lives of these nitroaromatics, the concentrations ob-

day. Beyond the first day, the rate of TNT loss is sig-

tained were plotted as the ln(*C*/*C*0) versus time (*t*),

nificantly slower. Because the concentration of TNT

where *C *is the concentration at time *t *and *C*0 is the

drops below 50% before the rate of loss slows, the first

initial concentration. For a first order rate process, a

half-life for TNT at room temperature was estimated

linear relationship should be obtained. If a process is

using the slope of the data plotted from days 0 through

first order, then the half-life can be determined easily

1. Additional half-lives will be estimated using the slope

using the simple rate equation:

of the data plotted from days 1 through 30. A similar

multi-stage process is indicated for TNT at 4C, but

ln (*C*/*C*0) = *kt*

the first stage is very short, lasting only through the

first 4 hours before the slower process takes over. The

where *k *is the rate constant equal to the slope of the

decrease in TNT concentration during this time is very

curve. The half-life is then calculated by dividing the

small, so the half-life here was estimated using the slope

natural log of 1/2 (0.693) by the rate constant. An

of the data plotted beyond the 4-hour mark. Data for

TNT at 4C, as well as for the other three key compo-

important point to note is that when the rate is first

order, the half-life is independent of the starting con-

nents at all temperatures, appear to be adequately de-

centration. Clearly, for the 22C data, the loss of TNT

scribed by a single first order processes. These data are

plotted for 4 and 4C in Figure 5. Rate equations de-

does not follow simple first order kinetics (Fig. 4). The

scribing the degradation rates for the primary analytes

plot also illustrates that the rate of degradation varies

at all three temperatures were established using the

greatly among the different compounds.

curve fitting functions in SigmaPlot, a commercial plot-

The data initially indicate that the rates for the deg-

radation of TNT at room temperature and at 4C are

ting package (Table 4). The half-lives for each com-

pound were then estimated by solving each of the func-

not first order. Close examination of the data at room

tions for time (*t*) when ln (*C*/*C*0) is equal to 0.693,

temperature reveals that the decay curve for TNT is

that is, one-half of the original concentration (Table 5).

composed of two linear sections, suggesting sequential

To demonstrate that the rate of degradation also de-

first order processes. Although the nature of the two

pended on soil type, we compiled the data collected by

processes is not understood at this time, the data indi-

Grant et al. (1993, 1995) for determining the pre-

cate that the first process is very rapid, causing a 50%

extraction holding times for soil samples (Tables A4 to

drop in the concentration of TNT in approximately 1

1

TNT

0

y = 0.573x 0.046

(days 0 to 1)

1

y = 0.068x 0.618

(days 1 to 20)

2

3

0

5

10

15

20

25

Time (days)

Figure 4. Plot of ln(*C/C*0) versus time (days) for TNT at room temperature (22 2C)

illustrating a multiple stage decay process.

8