Figure 2. Linear partitioning between chamber vapor and membrane-covered water samples

Table 4. Estimates of Henry's law constants (vapor/water) for VOCs at 11C

Table 5. Ratio (i.e., high:low) of analyte concentration (mg/kg) established for the soil samples

SR98_03
Page Navigation
6

Figure 2. Linear partitioning between

chamber vapor and membrane-

covered water samples.

Chamber Vapor Concentration (mg/L)

centrations has been observed. Therefore, not only

compounds, therefore creating a condition where

the water inside the membrane covered vials was

were these five experiments with the lowest ana-

exposed to an enriched vapor concentration.

lyte concentrations more susceptible to biodegra-

Aside from this experimental artifact, the relative

dation, but the rate of biodegradation was greater

standard deviation for the mean values was

than the rate of analyte diffusion. Because of the

always ≤17%, and often <10%. Therefore, consis-

impact of biodegradation on the aromatic hydro-

tent with the principles of Henry's law, constant

carbon, only the chlorinated compounds were

(or linear) partitioning was empirically estab-

evaluated for the soil samples in experiments 15.

lished over the concentration range tested (Fig. 2).

As with the water samples, the mean of the soil

Duplicate water samples and, in some cases soil

sample duplicates was used for subsequent data

samples, were included in the experiment in order

interpretations.

to assess the precision of this experimental

Close inspection of the analyte soil concentra-

approach. Overall, the differences among dupli-

tions reported in Appendix A shows that the effect

cates was small; however, differences among the

of moisture was not always consistent. Two gen-

duplicate water samples were generally smaller

eral trends were 1) as the moisture content

than for the soil samples. The discrepancy in preci-

increased, analyte concentrations decreased for

sion between these two matrices was probably

the CR-S soil, and 2) while the CR-D soil showed

due to inconsistencies associated with packing a

increasing analyte concentrations with increasing

soil into a small vial and losses due to the biologi-

moisture content. There were no trends with

cal degradation of the aromatic compounds (i.e.,

regard to moisture in the majority of cases for the

Ben, Tol, and p-Xyl).

Wis soil.

The amount of water remaining at the end of the

The inconsistencies in trends between soil

exposure period for the three moist soil conditions,

moisture and analyte concentrations may have

and movement of water vapor onto the initially air

been caused by either poor seals between the vial

dried soils, made all of the mineral surfaces

rim and membrane, or inconsistencies in vial

hydrated. The presence of moisture and oxygen

packing. To avoid this potential experimental arti-

created conditions conducive to microbiological

fact and suppress the influence of moisture

degradation processes (Atlas 1981). Consistent

altogether, the concept of using a mean value was

with these conditions, several very low or nonde-

considered. Before taking this step, the ratio of

tectable concentrations for Ben, Tol and p-Xyl

high to low analyte concentrations for a soil in

were established for soil samples from experi-

each experiment was first evaluated. The results of

ments 15 (App. A). Although not anticipated,

this analysis, with the omission of the saturated

especially for the CR-D and Wis soils, losses of aro-

condition for the CR-S soil, showed the ratio to be

matic hydrocarbons and persistence of chlorin-

a factor of three or less (75 out of 99 cases, Table 5)

ated compounds are consistent with an earlier

for most cases. Therefore the use of mean values

study (Hewitt 1996). Furthermore, the diminish-

could be justified by accepting an uncertainty of a

factor three (3). Further justification for omitting

ing of biodegradation losses at high analyte con-

Index

Privacy Statement
Press Release
Contact