EVALUATION OF THE USE OF LOW FLOW PASSIVE SAMPLING TECHNIQUE IN OFFSET PRINTING PLANTS

Transcription

1 Nofer Institute of Occupational Medicine, Łódź, Poland ORIGINAL PAPERS International Journal of Occupational Medicine and Environmental Health 2006;19(4): DOI /v EVALUATION OF THE USE OF LOW FLOW PASSIVE SAMPLING TECHNIQUE IN OFFSET PRINTING PLANTS MERJA HAUTAMÄKI 1,2, PENTTI KALLIOKOSKI 1, MARKO HYTTINEN 1, PERTTI PASANEN 1, JUHA LAITINEN 3, JUHANI KANGAS 3, RITVA LUUKKONEN 2, and STUART BATTERMAN 4 1 Department of Environmental Sciences University of Kuopio, Finland 2 Finnish Institute of Occupational Health Helsinki, Finland 3 Finnish Institute of Occupational Health Kuopio, Finland 4 Department of Environmental Health Sciences School of Public Health University of Michigan Ann Arbor, Michigan, MI, USA Abstract Objectives: The aim of this study was to evaluate the suitability and applicability of low-flow passive tubes for sampling of organic solvents in occupational environment. Materials and Methods: Laboratory and field studies were conducted to continue the evaluation of low-flow diffusive sampling combined with thermal desorption and gas chromatographic/mass spectrometric analysis in occupational hygiene surveys. Passive sampler tubes with Tenax GR adsorbent were employed to assess exposures to organic solvents in 10 small sheet-fed offset printing plants in Finland. The retention of the solvent compounds in the samplers was investigated using laboratory chamber studies. Active sampling with activated carbon tubes served as the control method. Results: The laboratory tests showed that passive tubes had good retention, stability and reproducibility for the solvent compounds. Parallel passive and active sampling in both breathing zone and area monitoring in the printing plants yielded similar measurements though passive sampling tended to give slightly higher concentrations (by 5 to 12%). The suitability of passive sampling to be done by workers was also confirmed. The study also shows that exposure to organic solvents can be effectively reduced by the use of vegetable oil-based s. Especially, since ventilation and other protective measures are often inadequate in small facilities, the use of toxic cleaning solutions should be avoided. Conclusions: Passive sampling tubes with capillary orifice analyzed with thermal desorption GC-MS are well suited for measuring occupational solvent exposure. Method is well suited for sampling done by workers. Key words: Passive sampling, Printing, Volatile organic compound INTRODUCTION A large number of printing industry workers are employed in small sheet-fed offset plants. Workers are exposed to organic solvents due to the use of s and 2-propanol as well as of other chemicals in the fountain solution of the press. The fountain solution moistens the plate cylinder of the press and usually contains approximately 10% of 2-propanol. Assessment of occupational exposure is mandatory within the European Union. However, occupational hygiene surveys can be expensive and This study was supported by the Finnish Work Environment Found. Received: September 29, Accepted: November 13, Address reprint requests to M. Hautamäki, MSc, Finnish Institute of Occupational Health, Topeliuksenkatu 41 a A, FI Helsinki, Finland ( merja. hautamaki@ttl.fi). 228

2 PASSIVE SAMPLING TECHNIQUE IN OFFSET PRINTING PLANTS ORIGINAL PAPERS laborious for small companies, such as the printing plants investigated in this study. Additionally, offset printing employees are a good example of a worker group exposed to a complex mixture of chemicals. Thus, simple sampling methods, e.g., colorimetric detector tubes, cannot be used and exposure assessment is generally based merely on the list of chemicals used and their safety data sheets. Diffusive samplers are easy to use and allow sampling done by the workers themselves. However, at least in Finland where the badge-type diffusive samples have been used almost exclusively, this possibility has not been widely utilized due to the risk of samples contamination. This risk is smaller for diffusion tubes. Low-flow passive sampling with thermal desorption tubes was found to allow convenient long-term sampling at an offset printing facility in Michigan [1]. The method was demonstrated to be capable of sampling compounds with concentrations spanning nearly five orders of magnitude. In the USA, aromatic hydrocarbons are often used as cleaning solvents in offset printing [2]. This was also the case in the plant studied in Michigan, USA [1]. The use of glycol ethers and chlorinated hydrocarbons has also been reported [3]. In Finland, solvent naphtha-based cleaning solutions containing 20 30% of 2-propanol, have traditionally been used. These solvents are being replaced by vegetable oil-based products, which still often contain solvent naphtha, but are generally free of 2-propanol. However, exposure of worker to 2-propanol continues even in plants using vegetable oil-based cleaning solutions due to its common use in the fountain solution of the press. The main aim of this study was to evaluate the suitability and applicability of low-flow passive tubes for sampling of organic solvents in occupational environments done by workers with low-flow passive tubes. The adsorption tube type employed in this study differs from that previously used in Michigan. Here, Perkin Elmer ADT 400 thermal desorption tubes, probably the most common type for occupational hygiene applications, were used as the basic tubes. The orifice used to resrtict the flow rate was also slightly different. Finally, this study provides information on the effect of cleaning solution type and exposure levels in 10 sheet-fed offset printing facilities, selected to be a representative sample of this types of workplace. Even though the contamination risk is very small for the tubes with capillary orifices used in this study, a small trial of sampling done by workers was performed to ensure this. MATERIALS AND METHODS Passive sampling and sample analysis Perkin Elmer stainless steel thermal desorption tubes were used in passive sampling. Tubes were loaded with 220 mg Tenax GR adsorbent. At the sampling end, the tube was equipped with a 6.35 mm diameter orifice (Swagelok SS ) to reduce the diffusive flux and to prolong the sampling time. The opposite end of the tube was sealed during sampling. Passive tubes were analyzed by an automated thermal desorption system (Perkin Elmer ATD400), gas chromatography (GC, Hewlett Packard 6890) and mass spectrometry (MS, Hewlett Packard 5973). Desorption conditions were as follows: tube desorption temperature was 250 C, helium flow was 50 ml min -1, the cold trap temperature was -30 C, and the secondary desorption temperature from the ATD cold trap was 200 C. A dimethylsiloxane copolymer column (HP5-MS) was used (50 m 0.2 mm, film thickness 0.5 mm). The GC oven temperature program started at 40 C (1 min hold) and was raised 5 C min -1 to 210 C and then raised 20 C min -1 to the final temperature of 280 C, which was maintained for 8 min. The identification of the collected compounds was accomplished by retention times, standards, and the GC/MS data library. Standards (1:1000, 1:500, 1:100) were made in methanol using the solvents used in printing plants. Standards were injected into the tubes, and analyzed as described above. Solvent naphtha was quantified using five main peaks of the cleaning solvents. Detection limit was measured by diluting target compounds so that the height of the peaks in the chromatogram was at least 5 times higher than the noise of the baseline (clearly identifiable peak). This was assured by four hour passive sampling test made in a chamber. Detection limits for a 4-h passive sampling for 2-propanol, hexane, heptane and undecane were 0.27 mg/m 3, 229

3 ORIGINAL PAPERS M. HAUTAMÄKI ET AL mg/m 3, 0.20 mg/m 3 and 0.58 mg/m 3, respectively. Diffusion coefficient for 2-propanol was obtained from literature [4]. For cleaning solvents, diffusion coefficients of the main compounds were calculated [4]. Passive sampling rate was determined for 2-propanol and each cleaning solvent as follows: (1) where: U m is the sampling rate (ml/min), D is the diffusion coefficient (cm 2 /s), A is the cross-section of the tube (cm 2 ), Z is the length of the air space in the tube (cm). Active sampling and sample analysis Active sampling with activated carbon tubes (SKC, ) was selected as the control method due to its wide use in occupational hygiene sampling in Finland. The air flow of the pumps (SKC, Model 222) was adjusted to 0.1 l min -1. Activated carbon tubes were desorbed with 2 ml carbon disulfide containing 5% 2-(2-butoxyethoxy)ethanol. Samples were analyzed by GC-FID (Hewlett Packard 5890 Series II) using a HP-20M Carbowax column, 50 m 0.32 mm, with a film thickness of 0.3 mm. Method conditions were as follows: the injector port and detector temperatures were 200 C and 250 C, respectively. The initial column temperature of 50 C was held for 5 min, increased to 100 C at 2 C min -1, and then raised to 140 C at 30 C min -1, where it was maintained for 15 min. The injection volume was 1 ml, and split-injection was used. The carrier gas was helium. Standards (0.2, 0.5 and 1 ml) from the cleaning solvents used in the printing plants were prepared in 2 ml carbon disulfide containing 5% 2-(2-butoxyethoxy)ethanol. Identification was accomplished using retention time, and quantification was based on the ratio of peak areas to an internal standard, 1-chloro-octane. The measured desorption efficiency of analyzed solvents was over 96%. The detection limit was calculated for ten parallel zero samples by adding three standard deviations of zero samples to the average zero signals. Detection limit for the cleaning solvents varied from 0.2 to 3.0 mg/m 3 in a 4-h air sampling. Different desorption and analyzing techniques were used to analyze passive and active samples, therefore, different detection limit definitions were used. Laboratory study Laboratory chamber tests were used to evaluate the sampling and recovery efficiency and back-diffusion of the passive Tenax GR tubes used. Eighteen tubes were put into a metal chamber (volume 120 l) and a traditional cleaning solvent, containing solvent naphtha, was fed by a diffusion vial (VICI Metronics Inc.) at a constant rate into the chamber for 4-h. The concentration inside the chamber was kept uniform by a fan. The air flow through the chamber was 6 l min -1 and it was cleaned by activated carbon. The relative humidity (RH) of the air was adjusted to 30 40% by mixing dry air and air moistened by bubbling it through deionized water. The temperature of the air was C. Immediately after sampling, six tubes were analyzed by TD/GC/MS as described earlier, and the remaining 12 tubes were kept open in a clean room. Six of these tubes were analyzed after 24 h and the remaining six tubes were analyzed after three days. Active sampling with activated carbon tube served as the control method. One 40-min active sample was taken during a 4-h passive measurement. The activated carbon tube was analyzed by GC-FID. Field study Field experiments were performed in 10 sheet-fed offset printing plants in Finland. At each plant, 4-h passive samples were collected in the breathing zones of one to three printers (personal samples) and at fixed sites (area samples) in the printing room. Area sampling was performed apart from printing machines; in the plant s office room if possible. Passive (Tenax GR) and active (activated carbon) samples were collected in parallel for personal and area samples. At least one 4-h active and one passive sample were collected in printers breathing zones and at fixed sites in each plant. Tubes were side by side during sampling. Temperature and relative humidity were measured during sampling (Vaisala temperature and humidity probe HM141) in each plant. In two plants, VOC samples were collected twice. The first sampling was done by the 230

4 PASSIVE SAMPLING TECHNIQUE IN OFFSET PRINTING PLANTS ORIGINAL PAPERS researchers and the second sampling was done by the workers on a subsequent day. The printing machines varied from small and old one-color machines to new four-color machines with automatic cleaning systems. The cleaning solvents and their main components are listed in Table 1. In six plants, printing blankets were cleaned using traditional cleaning solvents, i.e., solutions containing 20 30% of 2-propanol and solvent naphtha. Two types of solvent naphtha-based s were used, which differed from each other by the boiling point range and by the Finnish occupational exposure limit (OEL), which is the 8-h time weighted average exposure that should not be reached in the workplace. Type A contained more volatile solvent (boiling point range, C) with a lower OEL (350 mg/m 3 ) than Type B (boiling point range, C; OEL 1200 mg/m 3 ). Three types of vegetable oil-based s were used (Types C, D, and E) in four plants, all of which contained solvent Table 1. Main components of the cleaning solvents used in the studied printing plants Printing plant Cleaning solvent 3, 5 A 1, 2, 3, 5, 6, 7, 9 B Main components 2-Propanol 2-Methylhexane 3-Methylhexane Heptane Methylcyclohexane 2-Propanol Methylcyclohexane Nonane Propylcyclohexane Decane naphtha with high boiling points (> 110 C; OEL 900 mg/ m 3 ). These fluids did not contain 2-propanol. The fountain solution was also free of 2-propanol in one of the plants (#8) using vegetable oil-based s. The Finnish OEL for 2-propanol is 500 mg/m 3. All of the cleaning fluids had concentrations of aromatic compounds below 1% according to the product safety data sheets. Statistical analysis Student t-test was used to compare the mean concentrations in the samples analyzed on different days (immediately, one day and three days after sampling). Paired t-test was applied to compare parallel passive and active samples. The differences were considered statistically significant at a value of p < RESULTS Laboratory study Concentrations of the cleaning solvents compounds measured using passive Tenax GR tubes after the chamber exposure are presented in Figure 1. Averages of six tubes are shown. The measurements of solvent naphtha were highly reproducible (within 3%) over a 3-day period. Concentrations of 2-propanol showed a statistically significant decline (p < 0.05), about 12% per day, possibly due to back-diffusion losses of 2-propanol, which have a low boiling point (83 C). 4 C 7, 8 D 10 E Pentamethylheptane 3-Methyl-5-propylnonane Trimethylhexane Trimethylhexane 4.6-Dimethylundecane 4-Methyldecane Butylcyclohexane Dodecane Decane Undecane 1R,2T,4T,5C-Tetramethylcyclohexane 1,2,3,5-Tetramethylcyclohexane Trans-1,4-diethylcyclohexane 1-Methyl-3-propylcyclohexane Trans-decahydronaftalene Fig Propanol and solvent naphtha concentrations in passive tubes immediately after exposures, and after sampling for 1 and 3 days in clean space. 231

5 ORIGINAL PAPERS M. HAUTAMÄKI ET AL. Field study Temperature and relative humidity in printing plants were C and 42 58%, respectively. Solvent naphtha and 2-propanol concentrations measured by passive and active sampling methods from printers breathing zone and from area samples are presented in Table 2. The average ratios (passive/active) for personal samples were: 0.99 for 2-propanol, 0.98 for cleaning solvent A, 1.06 for cleaning solvent B, and 1.23 for cleaning solvent C E. Somewhat surprisingly, differences were less consistent for area samples. The corresponding ratios were 1.11, 1.16, 1.11 and This is probably due to low concentrations of area samples, which were close to the detection limit of the active method. Differences between parallel passive and active samples collected from the same plant were not statistically significant. Overall, the agreement between passive and active personal and static samples was 12% and 5%, respectively, which compares well with the relative overall uncertainty requirement of 50% given in EN 482 for workplace monitoring [6]. In the case of the presence of multiple hazardous impurities in the workplace air, the combined occupational exposure limit (OEL C ) represents additive exposures as follows: C 1 /OEL 1 + C 2 /OEL 2 + C 3 /OEL (2) where: C 1 is the measured concentration of compound 1, OEL 1 is the occupational exposure limit value of compound 1 etc. If the result is equal or greater than one, the combined OEL is exceeded and exposure level is considered to Table 2. Breathing zone and static concentrations of solvent naphtha and 2-propanol used in ten offset printing plants Ceaning solvent used in printing plant Concentration in breathing zone Concentration in area sample 2-Propanol Solvent naphtha 2-Propanol Solvent naphtha Passive Active Passive Active Passive Active Passive Active Naphtha-based Mean ± SD 59.7 ± 50.6 (n = 11) Vegetable oil-based 59.1 ± ± 44.4 (n = ± ± ± ± ± * * * * Mean ± SD 17.8 ± 20.6 (n = 5) Both types of s 16.0 ± ± 7.0 (n = 5) 11.9 ± ± ± ± ± (n =2) 86.7 (n = 2) 17.7 (n = 2) 19.4 (n = 2) 16.3 (n = 1) 18.3 (n = 1) 1.7 (n = 1) 1.9 (n = 1) Total Mean ± SD 48.7 ± 42.3 (n = 18) 48.9 ± 40.9 (n = 15) 58.7 ± 47.2 (n = 18) 55.8 ± 44.7 (n = 15) 19.2 ± ± ± ± 16.6 *Concentration under detection limit. SD standard deviation. 232

6 PASSIVE SAMPLING TECHNIQUE IN OFFSET PRINTING PLANTS ORIGINAL PAPERS be harmful. The calculated additive exposure effects of 2-propanol and solvent naphtha in printing plants, classified according to the type of solvent fluid used, are listed in Table 3. Even though none of the plants studied had local exhaust ventilation, solvent exposure levels remained below 50% of the combined OEL for all the plants. However, there was a clear difference between the plants using solvent naphtha and vegetable oil-based s. The mean OEL C were 0.25 and 0.06, respectively. As expected, the exposure level in the plant using both types of Table 3. The average combined solvent exposure effects in printing plants Cleaning solvent in printing plant Combined solvent exposure effect Passive sampling (n = 18) Active sampling (n = 15) Solvent naphtha-based Mean Vegetable oil-based Mean Both types of cleaning fluids Total mean Table 4. Breathing zone concentrations (±SD) of 2-propanol and solvent naphtha in sampling done by the workers Printing plant, Sampling day Concentration (mg/m 3 ) 2-Propanol Solvent naphtha 1, Day 1 (n = 2) 31.9 ± ± 2.3 1, Day 2 (n = 2) 20.6 ± ± 9.8 4, Day 1 (n = 2) 28.9 ± ± 7.6 4, Day 2 (n = 2) 36.3 ± ± 0.6 s was between these extremes and very close to the grand mean for all plants. The results of sampling done by workers are presented in Table 4. Two parallel personal samples were collected on two consecutive days by printers. Samples were collected by the workers and researchers on different days, thus concentrations in Tables 2 and 4 cannot be directly compared. DISCUSSION The laboratory study showed that passive Tenax GR tubes provided good retention and stability of the solvent compounds. Tubes kept open 3 days after sampling in a clean space still contained the same amount of solvent naphtha as the tubes that were analyzed immediately after sampling. While the concentration of the more volatile 2-propanol decreased by 25% over three days, this test represents a worst-case analysis in that exposures are likely to occur throughout the workday, and not solely at the beginning of the sampling period. The storage stability of closed tubes is very good. Earlier, Peng and Batterman [7] loaded 20 ng of VOCs in nine passive tubes and analyzed three of them immediately. The remaining tubes were analyzed after one and six weeks of storage. After 1 week, an average loss of 14% was determined for 51 VOCs from tubes wrapped in clean aluminum foil and stored in a glass jar. No further losses were observed. Thus, passive sorbent tube TD/GC/MS method well suits sampling done by workers because immediate analysis is not necessary and even poor fastening of tube caps does not fully destroy the samples. However, a positive bias may occur if poorly sealed tubes are stored in a high concentration environment. Of course, sample tubes should be carefully sealed and analyzed as soon as possible. The 2-propanol concentrations are in good agreement with the results reported by Svendsen and Rognes [8] who showed personal exposure levels of 2-propanol ranging from 2 to 179 mg/m 3 in seven offset printing plants. Concentrations of aliphatic and aromatic compounds were low in all the plants. The arithmetic mean of combined expo- 233

7 ORIGINAL PAPERS M. HAUTAMÄKI ET AL. sure was from 0.02 to 0.77 mg/m 3 [8]. In the United States, Wadden et al. [2] measured total VOC concentrations from 50 to 110 mg/m 3 in a sheet-fed offset printing plant. The combined solvent exposure ranged from 4 to 43% of OEL value. The highest combined exposures were measured in printing plants where Type A and B cleaning solvents were used. The two plants using exclusively vegetable oil-based s showed lower exposure levels, indicating that the use of this kind of s can effectively reduce exposure to solvents. This study also demonstrates that unnecessarily toxic cleaning solutions are used in US printing plants and in those of many other countries. The selection of inherently safe materials is especially important in small enterprises, such as the sheetfed offset printing plants studied here, because ventilation and other protective measures often remain inadequate. Passive sampling generally yielded slightly higher concentrations than active sampling in the field. This is probably due to limitations in the sensitivity of the traditional method at the relatively low concentrations investigated. Similar conclusions were also made by Batterman et al. [1]. The samples collected by workers showed concentrations of the same order of magnitude as determined by the researchers. Also, sample precisions (for replicate samples) were similar for both kinds of samples. Thus, the low-flow passive method appears suitable for sampling done by workers. CONCLUSIONS This study confirmed that passive sampling tubes with capillary orifice analyzed with thermal desorption GC-MS are well suited for measuring occupational solvent exposure. This method has many benefits compared to active sampling using activated carbon. Operation is very simple, and no pumps are needed. Thermal desorption eliminates the need for solvent desorption and due to its sensitivity, only small sample volumes are required. The method is especially useful in investigating complex mixtures, including solvent naphtha, at low exposure levels. In addition, the method is well suited for sampling done by workers. REFERENCES 1. Batterman S, Metts T, Kalliokoski P, Barnett E. Low-flow active and passive sampling of VOCs using thermal desorption tubes: theory and application at an offset printing facility. J Environ. Monit 2002;4: Wadden RA, Scheff PA, Franke JE, Conroy LM, Javor M, Keil CB, et al. VOC emission rates and emission factors for sheetfed offset printing shop. Am Ind Hyg Assoc J 1995;56: Crough KG, Gressel MG. The control of press cleaning solvent vapors in a small lithographic printing establishment. Appl Occup Environ Hyg 1999;14: Pannwitz KH. Diffusionskoeffizienten. Drägerheft 1983;327: Nelson GO. Gas mixtures: preparation and control. Chelsea (MI): Lewis Publishers; EN 482. Workplace atmospheres General requirements for the performance of procedures for the measurement of chemical agents. Brussels: European Committee for Standardization; Peng CY, Batterman S. Performance evaluation of a sorbent tube sampling method using short path thermal desorption for volatile organic compounds. J Environ Monit 2000;2: Svendsen K, Rognes KS. Exposure to organic solvents in the offset printing industry in Norway. Ann Occup Hyg 2000;44: