Abstract
Objectives: This study aims to develop and validate a sampler to measure workers’ exposure to the vapor and mist of benzyl alcohol. Methods: Recovery rate, extraction and desorption rates, breakthrough, and storage stability were tested using Slim-J connected to a glass fiber filter upstream (the connected sampler). The recovery rate of the connected sampler was compared with that of XAD-7. Benzyl alcohol on the filter was extracted, and that in Slim-J resin was desorbed by methanol with an internal standard (N,N-dimethylformamide). Benzyl alcohol was quantified through gas chromatography using a flame ionization detector (GC/FID). Air sampling was conducted by attaching the connected sampler to the chest of a worker during bridge paint film removal. Results: Calibration curves showed linearity with correlation coefficients >0.999. The lower limit of quantification was 0.54 mg/m3 of the airborne concentration at 5-mL desorption with 120-L air sampling. The recovery rates of the connected sampler were 101–103%, whereas those of XAD-7 were 72–78%. The average extraction efficiency from the filters was 105.3%, whereas those from Slim-J were 94.5%. No breakthrough was recognized by aeration at 1 L/min for 120 min. Benzyl alcohol in the sampler was stable for up to 7 days. The sampled air by the connected sampler during bridge paint film removal indicated an isolated peak by GC/FID. Conclusions: The connected sampler is reliable and suitable for measuring levels of personal exposure to benzyl alcohol in vapor and mist phases.
Introduction
Benzyl alcohol (CAS Registry Number®: 100-51-6), a colorless transparent liquid, has been recently used as a substitute for dichloromethane in removers for repainting structures, such as bridges1). According to the Technical Committee on Removal of Coating by Waterborne Release Agent, the benzyl alcohol contents of currently available removers range from small percentages to 75%2). Workers apply the paint remover onto the paint films of steel structures using a sprayer, roller, or brush, allowing it to remain for a designated period. After the paint film swells and softens, workers remove it with a spatula or scraper. However, fatal intoxication accidents among workers have been reported during the peeling of the bridge coatings, which is typically performed in enclosed areas during the removal process of the paint film from bridges3). For instance, Imada et al. reported eight cases of benzyl alcohol intoxication from March 2014 to July 20194). However, the exposure levels to benzyl alcohol in these cases have not been clarified. It is crucial to determine workers’ exposure to benzyl alcohol (vapor and mist) to prevent intoxication.
Several methods have been described to monitor personal exposure to benzyl alcohol in workplace air5,6,7). For example, the United States Department of Labor’s Occupational Safety and Health Administration (OSHA) method No. PV20095) aims to measure the vapor phase of benzyl alcohol. However, the mist from removers containing benzyl alcohol is widely recognized as the primary source of exposure in work environments, considering that workers frequently utilize sprayers for applying removers. In addition, these removers’ liquid viscosity is high, and their boiling point and saturation vapor pressure are 206°C and 0.125 hPa (25°C), respectively8).
In this study, we aimed to develop and validate a devised personal exposure sampler that simultaneously collects vapor and mist of benzyl alcohol. Two collecting devices (samplers) consisting of a filter and an absorption tube connected tandemly were tested. One was XAD-7 (SKC Inc., Eighty Four, PA, USA) with a built-in glass fiber filter, whereas the other was the connected sampler. The connected sampler consisted of an adsorbent (Slim-J) and a glass fiber filter upstream (GL Sciences Inc., Tokyo, Japan). The connected sampler was validated through comparison with XAD-7.
The analysis of benzyl alcohol was based on OSHA method No. PV20095). The occupational exposure limit for benzyl alcohol was set at 25 mg/m3 as the maximum allowable concentration by the Japan Society for Occupational Health (JSOH)9). The Deutsche Forschungsgemeinschaft (DFG) set the time-weighted average at 22 mg/m3. However, OSHA, the National Institute for Occupational Safety and Health, and the American Conference of Governmental Industrial Hygienists did not set the occupational exposure limit for benzyl alcohol10). Therefore, we aimed to measure benzyl alcohol concentrations from 2.5 to 50 mg/m3, 1/10 to twice the maximum allowable concentration by JSOH.
Methods
Chemicals
Benzyl alcohol was purchased from Tokyo Kasei Kogyo (Tokyo, Japan). Methanol (≥99.8% purity), N,N-dimethylformamide (≥99.5% purity), and benzoic acid (≥99.5% purity) were purchased from Wako Pure Chemical Industries, Ltd. (Tokyo, Japan). Methanol was for pesticide residue-PCB analysis grade. N,N-dimethylformamide and benzoic acid were of analytical grade.
Samplers
Two types of samplers were tested in this study. One sampler was XAD-7 (200/100 mg adsorbent with built-in a glass fiber filter) (catalog no. 226-57A; SKC Inc.). The other was an InertSep Slim-J AERO SDB400 collection tube (400 mg) (catalog no. 5010-65780; GL Sciences Inc.) (Slim-J), a luer device type collection tube using styrene–divinylbenzene adsorbent, attached with a glass fiber filter upstream. The glass fiber filter, which was specially prepared by GL Sciences Inc., was placed in a polypropylene holder and fixed on the top of Slim-J with a Teflon ring (the connected sampler) (Figure 1). SKC Universal Pumps (model no. 224-PCXR4; SKC Inc.) were used to draw air through both samplers.
Fig. 1. Comparison of samplers. (A) The connected sampler (a glass fiber filter+InertSep Slim-J AERO SDB400). (B) XAD-7 (200/100 mg, with a built-in glass fiber filter)
After sampling, the filters were removed from the holders and placed in 4-mL glass vials. Then, these were sonicated with 2 mL of extracting methanol with an added internal standard (N,N-dimethylformamide; 0.24 mg/mL) for 15 min and allowed to stand for 45 min. Slim-J was desorbed with 5 mL methanol with an added internal standard (N,N-dimethylformamide; 0.24 mg/mL) using a syringe at a flow rate of about 1 mL/min. The eluate was metered up to 5 mL. One microliter of the liquid was then injected into a gas chromatograph with a flame ionization detector (GC/FID).
Instruments
We used a 7890A GC/FID (Agilent Technologies, Palo Alto, CA, USA). The column was a dimethylpolysiloxane (50 m×0.53 mm I.D., 5 μm; catalog no. 007-1; Quadrex Corporation, Bethany, CT, USA). Helium was used as the carrier gas at a flow rate of 6.0 mL/min. The inlet and detector temperatures were set to 270°C, and samples (1 μL) were injected in a split mode (split ratio, 1:1). The oven temperature was held at 35°C for 3 min and then increased to 150°C at a rate of 10°C/min and to 200°C at 5°C/min.
Method validation
The proposed method was validated according to the guidelines of Japan’s Ministry of Health, Labour and Welfare (MHLW)11).
Calibration curves were prepared in five standard series ranging from 0.013 to 5.2 mg/mL, and the linearity of the calibration curves was checked. Five samples of 0.026 mg/mL of the standard (air concentration: 1.08 mg/m3 measured at a 120-L sampling volume) were analyzed to confirm the lower limit of quantification (LOQ).
A standard solution at a certain concentration was spiked onto the filter of a sampler. At the same time, room air (temperature, 21.2°C–23.5°C; relative humidity, <23.8%) was drawn through the sampler at a flow rate of 1 L/min for 120 min. For storage, the filter was sealed in a vial, Slim-J was tightly stoppered at both ends with the provided stoppers, and both were stored in a refrigerator (4°C).
The recovery rate of the connected sampler was compared to that of XAD-7. Benzyl alcohol was introduced to the XAD-7 filter and the filter of the connected sampler at amounts of 0.052 and 0.520 mg, respectively. Room air was drawn through them at 1 L/min for 10 min, using an SKC universal pump.
In the extraction and desorption rate tests, benzyl alcohol of 0.26, 2.6, and 5.2 mg (equivalent to 2.2, 21.7, and 43.3 mg/m3×120 L, respectively) were introduced separately to the filter and Slim-J, with a total of five replicates (n=5). The filter was not aerated; however, Slim-J was aerated at 1 L/min for 10 min. Both were stored in a refrigerator (4°C) for 1 day and then extracted or desorbed to check the extraction or desorption rate*1.
For the recovery tests, 0.26, 2.6, and 5.2 mg of benzyl alcohol were spiked to the filter of the connected sampler. The samples were aerated for 10 or 120 min, and the filter and Slim-J were then treated using methanol with an internal standard on the same day to check the recovery rate*2 (n=5). In addition, a breakthrough test was conducted using two Slim-J samplers connected tandemly and aerated for 120 min at 1 L/min by spiking 2.6 and 5.2 mg of benzyl alcohol on the filter (n=3).
For the storage stability tests, 0.26 and 5.2 mg of benzyl alcohol were spiked to the filter and Slim-J separately (n=3). The filter was not aerated, but Slim-J was aerated at 1 L/min for 10 min. After the addition of a standard solution, the filter was placed in a vial and capped. Slim-J was aerated, and both its ends were tightly sealed. They were stored in a refrigerator (4°C) for up to 7 days. Extraction and desorption were performed after 1, 3, and 7 days of storage and compared with the recovery rate on Day 0 of storage as the standard.
*1 extraction or desorption rate = benzyl alcohol in the solution/spiked benzyl alcohol on the filter or Slim-J
*2 recovery rate = benzyl alcohol in the solution after aeration/spiked benzyl alcohol on the connected sampler
Sampling at a workplace
A personal sampling of a worker engaged in removing bridge paint film was conducted using the connected sampler with a pump of AIRCHEK TOUCH (model no. 220-5000TC; SKC Inc.) at 1 L/min for approximately 120 min.
Results
Calibration curves and LOQ
We prepared a standard solution by adding benzyl alcohol in an extraction and desorption solution. The calibration curves showed linearity in the concentration range of 0.013–5.2 mg/mL of the standard solution, with a correlation coefficient of >0.999. The LOQ was defined as 10 times the standard deviation (SD) (n=5) of the peak area ratio (benzyl alcohol/IS) of 0.026 mg/mL of the standard solution (corresponding to 1.08 mg/m3 by aeration at 1 L/min for 120 min)11). The LOQ was assumed to be 0.00048 mg/mL from the calibration curves. However, the reproducibility of lower than 0.013 mg/mL was not precise: for example, 0.0026 mg/mL; 0.0408±0.0121% (relative SD [RSD] 29.7%) and 0.0065 mg/mL; 0.1030±0.0142% (RSD 13.8%), in contrast with 0.013 mg/mL; 0.1876±0.0021% (RSD 1.1%). Therefore, the actual LOQ was set to the lowest concentration of the calibration curve (0.013 mg/mL). This corresponds to 0.54 mg/m3 of the airborne concentration at an analytical condition of 5 mL desorption with 120 L air sampling, which is sufficient to analyze 1/10 of the maximum allowable concentration of benzyl alcohol proposed by JSOH9) (25 mg/m3).
Recovery rate of benzyl alcohol by the connected sampler compared with that of XAD-7
Table 1 shows the amount of each spike and the recovery rate of benzyl alcohol when the chemical was spiked on the filter of XAD-7 and that of the connected sampler. Whereas the recovery rate of XAD-7 was 70–80%, and that connected sampler was almost 100%.
Table 1. Comparison of benzyl alcohol recovery rates between XAD-7 (
n=5) and Slim-J attached with the glass fiber filter (the connected sampler) (
n=3)
| Sampler | Spiked amount (mg) | Mean±SD (%) | RSD (%) |
|---|
| XAD-7 | 0.052 | 78.0±5.9 | 7.6 |
| 0.520 | 72.0±7.0 | 9.7 |
| connected sampler | 0.052 | 102.7±2.9 | 2.8 |
| 0.520 | 101.4±1.7 | 1.7 |
SD, standard deviation; RSD, relative standard deviation.
Note: Solutions containing a specific amount of benzyl alcohol were introduced onto the XAD-7 filter and the connected sampler’s filter. Simultaneously, room air was drawn through the sampler at a flow volume of 1 L/min for 10 min.
The extraction and desorption rates of the filter and Slim-J were examined separately. The extraction rates from the filters was 105.3% (RSD 0.7–1.7%, n=5), whereas the desorption rates from the Slim-J was 94.5% (RSD 1.3–4.0%, n=5).
Recovery of benzyl alcohol from the connected sampler after sampling
Recovery rates were examined by spiking benzyl alcohol to the filter of the connected sampler. As shown in Table 2, their respective recovery rates varied depending on the added volume and aeration time. However, the total recovery rates per connected sampler (as the whole sampler) were over 89.5% (RSD 0.7–7.4%, n=5), which were excellent and stable.
Table 2. Recovery tests (
n=5) (sampling volume: upper row 10 L, lower row 120 L)
Spiked amount
(mg) | Aeration time
(min) | Filter recovery | Slim-J recovery | Total recovery |
|---|
| Mean±SD (%) | RSD (%) | Mean±SD (%) | RSD (%) | Mean±SD (%) | RSD (%) |
|---|
| 0.26 | 10 | N.D. | ― | 96.7±3.3 | 3.4 | 96.7±3.4 | 3.3 |
| 120 | N.D. | ― | 92.0±6.8 | 7.4 | 92.0±6.8 | 7.4 |
| 2.6 | 10 | 94.2±1.4 | 1.5 | 9.0±2.2 | 23.8 | 103.2±1.6 | 1.5 |
| 120 | N.D. | ― | 91.1±4.5 | 4.9 | 91.1±4.3 | 4.7 |
| 5.2 | 10 | 93.9±3.2 | 3.4 | 4.9±2.9 | 59.8 | 98.9±0.7 | 0.7 |
| 120 | 27.4±33.9 | 124.0 | 62.2±29.8 | 48.0 | 89.5±4.5 | 5.1 |
N.D., none detectable (<limit of quantification); RSD, relative standard deviation; SD, standard deviation.
Note: Solutions with a given amount of benzyl alcohol were spiked onto the filter of the connected sampler. Simultaneously, room air was drawn through the sampler at a flow volume of 1 L/min for 120 or 10 min. The spiked amounts corresponded to air concentrations of approximately 2.5–50 mg/m3 in a 120-L air sampling for benzyl alcohol.
In addition, two Slim-J samplers were connected in tandem and tested for breakthrough (Table 3). For the addition of 2.6 mg of benzyl alcohol (equivalent to 25 mg/m3), the total recovery of the filter and fore Slim-J ranged from 89.6% to 96.8%. No benzyl alcohol was detected in the rear Slim-J; that is, no breakthrough was observed. Meanwhile, the addition of 5.2 mg of benzyl alcohol (equivalent to 50 mg/m3) showed about 1.1–2.1% breakthrough. However, the stable total recovery rate of the filter and fore Slim-J in the front layer at over 90% suggests that the results can be corrected by the total recovery rate.
Table 3. Breakthrough tests (
n=3)
| Spiked amount (mg) | Recovery (%) |
|---|
| Filter+fore Slim-J | rear Slim-J |
|---|
| Mean±SD (%) | RSD (%) | Mean±SD (%) | RSD (%) |
|---|
| 2.6 | 93.0±3.6 | 3.9 | N.D. | ― |
| 5.2 | 91.3±2.0 | 2.1 | 1.6±0.5 | 31.4 |
N.D., none detectable (<limit of quantification); RSD, relative standard deviation; SD, standard deviation.
Note: Two Slim-J samplers were connected and a glass fiber filter was attached. The standard solution was spiked onto the filter. Simultaneously, room air was drawn through the sampler at a flow volume of 1 L/min for 120 min. The spiked amounts corresponded to air concentrations of approximately 25–50 mg/m3 in a 120-L air sampling for benzyl alcohol.
Table 4 shows the recovery of benzyl alcohol in the filter and Slim-J separately after 1, 3, and 7 days. The recovery was at least 90% for 7 days (SD 0.2–7.2%, RSD 0.2–7.5%, n=3) at 0.26 and 5.2 mg, which confirmed that they could be stored in a refrigerator at 4°C for up to 7 days. Although there was a concern that benzyl alcohol could oxidize to benzoic acid during the aeration and storage processes, the benzoic acid peak was confirmed, and no benzoic acid was detected in the extraction liquid.
Table 4. Storage stability (
n=3)
| Sampler | Spiked amount (mg) | day | Recovery (%) | RSD |
|---|
| Mean±SD | (%) |
|---|
| Filter | 0.26 | 0 | 100.0±0.6 | 0.6 |
| 1 | 99.9±1.3 | 1.3 |
| 3 | 99.1±1.4 | 1.4 |
| 7 | 97.8±0.2 | 0.2 |
| 5.2 | 0 | 100.0±0.8 | 0.8 |
| 1 | 101.5±0.7 | 0.7 |
| 3 | 98.2±1.6 | 1.7 |
| 7 | 100.8±0.4 | 0.4 |
| Slim-J | 0.26 | 0 | 100.0±1.7 | 1.7 |
| 1 | 100.2±4.9 | 4.9 |
| 3 | 96.8±7.2 | 7.5 |
| 7 | 99.9±3.2 | 3.2 |
| 5.2 | 0 | 100.0±5.0 | 5.0 |
| 1 | 100.4±4.0 | 4.0 |
| 3 | 101.7±4.8 | 4.8 |
| 7 | 94.5±6.8 | 7.2 |
RSD, relative standard deviation; SD, standard deviation.
Note: Filters and Slim-J were spiked separately with the standard solution. Simultaneously, room air was drawn through Slim-J at a flow volume of 1 L/min for 10 min. The spiked amounts corresponded to air concentrations of approximately 2.5–50 mg/m3 at 120-L air sampling for benzyl alcohol. The samplers were sealed tightly and stored in a refrigerator at 4°C for up to 7 days.
Typical chromatographs of a standard solution (A) and a collected air sample at the workplace, (B) a filter, and (C) Slim-J are shown in Figure 2. The peak of benzyl alcohol was isolated and not interfered with.
Fig. 2. Chromatograms of (A) a 0.26-mg/mL standard solution (equivalent to 10.8 mg/m3, 120-L sampling, 5 mL desorption), (B) a solution extracted from a filter, and (C) a solution desorbed from Slim-J. (B) and (C) were for the connected sampler, which was attached to a worker’s chest (breathing zone) during bridge paint film removal. Two chart examples of the field data showed a total of 41.3 mg/m3 of benzyl alcohol. Peak 1, Internal Standard: I.S.; N,N-dimethylformamide (0.24 mg/mL). Peak 2, benzyl alcohol. No inhibitory peaks were detected.
Discussion
In this study, the connected sampler showed a recovery rate of more than 90%, no breakthrough at 1 L/min for 120 min, and 100% storage stability for 7 days at 0.26 to 5.2 mg (equivalent to 2.5–50 mg/m3). The target concentration range (2.5–50 mg/m3) spans from 1/10 to twice the allowable concentration set by JSOH at 25 mg/m3. Ensuring precise measurement within this range is crucial for occupational health. When the benzyl alcohol concentration is 1/10 or less, the risk to workers is minimal. If it is twice or more, measurement is achievable by reducing the sampling time. The recovery rate of the connected sampler was better than that of the XAD-7. Moreover, the sampling by the connected sampler in a workplace during bridge paint film removal showed a clear peak by GC/FID analysis. Thus, the connected sampler is considered a suitable sampler for measuring personal exposure levels of benzyl alcohol.
OSHA PV2009 recommends the use of XAD-7 (100/50 mg) without a glass fiber filter5). However, we subjected XAD-7 (200/100 mg) to experimentation in this study. XAD-7 (200/100 mg) consisted of a built-in glass fiber filter upstream of two layers of adsorbent in the same tube to collect both vapor and mist. However, at 0.052 mg to 0.52 mg of benzyl alcohol addition and 10-L sampling, XAD-7 showed poor recovery (72.0–78.0%).
The DFG maximum workplace concentration (MAK, 2019, Vol 4, No. 3)6) was designed to collect benzyl alcohol vapor and mist, but the sampling system consisted of a quartz fiber filter and a collection stainless tube filled with 75 mg of Tenax TA (35–60 mesh) connected downstream for thermal desorption. In addition, the analysis was conducted using gas chromatography–mass spectrometry. Although we did not compare our method to that of DFG in this study, our method using the connected sampler was enough to analyze the collected vapor and mist in workplaces.
Conclusions
Using the connected sampler (Slim-J connected to a glass fiber filter upstream), we developed a method for measuring benzyl alcohol in both vapor and mist phases in working environments. This connected sampler conducted measurements that maintained a recovery rate of about 90% for 120 min of collection in the range equivalent to 1/10 to 2 times 25 mg/m3.
Acknowledgments
This study was commissioned by the MHLW as part of the FY2021 Project for the Promotion of Risk Assessment of Chemical Substances in the Workplace (Exposure Survey). We would like to sincerely thank Enago for their expertise in linguistic and textual editing and their assistance in the preparation of the manuscript.
Disclosure
Approval of the research protocol: This study did not require an ethical review because the subject of this study was neither human nor animal.
Conflicts of interest
The authors declare that there are no conflicts of interest.
References
- 1. Chemical Substances Control Division, Safety and Health Department, Labor Standards Bureau, Ministry of Health, Labor and Welfare (MHLW). Prevention of Occupational Accidents in Paint Stripping Work with Stripping Agents-No. 1. https://www.mhlw.go.jp/content/11300000/000940862.pdf. http://www.mhlw.go.jp/stf/seisakunitsuite/bunya/0000149924.html (in Japanese). Accessed December 28, 2022.
- 2. Technical Committee on Removal of Coating by Waterborne Release Agent. Aichi, Japan. Technical data-12 Prevention of Occupational Accidents in Paint Stripping Work Using Water-Based Paint Stripping Agents for steel Structures. https://www.c-wra.jp/_files/ugd/0eab65_f0bf9aa58b154d4584ce39c12cfcc971.pdf (in Japanese). Accessed March 28, 2023.
- 3. Ito C, Yamada E, Kasuya M, et al. A case of benzyl alcohol poisoning by inhalation of paint stripper. Nihon Kyukyu Igakukai Zasshi. 2018; 29: 254-259 (in Japanese).
- 4. Imada Y, Hatano Y, Ida K, et al. Acute health hazards associated with stripping operations using paint stripping agents containing benzyl alcohol. Jpn J Toxicol. 2020; 33: 94-98.
- 5. Occupational Safety and Health Administration. OSHA Method. http://www.osha.gov/sites/default/files/methods/osha-pv2009.pdf. Accessed December 22, 2022.
- 6. Tschickardt M, Krämer W, Schmitt R, Hebisch R, Brock TH, Hartwig A, MAK commission, et al. Benzyl alcohol–Method for the determination of benzyl alcohol in workplace air using gas chromatography after thermal desorption [Air Monitoring Methods, 2018]. MAK Collect Occup Health Saf. 2019; 4: 1667-1682.
- 7. Eiji T. Survey on Chemical Substances in the Environment (Initial Environmental Survey: Benzyl Alcohol in Air). http://www.city.sapporo.jp/eiken/annual/no35/documents/taikigaiyo.pdf. Accessed December 22, 2022.
- 8. Ministry of Health, Labour and Welfare (MHLW). Japan. Workplace Safety [Web site]: Various Teaching Materials and Tools. https://anzeninfo.mhlw.go.jp/anzen/gmsds/100-51-6.html (in Japanese). Accessed December 22, 2022.
- 9. The Committee for Recommendation of Occupational Exposure Limits, Japan Society for Occupational Health. Documentation of the proposed reasons for provisional occupational exposure limits (2021). Sangyo Eiseigaku Zasshi. 2021; 63(5): 179-211 (in Japanese).
- 10. American Conference of Governmental Industrial Hygienists (ACGIH). Guide to occupational exposure values. Benzyl Alcohol ACGIH 2022; 1: 22.
- 11. Ministry of Health, Labour and Welfare (MHLW). Japan. Guidelines For Exposure Assessment Of Workers To Hazardous Substances. http://www.mhlw.go.jp/shingi/2010/01/s0115-4.html (in Japanese). Accessed December 23, 2022.
