Document ZJRG6mD5oDexkjnb85GN9QK4J
The Use uml KflVcl of Mixed Slandards on the
Quaiilitiition of I'olyehloriiKited Uiphcii)Is
)iy F. I.m: }Ii:i:zjioi.u and Vimcima F. Srom VocifiC A'i\hfty I'roillicis Tccllllnlo^v Cvr.lcr A'OAA, ftiitimmi Mttntti' Ai.Wt/vy I'rtnlttc( Center k U.S. I>cpartmrnl oj Commerce ^. Seattle, ICtu-htni'jon
Polychlorinated biphenyl# (PCB's) have been recognized and well documented as global pollutants. An excellent euramary of the distribution of PCD'o in the global ecosystem has been pre pared by RIOMROUGll (1968), and a bibliography covering PCD Investigations irora lfiBl to 1971 has been compiled by QUINHY (1977 for the Oak Ridge National Laboratory.
Despite this history of Investigations, the analytical quantitation of PCB's la still largely a matter of approximation. Css chromatography, the moot common analytical procedure for these compoundn, is hampered by the lock of primary standatds of Individual rCD components. PCD's have most often been quanti
tated by measuring peak heights of prominent GLC peaks, as lUSBimOUCH (1969), Z1TK0 (1971a), and HANSEN (1971), or by mea suring total Area of all PCS peaks as ARMOUR and BURKE (1970) have; and relating these to s suitable standard PCD mixture, sue)! as Aroclor 1254 or 1260. This method may work In some cases ns it did for SKUENTNY (1971), but It has eeveYe short
comings .
If the material being analyzed for PCB's contained only one particular Aroclor, ouch as 1254, and If the chromatogram of the
auple matched that of Aroclor 1254, then measurement of peak heightn or p<ok areas would be satisfactory in the quantitation procedure, llnfoi tnnately, the samples we have been analyzing (fluh and Huh products) often contain widely varying mixtures of PCB's, and unless the chromatograms natch chose of n readily available standard, such no one of the Aroclors, an accurate quantitation of the chromatograms Is very difficult. Even In cases where a specific Aroclor Is fed to an experimental animal, KOFMAN (1969) and MUU1ERN (1971) found that the profiles of the PCD'n retained often do not match those of the original Aroclor.
This in Itself, would not be a problem except that, as re ported by 7.7TKO (1971b), and ns substantiated by us, the CLC re*, sponse Is not constant for the variously chlorinated Isomers found in n Pen mixture. As a result, the vsluee found when ana lyzing for Ten's can vary by a considerable amount depending on which Aroclor, or mixture of Aroclors, io used as a otondard.
The purpose of thin paper Is to (a) Illustrate the problem created by samples containing mixtures of PCB's not readily com parable to a particular Aroclor; (b) to present aon* figures to illustrate, in numerical terms, live effect that the choice of standards has on the final result; and (c) to offer ns a solution to the problem the use of standards ot mixed Arocloro. For this reason, a detailed section on Standards has been included in the following section.
10
ll'ill.titi id !.a irxnni.nt.O
A 1 utii .dn,. ,
V I, !*>, .Vi. I '.) IV..1 hi Springfr Vfil., .Sr. Y'lit Jin.
j: v.
Extr<
Florlf P.lutl* than t
PCD fit
1 sltici a sit; snhydi but tl only 1 of 2% sepatr
Chrrwj
> mi a 107. D( rnh ( i/as Unit. 50-60
(Juan rJ
1 corpst end fit Achiet an r.:v. peak s for th tJculs d l f f r Aracl.
Stonge
A a t amis of the 1260 t Lure w he co".
*Kerr v
**Prep C
MQNS 082786
METHODS
Extractloo and Cleanup
Sample# were extracted Into hexane and cleaned up on a Plorlnll (PR grade)* colonm, as reported by REINERT (1970). Elution from Florisil with neat hexane effected better cleanup then that achieved with the nixed solvent described by kelnert.
FOB Separation
FCB'e were separated froa DDT and its analogs by means of a allien gel column according to the method of REltfERT (1971) with a alight modification* Rclnert recommended using completely anhydrous silica gel activated at 2C0*C for 8 houro prior to use, but treating the allies gel this way resulted In the recovery of only 30-40Z of the PCB*s. Activation for 17 hours and addition of 2X HjO by weight to the activated allies gel gave 65-1007. separation of rCB'a and pesticides.
Chromatography Conditions
A Wllkena 60l)D go* chromatograph with a tritium detector and a 5'x 1/0 sllanlzod glass column containing equal parts of
10Z DC-200 on 80/100 mesh Cas-Chron Q** and 157. QF-1 on 80/100 meeh Cas-Chroti Q** was used for quantitation. Column tcirperature was maintained at 195*C by a Vartan Aerograph 328A Isothermal Unit. Injector temperature was 240*C and the N2 gas flow 50-C0 lol/win.
Quontlfntlon
The choice of appropriare mixed standard vnt made by visual
comparison of gas chrorcatograras of the extract from the ssnrple
and of varying proportions of PCB standards. Quantitation us
achieved by cutting out the chromfitoj/.mma end weighing them on
on analytical balance;. Such an Integrated measurement of total
peak aren offers the best possibility at present ot compensating
for the differences between e sample containing most of a par
ticular Aroclor (some components having bean either lost or
differentially metabolized) and a standard of that particular
Aroclor.
.
Standards
A detailed description may prove useful. Once separate standards of Aroclors 1254 and 1260 had been prepared, mixtures of the two were prepared at increments of about 10Z In Aroclor 1260 content. The total (1254 + 1260) concentration for each mix ture was kept approximately equal so that the chromatograms could be compared easily and an assessment could be made of the changes
Mention oi trade names and specific companies does not imply endorsement
Prepared by Applied Science Laboratories, Inc., State College, PA
11
MONS 082 78 7
in total pe\k area with the changes In Aroclor composition. Figure 1 illustrate* the change in total peak area aa tha com position of tha standard change*.
Once all the mixed standards had been run on the GLC, tha chromatogram* were cut out, weighed (to give Figure 1), and placed in a reference file. This allowed the GLC operator to match tepidly the profile obtained from a sample with that of
Compoeitlon of standard Figure l. Dependence of total peak area on the relative amounts of Aroclors 1254 and 1260 in a mixed standard. (Total peak area . pnpor/nc. FCB's injected). the appropriate mixed standard. It was then relatively easy to make aevetal appropriate dilutions of the mixed standard to obtain a standard curva. The same procedure can also be used for preparing mixed standards of other Aroclors, such aa 1242 and 1254. Mixed standards of three or move Aroclors are feasible but obviously more complicated.
RESULTS AND DISCUSSION Figure 2 illustrates the problem created by samples contain ing mixtures of PCB's not readily comparable to a specific Aroclor. Chromatograms f extracts of menhaden (tlrevoortla tyrannus) meal, channel catfish (Ictaluru* punctatua). ami silky shark"(Carcharhinus falclforals) liver*** are placed to indicate #**From the laboratory of Dr. George Harvey, Wood* Hole Oceanographic Institution
12
MQNS 082788
MOMS 081789
j
i
j
i
J
1 * J i 1
i. 1 !
*
TABLE
Variation of total FCB residues (Aroclor 1254 and 1260) as the coTTpesitlcn of the quantitating standard changes
1 1260
0
10 12 30 40 60
70 80
90 100
Species
Total IT3 content (ppn)
Silky shark liver
Channel catfish*
Goatish+
Menhaden
17.0 6.7
5.4 0.94*
--
6.2
---
-- ---
6.1 6.9 4.4
...........................................
--
0.80
--
10.2 10.9* 8.5 7.5 -
3.5* 3.6 2.9 2.6 -
3.4 3.5 2.8* 2.5 2.4
-
-- -
-- 0.43
Best approxication of true value --t.'ot calculated +Edible portion: iruscle and skin
i
i
I
i i J{
|
j I
f
l
MQNS 0 8 2 7 9 0
f(
"'C- SJ 7! -J r> jc -- le* <5 < <
<cs***o*if tc--<9r. >1 :3 'r 3 na-r9*
r p* n i( >- i ^ * * ;T ? 3.** * 2
0er'<3'o*<-0r-*''
--c * _ r>
3
" r* "! ~C
n
'
n U 3 O " <r *O n ,!
9 a. n .-r c ?
A J" 3 V "
n n n > 2.
tVe approximate composition In terms of Aroclors present. Although the chromatograms are somewhat crowded on the page, it can be aeen that neither the channel catfish nor the silky shark can be quantitated as Aroclor 1234 or 1260. The profiles more cloeoly resemble those of the nixed Aroclors.
The table illustrates what happens when a sample is quanti tated by Aroclora 1254, 1260, and various nixtures of the two. The asterisks Indicate the best approximation of the true values as calculated by the standards Judged most similar in profile to the narples. If one were to plot the values with the changing compositions of the standards, the curves would resemble those of Figure 1. Both Figure 1 end the table Illustrate the fact that the selection of e standard can alter the analytical values by more than a factor of 2. In tha case of the channel catfish, It can ba saen that the standard one chooses for quantitation could determine whether or not it Is over the 5.0 ppo temporary tolerance proposed by FDA for fish.
SUMMARY
A study was made of the problems encountered when the chro matogram of a sample being analysed for PCB's via CLC cannot readily be compared to a particular Aroclor standard. The use of mixed Aroclor standards was proposed and a study cade of the effece of mixed standards on the analytical results. It was found that tha choice of a standard could Biter the values by store than a factor of 2. In addition, the particular mixture used ut a standard should be reported alpng with resulting values.
Reforences
ARMOUR, J.A. and BURKE, J.A., J. Asa. Offic. Anal. Chem. 53, 761 (1970)
NANSLit, D.J. t bJ ., Bull, Environ. Content. Toxicol.
6,113 (1971)
*
K0EM/.U, J.H. et ]., Mature 2!!*1126 (1969)
MUIJIEIUI, B.
l.f J. Ass. Olrlc. Anal. Chcm. 54, 548 (1971)
QUIir.1V, G.E. cd., Polyehloroblphenyls (PCB's) and Related Chlorophenyla: Effects on Environment, Vol. I, Bibliography 1881-1971, (1972), Oak Ridge National Laboratory, Oak Ridge, Tennessee (Oak Ridge Order No. TIRC-1, ORNL-EIS-72-20)
RE1KERT, R.E., Pontic. Monit. J, 3,233 (1970)
HEIKF.RT, R.E. and SKYPER, D, Bull. Environ. Contom, Toxicol. 6,335 (1971)
RISEBROUGH, R.W. et al.t Nature 220,1098 (1968)
RISEIIROUCH, R.tf, fL il, Bull. Environ. Contan. Toxicol. 4,192 (19>9)
SKRKNTHY, R.F., Bull. Environ. Contam. Toxicol. 6.409 (1971) 2ITKO, V., Bull. Environ. Contna. Toxicol. 6,464 (I97ln)
2ITKO, V., Bull. Environ. Contnn. Toxicol. 6,160 (1971b)
15
MQNS