Phenol, 2-chlorophenol, 4-chlorophenol, 2,4-dichlorophenol, 2,6-dichlorophenol, 2,4,6-trichlorophenol, DMEQ- COCl, and general reagents were obtained from Wako Pure Chemical Industries (Osaka, Japan). Tap water was collected from the laboratory supply.

The HPLC system comprised a model L-6200 pump (Hitachi, Tokyo, Japan), a Rheodyne injection valve (Cotati. CA, USA) with a 20-μL loop, and a model RF-10A_XL fluorescence detector (Shimadzu, Kyoto, Japan) operating at an excitation wavelength of 400 nm and emission wavelength of 500 nm. The HPLC column (Inertsil ODS-4, GL Science Inc., Kyoto) was 50 × 2.1 mm i.d., packed with 5 μm particles of C₁₈ material. Quantification of the peaks was performed using a Chromatopac Model C-R3A integrator (Shimadzu). The mobile phase was prepared by the addition of acetonitrile (320 mL) to a solution of 680 mL containing trifluoroacetic acid (0.1 v/v%) in ultrapure water (Milli-Q water purification system, Simplicity^® UV, Millipore Corporation, Bedford, MA, USA). The samples were eluted from the column at room temperature at a flow rate of 0.5 mL/min.

Standard samples of phenol and chlorophenols were dissolved in ultrapure water and acetone, respectively, and adjusted to the concentration of 1 g/L. The standard mixture was diluted with ultrapure water. Borate buffer (0.1 M) was adjusted to pH 9.0 by the addition of NaOH. Borate buffer (100 μL) was added to diluted standard samples (100 μL, containing 0, 0.01, 0.02, 0.05, 0.1, 0.2, or 0.5 mg/L of each compound). DMEQ-COCl solution in acetonitrile (2 mg/mL, 100 μL) was added and vortexed. The mixture was allowed to react for 3 min at room temperature. Saturated L-aspartate solution (100 μL) was added to stop the reaction, and aliquots (10 μL) were injected into the HPLC system.

Water samples spiked with standard samples (lower limit of quantification, 0.05, 0.1, 0.2, or 0.5 mg/L of each compound) were passed through 0.45 μm filters (Cosmonice Filter S, Nacalaitesque) to remove suspended substances and immediately analyzed. Filtration was confirmed to have no effect on the analysis by analyzing a sample spiked with standards before and after filtration; no significant difference in recovery was observed.

Relative recovery was expressed as the ratio of the slope of the calibration curve prepared from a water sample spiked with standard sample to the slope of the standard calibration curve prepared as described above. Relative recovery data were used to assess the accuracy of the method.

For the time course study, the reaction time was set at 2, 3, 5, 10, and 15 min. Phenol and five chlorophenols (100 μL, each 0.1 mg/L), borate buffer (100 μL, pH 9.0)and DMEQ-COCl (100 μL, 2 mg/mL) were mixed as described in Materials and Methods. The maximum peak area or plateau level was reached within 3 mins (Figure 2).

Next, pH dependency (pH 7.5 to 10.0) was examined at the derivatization time of 3 min (Figure 3). Peak areas of DMEQ-phenol and DMEQ-4-chlorophenol at pH 7.5 were 33% and 82% of the maximum area, respectively. The peak areas of all DMEQ derivatives tended to be decreased at pH 10.0. However, the peak areas of

all compounds were stable (more than 95% of the maximum area) in the range of pH 8.5 to pH 9.3.Since, under pH 8.5, it is guessed that a little pH change may effect phenol detection, simultaneous determination of tested compounds is not better at pH 8.5.

In preliminary tests (data not shown), the peak areas of all derivatives obtained with 1 mg/mL of DMEQ- COCl at pH 9.0 and room temperature were less than about 60% of the control. Furthermore, when using 3 mg/mL of DMEQ-COCl, a large blank peak interfered with the peaks of phenol, 2-chlorophenol, and 4-chlo- rophenol. Thus, the derivatization time of 3 min at pH 9.0using 2 mg/mL of DMEQ-COCl was chosen for the assay.

Figure 4 shows typical chromatograms obtained from (A) blank and (B) standard sample (0.1 mg/L). In the analysis, full scale (128 mV) was automatically shifted to 64 mV at 20 min to clearly observe the DMEQ-2,4,6- trichlorophenol peak. The retention times of DMEQ-phenol, DMEQ-2-chlorophenol, DMEQ-4-chlorophenol,

DMEQ-2,6-dichlorophenol, DMEQ-2,4-dichlorophenol, and DMEQ-2,4,6-trichlorophenol were 3.6, 5.7, 7.9, 9.3, 13.6, and 23.9 min, respectively. The running time was 30 mins.

The standard curves of phenol, 2-chlorophenol, 4-chlorophenol, 2,6-dichlorophenol, 2,4-dichlorophenol, and 2,4,6-trichlorophenol were constructed by plotting integrated peak areas of derivatives vs. concentrations of phenolic compounds. The calibration data are summarized in Table 1. The plots were linear in the concentration range of 0.01 - 0.05 to 0.5 mg/L, with r² value ≥ 0.9967 for all compounds. The values of the lower limit of quantification were taken as the lowest concentration on the standard curve. The lower limits of detection for phenol and 2-chlorophenol were estimated as the concentrations giving a detectable peak, since these peaks were located close to a large blank peak. The lower limits of detection for 4-chlorophenol, 2,4-dichlorophenol, 2,6-dichlorophenol, and 2,4,6-trichlorophenol were taken to be the concentrations giving a signal-to-noise ratio of 3:1. The lower limits of detection for phenol, 2-chlorophenol, 4-chlorophenol, 2,4-dichlorophenol, 2,6- dichlorophenol, and 2,4,6-trichlorophenol were 0.003, 0.002, 0.001, 0.004, 0.002, and 0.008 mg/L, respectively (range of 0.0025 to 0.020 ng/20 μL injection as absolute amount).

The method using NBD-F previously developed in our laboratory showed relatively poor sensitivity [20] . The present method shows 1.2- to 5-fold greater sensitivity in terms of the lower limits of detection (8.5- to 33-fold improvement in terms of absolute amount).

Precision and accuracy for intra-day and inter-day assays of these derivatives are shown in Table 2 and Table 3. In the intra-day assay, the range of standard deviation was within 2.7% to 8.0%. The recoveries were within 96.5% to 104.2%. In the inter-day assay, the range of standard deviation was within 4.3% to 8.9%, and the recoveries were within 95.5% to 105.0%.

^*: Data are expressed as absolute amount (ng/20 μL injection).

Tap water samples were collected from the laboratory, and the proposed method was employed to determine phenol and chlorophenols in spiked tap water. As shown in Table 4, phenol and chlorophenols in tap water samples were below the lower limits of quantification in unspiked samples. Calibration curves prepared from tap water samples spiked with the six compounds showed a linear relationship between concentration and peak response, with r² ≥ 0.9933. The slopes of the calibration curves were similar to those of the standard calibration curves, and the relative recovery values were 89.8% to 111.0%. These results indicate that our method is capable of monitoring very low levels of contamination with phenol and chlorophenols in tap water.