Table 10. Correction factors based on two exposure conditions and
periods between analytes and the Pt. B soil as prepared for Aq-NaCl
sat'd-HS analysis.
TDCE*
CDCE
Ben
TCE
Tol
PCE
E-Ben p-Xyl
o-Xyl
Correction
factors†
1.47
1.92
2.00
2.36
3.19
3.00
4.67
5.20
7.10
Corrected
values**
6.73
6.57
4.18
5.92
3.76
3.72
2.90
2.81
3.76
Spiked
values††
8.26
8.45
5.83
9.98
6.54
9.10
6.23
6.22
6.71
* See Table 2 for full names.
† Correction factors based on analytes added to aqueous slurry 24 hours prior to
analysis.
** Corrected analytes estimates (g) from Table 5 of the aqueous treated Pt. B soil.
†† Analyte concentrations (g) in aqueous spike.
be much greater with regard to residence times
SUMMARY
and methods of analytecontaminant introduc-
tion.
This comparison of different sample prepara-
Proposed Methods 5021 and 5035 both recom-
tion methods for the analysis of VOCs in soils was
mend in-vial methods to solve the volatilization
made on samples that exclude the systematic error
and preservation issues that plague VOC deter-
associated with sample collection and handling.
minations in solid waste matrices. These two loss
Investigators who have taken precautions to elim-
mechanisms have been shown to often cause more
inate confounding effects all have concluded that
than a 90% reduction in VOCs between collection
MeOH extraction is the most robust method of
and analysis (Hewitt et al. 1995, Hewitt and
recovering VOCs from soil. To my knowledge,
Lukash 1996). Clearly, in-vial methods are neces-
papers that report contrary findings either made
sary for obtaining site-representative VOC concen-
no attempt to or failed to limit the error associat-
trations from vadose-zone samples. These two
ed with sample handling. Except in the case where
methods, along with the currently used Method
there are no matrix effects (i.e., little or no organic
5030, recommend an aqueous dispersionextrac-
carbon or clay content) such as was shown for Ott
tion method for low-level (less than 0.2 mg/kg)
sand, yields less than what can be obtained using
VOC determinations in soils and MeOH extrac-
MeOH extraction will often be obtained when
tion for high-level (more than 0.2 mg/kg) deter-
either purging a soilwater slurry or performing
minations. The water-based sample analysis pro-
an equilibrium HS analysis of a soil or soilwater
cedures for Methods 5030 and 5035 are performed
slurry. In general, soil matrix effects will increase
by dynamic purging, while static headspace is
with the analyte octanolwater partition coefficient
used for Method 5021.
and the organic carbon content of the soil. Fur-
As shown here and previously, MeOH is a
thermore, because analyteorganic carbon matrix
superior solvent in comparison to water for recov-
effects increase with increasing solution electro-
ering VOCs from soils (Hewitt et al. 1992, Askari
lyte concentrations, using a salting-out approach
et al. 1996, Minnich et al. 1996), and because of
with soils can create more problems than advan-
mass transfer, dynamic purging may be more effi-
tages.
cient than the static equilibrium methods (Hewitt
Any of the equilibrium HS methods described
et al. 1992). With regard to establishing represen-
here would be adequate for on-site screening
tative VOC values for soil samples, the discrepan-
applications, where the number of samples that
cies among sample preparation methods found
can be inexpensively processed is of greater impor-
here and elsewhere (Flores and Bellar 1993a,b) can
tance than the certainty of any single value. In
rival the error associated with volatilization and
addition, alternative solvents could be used to
biological degradation losses incurred during col-
reduce regulatory agency concerns and achieve
lection and handling.
lower detection limits than currently can be
18
Previous Page
