Figure 8. Continuous sampling of headspace vapor above moist Raritan Arsenal soil for TNT by alternating two polyacrylate SPME fibers at 20-minute intervals at 23C.

Figure 7. Continuous sampling of TNT vapor by alternating two polyacrylate SPME fibers at 20-minute intervals.

Figure 9. Mass of 2,4-DNT vapor sorbed on a polyacrylate SPME fiber at 23C.

SR99_21
Page Navigation
10

Total Sampling Time (min)

Figure 8. Continuous sampling of headspace vapor above moist Raritan Arsenal

soil for TNT by alternating two polyacrylate SPME fibers at 20-minute intervals

at 23C.

with a soil from Raritan Arsenal that was field

volume sampled (mL) = mass collected (g)

contaminated with a low concentration of TNT

vapor concentration (g/mL).

(about 1.4 mg/kg). A 3.02-g portion of the soil

(1)

was moistened with water (5% water, on a dry

weight basis) and allowed to equilibrate for 3

Because the vapor pressure of TNT is low and

days in a 40-mL amber vial. The headspace in the

TNT is known to be very sorptive to surfaces, try-

ing to establish a known vapor concentration by

vial was then sampled continuously for 200 min-

injection of a given mass of TNT into a known

utes using two polyacrylate fibers as described

volume of air is impractical. On the other hand,

above. The results are shown in Figure 8. Using

the vapor pressure of TNT as a function of tem-

the best fit linear equation, we find that the mass

perature has been measured by a number of

collected in the first 20 minutes was 20.4 pg, while

investigators (Pella 1977, Leggett et al. 1977, etc.).

the mass recovered in the second 20-minute sam-

Results from the various investigators vary some-

pling period was 19.8 pg. The reduction from the

what, but we believe the values published by Pella

first to the second sampling periods amounts to

(1977) are the most reliable. Pella used a liquid

only about 3% and, hence, the assumption that

solvent to trap the equilibrium vapor from a

the vapor concentration is constant over the sam-

pling period when the TNT is sorbed on soil par-

known volume of gas, minimizing sorption

ticles is valid.

effects, and reducing the problem to a chromato-

These results demonstrate that the approach of

graphic analysis against standard solutions in the

sampling the headspace using SPME fibers does

same matrix. This type of experiment should be

accurate and precise to a few percent.

not substantially reduce the vapor concentration

The vapor pressure of TNT (p) at 23C accord-

of TNT during a 20-minute exposure period.

ing to Pella is 6.71 106 torr ([log10 (p in torr) =

Since the mass of TNT collected with time is lin-

(12.31 0.34) (5175 105) T, where T is the abso-

ear, probably controlled by the diffusion rate of

lute temperature in kelvins]). This converts to a

TNT away from the surface of the fiber, it may be

vapor concentration of 8.25 108 g/L. When we

possible to calibrate the volume of vapor sampled

exposed a polyacrylate SPME fiber to the equilib-

as a function of exposure time for a given chemi-

rium vapor above military-grade TNT (Picatinny

cal. One way this could be done is to sample a

Arsenal, 1966 vintage) and determined the mass

headspace volume where the concentration of

of TNT sorbed by the fiber using GC-ECD in a

TNT in the vapor was known, and from mass cal-

20-minute exposure period, we collected 3.8 109

ibration of the TNT collected, the volume sam-

g of TNT. Computing the volume sampled from

pled as a function of time can be calculated from

To Contents

Index

Privacy Statement
Press Release
Contact