Determination of Benzene , Toluene and N-Hexane in Urine and Blood by Headspace Solid-Phase Microextration / Gas-Chromatography for the Biomonitoring of Occupational Exposure

A análise de tolueno, benzeno e n-hexano inalterados em urina e sangue humano é uma ferramenta para avaliar a exposição ocupacional a esses solventes. Um método simples usando microextração em fase sólida acoplada ao headspace (HS-SPME), usando fibras de carboxeno/ polidimetilsiloxano (CAR-PDMS) e PDMS, para análise simultânea por cromatografia gasosa de tolueno, benzeno e n-hexano em urina e em sangue foi desenvolvido com intuito de realizar a monitorização biológica de indivíduos expostos ocupacionalmente a esses solventes. O método foi aplicado na análise de amostras de trabalhadores que utilizavam colas, contendo solventes, na recuperação de calçados. O tolueno foi detectado em todas as amostras de sangue coletadas em sapatarias quando se utilizou a fibra de CAR-PDMS (16,0-55,2 mg L, n=7) e o n-hexano foi detectado em duas amostras (33,0 and 41,3 mg L). O benzeno não foi detectado em nenhuma das amostras de sangue e/ou urina. Nenhum dos solventes pode ser quantificado nas amostras de urina.


Introduction
Volatile organic compounds (VOCs) presents vapor pressures above 0.01 kPa and boiling points below 150 ºC. 1,2he determination of VOCs in inhaled air, blood, and urine is important from a toxicological, biochemical and forensic viewpoint. 3,4rganic solvents such as n-hexane, benzene, toluene and xylenes are present in a wide range of commercially available products because of their industrial use in the manufacturing of paints, adhesives, thinners, glues, plastic, pitch, bitumen, some pharmaceutical products and as intermediate in the formation of other chemical substances.
n-Hexane is one of the most important aliphatic compounds and is utilized as a solvent in several industries (i.e., pharmaceutical, chemical and shoe manufacturing) as ingredient in glues and during the manufacturing of polyolefins and rubber.][7] Exposure to benzene is less likely than that of n-hexane and toluene since its use as a solvent in industrial processes was banned in 1980. 8,9The International Agency for Research on Cancer (IARC) 10 classified it as a potential carcinogenic type I.Then, the use of benzene is restricted to two situations.First, in synthetic processes, when benzene is an essential reagent and it is not possible to exchange for another one.Another situation is in the petrochemical industry, when present in the process of petroleum cracking.In addition, benzene can be encountered in cigarette smoke and after the combustion of fossil fuels, such as gasoline.The toxicological effects related to benzene are nonlymphocytic leukemia, [11][12][13] aplastic anemia, chromosome aberrance [14][15][16][17] and bone marrow progressive degeneration.
Toluene is a very common chemical which is widely used as a solvent in many consumer products, such as paints, glues and adhesives.Toluene is also used as a synthetic intermediate and as an additive in fuel to increase gasoline octane rating.It produces central nervous system toxicity and fetotoxicity. 18uman contamination by VOCs is generally evaluated by monitoring metabolites in urine.The 2,5-hexanodione, trans, trans muconic acid and hippuric acid are the main biomarkers of n-hexane, benzene and toluene, respectively.However, this procedure presents some disadvantages.Firstly, external sources such as food, mainly fruits and vegetables, can also contain these substances.Secondly, the urinary excretion of the metabolites is influenced by several factors such as age, alcohol consumption and smoking habits.][21][22] In contrast, the use of unchanged solvents as biomarkers is more specific.The emergence and detection of solvents in urine and/or blood provides high selectivity since they are not endogenous substances. 18olvents in urine are present in low concentrations as compared to blood. 23Only about 0.06% of the absorbed toluene and 0.1-0.3% of benzene are excreted unchanged in the urine.Furthermore, the American Conference of Governmental Industrial Hygienist (ACGIH) 24 recommends the measurement of toluene in blood, where it is present in higher concentration, rather than in urine.
In toxicological analysis and other analytical procedures, sample preparation is an essential step to eliminate interferences and selectively extract analytes from the biological fluids.Solid phase microextraction (SPME) is a solventless extraction technique based on the sorption of the analyte in small diameter fibers coated with sorbent material. 19,23This technique involves two steps.First the analytes partition between the fiber coating and the sample matrix, followed secondly, by desorption of the concentrated extract into the analytical instrument. 25The headspace mode (HS-SPME) permits an efficient mass transfer and is better adapted for the analysis of volatile and semi-volatile chemical compounds, especially in complex matrices.
7][28][29][30][31] However, in all these methods unaltered n-hexane was not included.Furthermore, there is no study dealing with simultaneous determination of unchanged n-hexane, benzene and toluene in blood and/or urine.It should be noted that simultaneous occupational exposure occurs more frequently than single exposure in industries that use products containing solvents.
This study focused on the development of a simple, sensitive, and environmental friendly HS-SPME/GC-FID analytical method.Two different fiber coatings, polydimethylsiloxane (PDMS) and carboxen-PDMS (CAR-PDMS) were employed for the simultaneous determination of toluene, benzene and n-hexane in urine and blood.The method was applied to the analysis of the solvents in urine and blood of workers occupationally exposed in shoe repair shops.

Instruments
A manual SPME device (Bellefonte, PA, USA) was used to hold the fiber.Fibers coated with PDMS (100 µm thickness) and CAR-PDMS (75 µm thickness) were purchased from Supelco.A 1-cm magnetic bar (Supelco) and water bath with digital temperature control (B.Braun Biotech International GmbH, Melsungen, Germany) were used in the SPME procedure.Glass vials, 16 mL, sealed with silicone/PTFE septa were also obtained from Supelco and used for HS-SPME extraction.
A GC 1000 gas-chromatograph with a FID detector (Ciola Gregori Ltda., São Paulo, Brazil) equipped with a splitless injector inlet liner, interfaced to a PC with a DANI DS 1000 integrator (Dani Strumentazione Analitica, Monza, Italy) and IQ3 software for data acquisition was used.The injections were conducted in splitless mode.

HS-SPME variables
The HS-SPME variables were optimized to obtain the best resolution and efficiency.For each coating phase the following parameters were studied: heating temperature, stirring time, fiber headspace exposure time and sodium chloride concentration in the matrix.Desorption of analytes from the fibers (time and temperature of the injector port) was also established.

HS-SPME procedure
Eight milliliters of urine or two milliliter of blood, previous homogenized, were transferred to 16 mL vials containing a magnetic stir bar and an adequate amount of sodium chloride.The vials were sealed with a silicone/ PTFE septum to prevent any loss of the analytes and kept heated in a thermoblock under constant stirring.After reaching equilibrium the septum was pierced with a SPME needle and the SPME fiber was exposed to the headspace for sorption of the volatile compounds.Subsequently, the SPME fiber was transferred to the GC injector for desorption.
Calibration curves were obtained spiking blank urine samples with six different concentrations of each solvent (5 replicates per concentration): toluene between 92 and 560 µg L -1 , benzene between 2 and 420 µg L -1 and n-hexane between 10 and 540 µg L -1 when CAR-PDMS fibers were used.The same procedure was adopted for PDMS fibers, however the concentrations of toluene, benzene and n-hexane ranged from 92 to 2000 µg L -1 , from 2 to 6000 and from 10 to 10000 µg L -1 , respectively.
For blood, a similar procedure was used (six different concentrations of each solvent, 5 replicates per concentration).The concentrations ranged from 16.6 to 1000 µg L -1 for toluene and from 33.2 to 2000 µg L -1 for benzene and n-hexane when CAR-PDMS fibers were used.The same procedure was adopted for PDMS fibers.However, the concentrations of toluene and benzene varied between 50 and 1000 µg L -1 and n-hexane between 100 and 2000 µg L -1 .
The lowest and highest values of the linear range for each analyte were defined taking into account the LOQ (limit of quantification) and the detector sensitivity response to each solvent.Isooctane was used as an internal standard for urine samples and ethylbenzene for bloods samples.
The linear regression was calculated by the least mean square method.Quantification limit (LOQ) was established as the concentration that provides an analytical signal ten times higher than noise with adequate precision (RSD < 20%).The LOQ for each solvent was the first concentration in the calibration curves.The intra-assay and inter-assay precision was estimated using spiked samples of urine or blood in five replicates at three concentrations levels within the linearity range.

Biological samples
Urine and blood samples were collected from healthy, non-smoking volunteers unexposed to VOC's (blank samples) for method optimization and validation.Samples from workers handling leather glue in shoe repair shops were obtained from seven volunteers and collected at the end of their work shift.Spot urine specimens were collected directly from the donor in 50 mL polyethylene vials.Aliquots of 8 mL from each sample were immediately transferred to headspace vials containing the internal standard and salt and then sealed.Blood was collected in a disposable system (BD Vacuntainer ® , Franklin Lakes, NJ, USA) and 2 mL aliquots were taken and treated as described for urine.The collection of urine and blood was in agreement with criteria established by the Ethics Committee of the Federal University of Alfenas (Unifal-MG).

Results and Discussion
Most of the published methods for analysis of unaltered solvents in biological fluids use GC/MS.There are few reported methods using HS-SPME followed by GC/FID to evaluate occupational exposure, no one for n-hexane.
Lee et al. 27 evaluated the occupational exposure of benzene and toluene in urine and blood, but analyzing these compounds in rat blood and urine.Alegretti et al. 23 quantified benzene, toluene and xylenes in whole blood to evaluated extreme exposure in inhalants addicts.Asakawa et al. 26 measured the occupational exposure of toluene and xylenes in urine of plastic films industry worker's.Moreover, poor epidemiological data about occupational exposure of Brazilian's shoes workers were found in the literature.

Optimization of the chromatographic conditions
An adequate resolution was obtained using the ZB-1 megabore column (30 m × 0.53 mm i.d.× 5 µm film thickness).Oven temperature was kept at 80 ºC for 3 min and then heated at a rate of 10 ºC min -1 until reaching 150 ºC.The temperature was maintained at 150 ºC for 3.2 min and then raised to 220 ºC at a rate of 40 ºC min -1 and then held for 6 min.Desorption was set at 240 ºC for 5 min (PDMS fiber) or at 260 ºC for 10 min (CAR-PDMS fiber), which allowed a complete desorption, without carryover.
Better resolution among analytes and endogenous components from urine and blood as well as minimum peak tailing were observed under the optimized conditions.Moreover, ZB-1 presented an adequate film thickness and hydrophobic phase which allowed better separation when compared to the other columns tested.Figures 1 and 2 present chromatograms of blank urine and blood samples, respectively.In both chromatograms no interfering peaks from endogenous compounds were observed among the analytes retention times (6-12 min).
Figures 3 and 4 show chromatograms of spiked urine and blood, respectively.The peaks obtained with CAR-PDMS fibers show peak tailing, which can be attributed to the slower desorption rate from this fiber.However, significantly greater responses were obtained with CAR-PDMS fibers when compared with PDMS fibers.

HS-SPME conditions
Table 1 summarize the variables studied and the ones selected as optimum for each solvent in urine and blood respectively.
A great number of fiber phases are commercially available today.][34] The mechanism of extraction varies depending on the type of coating used.PDMS coating is one of the most widely used coatings for extracting volatile compounds because of the relative ease of extraction by absorption, which is not affected by competition between analytes.This process depends on the partition coefficient sample/ headspace.At the same time, coating thickness and molecular size of analyte will determine the capacity of absorption on the fiber. 1 On the other hand, the sensitivity of mixed-phase SPME coatings, such as CAR-PDMS, is reported to be much higher compared to PDMS for extracting volatile compounds. 32,33In addition, CAR-PDMS coated fibers  exhibit micro and mesopores with 2-20 Å diameter, appropriate for extracting compounds, such as solvents, containing 2-12 carbons in their molecular structures.These coatings extract primarily via adsorption into the fiber coating and require more time to reach equilibrium or saturation of the pores.The results showed in Table 1 demonstrate the higher sorption time of CAR-PDMS compared to PDMS fibers.Also, different desorption times and temperatures emphasize these properties: CAR-PDMS fibers required 10 min at 260 ºC for complete desorption, without carryover, while PDMS required only 5 min at 240 ºC.Despite this higher sorption time, greater amounts of volatile and semi-volatile compounds were extracted by CAR-PDMS compared to PDMS.The competitive adsorption and displacement effects make mass equilibrium and quantification particularly challenging, as observed in the narrow linear range of CAR-PDMS compared with PDMS fibers 35 (see Tables 2 and 3).

Method validation
Tables 2 and 3 report the validation parameters established for urine and blood respectively.For urine, R 2 values > 0.99 were obtained for all solvents, regardless of the fiber type.For blood samples R 2 values > 0.99 were obtained, except for toluene (R 2 = 0.98) in PDMS fibers.Variation of coefficients lower than 17% was observed for both fibers.These results are in accordance with ANVISA, 36 FDA 37 and IUPAC 38 requirements.
The higher slope associated with PDMS fibers indicates better sensitivity when using this coating for urine analysis.However, CAR-PDMS demonstrated better detectability and lower LOQ values compared to the PDMS fibers.The LOQ values for blood samples were higher than those for urine since blood contains endogenous hydrophobic substances, which have an affinity to these solvents, making volatilization of the analytes more difficult.The sample volumes of 8 mL for urine and 2 mL for blood were selected to obtain a lower LOQ.In the case of blood it is not recommended to use a large sample volume due to the invasive procedure of sampling.

Samples from occupationally exposed workers
The validated method was applied to the determination of toluene, benzene and n-hexane in urine and blood collected from seven shoe workers who handle glues containing solvents, mainly toluene and n-hexane.In blood samples, the use of CAR-PDMS fibers allowed the quantification of toluene (from 16.0 to 55.2 µg L -1 ) in all samples.Only one blood sample presented result for toluene when using PDMS fiber (53.9 µg L -1 ).The biological exposure index (BEI) for toluene in blood, as recommended by the ACGIH, is 50 µg L -1 .This represents the maximum allowable blood concentration which presents no risk for the majority of workers exposed to the solvent in an occupational environment.Only one worker presented blood toluene higher than the allowable BEI, while the other workers were within acceptable exposure levels.Figure 5 shows a chromatogram of the analysis of blood sample from a shoe worker.

Conclusions
The HS-SPME analytical method proposed for toluene, benzene and n-hexane analysis in blood and urine is highly sensitive and specific and can be performed with equipment available in most laboratories (HS/SPME-GC/FID).The LOQ, precision and linearity obtained support its usefulness in routine biomonitoring analysis, with CAR-PDMS fibers, having lower levels of detection for solvent exposure and is recommended for blood analysis.

Figure 2 .
Figure 2. Chromatogram of blank blood extracted by PDMS fiber (I) and by CAR-PDMS fiber (II), containing 50mL de methanol(a).

Table 1 .
Levels of the HS-SPME parameters evaluated during univariate optimization for urine and blood samples *Optimized values.

Table 3 .
Validation data obtained for the solvents analysis in blood samples

Table 2 .
Validation data obtained for the solvents analysis in urine samples