Introduction
Trace metals are known to occur in mollusc shells.
However, they occur in such low concentrations as to be
below the detection limits of most analytical methods.
The encyclopaedic work of Vinogradov (1953) on the ele-
mental chemical composition of marine organisms gave the
composition of mollusc shells as mostly calcium carbonate
with some magnesium carbonate, and the rest was collect-
ively termed 'ash'. More recently, Brooks et al. (1965),
using the spectrographic method, attempted a less macro-
scopic analysis of trace metals in molluscs. Except with
Mo, V and Cu, they were again limited by the detection
limits of their methods. They were only able to give es-
timated of the concentrations of many trace metals in
mollusc shells. The use of atomic absorption spectropho-
tometry in microanalysis offered a potentially very re-
warding analytical tool for looking at trace metals in
very low concentrations since the detection limits are
very much lower than all other methods used previously
(Pringle, 1968).
When looking at trace metals in mollusc shells with
the atomic absorption spectrophotometer, however, there
was so much matrix interference, especially at the lower
wavelengths, that the true readings from the trace metals
were completely masked in some cases. This was believed
to be due tothe carbonates in the shell. Various workers
have done work on se-water, which has relatively high le-
vels of both calcium and magnesium carbonate compared to
the concentrations of its trace metal components, by
50
passing it through an ion-exchange chelating column of
Chelex-100. They have, with remarkable success, been
able to effectively separate the trace metals from its
inorganic matrices (Riley et al. 1968).
The purpose of this paper is to introduce an as yet
imperfect technique for using the Chelex column in a si-
milar way to selectively extract certain trace metals,
concentrate the elutant from the column by as much as ten¬
fold (thereby increasing the effective detection limit of
the atomic absorption spectrophotometer by a factor of ten)
and then determining their concentrations using the atomic
absorption spectrophotometer.
The Bay Mussel, Mytilus edulis, was selected because
of its easy availability and its generally wide distribu-
tions.
50
- 3 -
I. Determination of the loss of metals from solution during
concentration by evaporation:
Materials and Methods.
Standard solutions containing Pb, Mn, Cu, Zn, Co
and Cd were prepared by serial delutions from parent
stock solutions (Harleco +7633) in the following concen-
trations: 0.5 ppm, 1.0 ppm, 5.0 ppm, 10.0 ppm. These
were all made in 4N nitric acid and were later referred
to as 0.5 ppm standard, 1.0 ppm standard, etc..
Fifty ml of a solution containing 1.0 ppm of Pb, Mn,
Cu, Zn, Co and Cd was made by diluting 5 ml of a 10.0 ppm
standard in a 50 ml volumetric flask. This i ppm solution
was then transferred to a 50 ml beaker and gently heated,
without boiling. To prevent contamination and to avoid
refluxing, a cleaned filter funnel, connected to an aspira-
tor with Tygon tubings, was inverted over the beaker. When
the solution reached a volume of slightly less than 10 ml,
it was removed from the heat, with the funnel still in-
verted over it, and allowed to cool to room temperature.
The volume of the solution in the beaker had reached a
level below 5 ml by this time. The contents of the beakei
was poured into a 5 ml volumetric flask. The sides of the
beaker were washed with 4N nitric acid and the washings
added to the 5 ml flask. More acid was then added to bring
the final volume to precisely 5 ml. This should result in
a 5 ml solution of 10.0 ppm, provided the evaporation had
not resulted in the loss of any metals from solution. The
entire procedure was repeated three times to obtain three
610
5 ml solutions of a theoretical concentration of 10.0 ppm:
Sample 1, Sample 2, Sample 3. These samples were then run
on the Perkin-Elmer 303 atomic absorption spectrophotome-
ter for the following metals: Pb, Mn, Cu and Cd.
Results and Discussions.
The results are shown in Table 1.
It is evident that the method for concentrating a sam-
ple solution ten-fold by means of evaporation under suction
had not resulted in any appreciable loss of Pb, Mn, Cu or
Cd from the solution; nor did the 4N nitric acid that was
used as a blank contain any measurable quantities of these
metals.
Thus the technique described provided an acceptable
way of concentrating a solution of trace metals down by a
factor of ten without any measurable loss of metals, thus
resulting in an effective increase of the detection limit
of the atomic absorption spectrophotometer by the same fac-
tor.
II. Collection and preparation of samples:
Materials and Methods.
Fifty-eight Mytilus edulis were collected from the in-
tertidal area of Coyote Point Recreational Area, San Mateo,
California on May 14, 1971. Fifty-five mussels were col-
lected from the wood pilings in the U.S. Coast Guard pier
at Monterey, California, on May 18, 1971. All of the mus-
sels were collected from the highest levels in their range
of distribution in both localities and they were all col -
lected from spots not more than ten feet from each other.
511
- 5 -
The mussels were cut open with a stainless steel blade
and the flesh was discarded. The shells were scrubbed with
a sponge pad and, when necessary, any barnacles and algae
were scraped off with the stainless steel blade. They were
then rinsed in tap water and oven-dried at 80°C overnight.
In all but two cases, only the left half of the mussel
shell was used.
Fifty-three half-shells from San Mateo were crushed in
a large beaker with the fingers to pieces of about 1 to 2
cm long and the dry weight of the shell fragments was de-
termined. The other five mussels from San Mateo were simi-
larly crushed and weighed, but individually. For the mus-
sels from Monterey, fifty were pooled and the remaining
five individuals were crushed and weighed in separate bea-
kers.
To dissolve the shells, first 12N hydrochloric acid
was added 5 ml at a time until there was no vigorous effev-
escence. Then 20N nitric acid was added, again 5 ml at a
time, to dissolve the remaining protein coat on the shells.
Towards the end, and only then, the mixture was warmed
slightly. A watchglass was placed over the beaker at all
times. A few drops of 33% hydrogen peroxide were then added
to oxidize any organic substances until the solution was
clear of visible particulate matter.
The solution of 53 half-shells from San Mateo was made
up to 500 ml with distilled water in a volumetric flask.
This was coded 'SM-stock'. The solution of 50 half-shells
from Monterey was made up to 250 ml, and this was coded
61.
6 -
'M-stock'. The five individual half-shells from San Ma-
teo were coded 'SM-1', 'SM-2', 'SM-3', 'SM-4' and 'SM-5'.
The five individuals half-shells from Monterey were simi-
larly coded as 'M-1', 'M-2', 'M-3', 'M-4', and 'M-5'.
The data on shell weight and the amounts of reagents
used are shown in Table 2.
III.Measurement of natural levels of Pb, Mn, Cu, Zn, Co and Cd
in shells of Mytilus edulis after passage through Chelex
column:
Materials and Methods.
Three aliquots of 25 ml each were withdrawn from SM-
stock and similarly three 25 ml aliquots were obtained from
M-stock resulting in the following sample solutions: SM-
stock 1, SM-stock 2, SM-stock 3; M-stock i, M-stock 2, and
M-stock 3. A blank was made with 10 ml of 12N hydrochloric
acid, 15 ml of 20N nitric acid and 18 drops of the 33% hy-
drogen peroxide (This corresponds to the total amount of
acids and H202 added to the M-stock aliquots and twice the
amounts added to the SM-stock aliquots.)
All the solutions (SM-stock 1 through 3, M-stock i
through 3; SM-1 through 5 and M-1 through 5) were neutra-
lized to a pH of 7.0 + 0.1 with 15N ammonium hydroxide and
then passed through a Chelex-100 column.
The chelating resin, Chelex-100, is a purified form
of Dowex A-1. A 50-100 mesh particle size was used. The
resin was purified by washing with distilled water, iN hy-
drochloric acid, more water, iN ammonium hydroxide, and
613
still more water. This procedure was repeated to give the
Chelex its final, purified, ammonium form. The Chelex in
its ammonium form was then packed into 12 mm diameter
glass tubes to give a column length of 6 cm.
The rate of flow of the neutralized samples through the
column was adjusted to be not more than one drop every two
seconds. The columns were then rinsed with 100 ml of dis-
tilled water and the 20 ml of a buffer solution of ammo-
nium acetate. This buffer was a one to one mixture of a 7.5N
ammonium hydroxide solution and a 50% aqueous solution of
glacial acetic acid. 40 ml of 4N nitric acid was then used
to elute the columns. The elutants were concentrated by
evaporation under suction as previously described so as to
obtain a final volume of 5 ml. The Perkin-Elmer 303 was
used in conjunction with a Perkin-Elmer 165 Recorder to mea-
sure the levels of the following metals in the final 5 ml
solutions: Pb, Mn, Cu, Zn, Co, Cd. The solutions were also
checked at a wavelength of 222 nm with the lead electrode
for any scattering.
Results and Discussions.
Part of theresults are shown in Table 3.
Cobalt could not be found in any of the samples, nor
in the blank. There was no scattering at all for all sam-
ples, including the blank. The following concentrations of
metals were found in the blank solution: .03 ppm Pb,2.3 ppm
Cu, 5.1 ppm Zn, and 0.7 ppm Cd. Since the blank readings
were proportionately deducted from the samples to obtain
the final adjusted levels in the dry shell, contamination
514
- 8 -
of the blank could very conceivably have resulted in aber-
rations of the copper and zinc concentrations especially.
Because of the high concentrations of Pb, Mn and Cd rela¬
tive to what was found in the blank, the values for these
metals are regarded to be dependable and accurate. However,
the values for these trace metals did not necessarily re-
present the absolute levels found in the shells since there
was no evidence to indicate that, given the conditions of
this experiment, the column had 100% retention and/or 100%
recovery.
IV. Determination of column efficiency:
A. Recovery of trace metals from a standard solution con-
taining known concentrations of Pb, Mn, Cu, Zn, Co and
Cd, but no shell material, after it has been passed
through a Chelex-100 column.
Methods.
Four 5 ml aliquots of a standard solution contain-
ing 5 ppm of all the relevant trace metals were with-
drawn and these were coded Standard 1, Standard 2, Stan-
dard 3 and Standard 4. These were neutralized to a ph
of 7.0 + 0.1 and then run through the Chelex column and
analyzed as before.
Results and Discussions.
The results are shown in Table 4(a).
The percentage recovery, obtained by comparing the
actual concentration to the expected concentration of 5
ppm, is shown for each metal in the last line of Table 5.
615
- 9
The generally consistent recoveries for the
four separate aliquots showed that the average re-
covery percentages were true indications o the co¬
lumn efficiency under the conditions of the experi-
ment with samples that did not contain shell material.
These figures disagree with the results of Riley and
Taylor (1968) and Davey et al. (1970) who reported 100%
or very close to 100% recovery for all the metals.
Riley et al. and Davey et al., however, used differ-
ent forms of the resin and under different experimen-
tal conditions (with generally higher ph for their
samples). The conditions under which this experiment
was carried out were those specified by J. Galloway
(personal communication to George Knauer) who held
that a sample pH of between 5.5 and 7.5 would not in-
terfere with column efficiency at the 100% level.
B. Recovery of trace metals from a standard solution con-
taining known concentrations of Pb, Mn, Cu, Zn, Co and
Cd, and known amounts of shell material, after it has
been passed through a Chelex-100 column.
Methods.
The remaining right halves of those three mussels
from which M-1, M-2 and M-3 were obtained were crushed
and dissolved as previously described, obtaining a
single solution of 100 ml. From this solution, three
20 ml aliquots were drawn. These were coded as Blank
1, Shell 1 and Shell 2.
Three 25 ml aliquots were then withdrawn from
the SM-stock solution and these were coded as Shell 3,
Shell 4 and Shell 5.
61
- 10 -
All the aliquots, except Blank i, were spiked with
5 ml of the same 5 ppm standard solution used earlier.
They were all neutralized, including Blank 1, and then
passed through the Chelex column and analyzed for Pb,
Mn, Cu, Zn, Co and Cd.
Results and Discussions:
The results are shown in Table 4(b).
Since Blank 1, Shell i and Shell 2 were drawn from
the same shell solution and since Blank i was not spiked
with standard solution, then Blank i was considered a va-
lid shell blank for Shell 1 and Shell 2 so as to correct
for the contributions from the shells. The amounts of
the various trace metals for the SM-stock, from which
Shell 3. Shell 4 and Shell 5 were obtained had already
been determined by three separate analyses (of SM-stock 1,
SM-stock 2 and SM-stock 3; see Table 3). The averages
for the various metals in SM-stock 1, SM-stock 2 and SM-
stock 3 were taken to be shell blank values for Shell 3,
Shell 4 and Shell 5. This average was coded as Blank 2.
Corrected values for each metal for the Shell i and Shell
2 samples were obtained by deducting the values for Blank
1 from the average values of Shell i and Shell 2; and by
deducting the values for Blank 2 from the averages of
Shell 3. Shell 4 and Shell 5. Then the corrected averages
for Shell i and Shell 2; and Shell 3, Shell 4, and Shell5
were pooled to obtain a final corrected average for all
samples with shell materials from which the percentage
recovery was calculated (see first line of Table 5).
517
O
- 11
The values do indicate that column efficiency (as
measured by the percentage recovery figures in Table 5)
was much reduced in the presence of shell materials, ex¬
cept in the case of lead. Since work done by Riley and
Taylor (1968) indicated that 100% recovery was possible
from the Chelex-100 column with large quantities of ses-
water, it is unlikely that the reduction in efficiency
was due to the carbonates present in the shell solutions.
A more likely reason is perhaps that the organic subs-
tances present in the mussel shells might have compli-
cated the chelating properties of the resin, resulting
in one or more of the following:
(a) metallic ions adhering to organic substances in the
shell samples so strongly that they were not able
to chelate with the resin;
(b) the ability of the column to retain ions was reduced
(perhaps by organic substances in the shell samples)
thus not all of the heavy metal ions were trapped
when the sample solution was passed through it.
(c) the very opposite of (b), namely, that the resin
bound so strongly withthe ions that they were not
completely eluded with the amounts of elutant used.
A possible explanation for the less than 100% re-
covery of the column in both cases (with and without the
shell material) is that at the sample pH (7.0 f 0.1),
the column did not have 100% retention capabilities.
V. Effects of sample pH on column efficiency:
Methods
Two 10 ml aliquots of a standard solution containing
518
- 12 -
10 ppm of Pb, Mn, Cu, Zn, Co and Cd, one at pH 7.0 and the
other at ph 7.8 (these were coded Sample A and Sample B
respectively), were run through the Chelex column accord-
ing to the same procedures outlined in III. The buffer
solution of ammonium acetate was used with Sample A but it
was not used in the case of Sample B.
After the columns had been eluted with 4N nitric acid,
they were eluted again with 20 ml of 2N hydrochloric acid.
The distilled water that had been passed through the columns
were kept and concentrated down; so were the 4N and 2N acid
elutants.
Results and Discussions.
The results are shown in Table 6.
Percentage recovery values can easily be obtained from
the 'Total' values in Table 6 by taking into consideration
the fact that the expected concentration should be 10 ppm
for all the metals involved.
It is obvious that the percentage recovery for Sample B
was much better than those for Sample A. The implications
of this are many:
(a) Distilled water is carrying off some copper and zinc so
perhaps its pH should be adjusted to near neutral or
slightly basic conditions (pH of distilled water used
was found to be between 5.4 and 5.8)
(b) In Sample B, where the column was exposed to a heavier
load, the 40 ml of nitric acid elutant did not seem to
be adequate for complete recovery of the trace metals
since appreciable amounts were founds in the second 2N
- 13 -
hydrochloric acid elutant.
(c) A sample pH of 7.8 was much closer to the optimum ph
for complete recovery than a pH of 7.0
VI. Determination of contamination inherent in the 15N ammonium
hydroxide:
Reagent grade ammonium hydroxide, of the same kind that was
used for neutralizing all the sample solutions, was analyzed
for Pb, Mn, Cu, Zn, Co and Cd. It was found that the 15N
ammonium hydroxide contained .4ppm Pb, .5ppm Zn, .2ppm Co
and.2ppm cadmium. Since ammonium hydroxide was extensive-
ly used in all samples which were subsequently concentrated
ten-fold, this inherent contamination by heavy metals could
be a very important source of contamination. Further-
more, the amounts of ammonium hydroxide needed varied from
sample to sample, thus adding an increased degree of variance
to the results.
Summary:
(1) Analyses of Mytilus edulis shell for the trace metals
Pb, Mn, Cu, Zn, Co and Cd using atomic absorption spectro¬
photometry is difficult because of scattering effects at
low wavelengths. To eliminate these interferences, experi-
ments were performed with the use of an ion-exchange resin
column of Chelex-100 with the result that scattering was
completely eliminated from the samples of shells.
(2) A method of concentrating a solution of trace metals by
evaporation under suction was developed. This method re-
sulted in a 100% recovery of all trace metals from a solu-
tion that had been concentrated ten times. This in effect
increased the detection limit of the atomic absorption spec-
320
-14.
trophotometer ten times.
(3) At a sample pH of 7.0, the Chelex column used had a
generally higher percentage recovery for solutions of trace
metals without shell material than for solutions with shell
material.
(4) Recovery percentage from the column for all the trace
metals involved was much higher at sample pH of 7.8 than
that at sample pH of 7.0.
(5) Reagent contamination resulted in some inaccuracies in
the values for some trace metals in the shells of Mytilus
edulis, especially copper and zinc. Nhen the concentra-
tions of some of the trace metals in the mussel shell are
so low, contamination becomes an all-important considera-
tion. Furthermore, the varying amounts of the inherently
contaminated ammonium hydroxide used in the samples added
an increased degree of variance to the results.
ACKNOWLEDGEMENT
I am deeply indebted to my advier, Dr. Donald P.
Abbott, for his unlimited enthusiasm, timely ad-
vices and much-needed encouragements throughout the
course of this work. I am also grateful to Dr. John
H. Martin and Mr. George A. Knauer for their expert
the atomic absorption spectro-
assistance in using
photometer and especially to Mr. Knauer for his un-
limited patience and overall guidance of the experi-
mental aspects of this work. Last but not least, I
would like to thank my colleague, Mr. David Graham,
for his much-needed inspirations at strategic inter-
vals during the ten-week course of this work.
This work was made possible in part by Grant GY-
8950 of the Undergraduate Research Participation Pro-
gram of the National Science Foundation.
50
ii
Literature Cited:
Brooks, R. R., and M. G. Rumsby. 1965. The biogeochemistry
of trace element uptake by some New Zealand bivalves.
Limnol. Oceabogr. 10: 521-527.
Davey, E. W., J. H. Gentile, Stanton, J. Erickson, and P.
J. Betzer. 1970. Removal of trace metals from marine
culture media. Limnol. Oceanogr. 15: 486-488.
Pringle, B. H., D. E. Hissong, E. L. Katz, and S. T. Mu-
lawka. 1968. Trace metal accumulation by estuarine
mollusks. J. Sanit. Eng. Div., 94: 355-475.
Riley, J. P., and D. Taylor. 1968. Chelating resins for
the concentration of trace elements from sea-water
and their analytical use in conjunction with atomic
absorption spectrophotometry. Anal. Chim. Acta 40:
479-485
Vinogradov, A. P. 1953. The elementary chemical composi-
tion of marine organisms. Sears Foundation Mar. Res.
Mem. 2, New Haven. 647 p.
52
iii
Table 1: Data on the shell solutions.
Reagents added
DRY WEIGHT
SAMPLE
HCI
H202
HNO3
180 drops
150 ml
100 ml
SMstock
66.272 g
NI
NI
2.626 g.
SM-1
18 drops
15 ml
10 ml
SM-2
1.369 g
NI
SM-3
4.580 g
15 ml
18 drops
10 ml
2.643 g
SM-4
15 ml
18 drops
10 ml
SM-5
1.923 g
180 drops
150 ml
100 ml
59.958 g
Mrstock
NI
NI
2.511 g
NI
M-1
NI
NI
2.613 g
NI
M-2
NI
NI
NI
2.179 g
M-3
NI
NI
NI
3.783
M-4
NI
NI
3.101
NI
M-5
* No information available.
*+ This weight represented both halves of the
shell.
5e
Table 2 : Efficiency of a ten-fold concentration
by evaporation.
Percentage Absorption
LEAD MANGANESE COPPER
CADMIUM
SOLUTION
10 ppm
51.9
44.5
21.0
75.6
Standard
51.2 45.0
75.3
Sample 1
20.6
75.3
51.2 44.7
Sample 2
21.0
51.4 44.9
Sample 3
21.2
75.0
4N nitric
acid blank
NOT DETECTABLE FOR ALL METALS
52
Table 3: Levels of some heavy metals in shell of
the Bay Mussel, Mytilus edulis
PARTS PER MILLION IN DRY SHELL
SAMPLE
COPPER
MANGANESE
CADMIUM
ZINC
LEAD
SM-stock 1
0.58
61.3
1.2
1.4
.09
.05
SM-stock 2
60.7
0.9
0.5
0.42
SM-stock 3
1.0
0.3
0.25
61.2
.01
SM- 1
.55
17.2
10.0
0.2
ND
SM- 2
ND'
22.0
.09
.03
SM-3
4.0
6.6
1.5
0.5
SM- 4
0.43
14.9
ND
.09
SM- 5
0.79
ND
.17
18.2
0.1
M-stock 1
ND
ND
2.9
0.5
0.1
ND
M-stock 2
3.9
0.6
0.3
M-stock 3
0.5
0.1
.02
3.7
0.1
M - 1
.06
0.6
0.49
0.4
0.7
M - 2
1.1
ND
.05
0.98
0.2
2.3
M - 3
ND
.09
0.4
1.0
M - 4
2.5
0.4
1.2
0.7
.13
M - 5
1.1
0.4
1.1
.08
1.6
Weighted
0.6 57.0
0.7
.08
1.0
0.5
0.1
averages:
3.3
0.2
.03
* Non-detection, after compensating for
blanks.
56
Table 4(a): Concentration of metals in samples
without shell material after being
passed through Chelex column.
concentration in parts per million
SAMPLE
Pb
Mn
Cu Zn
Co
Standard 1
2.42 3.38 4.05 3.6 3.07 3.31
Standard 2
2.08 3.48 4.35 6.4 3.35 3.63
Standard 3
2.19 3.36 4.44 3.9 3.29 3.48
Standard 4
2.57 4.40 4.40 3.91 lost 3.75
Av(all
2.40 3.91 4.34 4.27 3.24 3.61
standards
2
vii
Table 4(b): Concentrations of metals in samples of
shell material spiked with known amounts of
standard solutions, after passing through a
Chelex column.
SAMPLE
ppm for sample solution of 5 ml
SOLUTION
Pb
Mn
Cu Zn Co Cd
3.25 2.96 4.42 3.80 0.41 3.56
Shell 1
Shell 2
2.43 3.26 3.92 3.11 0.27 3.02
Av (Shell 1, Shell 2)
2.84 3.11 4.07 3.45 0.34 3.29
Blank 1
ND* 0.06 0.72 0.93 ND 0.02
Corrected average
2.84 3.05 3.35 2.52 0.34 3.27
for Shell 1, Shell 2
Shell 3
2.48 39.4 2.87 1.09 1.33 3.0
Shell 4
2.42 44.0 3.52 1.32 1.03 3.4
Shell 5
2.77 45.1 3.38 1.41 0.43 3.3
Av(Shell 3, 4 & 5)
2.56 42.8 3.26 1.27 0.93 3.23
Blank 2
0.27 40.4 0.70 0.50 ND 0.05
Corrected average
2.19 2.37 2.56 0.77 0.93 3.18
for Shell 3,4, & 5.
Corrected average for
2.52 2.71 2.96 1.65 0.64 3.23
Shell 1,2,3,4 & 5.
* Not detectable.
52.
O
viii
Table 5: Recovery percentages for metals using Chelex
column at sample pH = 7.0 + 0.1
percentage recovery
SAMPLE
Cd
Co
Zn
Cu
33.0 12.8 64.6
59.2
With shell
50.4
54.2
Without
85.4 65.8 72.2
48.0 78.2 86.8
shell
50
ix
Table 6: Levels of concentriion of heavy metals in
various effluent portions from the Chelex
column at two different pH's.
ppm in a volume of 10 ml
CONC
PORTIOI
SAMPLE
FACTOR
Pb Mn Cu Zn Co Cd
ND* ND 0.6 0.1 ND ND
Dist'd
ND ND 0.2 0.2 ND ND
water
ND ND ND ND ND ND
Buffer
1.2 1.5 1.4 2.1 1.6 1.6
Nitric
1.6
7.5 9.1 9.3 10.0 4.4 10.3
Acid
1.6
ND ND ND ND ND ND
HCI
0.4 0.7 0.6 0.7 0.7 0.8
elutant
2.5
1.2 1.5 1.4 2.1 1.6 1.6
Total
7.9 9.8 9.9 10.7 5.1 11.1
Non-detection
Total of concentrations in the nitric and
hydrochloric acid elutants.
530