 overpotential usually  is the greatest barrier, it also is
 the one that can be dealt with best. Because we are not
 concerned with other reductions (including reducing
 water  to hydrogen), the only barrier is the limit of a
 negative potential that is practical  and safe to apply.
 Fortunately, most of the materials  in raw water are
 reducing agents. Although some may be affected by
 electroreduction, this probably does not present a sig-
 nificant obstacle. To date, this option has not been
  explored for low-concentration treatment at anything
  90
                                                                                                Urbansky

-------
  approaching pilot scale. Although electrochemical tech-
  nologies are well established for other industries (e.g.,
  electroplating of metals, electrolysis of brine), they
  have not yet found a place in drinking water treatment.
      In this category, it appears that the most success-
  ful strategies for remediating perchlorate contamina-
  tion will utilize metal cation-catalyzed reduction by
  either chemical or electrochemical means. Several metal
  chelates have potential at this point, especially if em-
  bedded in an electrode for use in electrochemical  re-
  duction.

 Biological and Biochemical Techniques. Bio-
 remediation is another matter entirely, and it may prove
 to be the most practical approach. A number of bacte-
 ria that contain nitrate reductases (Payne, 1973) are
 capable of reducing perchlorate (Schilt, 1979). Staphy-
 lococcus epidermidis is capable of reducing perchlor-
 ate in the absence of nitrate. Cell-free extracts of ni-
 trate-adapted Bacillus cereus also reduce perchlorate
 (and chlorate) (Hackenthal,  1965). As  would be ex-
 pected, sodium perchlorate, especially in higher con-
 centrations,  has  been shown to be toxic to several
 species  of bacteria. Unfortunately, S. epidermidis is
 also pathogenic; it increasingly is encountered as a
 source of nosocomial infections, especially opportu-
 nistic infection with in-dwelling intravenous or uri-
 nary catheters (Archer, 1995). It is encountered with
 other medical apparatus such as prosthetic joints, pace-
 makers, heart  valves, and breast implants (Archer,
 1995). Like S. epidermidis, B. cereus is pathogenic. B.
 cereus is known for food poisoning, ocular infections,
 and pneumonia with other sites sometimes affected
 (Tuazon, 1995).
     Rikken et al. (1996) reported that perchlorate and
 chlorate are reduced to  chloride  by  Proteobacteria
 with acetate as a nutrient (reductant) at near-neutral
 pH. While they did show loss of perchlorate and chlo-
 rate, their mechanisms failed to include  contributions
 from uncatalyzed reactions.  Specifically, they con-
 cluded that a dismutase is responsible for all elimina-
 tion of toxic chlorite from the cell, catalyzing its dis-
 proportionation to dioxygen and chloride. However,
 the uncatalyzed disproportionation of chlorite to chlo-
 ride and chlorate is not necessarily negligible. Korenkov
 et al. (1976) patented Vibrio dechloraticans Cuzensove
 B-1168 for perchlorate reduction; V. dechloraticans is
 nonsporulating, motile, and gram negative. Malmqvist
 et al. (1994)  showed that Ideonella dechloratans can
reduce chlorate, but they did not test for perchlorate
reduction.
     Over the past 8 years, work at AFRL has shown
that perchlorate is metabolized by  Wolinella
succinogenes, strain HAP-1 (U.S. Air Force,  1994;
  Wallace and Attaway, 1994). W. succinogenes is ca-
  pable of using either chlorate or perchlorate to oxidize
  Brewer's yeast. Pilot-scale systems at Tyndall AFB,
  Florida showed that perchlorate levels could be re-
  duced from 3000 |J,g mL-1 to below 0.5 |j,g mlr1 (Hurley
  et al., 1997). HAP-1  was first isolated from a munici-
  pal anaerobic digester. The bacterium is an antibiotic-
  resistant, nonsporulating, motile, Gram-negative, obli-
  gately anaerobic bacillus (Wallace et al., 1996). This
  sort of remediation may be  effective at a site where
  perchlorate concentrations in water  are high, but it
  very likely would be impractical for  the treatment of
  drinking water unless it can be demonstrated to reach
  even lower perchlorate concentrations. AFRL's efforts
  have led to the implementation of a production-scale
 bioreactor in Utah for meeting perchlorate discharge
 requirements (Hurley, 1998).
     Very little research has been done on perchlorate
 reductases. It may be possible to isolate these from
 bacteria and use them directly as reagents without the
 parent organisms. The mechanisms of these catalysts
 are not well understood, and the reductases themselves
 have not been well characterized. It may be possible to
 synthesize an analogous catalyst based on the reduc-
 tase, but only if the fundamental bioinorganic chemis-
 try is understood. Although nitrate reductases are based
 on molybdenum (Coughlan,  1980), it  has not been
 verified  whether this is also  true for the perchlorate
 reductases.
     Several projects are ongoing  in the affected areas
 of EPA's Region 9 and were described at a recent
 meeting. Catts (1998) reported that  a pilot-scale
 bioreactor has been constructed for the Baldwin Park
 Operable Unit in California  using microbes derived
 from the food-processing industry. Operation of this
 pilot unit over a period of several months showed that
 perchlorate and nitrate could  be reduced to undetect-
 able levels, i.e., [ClOj] < 4 ng mL-1. Ethanol was used
 as a food source and minerals were added to the sys-
 tem. The perchlorate-reducing microbes were not iso-
 lated or characterized. Sase (1998) reported on a spe-
 cially designed anion exchange system with alternately
 regenerating columns  that is undergoing testing by the
 Main San Gabriel Basin Watermaster.

 Conclusions

 Bioremediation and biological or biochemical  treat-
 ment appear to  be  the most economically feasible,
 fastest, and easiest means of dealing with perchlorate-
 laden waters at all concentrations. Although other tech-
 niques may find  application  to select systems, e.g.,
point-of-use or small utilities, it appears that biological
and biochemical approaches will play the greatest role
Perchlorate Chemistry: Implications lor Analysis and Remediation
                                                                                                     91

-------
in solving the perchlorate problem. Some situations
may require a combination  of technologies to best
meet unique needs. This is a complex problem, and
many of the standard technologies that have domi-
nated the drinking water industry for the past several
decades will not work for this contaminant when used
in the conventional ways. Many of the possibly effec-
tive technologies have not been applied to drinking
water specifically, and the interplay with other treat-
ment technologies required for regulation compliance
must be assessed. In addition to rapid implementation
of effective and workable technologies, ongoing de-
velopment will be required to find new technologies
and to make them affordable and assimilable into the
industry.


References

Abu-Omar, M.M., and J.H. Esperison. 1995. "Facile abstrac-
   tion of successive oxygen atoms from perchlorate ions by
   methylrhenium dioxide." Inorg. Chem., 34:6239-6240.
Abu-Omar, M.M., E.H. Appelman, and J.H. Espenson. 1996.
   "Oxygen-transfer reactions of methylrhenium oxides."
   Inorg. Chem., 35(26):7751-7757.
Alegret, S., A. Florido, J.L.F.C. Lima, and A.A.S. Machado.
    1986. Quim. Anal (Barcelona) 5:36-48.
Archer, G.L. 1995. In:  G.L. Mandell, I.E. Bennett, and R.
   Dolin(Eds.),Ma/HfeW, Douglas, andBennett's Principles
   and Practice of Infectious Diseases, 4th ed. Churchill
   Livinstone, New York, NY. Ch. 174.
Avdalovic, N., C.A. Pohl,  R.D. Rocklin, and J.R. Stillian.
    1993. "Determination of cations and anions by capillary
    electrophoresis combined with suppressed conductivity
    detection." Anal. Chem. 65:1470-1475.
Baczuk, R.J., and R.J. Dubois. 1968. Anal. Chem. 40:685-
    689.
Barrow, G.M. 1988. Physical Chemistry, 5th ed. McGraw-
    Hill, New York, NY. Appendix B, Table 2.
 Biesaga, M., M. Kwiatkowski, and M. Trojanowicz. 1997.
    "Separation of chlorine-containing anions by ion chro-
    matography and capillary electrophoresis."/. Chromatogr.
    A, 777:375-381.
 Bredig, G., and J. Michel. 1922. Z. Physik. Chem., 100:124.
 Brown, G. 1986. "The reduction of chlorate and perchlorate
    ions at an active titanium electrode."/. Electroanal Chem.,
    798:319-330.
 Brown, Gilbert. 1997. Personal communication from Brown,
    U.S. Department of Energy, Oak Ridge National Labora-
    tory.
 Bruton, J.D., J. Lannergren, and H. Westerblad. 1995.
    "Mechano-sensitive linkage in  excitation-contraction
    coupling in frog skeletal muscle." /. Physiol. Land., 484
    (Part 3):737-742.
 Buchberger, W., and K. Haider. 1997. "Studies on the com-
    bination of ion chromatography-particle-beam mass spec-
   trometry with capillary columns." J. Chromatogr. A,
   770:59-68.
Cakrt, M., J. Bercik, and Z. Hladky. 1976. Fresenius' Z.
   Anal. Chem., 257:295-298.
Cal DHS, see California Department of Health Services.
California Department of Health Services. 1997a. "Perchlo-
   rate in California Drinking Water." California Depart-
   ment of Health Services, September, http://www/
   dhs.cahwnet.gov/prevsrv/ddwem/perchl.htm#advice.
California Department of Health Services. 1997b. Determi-
   nation of Perchlorate by Ion Chromatography, Rev. 0.
   June 3. Sanitation and Radiation Laboratories Branch.
Capen, C.C., and S.L. Martin. 1989. Toxicol. Pathol, 17:266-
   293.
Catts, J.. 1998. "Biological Treatment of Low Level Perchlo-
   rate Contamination." Presented at the Perchlorate Stake-
   holders Forum, May 19-21, 1998, Henderson, NV.
Chaudhuri, P., and H. Diebler. 1986. "Kinetics and equilibria
   of 1:1 complex formation of hexa-aquatitanium (III) with
   malonic and methylmalonic acid." J.  Chem. Soc. Dalton
   Trans., 1693-1695.
Chiovato,  L., F. Santini,  and  A. Pinchera.  1997.  'Treat-
   ment of Hyperthyroidism,"  Pisa, Italy, http://www.
   thyrolink.com/thyint/2-95int.htmffintro.
Ciavatta, L., M. luliano, and R.  Porto.  1989. Ann. Chim.,
   79:319-333.
Collinson, W.J., and D.F. Boltz. 1968. "Indirect spectropho-
   tometric and atomic absorption spectrometric methods
   for determination of perchlorate." Anal. Chem., 40:1896-
    1898.
Cooper, D.S. 1991. 'Treatment of Thyrotoxicosis." In: L.E.
   Braverman and R.D. Utiger (Eds.), The Thyroid: A Fun-
   damental and Clinical Text, 6th ed. J.B. Lippincott, Phila-
   delphia, PA. pp. 887-916.
Cooper, D.S. 1996. 'Treatment of Thyrotoxicosis." In: L.E.
   Braverman and R.D. Utiger (Eds.), Werner and Ingbar's
    The Thyroid, 7th ed. Lippincott-Raven, Philadelphia, PA.
    Ch. 53.
Cope, V.W., R.G. Miller, and R.T.M. Fraser. 1967. 'Tita-
    nium III ion as a reductant in electron-transfer reactions."
    J. Chem. Soc. (A), 301-306.
Corr, J.J., and J.F. Anacleto. 1996.  "Analysis of inorganic
    species by capillary electrophoresis-mass spectrometry
    and ion exchange  chromatography-mass spectrometry
    using  an iori spray source." Anal. Chem., 68:2155-
    2163.
 Cotton, F.A., and G. Wilkinson.  1988. Advanced Inorganic
    Chemistry, 5th ed.  Wiley, New York, NY. p. 668.
 Cotton, F.A.,  G. Wilkinson, and P.L.  Gaus. 1987. Basic
    inorganic Chemistry, 2nd ed. Wiley, New York, NY. p.
    219.
 Coughlan, M., Ed. 1980. Molybdenum and Molybdenum-
    Containing Enzymes. Pergamon, New York, NY.
 Crowell, W.R., D.M. Yost, D.M.,  and J.M. Carter. 1929.
    "Catalytic  effect of ruthenium salts  on the reduction of
    perchloric acid by hydrobromic acid." J. Am. Chem. Soc.,
    57:786-794.
  92
                                                                                                     Urbansky

-------
 Crowell, W.R., D.M. Yost, and J.D. Roberts,  1940. 'The
    catalytic effect of osmium compounds on the reduction of
    perchloric acid by hydrobromic acid." J. Am. Chem. Soc.,
    (52:2176-2178.
 Daignault, L.G., D.P.  Rillema, and D.C. Jackman. 1990.
    "Retention  behavior of ions  on high  performance  ion
    chromatography columns, part II: Retention behavior and
    polarizability." J. High Res. Chromatogr., 293-294.
 Dean, J.A., Ed. 1985. Lange's Handbook of 'Chemistry, 13th
    ed. McGraw-Hill, New York, NY. Table 9-1.
 Dictionary of Organic Compounds, 6th ed. 1996. Chapman
    & Hall, London. "Nitron, N-0-01238."
 DOC, see Dictionary of Organic Compounds.
 Dollarhide, J.S. 1992. "Provisional Non-Cancer and Cancer
    Toxicity Values for Potassium Perchlorate." Memoran-
    dum to Dan Stralka, U.S. Environmental  Protection
    Agency.
 Dollarhide, J.S. 1995. Correspondence to Mike Girard, U.S.
    Environmental Protection Agency.
 Duke, F.R., and P.R. Quinney. 1954. "The kinetics of reduc-
    tion of perchlorate ion by Ti(HI) in dilute solution." /.
    Am. Chem. Soc., 76:3800-3803.
 Efstathiou, E.G., andT.P. Hadjiioannou. 1911 &. Anal. Chem.,
    49:410-418.
 Efstathiou, E.G., and T.P. Hadjiioannou.  1977b. Talanta,
    24:270-272.
 Elmosalamy, M.A.F., G.J. Moody, J.D.R. Thomas, and S.M.
    Hassan. 1987. Anal. Lett., 20: 1541-1555.
 Emsley, J. 1989. The Elements. Clarendon, Oxford, pp. 48,
    134.
 European Communities. 1982. "EC Directive Relating to the
    Quality of Water Intended for Human Consumption." 80/
    778/EEC, Office for Official Publications of the Euro-
    pean Communities, 2 rue Mercier, L-2985, Luxembourg.
 Fogg, A.G., A.A. Al-Sibaai, and K.S. Yoo. 1977. Anal. Lett,
    10:173-182.
 Foye, W.O., Ed. 1989. Principles of Medicinal Chemistry,
    3rd ed. Lea & Febiger, Philadelphia, PA. pp. 612-613.
 Fruen, B.R., J.R. Mickelson, TJ. Roghair, H.L. Cheng, and
    C.F. Louis.  1994. Am. J. Physiol, 266 (Part  1):C1729-
    1735.
 Gonzalez, A., and E. Rios. 1993. J. Gen. Physiol., 102:373-
   421.
 GonzalesTejera,MJ.,andF.ColomPolo. 1984.Ann. Chim.,
   80:219.
 Hackenthal, E. 1965. Biochem. Pharm., 14:1314-1324.
 Haight, G.P. Jr., and W.F. Sager. 1952. "Evidence for pref-
   erential one-step divalent changes in the molybdate-cata-
   lyzed reduction of perchlorate by stannous ion in sulfuric
   acid solution." J. Am. Chem. Soc., 74:6056-6059.
Harris, D.C. 1991. Quantitative Chemical Analysis, 3rd ed.
   Freeman, New York, NY. p. 146, 722-723.
Hassan, S.S.M., and M.M. Elsaied. 1986. Talanta, 33:679-
   684.
Hauser, P.C., A.P.C. Hong, andN.D. Renner. 1995. "Surface
   charge reversal for inorganic anion determination in cap-
   illary electrophoresis with an ion-selective microelec-
   trode as detector." J. Cap. Elec.,  2:209-212.
 Hiiro, K., A. Kawahara, and T. Tanaka. 1979. Anal. Chim.
    Acta, 110:321-324.
 Hills, E.F., C. Sharp, and A.G. Sykes. 1986. "Kinetic studies
    on the perchlorate oxidation of aqua Mo3+ and Mom2."
    Inorg. Chem., 25:2566-2569.
 Hobson, Q.J.G. 1961. Br. Med. J., 7:1368-1369.
 Hopirtean, E., M. Preda, and C. Liteanu. 1976. Chem. Anal.
    (Warszawa), 27:861-866.
 Hopirtean, E.,  and E. Stefaniga. 1978.  Rev. Roum. Chim.,
    23:137-142.
 Hopirtean, E., E. Veress, and V. Muresan. 1977. Rev. Roum.
    Chim., 22:1243-1250.
 Horanyi, G., and G. Vertes. 1974. Inorg. Nucl. Chem. Lett,
    10:767.
 Horanyi, G., and G.  V6rtes.  1975a. J. Electroanal. Chem.,
    63:359.
 Horanyi, G., and G.  Vertes.  1975b. J. Electroanal. Chem.,
    64:252.
 Hseu, T.M., and G.A. Rechnitz. 1968.  Anal. Lett., 1:629-
    640.
 Hurley, J. 1998. "Biological Treatment of Wastewater." Pre-
    sented at the Perchlorate Stakeholders Forum, May 19-
    21, 1998, Henderson, NV.
 Hurley, J.A., W. Wallace, and E. Coppola. 1997. "Prototype
    Demonstration of Ammonium Perchlorate Biodegrada-
    tion." http://www.afcesa.af.mil/afcesa/ce-magazine-
    fal-story21-htm, "Demonstration of Ammonium Perchlo-
    rate  Degradation." http://www.brooks.af.mil/hsc/al/eq/
    prod!3.html, "Ammonium  Perchlorate  Biodegradation
    for Industrial Wastewater Treatment." http://estcp.
    xservices.com/projects/comply/comp_al.htm
 Imato, T., A.  Jyo, and  N. Ishibashi. 1980. Anal  Chem.,
    52:1893-1896.
 Ishibashi, N., and H.  Kohara. 1971. Anal. Lett., 4:785-792.
 Ishibashi, N., A. Jyo, and K. Matsumoto. 1973. Chem. Lett.,
    1297-1298.
 Jain, A., M. Jahan, and V.  Tyagi.  1987.  Analyst (London),
    772:1355-1357.
 James, H.J., G. Carmack, and H. Freiser. 1972. Anal.  Chem.,
   44:856-857.
 Jandik, P., J.B. Li, W.R. Jones, and D.T. Gjerde.  1990. "New
   method of background eluent conductivity elimination in
   gradient ion chromatography." Chromatographia, 30:509-
   517.
Jasim, F. 1979.  Iraqi J. Sci., 20:430-447.
Jong, D.S., K. Stroffekova, and J.A. Heiny. 1997. J. Physiol.,
   499 (Part 3):787-808.
Jyo, A., T. Imato, K. Fukamachi, and N. Ishibashi. 1977. "A
   new type of ion-selective electrode based on an anion
   exchange resin with a hydrophobic site." Chem. Lett.,
   815-816.
Jyo, A.,  T.Jmato, H.  Kohno, and N.  Ishibashi. 1983. Bull.
   Chem. Soc. Jpn., 56:3177-3178.
Kallen, T.W., and J.E. Barley. 1971. "Reduction of the per-
   chlorate  ion by aquoruthenium(II)." Inorg. Chem.,
   10:1152-1155.
Kataoka, M., andT.J. Kambara.  1976. J. Electroanal. Chem.
   Interfacial Electrochem., 73:279-284.
Perchlorate Chemistry: Implications for Analysis and Remediation
                                                                                                            93

-------
Kawasaki, M., T. Omori, and K. Hasegawa. 1993. "Adsorp-
   tion of pertechnetate on an anion exhange resin."
   Radiochim. Acta, 63:53-56.
Khammari, A., S. Baudot, and J. Noireaud. 1996. Acta Physiol.
   Scand., 756:447-456.
King, J.H. 1995. "Effects of Ammonium Perchlorate on the
   Thyroid Hormone Levels of the Sprague-Dawley Rat."
   Master's Thesis, Wright-Patterson Air Force Base, Air
   Force Institute of Technology.
King, W.R., Jr., and C.S. Garner. 1954. "Kinetics of the
   oxidation  of vanadium(II) and vanadium(m) ions by
   perchlorate ion." J. Phys. Chem., 58:29-33.
Kiss, L., M.L. Varsanyi, and E. Dudas. 1973. Acta Chim.
   (Budapest), 79:13.
Korenkov,  V.N.,  V. Ivanovich, S.I. Kuznetsov, and J.V.
   Vorenov. 1976. U.S. Patent 3,943,055.
Krokhin, O.V.,  H. Hoshino, O.A. Shpigun,  and T.
   Yotsuyanagi. 1997.  "Use of cationic polymers for the
   simultaneous determination of inorganic anions and metal-
   4-(2-pyridylazo) resorcinolato chelates in kinetic differ-
   entiation-mode capillary electrophoresis." J. Chromatogr.
   A, 776:329-336.
Liu, B.-Y., P.A. Wagner, and I.E. Barley. 1984. "Reduction
   of perchlorate ion by  (N-(hydroxyethyl)ethylene-
   diaminetriacetato)aquotitanium(III)." Inorg. Chem.,
   23:3418-3420.
Ma, J., K. Anderson, R. Shirokov, R. Levis, A. Gonzalez, M.
   Karhanek, M.M. Hosey, G. Meissner, and E. Rios. 1993.
   J. Gen. Physiol., 702:423-448.
Malmqvist, A., T. Welander, E. Moore, A.  Ternstrom, G.
   Molin, and I.-M. Stenstrom. 1994. Sys. Appl. Microbiol,
   77:58-64.
Manahan, S.E., M. Smith, D. Alexander, and R. Hamilton.
   1970. U.S. Clearinghouse Federal Scientific and Techni-
   cal Information, PB Rep., No. 192809, 70.
Masuda, Y.,  J. Liu, and E.  Sekido. 1991.  /. Electroanal.
   Chem. Interfacial Electrochem., 373:95-107.
Matei, F.,  G. lonescu, and  A.  Duca. 1986. Chem. Anal.
   (Warsaw), 37:351-60.
Mathieu, J.B., and D. Landolt. 1978. /. Electrochem. Soc.,
   725:1044.
Mathieu, J.B., HJ. Mathieu, and D. Landolt. 1978. J.
   Electrochem. Soc., 725:1039.
Maurino, V., and C. Minero. 1997. "Cyanuric acid-based
   eluent for suppressed anion chromatography." Anal.
   Chem., 69:3333-3338.
Mouhandess, M.T., F. Chassagneux, and C.R. Vittori. 1980.
   Acad. Sci. (Paris) CR Ser. C, 290:417.
Nabar, G.M., and C.R. Ramachandran. 1959. Anal. Chem.,
   37:263-265.
Neuhold, C.G., K. Kalcher, X. Cai,  and G. Raber. 1996.
    "Catalytic determination of perchlorate using a  mod-
    ified carbon  paste electrode." Anal. Lett., 29:1685-
    1704.
 Nikolskii, B.P., E.A. Materova, and A.L. Grekovich.  1977.
    In: E. Pungor and  I. Buzas (Eds.), Ion-Selective  Elec-
    trodes, 2nd Symposium. Akad. Kiado, Budapest, Hun-
    gary.
Orgiassi, J., and R. Mornex. 1990. "Hyperthyroidism." In:
   M.A. Greer (Ed.), The Thyroid Gland. Raven Press, New
   York, NY. pp. 405-495.
Pathan, A.S., and A.G. Fogg. 1974. Proc. Soc. Anal. Chem.,
   77:162-163.
Payne, W.J. 1973. Bacterial. Rev., 37:409-452.
Percival, A.L., A.J. Williams, J.L. Kenyon, M.M. Grinsell,
   J.A. Airey, and J.L. Sutko. 1994. Biophys. J., 67:1834-
   1850.
Pereon,  Y., A.  Khammari,  and J. Noireaud. 1996. Acta
   Physiol. Scand., 758:287-294.
Rikken, G.B., A.G.M. Kroon, and C.G. van Ginkel. 1996.
   "Transformation of (per)chlorate into chloride by a newly
   isolated bacterium: reduction  and  dismutation." Appl.
   Microbiol. Biotechnol, 45:420-426.
Rillema, J.A., and D.L. Rowady. 1997. Proc. Soc. Exp. Biol.
   Med., 275:366-369.
Rohn, T.J., and G.C. Guilbault. 1974. Anal. Chem., 46:590-
   592.
Rothmund, V. 1909. Z. Anorg. Chem., 62:109.
Sase, R. 1998. "Perchlorate Treatment by Ion Exchange."
   Presented  at the Perchlorate Stakeholders Forum, May
   19-21,  1998, Henderson, NV.
Schilt, A.A.  1979. Perchloric Acid and Perchlorates. GFS
   Chemical Company, Columbus, OH.
Selig, W.  1977. Microchem. J., 22:1-6.
Selig, W.  1979. Mikrochim.  Acta, 2:437-445.
Selig, W.  1980a. Precipitation Titration of Perchlorate Us-
   ing New Titrants, UCRL-83153 (Rev. 1),  Lawrence
   Livermore Laboratory, University of California: Liver-
   more, CA.
Selig, W,  1980b. Talanta, 27:357-360.
Selig, W.S. 1988. Fresenius'  Z. Anal. Chem., 330:127-
   129.
Shahine, S., and N. Ismael. 1976. Mikrochim. Acta, 2:75-59.
Shahine, S., and S. Khamis. 1979. Microchem. J., 24:439-
   443.
Sharp, M. 1972. Anal. Chim. Acta, 62:385-392.
Silber, H.B., and Y. Zhang.  1991. Eur. J. Solid State Inorg.
   Chem., 28 (Suppl.):267-270.
Stanbury,  J.B., and J.B. Wyngaarden.  1952. "Effect of per-
   chlorate on the human thyroid gland." Metabolism, 1:533-
   539.
Sykut, K., J. Dumkiewicz, and R. Dumkiewicz. 1979. Ann.
   Univ. Mariae Curie-Skodowska, Sect. AA: Phys. Chem.
   (Vol. Date 1978), 33:1-16.
Takeuchi, T., E. Suzuki, and D. Ishii. 1988. "Ion chromatog-
   raphy  with micropacked alumina columns." Chromato-
   graphia, 25:480-482.
Tateda, A., J.S. Fritz, and S. Itani. 1970. Mem. Fac. Sci.,
   Kyushu Univ., Ser. C., 7:147-154.
Taube, H. 1982. In: D.B. Rorabacher and J.F. Endicott (Eds.),
   Mechanistic Aspects of Inorganic Reactions. ACS Sym-
   posium Series, No. 198, p. 151.
Tsinman, A.L, L.M. Pischik, and G.L. Makovei. 1975. Sov.
   Electrochm., 77:1598.
 Tsui, D., and S. Pia.  1998. "Analytical Techniques and
   Pending Interlaboratory Study." Presented at the Perchlo-
 94
                                                                                                      Urbansky

-------
   rate Stakeholders Forum, May 19-21, 1998, Henderson,
   NV.
Tuazon, C.U. 1995. In:  G.L. Mandell, I.E. Bennett,  and
   R. Dolin (Eds.), Mandell, Douglas, and Bennett's Prin-
   ciples and Practice  of Infectious Diseases, 4th ed.
   Churchill Livinstone, New York, NY. Ch. 187 and refer-
   ences therein.
UCLA, see University of California, Los Angeles.
University of California, Los Angeles. 1997.  "Perchlorate
   Discharge."  http://www.nuc.ucla.edu/cases/protocols/
   perchlor/html.
Urbansky, E.T., B.T.  Cooper, and D.W. Margerum. 1997.
   "Disproportionation kinetics of hypoiodus acid as cata-
   lyzed and suppressed by acetic acid-acetate buffer." Inorg.
   Chem., 36(7): 1338-1344 and references therein.
U.S. Air Force. 1994. U.S. Patent 5,302,285.
Vasina, Y.S., and O.A. Petrii. 1969. Sov. Electrochem., 6:
   231.
Verpoorte, E.M.J., and D.J. Harrison. 1992. J. Electroanal
   Chem., 525:153-166.
Vosta, J., and J. Havel.  1973a. Scr.  Fac. Sci. Nat. Univ.
   Purkynianae Brun., 3, 99-107.
Vfrtes, G., and G.  Horanyi. 1974. /. Electroanal.  Chem.,
   54:445.
Vosta, J., and J. Havel.  1973b. Scr. Fac. Sci. Nat. Univ.
   Purkynianae Brun., 3:109-115.
Wallace, W., T. Ward, A. Breen, and H. Attaway. 1996.
   "Identification of an anaerobic bacterium which reduces
   perchlorate and chlorate as Wolinella succinogenes." J.
   Ind. Microbiol, 16:68-72.
Wallace, W., and H. Attaway. 1994. Abstr. Gen. Meet. Am.
   Soc. Microbiol., 421.
Wang, E., and S. Kamata. 1992. Anal. Chim. Acta, 261:399-
   404.
Weiss, J.A., and J.B. Stanbury. 1972. Anal. Chem.,44:619-620.
Welcher, F.J. 1947. Organic Analytical Reagents, Vol. 3.
   Van Nostrand, New York, NY.  pp.  138-146 and refer-
   ences therein.
Wilson, A.C. 1979. Ph.D. Dissertation, Washington State
   University.
Wilson, A.C., and K.H. Pool.  1976. Talanta, 23:387-388.
Wirt, K., M. Laikhtman, J. Rohrer, and P.E. Jackson. 1998.
   Am. Env. Lab., 10, 1 ff.
Wolff, J. 1964. Physiol. Rev., 44:45-90.
Wolff, J., and J.R. Maurey.  1963. Biochim. Biophys. Acta,
   69:58-67.
Yano, M., R. El-Hayek, and N. Ikemoto. 1995. Biochem,,
   34:12584-12589.
Perchlorate Chemistry: Implications for Analysis and Remediation
                                                   95

-------

-------