Ann. occup. Hyg., Vol. 46, No. 1, pp. 89-96, 2002
© 2002 British Occupational Hygiene Society
Published by Oxford University Press
Article |
Generation of Controlled Atmospheres for the Determination of the Irritant Potency of Peroxyacetic Acid
Institut National de Recherche et de Sécurité, BP 27, 54501 Vandoeuvre les Nancy Cedex, France
Received 12 March 2001; in final form 28 June 2001.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONINTERFERENCE STUDYGENERATION OF CONTROLLED...CONCLUSIONREFERENCES |
|---|
Given the physical properties of peroxyacetic acid, which decomposes into acetic acid and hydrogen peroxide, the generation and analysis of controlled atmospheres used to test the irritant potency of this peracid in mice require specific developments. The sampling and analytical method was based on the simultaneous sampling on a titanyl sulphate-impregnated silica gel tube (allowing the determination of total peroxides, peroxyacetic acid and hydrogen peroxide) and in an impinger containing a methyl-p-tolyl sulphide solution (of which the analysis yields the concentration of total acids, peroxyacetic acid and acetic acid, and peroxyacetic alone). From these results the concentrations of the different products can be inferred without interference. A special device composed of inert materials was designed for the generation of the controlled atmosphere. Buffering the peroxyacetic solution at pH 7 with a phosphate buffer allowed the generation of peroxyacetic acid atmospheres with negligible concentrations of acetic acid and hydrogen peroxide.
Keywords: controlled atmosphere; peroxyacetic acid; irritant potency
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONINTERFERENCE STUDYGENERATION OF CONTROLLED...CONCLUSIONREFERENCES |
|---|
Low concentrations (1–15%) of peroxyacetic acid (PAA) are used as sanitizers, disinfectants and sterilants in the food, beverage and medical industries (ECETOC, 2001). The increasing use of peroxyacetic acid has pointed out the necessity for the occupational hygienist to derive a limit value in order to prevent the irritation felt by workers employed in premises where this disinfectant is used. In a companion article, Gagnaire et al. (Gagnaire et al., 2002) present an experiment carried out to establish such a value. Since commercial solutions of PAA are produced from the reaction between acetic acid (AA) and hydrogen peroxide (HP):
CH3COOH + H2O2 CH3COOOH + H2O
and consequently decompose into these original materials, which themselves have an irritant potency, the choice of generation and analytical methods, and the study of possible interfering factors are particularly important aspects of the study described by Gagnaire et al. (Gagnaire et al., 2002). It is indeed particularly important that the mice are exposed to pure compounds (or to a mixture with an accurately determined composition). A study described in the present paper was carried out specifically to determine the optimal operating conditions for generating (and analysing the composition of) controlled atmospheres to determine the irritant potency of AA, HP, PAA and commercially available mixtures of these three compounds (sold under the name of peroxyacetic acid).
|
INTERFERENCE STUDY |
|---|
|
TOPABSTRACTINTRODUCTIONINTERFERENCE STUDYGENERATION OF CONTROLLED...CONCLUSIONREFERENCES |
|---|
Materials and methods
Analytical methods
The following reference sampling and analytical methods are currently used in our laboratory for the compounds considered individually.
HP (Hecht et al., 1999): sampled on silica gel coated with titanium oxysulphate. The silica gel is then desorbed by percolation with a sulphuric acid molar solution. The solution is then analysed by spectrophotometry at 410 nm. Peroxyacetic acid was shown to be a positive interferent in the determination of hydrogen peroxide.
AA (Simon et al., 1989): sampled on Florisil. The sorbent is then desorbed in water and the solution analysed by liquid chromatography with conductivity detection. As PAA decomposition leads to the formation of AA, positive interference will occur in the determination of AA in the presence of PAA.
PAA: Effkemann et al. (Effkemann et al., 1999) described a specific method for the determination of atmospheric concentration of PAA. It is based on the following reaction between PAA and methyl-p-tolyl sulphide (MTS) with the formation of methyl-p-tolyl sulfoxide (MTSO) and AA:
CH3COOOH + CH3C6H4SCH3 
CH3C6H4SOCH3 + CH3COOH
Pinkernell et al. (Pinkernell et al., 1994) showed that interference due to HP when determining PAA by means of MTS is very low, in so far as the solution is analysed immediately after the reaction, the reaction between HP and MTS being kinetically disadvantaged.
Reagents
The following products were used to study these interferences: MTS 99% (Aldrich ref. 27595-6); MTSO (Fluka ref. 69422); titanium oxysulphate (Riedel de Haen ref. 93100); Florisil" 30–60 mesh (Fluka ref. 46384); silica gel 0.2–0.5 mm (Merck ref. 1.07733.1000); PAA 
39% (Fluka ref. 77240); HP 30% (Merck ref. 23.615.245).
Spectrophotometric measurements
A Perkin-Elmer Lambda 11 spectrophotometer was used. The reading cell (Hellma, 390 µl volume, 10 mm optical path) supplied a continuous 3 ml/min flow by a peristaltic pump. The reading was taken at 410 nm after 1 min. This delay is necessary for equilibrating the flow cell.
Liquid chromatography analysis
MTS and MTSO were analysed by high-performance liquid chromatography (Shimadzu LC-10 ATvp pump; Rheodyne 7125 injection valve; Alltech 250 mm x 3 mm, 5 µm, C18 Alltima column; Shimadzu SPD-6A UV detector; Spectra Physics integrator). The mobile phase was an acetonitrile/water (60/40) mixture at a flow rate of 0.8 ml/min. The wavelength of the UV detector was set to 230 nm.
Liquid chromatography was also used to analyse AA (same pump, valve and integrator; Biorad 300 mm x 7.8 mm Aminex HPX 87 H organic acid column; Metrohm IC 732 conductivity detector). The 2.5 x 10–4 M sulphuric acid eluent was pumped at a flow rate of 0.8 ml/min.
Results
Interference of HP on the determination of PAA
The principle of the sampling method is based on bubbling at a flow rate ranging from 0.1 to 1 l/min in a 2.7 mM MTS solution in an ethanol/water (75/25) mixture. The first stage consisted of verifying the retention capacity of the solution in normal experimental conditions, i.e. bubblers filled with 15 ml of solution and a 4 h test. The results showed a loss of MTS of about 6% per sampling hour. Consequently, with a security factor of 3, the sampling device can measure an atmospheric concentration of PAA of 4 p.p.m. for 4 h at a bubbling flow rate of 0.25 l/min. The MTS bubbling solution thus defined was used for all the tests carried out in this study.
The efficiency of the sampling method was also tested. The quantity of PAA sampled in midget fritted-glass impingers containing MTS at a flow rate of 0.25 l/min was compared with the same quantity of PAA introduced directly into reference vials containing a quantity of MTS equal to that contained in impingers. To prepare the impingers, the PAA solution was deposited on a quartz fibre filter (Whatman QMA 37 mm) contained in a cassette (Millipore M000037A0) connected to the inlet of the impinger, itself connected to a sampling pump. The recovery in the bubbling solutions was 97% (correlation coefficient r2 = 0.999) of the quantity introduced directly into the vials (five series of three impingers and three references for quantities of PAA ranging from 1.6 to 8.5 µmol/sample).
A similar experiment was carried out to estimate any interference caused by HP when determining PAA. Three kinds of samples were prepared:
1. 50 µl of a 0.1 M HP solution was deposited on a quartz fibre filter (according to the technique described previously for PAA deposits) and vaporized in an impinger containing a MTS solution;
2. 50 µl of the same HP solution was vaporized in the same way in a titanium oxysulphate coated silica gel tube, prepared according to the method described previously (Hecht et al., 1999);
3. 50 µl of this HP solution was introduced directly into a silica gel tube of the same type (reference sample).
Considering tube (3) as a 100% reference, the recovery on tube (2) was ~97%, confirming the results of Hecht et al. (Hecht et al., 1999). On the other hand the analysis of MTSO after a 3 day reaction between HP and MTS (impinger 1) showed that <1% of HP had reacted. It can therefore be considered that the presence of HP does not interfere with the atmospheric analysis of PAA.
Interference of PAA on the determination of HP
The reaction between PAA and titanium oxysulphate was studied first, two experiments being carried out in parallel.
1. 50 µl of a 1% PAA solution was introduced into 50 ml of a 5 mM absorbing solution of titanium oxysulphate (0.5 g of titanium oxysulphate in 5 ml of sulphuric acid made up to 500 ml with deionized water). This mixture was stirred constantly and allowed to flow continuously through a spectrophotometric cell by means of a peristaltic pump. The absorbance was read at 410 nm every 3 min for 6 h. This experiment was carried out three times with three different PAA solutions.
2. 50 µl of the same PAA solutions was added directly to the usual working MTS solution. Five samples were prepared for each of the three PAA solutions. The solutions were then analysed by the MTS/MTSO liquid chromatography method described previously.
The reaction kinetics of PAA with titanium oxysulphate are shown in Fig. 1. Immediately after introducing PAA into the measurement system, the absorbance increased suddenly, corresponding to the reaction of the HP contained in the PAA solution, and then continued to increase much more slowly to reach a maximum after 5–6 h of reaction, this second phase corresponding to the reaction of PAA itself. The results are summarized in Table 1, which shows that the difference between the final concentration and the initial concentration determined by the spectrophotometric method corresponds to the PAA, which reacts slowly with titanium oxysulphate. The concentration measured by liquid chromatography can be considered as a reference, as it is based on a reaction specific to PAA subject to only slight interference by the presence of HP. It can be seen from these results that the reaction between PAA and titanium oxysulphate is complete after a few hours: for the following experiments a 24 h delay was planned before the analysis was carried out. From these results with the titanium oxysulphate method, which gives the sum of the peroxides, and as the MTS method is specific to PAA, the difference between the two figures leads to the HP concentration.
|
|
The next step of the experiment consisted in verifying these results in atmospheric samples, i.e. with the products vaporized. Six series of samples were prepared, each series including:
1. three reference samples prepared by introducing a defined quantity of PAA directly into the MTS bubbling solution;
2. three samples resulting from vaporizing the same quantity of PAA in the MTS solution contained in an impinger;
3. three reference samples prepared by introducing a defined quantity of PAA directly into the titanium oxysulphate solution for series 1–3, and on a titanium oxysulphate tube for series 4–6;
4. three samples resulting from vaporizing the same quantity of PAA in the titanium oxysulphate solution contained in an impinger for series 1–3, and on a titanium oxysulphate tube for series 4–6.
The results of this experiment are summarized in Table 2. The efficiency of the bubbling in MTS solutions is good, with a mean recovery rate of 96.1% [95% confidence interval (CI) = 91.3–100.8%]. The corresponding values for the sampling on TiOSO4 tubes are respectively 95.1% and 90.0–100.3%. On the other hand, the efficiency of the sampling in an impinger containing the titanium oxysulphate solution is poor (mean recovery rate = 85.8%; 95% CI = 74.1–97.4%). In conclusion, sampling the atmosphere simultaneously in an impinger containing the MTS solution and on a titanium oxysulphate-coated silica gel tube permits the determination of PAA and the total peroxides, respectively, and consequently PAA and HP.
|
Determination of AA in the presence of PAA
As one degradation product of PAA is AA, clearly the sampling and analysis method to determine the atmospheric concentration of AA is subjected to interference by the presence of PAA. The efficiency of sampling PAA on a Florisil tube was firstly studied according to the method described by Simon et al. (Simon et al., 1989). The PAA was vaporized, and after analysis of the tube, the recovery ranged from 40 to 80%, making it impossible to use Florisil for a method by difference similar to that used for the simultaneous determination of PAA and HP [total acids (PAA + AA) being measured on Florisil and PAA in the MTS solution].
As the reaction between PAA and MTS is total and leads to the formation of a molar quantity of AA equal to the molar quantity of PAA (see equation 2), it was hypothesized that the determination of AA in the MTS bubbling solution would provide the total concentration of the acids (PAA + AA). In addition, since the analysis of MTSO allows the concentration of PAA itself to be measured, the difference between the total acid concentration and the concentration of PAA is equal to the concentration of AA. Six series of samples were prepared, each series including:
1. three reference samples prepared by introducing a defined quantity of PAA directly into the MTS bubbling solution for series 1 and 2; for series 3–6, 100 µl of a phosphate buffer were added to maintain the pH of the solution at ~7;
2. three samples resulting from vaporizing the same quantity of PAA in the MTS solution contained in an impinger. For series 3–6, 100 µl of the phosphate buffer solution were added to this MTS solution to bring its pH to ~7.
MTSO was analysed by the method described by Effkemann et al. (Effkemann et al., 1999) and total acids by the methods proposed by Simon et al. (Simon et al., 1989). The results of this experiment are summarized in Table 3. AA collection is considerably improved by adding buffer solution to the MTS solution (up from 85 to 95%): poor recovery rates and high coefficients of variation were recorded for unbuffered solutions (series 1 and 2).
|
Consequently, the simultaneous determination of the atmospheric concentrations of PAA, HP and AA is based on the principle described in Table 4: simultaneous sampling on a TiOSO4-impregnated silica gel tube and in an impinger containing an MTS solution, and analysis of total peroxides, total acids and PAA (by determining the MTSO).
|
|
GENERATION OF CONTROLLED ATMOSPHERES |
|---|
|
TOPABSTRACTINTRODUCTIONINTERFERENCE STUDYGENERATION OF CONTROLLED...CONCLUSIONREFERENCES |
|---|
Materials and methods
Although the analytical difficulties involved had been resolved, a solution to the problem of generating PAA alone still had to be found, since this product does not theoretically exist alone (see equation 1). A special device composed only of ‘inert’ materials (glass or PVC) was designed for this study, the initial results in the usual aluminium and stainless steel controlled atmosphere devices having shown that 99% of the generated HP was destroyed before arrival in the inhalation chamber. The generation device (shown in Fig. 2) functioned as follows: a motorized syringe (KD Scientific, Bioblock, Illkirch, France) delivered a constant flow of the product to be generated in a vaporization chamber composed of a column filled with a glass bed. In this vaporization chamber, a stream of previously heated air flowed counter to the stream of generated product. After being mixed with moistened air, this air directed the generated product towards the inhalation chamber, where a bypass installed at the inlet enabled a constant flow to be maintained inside.
|
The following operating parameters were kept constant during generation of the different compounds:
• air flow in the vaporization chamber: 10 l/min;
• temperature of the vaporization chamber: 55°C;
• flow rate of the solution of the generated product: 50 µl/min;
• flow rate of moistened air: 10 l/min;
• flow rate in the inhalation chamber: 5 l/min.
With these parameters kept constant, the different concentrations of the compounds in the inhalation chamber were obtained by varying the concentrations of the vaporized solutions.
Validation of the generation device
Two parameters were checked: namely, the stability of the concentration generated and the homogeneity inside the inhalation chamber were verified to ensure that all the animals were exposed to the same constant concentration in the Gagnaire et al. (Gagnaire et al., 2002) experiment. HP was chosen for this validation as its oxidation–reduction properties are close to those of PAA and as it is stable in aqueous solution.
The generation device was started with a 3 M solution of HP in the operating conditions described previously. The concentration inside the inhalation chamber was measured every hour for 7 h with titanium oxysulphate-coated silica gel tubes. Four tubes were sampled at a time, each at a different exposure inlet of the inhalation chamber. The sampling duration ranged from 6 to 15 min. From the results reported in Fig. 3, it can be seen that the atmospheric concentration inside the chamber stabilizes after only 3 h. In practice, the system was allowed to stabilize for 24 h before carrying out the inhalation tests. The motorized syringe was refilled twice a day (07:00 and 17:00 h).
|
A theoretical calculation of the inhalation chamber concentration shows that it should be 167 p.p.m., whereas the experimental value is 109 p.p.m. (65% of the theoretical value). It was hypothesized that a part of the HP generated was destroyed in the generation device and/or adsorbed on the walls. Besides, on emptying the motorized syringe after stoppage of product vaporization following a series of generation operations, a significant concentration of HP was measured in the inhalation chamber, probably resulting from a release of HP from the walls of the device. The hypothesis of the formation of ‘super-oxidized compounds’ was also examined (which might explain the difference between the theoretical and the actual HP concentration): no trace of ozone or nitrogen oxides was found in the inhalation chamber atmosphere.
The difference in concentration between the four exposure inlets of the inhalation chamber was also studied. Fourteen series of four samples (one per exposure inlet, numbered 1–4 from the lower part to the upper part of the inhalation chamber) were taken in succession, some at 70 p.p.m., the remainder at 140 p.p.m. The results of the analysis of these samples are shown in Table 5. A multiple-sample comparison shows that there are two homogeneous groups, one constituted of nos 1 and 2, the other of nos 3 and 4. Assigning a 100 value to the mean value of each series, the respective mean value of each sampling point considering the 14 series is 101.9 for no. 1, 101.6 for no. 2, 98.8 for no. 3 and 97.7 for no. 4. The difference between the four sampling points is consequently low, and a ‘mean’ sampling point was taken between points 2 and 3 for the experimental study performed by Gagnaire et al. (Gagnaire et al., 2002).
|
Generation of PAA alone
As previously mentioned, PAA does not exist alone in solution, but is always accompanied by HP and AA. The aim of this work was to determine the conditions under which PAA could be generated alone in the atmosphere. From the physicochemical properties of the three products considered, which are presented in Table 6, it appeared that buffering the solution at pH 7 with a phosphate buffer might be a good way of generating PAA alone if its vapour pressure (not available in the literature) was greater than that of HP. Preliminary tests showed that it was not possible to buffer the PAA solution before generation as a large quantity of gas is given off suddenly, which does not allow a constant feed of the generation device. The device had to be modified by adding a peristaltic pump, as shown in Fig. 4. Feeding the vaporization chamber with constantly renewed solutions of buffer and PAA allowed regulation of the emission of gaseous PAA without impoverishing the reaction medium.
|
|
At the beginning of generation, the vaporization chamber was filled with 7.5 ml of the PAA solution and 7.5 ml of the buffer solution. This chamber was fed by 0.125 ml/min of both PAA and buffer solution, with parallel extraction of 0.25 ml/min of the mixture to keep the level constant. The PAA generated was stripped from the vaporization chamber by means of a 0.25 l/min flow supplied by a mass flow controller.
A number of examples of the atmospheres generated using this system are given in Table 7. Each mean concentration represents five measured values. It should be noted that certain values (or certain limits of the 95% CI) of the HP or AA concentrations could be negative. This is due to the method of calculation as the concentrations of HP and AA result respectively from the difference between the total of peroxides and the concentration of PAA, and from the difference between the total of acids and the concentration of PAA. Given the low atmospheric concentrations of HP and AA and the uncertainties on the measurements, these negative values are hardly surprising. Besides, from the results summarized in Table 7, it would clearly appear that the objective has been achieved: it is possible in precise conditions to generate PAA alone with negligible concentrations of AA and HP.
|
|
CONCLUSION |
|---|
|
TOPABSTRACTINTRODUCTIONINTERFERENCE STUDYGENERATION OF CONTROLLED...CONCLUSIONREFERENCES |
|---|
The need for controlled atmospheres with an accurately known composition required the design of a special generation device. In addition, special emphasis was placed on the sampling and analytical methods, and problems of interference between compounds with similar properties had to be taken into account. Finally, the importance of generation conditions has been highlighted. In certain precise conditions it is indeed possible to generate PAA alone. In addition to this aspect of the study, which allowed the generation of controlled atmospheres, industrial hygienists now have at their disposal a method which allows a global (PAA, AA and HP) evaluation of exposure of workers employed in factories using commercial PAA solutions.
|
FOOTNOTES |
|---|
* Author to whom correspondence should be addressed.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONINTERFERENCE STUDYGENERATION OF CONTROLLED...CONCLUSIONREFERENCES |
|---|
ECETOC. (2001) Joint assessment of commodity chemicals, no. 40. Peracetic acid and its equilibrium solutions. Brussels: European Center for Ecotoxicology and Toxicology of Chemicals.
Effkemann S, Broodsgaard S, Mortensen P, Linde S-A, Karst U. (1999) Determination of gas phase peroxyacetic acid using pre-column derivatization with organic sulfide reagent and liquid chromatography. J Chromatogr A; 855: 551–61.[Medline]
Gagnaire F, Marignac B, Hecht G, Héry M. (2002) Sensory irritation of acetic acid, hydrogen peroxide, peroxyacetic acid and their mixture in mice. Ann Occup Hyg; 46; 97–102.
Hecht G, Aubert S, Gérardin F, Héry M. (1999) Workplace monitoring of hydrogen peroxide using titanyl-coated sorbents. J Environ Monit; 1: 149–52.
Klopotek B-B. (1998) Peracetic acid methods of preparation and properties. Chim Oggi; 16: 33–7.
Pinkernell U, Karst U, Cammann K. (1994) Determination of peroxyacetic acid using high performance liquid chromatography with external calibration. Anal Chem; 109: 985–7.
Simon P, Brand F, Lemaçon C. (1989) Florisil sorbent sampling and ion chromatographic determination of airborne aliphatic carboxylic acids. J Chromatogr; 479: 445–51.[Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  G. HECHT, M. HERY, G. HUBERT, and I. SUBRA Simultaneous Sampling of Peroxyacetic Acid and Hydrogen Peroxide in Workplace Atmospheres Ann. Hyg., November 1, 2004; 48(8): 715 - 721. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. GAGNAIRE, B. MARIGNAC, G. HECHT, and M. HERY Sensory Irritation of Acetic Acid, Hydrogen Peroxide, Peroxyacetic acid and their Mixture in Mice Ann. Hyg., January 1, 2002; 46(1): 97 - 102. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||