Document 1QBEevZG15akmO1b837QpjVrj
AR226-2729
E. I. du Pont de Kemours 4 Co. (Inc.) Polymer Products Department Experimental Station Laboratory
Lab No. ER-7^9 Page 1 of 11
Determination of Perfluorooctanoic Acid in Vater Gas Chromatographie Method
I. Applicability
This method has been developed for the determination of ammonium perfluorooctanoate (Cg-APFC) and perfluorooctanoic acid in aqueous solution, at concentrations from 0.01-10 yg/mL. The procedure described here is for the analysis of aqueous solutions from air impingers. The method has also been used for determination of Cn in blood samples and, depending on specific interferences, could be applied to other aqueous systems. A modification vhich eliminates the drying step can be used for appropriate solid samples (such as Nuclepore filters used in air sampling) and for methanol solutions.
II. Unusual Safety Considerations
The 3M Company has found in preliminary studies that Co caused birth defects
when fed to rats in a laboratory experiment. Female employees of child bearing capability should be restricted from any portion of this procedure vhich offers significant chance for exposure. Although no health problems are knovn for workers exposed to Cg, it has been found in blood samples and may be very
slowly eliminated from the body (Reference 3 ). Standard laboratory safety
T for handling toxic, emoroyotoxic, etc. materials and for corrosives should be used in working with these perfluorinated acids and their solutions.
The electron caputre detector contains a radioactive source (^Ni), and the manufacturer's instructions for safe operation must be observed.
Care should be taken to avoid contact in working with the low temperature baths used for freeze--drying.
III. Principle
Freeze-drying (lyophilization) is used to remove the water from sample aliquots and permit derivatization. Addition of methanolic HC1 to the dried residue, along with perfluorodecanoic acid (C..) internal standard, converts the acids to more volatile methyl esters. 1These are extracted from the reaction mixture into hexane solution for GC analysis. An electron capture detector (ECD) is used for sensitivity and selectivity.
IV. Fundamental Equations
C7F15COO~M+ + CH30H ~3* HC1> C.^1 COOSlf + MOH
(M+ * V , Ha4 H+ )
(acid not amenable to GC analysis)
> (non-polar, volatile ester for GC)
Page 2 of 11
P? V. Interferences
As in any GC analysis, compounds with the same retention time as
the compound of interest vill not he distinguished from it. These
can include impurities present in the reagents or other components
in the samples themselves.
Interfering peaks from solvents, derivatizing reagent, etc., are not
generally observed in this laboratory with the reagents specified
below (Note l), but each new batch should be checked and a blank
included in every analysis. A small peak at the Cg ester position
. is found in reagent blanks containing perfluorodecanoic acid internal
standard however, probably from Cg present as an impurity in that
material. This will give an intercept slightly greater than zero in
the calibration plot, but is a significant contribution to the Cg
peak area only at the lowest concentrations (ca l/3 of the total for
0.01 ppm Cg, when ca 1 ppm
is added).
In the air impinger samples examined to date, no obvious interference or
contribution from compounds other than Cg has been observed. Preliminary
analyses of FEP "in-situ" dispersing agent indicate that while one
^
component of that mixture would co-elute with Cg, it should be accompanied
by other nearby peaks.
VI. Sensitivity, Precision and Accuracy
The method has been used over the concentration range 0.01-10 ug/mL Cg.
) Adjustments in the volumes of sample or reagents could probably be maae
' to extend this somewhat in either direction, if necessary. Due to the relatively narrow linear range of the ECD, however, the calibration
curve must always cover the region of interest.
Prom the limited data available at this time, precision is estimated to
be 10$ relative, with quantitative recovery of spikes within this
` uncertainty. Accuracy will also be affected by the choice of Cg standard,
which should correspond to the fluorosurfactant composition of the
samples (Note 2).
VII,
Apparatus
Instruments and equipment used in this laboratory are specified here; equivalent apparatus can be substituted.
1 . Gas Chromatograph and Supplies
Hewlett-Packard 5830 GC with HP 18803 Electron Capture Detector, equipped for on--column injection with glass packed columns
10 ft x 2 mm id glass columns (HP configuration 5 -with ECD adapter),
packed with 10$ 0V-210 oh 70/80 mesh Chromosorb W.AW.DMCS. and
conditioned at 200C
Hamilton 7 0 M 10 pL syringe
Page 3 of 11
2 . Lyophilizer
Labconco No. 75352 bench-top freeze dryer (12-port, dry ice cooled)
Labconco No. 75^06 (150 mL) or No. 75^08 (300 mL) Fast-Freeze flasks, with Ho. 75^76 stainless steel adapters
Vacuum pump with auxiliary cold trap
McLeod gage (or electronic vacuum gage) reading in the
5-0.005 Torr range.
'
3. Thermostated Reaction Block
Pierce No. 18802 Reacti-Therm Heating Module, with Reacti-Block to
hold 2-dram vials and thermometer (Block C, No. 18803, has 12 holes
but they must be enlarged to ca 19 mm to accomodate the 2-dram vials.
Block B, No. 18802, can be used as is, but will hold only 9 vials
and provides poorer thermal contact).
it. Vials and Septum Caps
Wheaton No. 222+881* or Pierce No. 13028 2-dram (ca 7 mL) screw cap septum vials (borosilicate glass)
Pierce No. 12713 Teflon*-Silicone Septa and No. 13216 open top screw caps
5. Pipets and Dispensers
Gilson P200 Pipetman, micropipet for quantitative delivery of 50 pL aliquots
1 and 2 mL volumetric pipets for measuring sample and reagent aliquots. (Brinkman Dispensette Bottle-Top Dispensers are a great convenience for repetitive delivery of solvents.)
6. Analytical Balance
7 . Common laboratory equipment
VIII.
Reagents
1. Cg standard, FC-ll+3 ammonium perfluorooctanoate (3M) or material
in use in the area to be monitored (Note 2)
A
2. Perfluorodecanoic acid, PCR Research Chemicals
3. Methanol, Fisher HPLC (A-it52) or Fisher Certified ACS (A-l+12) (Note l)
* Reg. U.S. Pat. & Tm. Off.
Page b of 11
It. Derivatization Reagent, 3% HC1 in methanol, prepared from
Applied Science No. 18053 Instant Methanolic HC1 Kit, hut substituting Fisher HPLC or ACS Methanol for the Lipopure Methanol provided (Note 1).
The reagent should he stored in the refrigerator, and can
he kept for ca 1 month.
5 Hexane, Phillips Spectrograde or Applied Science Lipopure (Note 1)
6. Water, distilled or deionized
'
7. Sodium hydroxide, 0.05 M aqueous solution
IX. Procedure
A. Preparation of Standard Solutions
Only final solution concentrations are given here without detailed preparation procedures, since amounts needed, concentrations of interest, etc., will vary between laboratories. Preparations should he such that concentrations can be calculated to three significant figures.
1. Perfluorodecanoic acid in methanol: ca 20 yg/mL
2. Ammonium perfluorooctanoate in water: At least four solutions
should he prepared which cover the concentration range of
interest. For the example in the chromatograms and calculations
below (0.1 -1.0 ppm), standard solutions were prepared containing
0.1, 0.25, 0.5, 0.75, and 1.0 yg/mL FC-1^3.
B. Preparation of Samples and Standards for Analysis (Note 3)
For each sample or standard solution, pipet a 1 mL aliquot into a 2-dram vial. Prepare a blank similarly, using distilled water. Add 50 yL 0.05 M NaOH to each and mix gently, keeping the solution in the bottom of the vial to avoid loss to the cap, sides, etc..
C. Freeze-Drying
1. A schematic diagram of the feeeze-drying apparatus is shown in Figure 1. Each of the 12 ports has a separate valve, so that flasks can he attached or removed while the system is under vacuum. The apparatus should he pumped down to an acceptable vacuum (Note ^ ) and the cold traps filled before attaching the drying flasks'.
Page 5 of 11
2. To prevent "bumping, the samples must be frozen before they ^
are placed under vacuum. To reduce the chance of contamination
or sample loss, cover each vial with a light filter beforehand,
a small piece of Kimwipe held in place with a rubber band works
well. Freeze the solution in each (Note 5)
keep the vials
at dry ice temperature until all of them are ready.
3. Place up to four vials in each drying flask and attach to the drying chamber. The pressure should drop below 0.5 Torr again within a few minutes if no leaks are present, and the samples should remain frozen as sublimation takes place. Allow them to dry for ca ^ hours, or until no ice is left in any of the
vials.
D. Derivatization
1 To the dried residue in each vial, add 1 mL methanolic HC1 ' derivatization reagent and 50 yL C standard solution (ca 20 yg/mL in methanol). Close with a septum screw cap and shake well.
2. Thermostat for 1 hour at 65C.
(At this point, the samples can be cooled to room temperature and left overnight to analyze the next day.)
3. Cool to room temperature, add 1 mL each hexane and distilled
water, and stiake f*or ca 2 siinutes.
U After the phases separate, the upper hexane layer can be " sampled directly for GC analysis. The solutions are stable in this form for at least several days, although hexane will gradually evaporate after the septum has been pierced.
GC Analysis
1. GC Conditions -- Instrument and column as above.
Temperatures: Injection port Detector Column oven
200C 325C 100C, isothermal
Carrier Gas:
90% argon/10/S methane (or nitrogen if
recommended for the instrument)
flow ca 30 mL/min
Injection Volume:
Sensitivity:
Him Time:
2 yL
attenuation as needed to keep peak height measurable,'2 -- 2 on HP 5830.
11-20 minutes, depending on sensitivity (Note 6)
Page 6 of 11
After installing the coluicn and establishing the conditions above,
check the ECD base frequency and noise level according to the
manufacturer's instructions, to insure that they are stable and
vithin acceptable limits.
_
2. Each solution should be run twice (or duplicates run once each),
bracketing samples with standards of similar concentration
in the second series of injections.
'
Representative chromatograms are shown in Figure 2 for blanks, standards and samples. Either peak height or peak area can be measured for quantitation.
X. Calculations (Note 7)
1. Normalized Cg Peak Values
A tabulation of peak heights for analysis of a series of five standards and two samples (each prepared in duplicate) is shown in Figure 3.
To correct for any variations in injection volume, reagent volumes, etc., the raw Cg peak heights (or areas) are first normalized relative to the internal standard average
C^0 average
CD B
(corrected)
=
C8D
x
--- r------C1Q
(Calculation of the C1Q average also provides a measure of the precision in sample preparation and GC analysis).
2 . Calibration Plot
A calibration curve obtained by plotting corrected Cg peak height vs. concentration of the standards is also shown in Figure 3.
For non-linear plots such as this, a smooth curve should be drawn through the data points. In some cases, a straight line can be fit through several points, and a linear least squares calculation made for the slope and intercept in that region.
3. Calculation of Sample Concentrations
Using the corrected Cg peak height's (or areas) for the samples, corresponding concentrations can be read directly from the calibration plot as shown in Figure 3- Average the values from duplicates for each sample.
Page 7 of 11
XI. Notes
1. The solvents specified in Section Till were found preferable to
several tried for this analysis> but others may also be satisfactory.
Even with the same reagents, the appearance of the blank may depend
on the condition of the detector. With the ECD used here,
contamination results in a significant decrease in selectivity
and increased reagent interference.
.-
2. The PC-1^3 samples examined to date appear to be a mixture of
ca 70-75$ straight chain material with several secondary components
(including branched isomers) from 1-9$. (Determined by NMR and
capillary GC with FID peak area). Under the GC conditions used
here, this results in a group of poorly resolved peaks at the Co
ester position, as seen in Figure 2. In contrast, commercial
perfluoro-n-octanoic acid (PCR) and some other ammonium salts
contain^95$ straight chain Cg, and appear more nearly as a
single peak in the analysis. The response factor determined
relative to the C^q internal standard also differs for the two
types of material, so it is important to choose the standard
most similar to the samples to be analyzed. All of the aqueous
samples examined thus far closely resemble FC-1^3 standards, but
for chromatograms differing greatly in appearance it may be more
appropriate to substitute a straight-chain standard.
3 At least for initial analyses using this procedure, duplicates should be prepared as a check on precision.
U . For this freeze-drying application pressures less than 0.5 Torr
seem to be satisfactory, although 0.025-0.05 may be achieved.
Readings will also depend on the type of gage used, since the
McLeod gage does not read the pressure of water vapor as electronic
gages d o .
5. A small dry ice/acetone bath or the center cold well of the drying chamber can be used for rapid freezing of the solutions in the bottom of the vials. As noted in Section II, care must be taken to avoid contact with the cold bath.
6. As seen in the chromatograms in Figure 2, some small, broad peaks
appear late in the run. (These appear to come from the Cin
standard, and might be eliminated in purified material), when
running at high sensitivity for low Cg concentrations, these
can interfere with the following run and must be allowed to elute
before the next injection.
,
4
Page 8 of 11
7. From the calibration plot in Figure 3 , it can be seen that the
ECD response is not linear with concentration over this range. Linear plots may be obtained over other ranges, but still with concentration--dependent response factors. The calculations described here thus make use of a calibration curve, where the internal standard is used to normalize raw peak values before plotting. _
Calculations could be done by a general internal standard
procedure instead for an ECD which -shows a more linear response
. and constant relative response factors. .{A calibration plot of
this sort has been observed with a Varian 3700 GC under similar
conditions).
.
XII. References
1. Research Notebooks E20029, E22306, E22k2b (S. Stafford)
2. J. Belisle and D. F. Hagen, "A Method for Determination of
Perfluorooctanoic Acid in Blood and other Biological Samples" ,
Analytical Biochemistry 101, 369 (1980)
3- F. A. Ubel, S. D. Sorenson, & D. E. Roach, "Health Status of
Plant Workers Exposed to Fluorochemicals - A Preliminary Report"
American Industrial Hygiene Association Journal 111, 58U (1980)
Prepared by S. S. Stafford, 8/8/80 Revised by S. S. Stafford V 3 / 8 1
Approved by L. J. Papa
FIGURE 1 - APPARATUS FOR FREEZE-DRYING SAMPLES
Page 9 of 1 1
(b) Vacuum Port, Drying Flask, and Samples
/*Vi Valve- n (vacuum or YSnt)
> Vacuum (Drying Chamber)
Vacuum Port Adapter (Vacuum port to flask)
2-Pram Vials, tissue over top
Drying Flask - Rubber top fits over glass bottom for a vacuum seal
Frozen Sample (lmL aliquot)
FIGURE 2 - CHROMATOGRAMS FOR CD DETERMINATION *
__________________________
o
rags j.u OI j_L
.. 7 GC conditions as on p. 5* Attenuation 2 `, chart speed 1 cm/min. Retention time (minutes) indicated "by each peak.
*The results shown here are for solutions prepared using 2 mL each derivatizing
reagent, hexane, and water, rather than 1 xnl as in the method. Relative peak heights
will not he affected, hut attenuation is 2x lower. .
. _
FIGURE 3 - CALCULATION OF CORRECTED PEAK HEIGHTS AND CALIBRATION`PLOT *
(a) Peak Height values for runs of ten standard and four sample solutions (duplicate preparations at each concentration)
(b) Calibration Plot, Corrected C peak height vs. concentration
Spin.*-!
1 "hJL - 6-1
e-j. " C-l
C '.i O ) " O'! - -1 M
*#
|?fHt4 jjj fIatfnliOIJC.ihM. , } C<i J Cu
"
K-tJ * (co/tr*fffifcoitTJe
101
I*
7- n if
isr '
0 13) itv'. i \ V
i
u .n t f i 10?, X ip l.i 10
\*\ \0 b lO li io7
o S o S n W \0> i* k ) fe n r 1Qo 7* A
o J \ 3 \0 *|4 ,DI S 4 1 4 H . 2 4 4 .J . lP l.4 4 4 .
o.ioJ
It
ti iJ.
ai a^r
iC>i |0 k
a h '1) at>.7
80-4 v'Sk-1 8o- 4 80-jS7-1 80-4^ $7 'i
Of-MO o .4 4 r o,2 tD o.JoD
i\?
70 10
7o.o 10 7 0 -7
ris> M S
ioX? i l f i 31*7
ffsVCs .
Calculated concentrations for duplicate sample preparations read from the curve
#The results shown here are for solutions prepared using
2 mL each derivatizing reagent, hexane, and water, rather <jjj
than 1 mL as in the method. Relative peak heights will
not be affected, but attenuation is 2x lower.
H
h
o
