4672 J. Agric. Food Chem., Vol. 48, No. 10, 2000
Schwartz et al.
Beckman model 6000 or 6500) using either Fisher Scintisafe
or Scintisafe Gel scintillation cocktail with a background value
determined by assaying an equivalent volume of cocktail in
absence of sample. Fisher Scintisafe Gel (10 mL) scintillation
cocktail was used as part of the background for TLC quanti-
tation. Aliquots of the cyclohexylamine volatile trap solutions
were mixed with 5 mL of Fisher Scintisafe scintillation
cocktail. Aliquots of the potassium hydroxide volatile trap
solutions were mixed with 10 mL of Packard Hionic scintil-
lation cocktail. Counting efficiencies were determined by
external standardization. Limits of quantitation and detection
were established by statistical methods of radioactivity count-
ing.
The radiolabel material balance was determined for each
sample on the day of harvest. The aqueous fractions and
volatile traps were radioassayed by LSC. The total radio-
chemical balance for each sample is equivalent to the sum of
the percent total dose for each of the aqueous and the combined
volatile fractions.
F igu r e 1. Representative Volatile Trapping System: CHA
) cyclohexylamine traps, KOH ) potassium hydroxide traps
The derivatization traps were either a 1% or a 10% cyclo-
hexamine solution in absolute ethanol. In this solution, the
volatile photolysis products, isocyanic acid and carbonyl
sulfide, are derivatized and trapped as the cyclohexyl thiourea
and urea derivatives. The 6 and 12 h and 1-3 day samples
were purged into 1% solutions of the cyclohexylamine. The 6
and 12 h and 1 day samples were purged into two cyclohexy-
lamine traps in series followed by the two potassium hydroxide
traps. A third cyclohexylamine trap was added in line for
purging the headspace of the 2 day samples. A fourth cyclo-
hexylamine trap was added in line for purging the headspace
of the 3 day samples. On the fifth day, the concentration of
the cyclohexylamine traps was increased to 10% in absolute
ethanol, and a fifth cyclohexylamine trap was added to the
series of traps. From days 7 through 30, four traps with 10%
cyclohexylamine in absolute ethanol were followed by two
KOH traps in series for the trapping system. The additional
traps and increasing the cyclohexylamine concentration en-
sured maintenance of material balance above 90%.
Between the nitrogen gas line and the sample was a sterile
Bactivent filter (0.1 µm pore size, 37 mm diameter, Gelman
Sciences) to ensure the incoming nitrogen was sterile. Filters
were replaced every 7 days or less just prior to collecting
volatiles. The cyclohexylamine traps consisted of a 40 mL vial
containing 20 mL of the 1% or 10% v/v absolute ethanol
solution. The KOH traps consisted of two 40 mL vials contain-
ing 20 mL of a 10% w/v aqueous KOH solution. Each volatile
trap was sealed with an open-top screw cap and a Teflon (Tuf-
Bond Disks, Pierce Chemical Co.)-coated septum. These vola-
tile traps were connected by Peek (poly ether ether ketone,
Alltech Associates, Inc.) 0.02” ID tubing from the headspace
of the first trap into the solution of the next consecutive trap.
The 18 gauge needle was inserted through the septum of the
final KOH trap to provide an outlet for the positive flow of
nitrogen through the trapping system.
Th ia m eth oxa m Ap p lica tion . An aliquot of the 14C-[thia-
zolyl]-thiamethoxam in acetonitrile was added to the pH 5,
0
.01 M sodium acetate buffer to prepare a 9.41 ppm dose
solution. The dose rate was calculated based on nine radioas-
says at pre-, mid-, and postdose. The dose solution was filtered
through a sterile 0.2 µm disposable filtering unit for steriliza-
tion. The dosing area was sterilized by spray with an ethanol/
water (70:30) solution. All sample vials, glassware, and any
utensils needed for dosing were autoclaved prior to use. The
dose solution (10 mL) was aseptically pipetted into each sample
vial. Aseptic conditions were monitored at selected times
throughout the experiment by inoculation with treated sur-
rogate samples onto plate count agar (Difco) Petri dishes.
Sa m p le In cu ba tion . Irradiated sample vial caps were
sealed with a glue gun (All Purpose Hot Melt Glue Stix, Arrow
Fastener Company, Inc.) and wrapped with Parafilm to ensure
a tight seal. The sample vials were placed on their sides in a
water bath to allow maximum exposure to the artificial light
source. A recirculating pump to maintain the temperature at
2
5 ( 1 °C fed the water bath. The temperature of the water
bath was monitored by an Omega thermocouple, an Omega
data logger, and computer support. The Omega thermocouple
was inserted into a surrogate vial containing 10 mL of sodium
acetate buffer. The samples were irradiated for 12 h per day,
for up to 30 days.
Dark control (nonirradiated) samples were sealed in the
same manner as the irradiated set and wrapped in aluminum
foil. The samples were placed in an aluminum foil lined box
to ensure no light reached the samples. The nonirradiated
samples were incubated up to 30 days in the Environmental
Specialties Incubation Chamber (model 9-19 TC/GC). The
temperature and humidity were monitored with a Honeywell
hydrothermograph, and the temperature was also monitored
by an Omega thermocouple, an Omega data logger, and
computer support.
Ch r om a togr a p h ic Meth od s. The aqueous fractions from
both the irradiated and nonirradiated set of samples were
assayed by TLC and LSC on the day of harvest. Aliquots (25
µL) of each sample were applied to the origin of two TLC plates
Sa m p le Ha r vest. Duplicate irradiated and nonirradiated
samples were harvested at 0, 6, and 12 h and on days 1, 2, 3,
(
2
0.25 mm thick silica gel, 60F fluorescence indicator, 254 λ,
0 cm × 20 cm, Merck). One plate was prepared for quanti-
5
, 7, 14, 21, and 30. The duplicate irradiated and nonirradiated
tation, and the second plate was prepared for cochromatog-
raphy with reference standards. An aliquot (2-3 µL) of
reference standard solution was applied to the origin and the
margins of the second TLC plate. Each plate was developed
in dichloromethane/methanol (90:10, v/v) (system 1) and
chloroform/methanol/ammonium hydroxide/water (80:30:4:2,
samples were purged for approximately 20 min postharvest
to collect volatiles. Following volatile collection, the volumes
of each sample were measured and each sample was radio-
assayed. The volumes of the volatile traps were measured and
each volatile trap was radioassayed. The radiochemical balance
was calculated. The volumes were measured by drawing the
entire sample into a sterile 10 mL disposable glass serological
pipet. Aliquots (1 mL) of each sample were transferred to a 7
mL vial in order to determine the pH of each sample. Samples
were stored below -2 °C when not in use.
Vola tile Collection . Zero hour samples were not purged
for volatile collection. All other samples were purged im-
mediately after harvest using the purge system shown in
Figure 1. The purge system consisted of a gang valve with four
ports, the sample vial, an empty trap, two to five cyclohexyl-
amine traps, and two KOH traps (10% aqueous). Nitrogen was
sparged through the headspace of the sample and into the trap
solutions. The flow rate ranged from 115 to 175 mL/min.
v/v/v/v) (system 2). The R
f
values of thiamethoxam were 0.31
in system 1 and 0.46 in system 2. Radioactive zones were
visualized with a bioimaging analyzer, BAS 2000 (Fuji, Inc.).
The mass spectral analysis was performed on a Finnigan TSQ-
7
000 LC/MS in the infused electrospray positive ionization
mode. The mobile phase started at 95% water with a linear
gradient to 100% MeOH (0.1% HCOOH) at a flow rate of 0.4
mL/min in a 2 × 150 mm Inertsil C8 column.
RESULTS AND DISCUSSION
Ma ter ia l Ba la n ce. Tables 1 and 2 contain the
irradiated and nonirradiated radiolabel material bal-
Schwartz
