phosphate, pH 7 containing 9 g/ L NaCl) and then activated
with glutaraldehyde (3 mL, 2.5% w/ v) in PBS for 3 h at room
temperature. The resulting orangeish-pink beads were then
washed to remove excess glutaraldehyde. BOX (50 units, 1.6
mg protein) in 20 mM phosphate buffer (pH 7) was slowly
rotated with the activated beads in a capped tube overnight
at 4 °C. The resulting beads were washed and then packed
into a 3 mm microbore borosilicate glass chromatography
column from Omnifit (Toms River, NJ). The 500 mg of
immobilized bilirubin oxidase beads could be packed into
two 10 cm columns. It should be noted that immobilized
bilirubin oxidase beads are often stable for up to 1 year when
stored at 4 °C.
measured directly by CZE. In brief, a P/ ACE 5500 CE system
(Beckman, Fullerton, CA) equipped with a UV detector
module set at 214 nm was used for the detection of PCP.
Separation was achieved using a polyimide-coated fused-
silica capillary (length to detector: 40 cm; total length: 47
cm; 50 µm ID) provided by Polymicro Technologies (Phoenix,
AZ) and a separation buffer containing 20% acetonitrile and
80 mM phosphate pH 7.5. Samples were injected by pressure
for 5 s, and the separation was conducted at 25 kV with a
controlled temperature of 30 °C. PCP was detected at about
11-12 min, and the relationship between concentration and
integrated area was linear between 5 and 200 µM. For FI
biosensor analysis, the PCP partially purified extracts were
diluted appropriately (10-100 fold) with the tartrate buffer
containing TCA and treated as above with BTFAIB to oxidize
any PCP present. The PCP concentration determined by the
biosensor was then compared to the values obtained by CZE.
For soil samples containing PCP below 80 ppm, the con-
centrated methanol extract could be used directly without
dilution for CZE analysis, since the detection limit was much
higher than that of the biosensor system. For the biosensor
system, the methanol extract (1 mL) was diluted directly to
PCP Oxidation Using Bis(trifluroacetoxy)iodobenzene.
Based on the method reported by Saby and Luong (21, 22),
the oxidation of PCP was performed in 100 mM trichloroacetic
acid, pH 1.0 containing 50 mM L-tartaric acid using PCP (10
mM in methanol) and freshly prepared BTFAIB (100 mM in
methanol) stock solutions. The PCP reactions (20 mL, 25
nM-10 µM) containing 500 µM BTFAIB were carried out at
room temperature, with light protection for 1 h. At the end
of the reaction, 500 µM hydrogen peroxide was added to
neutralize any unreacted BTFAIB, followed by the addition
of zinc powder (100 mg) to reduce the 1,4-TCBQ to 1,4-
TCHQ. The use of tartrate in the buffer was necessary to
maintain the pH below 4 upon zinc addition, otherwise the
reduction reaction could not proceed to completion. The
pH was altered to about 3 by the addition of 200 µL of 8 M
NaOH. The 1,4-TCHQ sample was then passed through a
2
0 mL with the tartrate buffer containing TCA. The sample
preparation technique using the Sep-Pak tC18 cartridges
resulted in recoveries of PCP of greater than 95%.
Apparatus. The FI biosensor system consisted of a
peristaltic pump (FIA Pump 1000, Eppendorf North America,
Madison, WI) that delivered the sample, buffer, and NADH
solution at a preset flow rate (Figure 1). A 100 µL sample was
injected into a 25 mM phosphate buffer (pH 3.0, containing
10% methanol) stream by a motorized injection valve (EVA
injector, Eppendorf). The loading and injection times were
controlled from the EVA injector. The resulting sample stream
then merged at a T-joint with the NADH containing buffer
(25 mM phosphate, pH 7.5), and homogeneous mixing (final
pH of about 6.8) was effected by a mixing coil just before the
sample entered the immobilized enzyme (BOX) column. The
NADH consumption was monitored by following the ab-
sorbance at 340 nm (Waters LC spectrophotometer, model
481, Milford, MA) or fluorescence (Waters scanning fluo-
rescence detector, model 747, with excitation at 345 nm,
emission 450 nm and a gain setting of 100). Signals were
digitized using an IBM-AT type computer equipped with a
DAS-8 A/ D card and custom software to obtain both peak
height and peak area. The output of the detectors after
conversion to voltage was also recorded on a strip chart
recorder.
Optim ization of FI Biosensor. The effects of buffer
strength, pH, flow rate, NADH concentration, and column
length on the signal response were investigated. The response
time, sensitivity, and reproducibility of the 1,4-TCHQ (1,4-
TCBQ reduced by zinc powder) signal were examined as well
as the stability of the enzyme column. For fluorescence
detection the NADH concentration needed to be reoptimized,
whereas the other operating parameters remained the same.
The optimized systems were then applied to samples of PCP,
which had been oxidized by BTFAIB as described earlier.
0
.45 µm Millex-HV filter (Millipore, Bedford, MA) to remove
the zinc powder. Directly after this filter was a Sep-Pak C18
cartridge which had been prewashed with methanol and then
a 100 mM TCA solution, pH 3.0 containing 50 mM L-tartaric
acid. Notice that if the cartridge was washed with TCA at pHs
below 3, a precipitate would be released during the elution
of 1,4-TCHQ. After loading of the 1,4-TCHQ sample, the
cartridge was washed extensively with water to remove all
excess TCA and hydrogen peroxide, which are known to
inhibit the bilirubin oxidase reaction (18). The cartridge was
submerged into a vial containing 10 mL of 50 mM phosphate
pH 3.0. The absorbed 1,4-TCHQ was then eluted by passing
2
mL of methanol through the cartridge followed by 8 mL
of water to arrive at a final concentration of 25 mM phosphate
pH 3.0 containing 10% methanol. This protocol ensures that
the running buffer and sample have the same ionic strength
and solvent concentration. However without this precaution,
injection artifacts were noticed which hampered the accuracy
of 1,4-TCHQ determination at concentrations near the
detection limit. It should be noted that the recovery of
standard 1,4-TCHQ was always greater than 95% using this
method.
Preparation of Soil Extracts. Soil samples (1 g) were
extracted with 50 mL of 10 mM NaOH for 4 h under light
protection. After centrifugation, the supernatants were
collected, and the PCP was partially purified and concentrated
if necessary by using Sep-Pak tC18 cartridges (Waters Corp.,
Milford, MA) employing the method of Male et al. (24). In
brief, cartridges were wetted with methanol and then washed
with 50 mM phosphate (pH 7.5). The samples (20 mL) were
neutralized (2 mL of 500 mM phosphate, pH 7.5) and passed
through the cartridges. After washing with phosphate buffer,
contaminants which interfered with the biosensor measure-
ment using either absorbance or fluorescence were released
from the cartridges employing 2 mL of a 50% methanol
solution containing 50 mM phosphate (pH 7.5). The PCP
was only eluted from the cartridges when the methanol
concentration was increased to 100%. The samples (1 mL)
were then diluted back to the original volume with 10 mM
NaOH for analysis using capillary zone electrophoresis (CZE)
or the biosensor system. The concentration of PCP was
Results and Discussion
Optim ization of FI Biosensor Detection for 1,4-TCHQ. The
biosensor system used a 10 cm column of immobilized
bilirubin oxidase and a T-joint to ensure a constant con-
centration of NADH (Figure 1). The signal responses for 1,4-
TCBQ and 1,4-TCHQ (1,4-TCBQ treated with zinc powder)
were virtually identical; therefore, even if the reduction of
1,4-TCBQ by zinc powder was not 100% complete it would
not effect any decrease in the signal response. All lines in
contact with the sample were in stainless steel 316 to
minimize adsorption to the lines that could result in a tailing
effect on the peaks. Based on such considerations, 1,4-TCHQ
3
2 9 2
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 15, 2000