DTIC ADA231763: The Octanol/Water Partition Coefficient of Methylmercuric Chloride and Methylmercuric Hydroxide in Pure Water and Salt Solutions

4 Pages·0.23 MB·English

by  Defense Technical Information Center

MS 579
Ir , nhe I, fu e und l( ,h en rv, Vol, pp. -1 19-)1, ", 0130-7268/9$1 30) (i t)
En vironmental Chemistry
Short Communication
THE OCTANOL/WATER PARTITION COEFFICIENT I',
OF METHYLMERCURIC CHLORIDE AND
METHYLMERCURIC HYDROXIDE IN
PURE WATER AND SALT SOLUTIONS "/
MICHAEL A. MAJOR* and DAVID H. ROSENBLATT -",
U.S. Army Biomedical Research and Development Laboratory,
Ft. Detrick, Frederick. Maryland 21701-5010
KAREN A. BOSTIAN
U.S. Army Medical Research Institute of Infectious Diseases, Frederick, Maryland 21701-50v
(Received 8 January 1990; Accepted 9 April 1990)
Abstract-The I-octanol/water partition coefficient (K,.) of methylmercury was determined at pH
7 in water and in aqueous solutions ranging from 0.0003 to 0.6000 mi n sodium chloride. It was also
determined over a pH range of 2 to 10 with a fixed sodium chloride concentration of 0.0045 M. In
these experiments, K,, was seen to increase with increasing chloride concentrations in the range 0.0
to 0.075 m and to decrease with increasing pH above 7. Results were similar whether low levels of
methylmercuric chloride or methylmercuric hydroxide were used as the starting materials. Partition-
ing wai not affected by changes in ionic strength.
Keywords-Octanol/.ater partition coefficient Methylmercury K.,,,
INTRODUCTION been found for the partition coefficient of methyl-
Mercury is found in many aquatic environments mercuric chloride in the diethyl ether/water system
in the methylated form. Inorganic mercury can be [7]. However, a search of the literature has failed
methylated both by microorganisms and abiotically to find K0,, values for either methylmercuric chlo-
II ].M ercury-contaminated freshwater systems usu- ride or methylmercuric hydroxide.
ally contain both niethylmercuric hydroxide and In this paper we present K,., values for methyl-
methylmercuric chloride, but in seawater the latter mercury(Il) starting as the chloride and as the hy-
predominates [2]. Due to its lower polarity, meth- droxide. Values were determined at pH 7 in
ylmercuric chloride would be predicted to have a ultrapure water and at 13 different salinities. They
higher K,... than the hydroxide [3] and also a were also determined at fixed salinity at nine dif-
greater rate of uptake by aquatic organisms [4]. ferent pH values.
The carbon-mercury bond is stable in methylmer- MATERIALS AND METHODS
cury compounds, whereas anions associated with
the metal exchange easily; thus, the relative con- I-Octanol was obta
centrations of the chloride and hydroxide forms re- ical Company (guaranteed purity 990) and frac-
flect environmental salinity [21, and addition of tionrIly redistilled. Ultrapure water was purchased
either methylmcrcuric chloride or methylmercuric from New England Reagent Laboratory, and meth-
hydroxide to buffered saline should result in very ylmercuric chloride and methylmercuric hydroxide
similar solutions. were purchased from the Alfa Chemical Division
Log K,,w values of 0.62 for mercury metal (51 ot Morton Thiokol Corporation. Aqueous solu-
and 2.26 for dimethylmercury (61 have been re- tions of the two methylmercuries were prepared
ported, In addition, a value for log K of 0.6 has at 1,000 ppm. Solutions of each of the two were
checked for purity by HPLC and diluted to 10 ppm
with ultrapure water or with sodium chloride volu-
'To whom correspondence may be addressed. tons of known concentration. The ultrapure wa-
91 2 05 096
h M. A. MAJOR Pr AL.
ter and the saline solutions that were used for all
2.
of the testing except the highest pH experiment
were prepared with the appropriate buffer at a
concentration of 0.01 m. The buffers used were: " 1
phosphate at pH 2, oxalate at pH 3, succinate at & 1.4
pH 4, acetate at pH 5, succinate at pH 6, phos- 2
phate at pH 7 and 8, and borate at pH 9 and 10. N
The ultrapure water and saline solutions for the
high-pH experiment were adjusted to pH 12 with .
NaOH and used without buffering. The effect of .
ionic strength on K,,, was tested by holding the sa- -6
linity at 0.05 m and increasing the ionic strength 0
with additions of NaCIO4. Perchlorate concentra- .2
tions were 0.01 M, 0.05 M, 0.1 M and 0.6 M.
The concentration of mercury in octanol could
not be assayed directly. To overcome this problem, .1 .2 .3 .4 .5 .6
the following procedure was used: Three aliquots Molar Concentrmtion of NaCl
of each aqueous test solution were removed for Fig. I. The octanol/water partition coefficients of
analysis, and three additional aliquots were ex- methylmercuric hydroxide, e,a nd methylmercuric chlo-
tracted with equal volumes of octanol by shaking ride, s, in pure water and in 0.075 M,0 .15 m,0 .30 Ma nd
for 20 min. The mixtures were allowed to separate 0.60 M NaCI.
for 10 h, the octanol was remove,,, and the aque-
ous portions were aniy.:ed. The concentration of
methylmercury in an octanol extract was calculated ultrapure water and increased to 1.7 ± 0.2 at chlo-
as. the difference in concentration between the orig- ride levels above 0.075 m. Subsequent experiments
inal and octanol-extracted aqueous samples. with methylmercuric hydroxide in the salinity range
To ensure that the mercuric compounds were up to 0.037 M NaC1 reveal a zone of nearly linear
partitioning as expected, and that adsorption or increase in K,,, with salinity between 0.0000 m and
other phenomena were not producing erroneous 0.0047 m NaCI. At higher levels, increases in salin-
results, the octanol fractions from these experi- ity produced smaller changes in K,, (Fig. 2).
ments were pooled and back-extracted with four Increase in pH of the system from 2 to 6 had lit-
successive additions of pure water. Analyses of
aqueous fractions and back-extracts were per-
formed on a Beckman SpectraSpan V sequential 1.1
plasma emission spectrometer. Calibration stan-
dards were prepared from stock solutions of mer- ,I
curic chloride (Fischer Scientific) in water. The
samples and standards were adjusted to 5,000 ppm C,
lithium to obscure sodium or other metal emis-
2
sions. The emissions of thesamples at 253.652 nm -
were recorded and mercury concentrations calcu-
lated by interpolation.
* 4
RESULTS .
The values of K,. for methylmercuric species "
as a function of chloride concentration are shown
in Figure 1. Pairwise comparisons between the 0.
K,.m eans, when either methylmercuric chloride l AS .03 .011 .04
or methylmercuric hydroxide was used as the Mor Oenentrtllon of 1480
source of methylmercury, show no statistically sig- Fig. 2. The octanol/water partition coerliclents of
nificant differences (p > 0.10), according to the methylmercuric hydroxide in the salinity range 0.0000 to
Student's I test. The mean K,, was 0.07 ± 0.05 In 0.0037 u NaCI,
Octanol/water partition wt methylmerctry 7
S- the methylmercuric cation. Similarly, the near-zero
K,, of methylmercury in pure water and in saline
solutions at high pH is attributed to a high propor-
",._ tion of methylmercuric hydroxide. From the be-
havior of the Kow/salinity curve, we believe that
4
1.2 the K. of methylmercury hydroxiJe is very nearly
equal to the K,,,, we observed with either species
in ultrapure water and in test solutions at pH 12
.- (0.07 ± 0.05). Conversely, the K(,,, of methylmer-
curic chloride (1.7 ± 0.2) is equal to the K,,, of the
omethylmercury equilibrium mixture at high salinity.
V ,A. \The constancy of K,,. values for methylmercu-
, ric chloride in the ionic strength range 0.0047 to
.
0.6 m may be attributed to the compound's high
0 2 3 " a r * i to it water solubility and small size.
PH The difference between the K, values for
Fig. 3. The octanol/water partition cocfficients of nchtelothryidlem seorcluutriiocn si owna sin + p1u.r6e ±w:0ta. 2te. r Aa nddiff einre nsctreo nogf
methylmercuric chloride in the pH range of 2 to 10 at a chi de s oul as effeon o
fixed salinity of O.(X)45 st NaCI. this magnitude should have some effect on biocon-
centration. Bioconcentration factors of 0.03 and
0.84 were calculated for methylmercuric hydroxide
and chloride in fish by means of the Arthur D. Little
tie effect on K,,,, but further increases caused a CHEMEST program [I11. This program is useful
sharp decrease in K,,, (Fig. 3). Samples at pH 12 in estimating environmental behavior of corn-
containing 0.00i ma nd 0.01 MN aCI had K,,s sta- pounds from their physical/chemical properties.
tistically indistinguishable from those of pure Thus, the absorption of methylmercury corn-
water. pounds by aquatic organisms should be enhanced
An additional experiment was designed to as- by increased salinity and by decreased pH. How-
sess the effect of changes in ionic strength. Raising ever, current estimations of bioconcentration were
the ionic strength of a 0.0047 Mc hloride solution established with organic compounds that accumu-
with potassium perchlorate in the ionic strength late primarily in the fat, whereas mercury can also
range of 0.01 stto 0.6 M did not affect the parti- bind to sulfur. This property may provide additional
tioning behavior, sites of accumulation of mercury on sulfhydryls in
Back-extractions of octanol pooled from these protein and prove estimates of bioconcentration
experiments confirmed that the octanol contained inaccurate.
a quar.:ity of mercury consistent with that lost from
the water and salt solutions through partitioning. REFERENCF S
DISCUSSION 1. Mitchell, B.E. and B. Richards. 1986. Levels of
chemical versus biological methylation of mercury in
The salinities chosen in this study approximate sediments. Bull. Environ. Contam. Toxicol. 36:401-
the range of chloride concentrations encountered 2.S tumm, W.a nd J3J. Morgan. 1981. Aquatic Chem.
in natural ecosystems. Altman and Dittmer 181 r?- istry, 2nd ed. John Wiley & Sons, New York, NY.
port a chloride level of 8 ppm (0.0002 M)a s a mean 3. llanoch, C. and A.J. Leo. 1979. Substltuent Con-
for North American river water; seawater i5 ap- stants for Correlation Analysis in Chemistry and Bi-
proximately 19,000 ppmn (0.54 m) in chloride [91. olog. John Wiley & Sons, New York, NY.
4. Lyman, '.,.J., W.F. Reehl and D.l. Rosenblatt.
Aquatic environments in the pH range of 6 to 8 are 1982. Handbook of Chemical Property Estimation
generally optimal for fish [101, and this is also the Methods: Environmental Behavior of Organic Corn.
range in which methylmercury exhibits Its greatest pounds. McGraw.Hill, New York, NY.
change inK w. 5. Shokhi, 0. and S.S okchl. 1985. The I.octanol/water
partition coefflcient of mercury. Bull. Chem. Soc.
Observed increases in K,,, with increasing ohio- Jpn. 58:3401-3402.
ride concentration and decreasing pH are consis- 6. Wasik, S.P. 1978. Partition of nrganoelements ino c-
tent with displacement of hydroxide by chloride on tanol/water/air systems. InF E. BDlnckman and
M. A. MAJOR i:r AL.
J.M . Bellama, eds.. Organotnetals and Organontet- 9. CRC Press. CRC Handbook oj Chemistrv and Phys-
alloids. ACS Symposium Series, Vol. 82, American ics. 1985. CRC Press, Boca Raton. Ff.,
Chemical Society, Washington, DC, pp. 314-326. 10. Connell, I).W. and €.J. Miller. 1984, Chemistry and
7. Corner, E.S. and B. Sparrow. 1957. Bioconcentra- Ecotoxicology of Pollution, John Wiley & Sons, New
tion of mercury and methylmercury in marine fish. York, NY.
J. Mar. Biol. Assoc. U.K. 36:459-469. II. CftEMEST: Computer program for estimation of
8. AIlman, P. and l).S. DItnmer. 1966. Environmental physical and environmental properties of organic
Biology. Federation of American Socicties for Exper- compounds. 1985. Arthur ). Little. Cambridge,
imental Biology, Bethesda, MD. MA.
Acoession For
NTIS GRA&I
DTIC TAB
Unannounced
Justification
By
Distrlbution/_
-AvailabilityC odes
IAvai l and/or
Dist I Special