An approach based on liquid chromatography/electrospray ionization–mass spectrometry to detect diol metabolites as biomarkers of exposure to styrene and 1,3-butadiene

Abstract

Styrene and 1,3-butadiene are important intermediates used extensively in the plastics industry. They are metabolized mainly through cytochrome P450-mediated oxidation to the corresponding epoxides, which are subsequently converted to diols by epoxide hydrolase or through spontaneous hydration. The resulting styrene glycol and 3-butene-1,2-diol have been suggested as biomarkers of exposure to styrene and 1,3-butadiene, respectively. Unfortunately, poor ionization of the diols within electrospray mass spectrometers becomes an obstacle to the detection of the two diols by liquid chromatography/electrospray ionization–mass spectrometry (LC/ESI–MS). We developed an LC/ESI–MS approach to analyze styrene glycol and 3-butene-1,2-diol by means of derivatization with 2-bromopyridine-5-boronic acid (BPBA), which not only dramatically increases the sensitivity of diol detection but also facilitates the identification of the diols. The analytical approach developed was simple, quick, and convincing without the need for complicated chemical derivatization. To evaluate the feasibility of BPBA as a derivatizing reagent of diols, we investigated the impact of diol configuration on the affinity of a selection of diols to BPBA using the established LC/ESI–MS approach. We found that both cis and trans diols can be derivatized by BPBA. In conclusion, BPBA may be used as a general derivatizing reagent for the detection of vicinal diols by LC/MS.

Epoxides are common metabolites of chemicals possessing carbon–carbon double bonds after biotransformation primarily by cytochrome P450 in animals and humans. Epoxides are usually cytotoxic, mutagenic, and carcinogenic due to their electrophilic reactivity to nucleophilic biomacromolecules, such as nucleic acids and proteins, to form covalent binding adducts [1–3]. Hydration of epoxides to diols is one of the major detoxification metabolic pathways, and the resulting diol metabolites are often found to correlate with epoxide intakes. This provides the basis to use the diol metabolites as biomarkers for human exposure to those xenobiotics with epoxidation potential.

Styrene (1) and 1,3-butadiene (4) are important industrial intermediates in the polymer production industry. They are widely used for the production of plastics and resins such as butadiene rubber, styrene rubber, adiponitrile, polychloroprene, nitrile rubber, and styrene butadiene latex [4]. Both styrene and 1,3-butadiene are metabolized to the corresponding epoxides 2 and 5, reportedly responsible for the toxicities induced by the parent compounds. In addition, these epoxide metabolites are sequentially hydrated to diols, namely styrene glycol (3) and 3-butene-1,2-diol (6), which have been recognized as biomarkers of styrene and 1,3-butadiene exposure, respectively (Scheme 1) [5–7].

Scheme 1.

Diol metabolites of styrene and 1,3-butadiene formation in vivo.

Development of sensitive and convincing analytical approaches for identification and quantification of such diols is pivotal to the success of biomarker and exposure studies. Styrene glycol and 3-butene-1,2-diol were reportedly analyzed by gas chromatography coupled with mass spectrometry (MS)¹ [8–10], flame ionization detector [11–13], or electron capture detector [14] as respective detectors, often requiring tedious derivatization. High-performance liquid chromatography with ultraviolet detection (HPLC–UV) was also applied to analyze styrene glycol [15,16]. Unfortunately, the low sensitivity along with high background limits its application for the detection of styrene glycol in a complex biological matrix. HPLC–UV is unable to analyze 3-butene-1,2-diol because it lacks a chromophore for UV absorbance. Liquid chromatography (LC)/MS has become a conventional tool for analysis of biomedical molecules. In this study, we established an LC/MS-based analytical approach to detect styrene glycol and 3-butene-1,2-diol, and the established approach may be applied to analyze many other vicinal diols. The approach not only allows us to detect diols as intact molecules but also facilitates the identification of the diols of interest.

Materials and methods

Apparatus

MS analyses were performed on an LC/electrospray ionization (ESI)–MS system that included an Agilent 1100 HPLC system coupled with a Sciex API 2000 tandem quadrupole mass spectrometer with an ESI TurboIonSpray source (Applied Biosystems, Foster City, CA, USA). The precolumn (C18, 2 μm) and analytical column (C18, 2.1 × 100 mm, 3 μm) were purchased from Alltech (Deerfield, IL, USA). Analyst Software (version 1.4.1, Applied Biosystems) was used to control the LC/MS system and conduct quantitative analysis. HPLC–UV analysis of styrene glycol was performed on a Shimadzu 10Avp system (Tokyo, Japan). A 300-MHz nuclear magnetic resonance (NMR) spectrometer (Varian, Palo Alto, CA, USA) was used for structural characterization of synthetic chemicals.

Chemicals and reagents

2-Bromopyridine-5-boronic acid (BPBA, 11), styrene glycol (1-phenyl-1,2-ethanediol, 3), 3-butene-1,2-diol (6), cis-1,2-cyclohexanediol (8), trans-1,2-cyclohexanediol (9), 3,4-hexanedione, and 1,2-naphthoquinone were purchased from Sigma–Aldrich (St. Louis, MO, USA). Optima-grade acetonitrile and methylene chloride (CH₂Cl₂) were purchased from Fisher Scientific (Fair Lawn, NJ, USA). The other chemicals were either optima or analytical reagent grade.

Synthesis

3,4-Hexanediol (7)

To a solution of 3,4-hexanedione (92 mg, 0.8 mmol) in ethanol (10 ml) at 0 °C, sodium borohydride was added (57 mg, 1.5 mmol) in small portions and then the reaction mixture was stirred at room temperature for 1 h. The solution was neutralized with 1 N HC1 and evaporated to dryness. The residue was extracted several times with ethyl acetate, and the pooled organic phases were washed with water a few times and dried over anhydrous Na₂SO₄. The solvent was removed under reduced pressure to give colorless oil that was chromatographed on silica gel (grade 60, 230–400 mesh, Fisher Chemical) to afford diol 7 (59 mg, 64%). ¹H NMR (CDCl₃): δ 0.99–1.04 (t, J = 7 Hz, 6H), 1.47–1.57 (m, 4H), 1.81 (s, 2H), 3.54–3.57 (m, 2H).

trans-1,2-Dihydronaphthalenediol (10)

To a stirred suspension of sodium borohydride (50 mg, 1.3 mmol) in ethanol (5 ml), 1,2-naphthoquinone (16 mg, 0.1 mmol) was added in small portions and then the mixture was stirred at room temperature for 24 h. The resulting light-colored suspension was poured into ice water (10 ml), acidified with 1 N HC1, and extracted several times with CH₂C1₂. The organic phase was washed with water, dried over anhydrous Na₂SO₄, and evaporated to dryness. The residue was recrystallized in hexane to afford trans-1,2-dihydro-naphthalenediol (6.5 mg, 40%) as white crystal solid [17]. ¹H NMR (CDCl₃): δ 4.50–4.55 (m, 1H), 4.84–4.87 (d, J = 10.5 Hz, 1H), 5.98–6.02 (dd, J = 8.9 Hz, 1H), 6.44–6.48 (dd, J = 9.9 Hz, 1H), 7.10–7.14 (m, 1H), 7.28–7.31 (m, 2H), 7.56–7.59 (m, 1H).

BPBA styrene glycol ester (12)

Equal molar amounts of styrene glycol were mixed with 20 mg of BPBA (0.1 mmol) in 2 ml of acetone, followed by the addition of 0.2 g anhydrous MgSO₄. The mixture was stirred at room temperature for 1 h, and thin layer chromatography (TLC) analysis showed a new spot. The solvent was removed by rotary evaporation. The remaining product was chromatographed on silica gel, yielding pure BPBA styrene glycol ester. ¹H NMR (CDCl₃): δ 4.23–4.28 (m, 1H), 4.75–4.79 (m, 1H), 5.61–5.66 (m, 1H), 7.28–7.39 (m, 5H), 7.55–7.58 (d, J = 7.9 Hz, 1H), 7.98–8.01 (d, J = 7.9 Hz, 1H), 8.80 (s, 1H). LC/MS: m/z 304/306 [M+1]⁺.

BPBA 3-butene-1,2-diol ester (13)

BPBA 3-butene-1,2-diol ester was synthesized by following a similar protocol for the synthesis of BPBA styrene glycol ester as above. ¹H NMR (CDCl₃): δ 4.06–4.11 (t, J = 9.1 Hz, 1H), 4.55–4.58 (t, J = 9.1 Hz, 1H), 5.02–5.10 (q, J = 14.8 Hz, 1H), 5.30–5.34 (d, J = 10.1 Hz, 1H), 5.39–5.46 (d, J = 17.1 Hz. 1H), 5.91–6.00 (m, 1H), 7.52–7.55 (d, J = 7.9 Hz, 1H), 7.92–7.95 (d, J = 7.9 Hz, 1H), 8.74 (s, 1H). LC/MS: m/z 254/256 [M+1]⁺.

Human urine sample preparation

Human urine samples (0.2 ml) were spiked with authentic standard styrene glycol or 3-butene-1,2-diol at various concentrations. Then 0.5 ml of CH₂Cl₂ was added to the urine samples, and the resulting samples were vortexed for 2 min and centrifuged at 5000 rpm for 5 min. The organic phase was pooled and blown to dryness by nitrogen gas. The remaining residue was redissolved in 100 μl of acetonitrile and ready for derivatization as below.

Derivatization of diols by BPBA

Diols, including those purchased from suppliers, synthesized in our laboratory, and extracted from biological samples, were dissolved in acetonitrile (200 μl for authentic standard samples, 100 μl for biological samples), followed by the addition of BPBA to a final concentration of 200 μM. The resulting mixtures were vortexed and analyzed by LC/MS.

LC/MS analyses

For these analyses, 5 μl of each sample was injected into HPLC coupled with a mass spectrometer without a split at a flow rate of 200 μl/min. The mobile phase was acetonitrile (0.1% trifluoroacetic acid [TFA])–water (0.1% TFA) (95:5, v/v). Column temperature was set at room temperature. MS was performed with a positive TurboIonSpray ion source. The corresponding two ions for the BPBA diol esters of test at a 1:1 ratio resulting from the isotopic bromine were monitored under selected ion monitoring (SIM) mode. The total areas under the two isotopic peaks were used for quantitative analyses. The MS operation conditions were as follows: source temperature, 350 °C; curtain gas, 20 psi; nebulizer gas, 25 psi; ion spray voltage, 4200 V; declustering potential, 20 V; focusing potential, 200 V; and entrance potential, 10 V.

Results

The effort to develop a sensitive, convenient, and convincing analytical approach to detect vicinal diols was initiated by the demand for analysis of styrene glycol as a part of our current mechanistic studies of respiratory toxicity induced by styrene. We failed to detect styrene glycol in a biological matrix by HPLC–UV or LC/MS with the sensitivity required for the mechanistic investigation. Authentic standard styrene glycol was spiked into human urine samples and extracted with CH₂Cl₂, followed by HPLC–UV analysis. The eluate was monitored by a UV detector at 210 nm. A high background was found to interfere with the visibility of the peak responsible for styrene glycol spiked in the urine samples (Fig. 1A). As shown in Fig. 1B, the peak of styrene glycol was buried by the UV absorbance attributed by other components excreted in the urine. The sensitivity for the detection of styrene glycol, even injected with pure styrene glycol solution, was also found to be low, and the peak of styrene glycol was barely visible in the chromatogram monitored at 210 nm injected with 5 μl of 10 μM styrene glycol solution (Fig. 1C). In addition, we attempted to analyze styrene glycol using LC/tandem mass spectrometry (MS–MS) by monitoring ion m/z 121 (the molecular ion of styrene glycol, [M–OH]⁺), but we failed to detect the diol even by loading with 10 times higher concentrated styrene glycol solution (100 μM [data not shown]) than that analyzed by HPLC–UV. Poli and coworkers reportedly detected styrene glycol by monitoring m/z 121 → 103 in MS/MS mode [18]. However, for some reason we failed to detect styrene glycol by our mass spectrometer.

Fig. 1.

HPLC–UV chromatograms (λ = 210 nm) of blank human urine (A), human urine spiked with styrene glycol (final concentration 10 μM) (B), and authentic standard styrene glycol in acetonitrile (final concentration 10 μM) (C) and mass chromatograms of human urine spiked with styrene glycol without BPBA (D), human urine spiked with styrene glycol with BPBA derivatization (final concentration 10 μM) (E), and authentic standard styrene glycol in acetonitrile (final concentration 10 μM) (F) with BPBA derivatization. - - -, m/z 304; —, m/z 306. Urine samples were extracted by CH₂Cl₂.

Boronic acids are known to reveal high affinity to vicinal diols by formation of boronic esters. The esterification reaction is fast and straightforward [19,20]. This allowed us to provide the possibility to enhance the sensitivity for the LC/MS detection of vicinal diols by use of boronic acid chemistry. We chose BPBA to derivatize the diols of our interest (Scheme 2). Our rationale for the use of BPBA as the derivatizing reagent included the fact that (i) the nitrogen incorporated into the boronic diol esters increased the efficiency of ionization in positive mode in electrospray MS and (ii) the incorporation of bromine in the complex facilitated the identification of diols due to the characteristic 1:1 isotopes resulting from natural isotopes of ⁷⁹Br and ⁸¹Br [21]. In addition, BPBA is chemically stable and commercially available, having great potential for general use as a derivatizing reagent of diols for MS analysis.

Scheme 2.

Structures of diols, BPBA, and formed corresponding BPBA esters.

Esterification of styrene glycol and 3-butene-1,2-diol by BPBA

As expected, both styrene glycol and 3-butene-1,2-diol reacted with BPBA to produce BPBA styrene glycol ester (12) and BPBA 3-butene-1,2-diol ester (13). The mass spectra of the two boronic esters showed the respective 1:1 ratio of molecular ions at m/z 304/306 (Fig. 2A) and m/z 254/256 (Fig. 2B), resulting from natural isotopes of ⁷⁹Br and ⁸¹Br. The structures of BPBA styrene glycol ester and BPBA 3-butene-1,2-diol ester were also confirmed by their NMR spectra (refer to Materials and methods section).

Fig. 2.

Mass spectra of styrene glycol (A) and 1,2-butene-3-diol (B) derivatized by BPBA.

LC/ESI–MS analyses of BPBA styrene glycol and 3-butene-1,2-diol esters

BPBA esters were prepared in acetonitrile and analyzed by LC/MS. Ion pairs of m/z 304/306 and m/z 254/256 were monitored in SIM mode. The former ion pair represents the [M+1]⁺ ions of BPBA styrene glycol ester, and the latter stands for the [M+1]⁺ ions of BPBA 3-butene-1,2-diol ester. Fig. 3A shows the mass chromatograms of BPBA styrene glycol ester (12). The peak, representing m/z 304, was found to overlap with that of m/z 306, and the superimposed ion peaks at m/z 304 and m/z 306 resulted from the 1:1 natural isotopes of bromine. The analysis of BPBA 3-butene-1,2-diol ester (13) was performed separately. As expected, the same characters in ion chromatography were observed in the analysis of BPBA 3-butene-1,2-diol ester (Fig. 3B) as those for BPBA styrene glycol ester.

Fig. 3.

Mass chromatograms of styrene glycol treatment with BPBA (A) and 3-butene-1,2-diol treated with BPBA (B). (A) - - -, m/z 304; —, m/z 306. (B) - - -, m/z 254; —, m/z 256.

To maximize the sensitivity for the detection of BPBA-modified diols, we optimized the mobile phase for LC. We tested a selection of water-abundant mobile phases. Eluted by these mobile phases, the detection sensitivity was low and peak shape, as well as calculated peak area, lacked reproducibility, possibly due to the hydrolytic effect of water in the eluting systems. In contrast, mobile phases with an abundance of acetonitrile produced much higher detection sensitivity and better reproducibility in peak area. In addition, the presence of TFA (0.1%) in the mobile phase was found to be necessary, whereas the peak height was significantly decreased when eluted with a solvent system without TFA. The presence of TFA facilitates the protonation of the pyridine moiety and prevents the boron atom from forming an anion [20], increasing the efficiency of ionization in positive mode. The acetonitrile-abundant mobile phase allowed the unreacted excess of BPBA to coelute with the BPBA esters (see Fig. S-2 in supplementary material). The coelution of BPBA along with the presence of abundant acetonitrile suppressed the hydrolysis of boronic esters. The mobile phase with 95% acetonitrile (0.1% TFA) in water (0.1% TFA) was found to be the best for the analysis of these BPBA esters by LC/MS. In addition, methanol was tested as an optional organic mobile solvent to elute BPBA esters. Unfortunately, it failed to offer the same detection sensitivity as acetonitrile did.

Solvent effect on derivatization of diols with BPBA

To identify a solvent system that facilitates the esterification of diols, we tested a variety of solvent systems in which BPBA reacted with styrene glycol. The solvent systems tested included water, acetonitrile/water (1:1, v/v), absolute acetonitrile, and absolute methanol. Styrene glycol derivatized by BPBA in these systems was analyzed by LC/MS. Not much difference in peak area representing the formation of BPBA styrene glycol ester was observed among the solvent systems tested, and the total ion peak area of the boronic ester was found to be 2.65 ± 0.08 (counts, × e⁶, n = 3) in water (Fig. 4A), 2.62 ± 0.05 in water/acetonitrile (1:1, v/v) (Fig. 4B), 2.29 ± 0.03 in methanol (Fig. 4C), and 2.50 ± 0.03 in acetonitrile (Fig. 4D). Acetonitrile was found to give the best peak shape and peak height (Fig. 4D), whereas water produced poor peak shape and decreased peak height (Fig. 4A). In a separate study, a similar observation was obtained, with acetonitrile producing the best peak shape and height for the analysis of 3-butene-1,2-diol (data not shown). Clearly, acetonitrile as a reaction solvent system offered a better environment to derivatize styrene glycol and 3-butene-1,2-diol relative to water and methanol. This led us to select acetonitrile as the reaction solvent in the rest of the study.

Fig. 4.

Mass chromatograms of styrene glycol derivatized by BPBA in water (A), water/acetonitrile (1:1 v/v) (B), methanol (C), and acetonitrile (D). - - -, m/z 304; —, m/z 306.

Detection of BPBA diol esters in acetonitrile and human urine

Stock solutions of styrene glycol and 3-butene-1,2-diol were prepared in acetonitrile at a concentration of 10 mM and stored at 4 °C. A serial dilution of styrene glycol and 3-butene-1,2-diol was made from the corresponding stock solution to the concentrations of 1.0, 5.0, 10.0, 25.0, and 50.0 μM (styrene glycol) and 0.1, 0.5, 1.0, 5.0, and 10.0 μM (3-butene-1,2-diol). The diols were derivatized with BPBA as described above and subjected to LC/MS analysis.

A human urine sample, obtained from a healthy male volunteer, was spiked with various amounts of styrene glycol and 3-butene-1,2-diol (urine/spiked solution, 9:1, v/v) to the same final concentrations as those for the preparation of standard solutions in acetonitrile above. The resulting urine samples (0.2 ml each) were mixed with 0.5 ml of CH₂Cl₂ for the extraction of the diols spiked. The diols extracted were derivatized with BPBA and subjected to LC/MS analysis.

Two isotopic peaks of the boronic esters with identical shape and height were observed, and the signal intensity increased with the increase of diol concentrations. The peak areas of the boronic esters derived from styrene glycol and 3-butene-1,2-diol were found to have a good linear relationship with the diols either dissolved in acetonitrile or spiked in human urine at concentrations in ranges of 1.0–50.0 and 0.1–10.0 μM, respectively. The calculated regression coefficients (R²) were 0.9995 and 0.9991 in acetonitrile and 0.9977 and 0.9992 in the human urine samples for styrene glycol and 3-butene-1,2-diol, respectively (see Fig. S-1 in supplementary material). Three independent experiments in parallel were performed to evaluate recovery. Table 1 shows the MS responses and calculated recoveries of styrene glycol and 3-butene-1,2-diol obtained from acetonitrile and human urine samples at three concentrations (high, medium, and low). It needs to clarify that the recoveries listed in Table 1 include the processes of extraction, derivatization, and response of MS analysis. Fig. 1D shows the mass chromatogram of a human urine sample spiked with styrene glycol (10.0 μM) without BPBA derivatization, and Fig. 1E shows the mass chromatogram of the same sample but derivatized with BPBA. Two new peaks with m/z 304/306 that were superimposed were observed after derivatization with BPBA. Fig. 1F shows the mass chromatogram of authentic standard styrene glycol (10 μM) derivatized with BPBA.

Table 1.

MS responses and recoveries of styrene glycol and 3-butene-1,2-diol at three concentrations.

	Spiked concentration (μM)	Peak area (cps, e5)		Recovery (%)
		Acetonitrile	Human urine
Styrene glycol	1.0	2.13 ± 0.13	2.56 ± 0.17	60.1 ± 7.5
	10.0	11.02 ± 0.14	7.13 ± 0.25	32.4 ± 1.6
	50.0	50.77 ± 1.10	35.60 ± 1.16	35.1 ± 1.9
3-Butene-1,2-diol	0.1	1.70 ± 0.10	1.57 ± 0.09	46.6 ± 5.7
	1.0	8.94 ± 0.39	8.46 ± 0.22	47.5 ± 3.8
	10.0	94.99 ± 3.13	89.63 ± 2.11	47.3 ± 82.7

Note. Data represent means ± standard deviations from three independent experiments in parallel.

The limits of detection (LODs) for styrene glycol and 3-butene-1,2-diol in acetonitrile were assessed and found to be 1.0 and 0.1 pmol, respectively (signal/noise = 3). Certainly, LOD also depends on the sensitivity of MS used.

Configurational effect on the formation of diol-derived BPBA esters

3,4-Hexanediol (7), cis-1,2-cyclohexanediol (8), trans-1,2-cyclo-hexanediol (9), and trans-1,2-dihydronaphthalenediol (10) were selected for concentration–response studies to obtain insight into the stereochemistry of BPBA reactions with vicinal diols. Several previous studies have discussed the relationship between the diol structures and their affinity to boronic acids [20,22,23]. The purpose of the current structure–affinity relationship study was to investigate the feasibility of BPBA for the derivatization of vicinal diols, particularly those with toxicological and environmental importance. A serial dilution of the diols selected was prepared in acetonitrile to the final concentrations of 1.0, 5.0, 10.0, 25.0, and 50.0 μM. After derivatization with BPBA, the resulting boronic esters were analyzed by LC/MS. Each diol produced a concentration-dependent response to esterification with BPBA along with a slope that reflects the affinity of BPBA to the diol of test. 1,2-Hexanediol was found to show the highest affinity to BPBA, followed by cis-1,2-cyclohexane-diol, trans-1,2-cyclohexanediol, and trans-1,2-dihydronaphthalene-diol (Fig. 5).

Fig. 5.

Concentration-dependent boronic ester formation in the reaction of BPBA with 3,4-hexanediol (◆), cis-1,2-cyclohexanediol (■), trans-1,2-cyclohexanediol (×), and trans-1,2-dihydronaphthalenediol (▲).

Discussion

Several research groups have reported the application of boronic acid chemistry for detection of vicinal diols by LC/MS. Higashi and coworkers synthesized a selection of boronic acids to detect 24,25-dihydroxyvitamin D₃ using LC/atmospheric pressure chemical ionization (APCI)–MS technology [24]. They found that the detection sensitivity of the steroid was significantly enhanced after boronic acid derivatization. Derivatizing by phenylboronic acid, Chen and coworkers detected some biologically important cis-diol-containing molecules, such as saccharides and steroids, and certain phenols, namely flavanols and catecholamines, using reactive desorption electrospray ionization (DESI)/MS [25].

The uniqueness of the approach we developed in this study includes the introduction of the elements nitrogen and bromine into boronic acid for the derivatization of vicinal diols. The incorporation of nitrogen facilitates the ionization of the resulting boronic esters analyzed by LC/MS in positive mode. Our MS analysis indicated that derivatization of styrene glycol and 3-butene-1,2-diol with BPBA not only made it possible to detect the two diols by LC/MS but also offered high sensitivity for the detection. Of equal importance is that the characteristic mass doublets, resulting from natural isotopes of bromine introduced into the boronic esters, allow us to postlabel the vicinal diols, enabling us to reduce false-positive errors that LC/MS analysis often experiences.

Oxygen-containing protic solvents, such as water and alcohols, are known to react with boronic esters to form the corresponding boronic acids or esters [19]. In this study, we tested a selection of solvent systems—water, water/acetonitrile (1:1, v/v), absolute acetonitrile, and absolute methanol—to identify a solvent system that is best for the derivatization of styrene glycol by BPBA. Absolute acetonitrile was found to be the best as a solvent used for the derivatization of the diol by BPBA. With this solvent system, LC/MS analyses produced the best peak shape and peak height, providing the highest sensitivity for the detection of styrene glycol. However, the presence of either water or methanol as a derivatizing solvent was found to ruin the peak shape, perhaps due to the solvolytic effect (hydrolysis or transesterification) of the two protic solvents on the boronic ester.

Given that BPBA showed a high affinity to styrene glycol and 3-butene-1,2-diol, we investigated the feasibility of BPBA to derivatize vicinal diols. A total of four diols as model compounds—3,4-hexanediol, cis-1,2-cyclohexanediol, trans-1,2-cyclohexanediol, and trans-1,2-dihydronaphthalenediol—were selected for the study. The affinity of BPBA to the diols of test was evaluated by monitoring the concentration-dependent response of each diol to BPBA to form the corresponding boronic ester.

As expected, cis-1,2-cyclohexanediol showed higher reactivity toward BPBA than did trans-1,2-cyclohexanediol. Among the four diols tested, 3,4-hexanediol was the only linear molecule without any ring systems and where the two hydroxyl groups can rotate freely, so that 3,4-hexanediol can readily react with BPBA. The O–C–C–O dihedral angle between two hydroxyl groups of trans-1,2-cyclohexanediol is bigger than that of cis-1,2-cyclohexanediol, possibly making trans-1,2-cyclohexanediol less reactive toward BPBA than cis-1,2-cyclohexanediol. As expected, trans-1,2-dihydronaphthalenediol was found to reveal the least affinity to BPBA. The fused phenyl ring of trans-1,2-dihydronaphthalenediol causes higher molecular rigidity relative to the cyclohexane ring of trans-1,2-cyclohexanediol. It costs more energy for the two hydroxyl groups of trans-1,2-cyclohex-anediol to rotate to a proper angle to react with BPBA than trans-1,2-cyclohexanediol. Even though BPBA showed relatively low affinity to the trans diol, the BPBA-based MS analysis allowed us to detect trans-1,2-dihydronaphthalenediol to at least the nearest 0.1 pmol. The analytical approach is sensitive enough to detect urinary excretion of trans-1,2-dihydronaphtha-lenediol as a biomarker of exposure to naphthalene, a known pneumotoxic compound [26].

In conclusion, we have successfully developed an analytical approach to detect styrene glycol and 3-butene-1,2-diol. This approach is convenient, sensitive, and highly convincing.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.ab.2008.12.007.