Determination of trace amounts of formaldehyde in acetone

Transcription

1 analytica chimica acta 604 (2007) available at journal homepage: Determination of trace amounts of formaldehyde in acetone X.H. Hilda Huang a, H.S. Simon Ip b, Jian Zhen Yu a,b, a Atmospheric, Marine and Coastal Environment Program, Hong Kong University of Science & Technology, Clear Water Bay, Hong Kong, China b Department of Chemistry, Hong Kong University of Science & Technology, Clear Water Bay, Hong Kong, China article info abstract Article history: Received 5 July 2007 Received in revised form 6 October 2007 Accepted 8 October 2007 Published on line 17 October 2007 Keywords: Formaldehyde Hydroxymethanesulfonate 2,4-Dinitrophenylhydrazine Carbonyls A method to quantify sub-ppm levels of formaldehyde in acetone has been developed and it is reported here. In this method, the different reactivities and stabilities of sulfite with formaldehyde and acetone are used to separate the two carbonyl compounds. Sulfite reacts with formaldehyde to form hydroxymethanesulfonate (HMS), the non-volatile and stable nature of which allows its separation from bulk acetone solvent. The resulting HMS is then converted back to formaldehyde under basic conditions, and formaldehyde is derivatized with 2,4-dinitrophenylhydrazine (DNPH) and quantified in its DNP hydrazone form using high-performance liquid chromatography-uv detection. The method detection limit at the 99% confidence level was mg L 1. A batch of samples can be processed within 4 h. The method has been applied to quantify the amount of formaldehyde in an analytical-grade acetone and in a commercial nail polish remover and the level of formaldehyde was found to be and mg L 1, respectively Elsevier B.V. All rights reserved. 1. Introduction Acetone is widely used in different industry sectors. It is an intermediate in chemical production and a solvent for printing, resins, electroplating, paints, inks, varnishes, lacquers, thinners, adhesives and product cleanup, drugs, vitamins, and cosmetics. Acetone also finds applications in extracting oils and fats in food processing and serves as a precipitation agent in the purification of starches and sugars. Acetone is also used in household products, such as nail polish remover [1]. In 1994, the United States Environmental Protection Agency (USEPA) listed acetone to be an acceptable substitute for ozone-depleting substances for certain uses under the program known as the Significant New Alternatives Policy Program [2]. Due to the widespread use of acetone, it is possible that harmful impurities in acetone could pose a health hazard in occupational settings and indoor environments where large quantities of acetone are used. One such impurity is formaldehyde, the simplest form of aldehyde. The toxic health effects of formaldehyde are well documented. It is a probable human carcinogen and causes nose and throat irritation [3,4]. The ability to quantify formaldehyde in acetone-based solvents, inks and other commercial products is important in assessing exposure risk related to the use of acetone-based products. The commonly used analytical methods to detect carbonyls typically utilize a derivatization agent to improve chromatographic behavior and/or detectability. Examples of common derivatizing agents include 2,4-dinitrophenylhydrazine (DNPH) [5] and o-(2,3,4,5,6-pentafluorobenzyl)-hydroxylamine [6]. These agents derivatize both formaldehyde and acetone. As a result, they cannot be readily used to determine trace levels of formaldehyde in the presence of a large amount Corresponding author at: Department of Chemistry, Hong Kong University of Science & Technology, Clear Water Bay, Hong Kong, China. Tel.: ; fax: address: chjianyu@ust.hk (J.Z. Yu) /$ see front matter 2007 Elsevier B.V. All rights reserved. doi: /j.aca

2 analytica chimica acta 604 (2007) of acetone. Derivatization by the Nash reagent is specific to formaldehyde; however, the stability of the derivative and the reagent is not robust and extra caution is needed to avoid degradation by oxidation [7,8]. The above analytical difficulty can be overcome by utilizing the different stabilities of sulfite towards formaldehyde and acetone. Formaldehyde is known to readily react with sulfite to form hydroxymethanesulfonate (HMS), which is resistant to oxidation and highly stable in neutral and acidic media [7]. For this reason, formaldehyde has been used as a stabilizer for sulfite because one major problem in the determination of sulfite levels originates from the instability of sulfite in the presence of oxygen [9,10]. In comparison, the acetone sulfite complex shows weak stability [9]. The difference in stabilities between the two carbonyl-sulfite adducts allows isolation of HMS from bulk acetone solvent. This can be achieved by mixing sulfite into the sample to convert formaldehyde into HMS, followed by evaporation of the bulk acetone solvent. It is not feasible to detect formaldehyde through ion chromatographic analysis of the resulting HMS since elution of HMS requires an alkaline mobile phase but HMS decomposes to sulfite under basic ph conditions. As a result, a less direct approach has to be used. We have developed the following approach to quantify sulfite-bound formaldehyde. NaOH is added to break down HMS into formaldehyde and sulfite, followed by oxidization of sulfite to sulfate using H 2 O 2. Subsequently, the released formaldehyde is derivatized by DNPH under acidic conditions. The formaldehyde DNP hydrazone is then extracted with chloroform, reconstituted in acetonitrile, and quantified using HPLC. The analytical scheme is illustrated in Fig. 1. Below, we report the details of this method. 2. Experimental 2.1. Chemicals and reagents Sodium sulfite, DNPH, paraformaldehyde, acetone (HPLC grade), hydrochloric acid (HPLC grade), and acetonitrile (HPLC grade) were purchased from Aldrich (Milwaukee, WI, USA) and used without purification. UHP grade N 2 (99.999% purity) was from Hong Kong Special Gas Company (Hong Kong, China) Preparation and HPLC analysis of standards A formaldehyde stock solution (1000 mg L 1 ) was prepared by dissolving paraformaldehyde in hot water ( 80 C). Standard solutions were prepared at five concentration levels (0.300, 1.00, 4.00, 7.00, and 10.0 mg L 1 ) by diluting various volumes of the stock solution with acetone. The resulting solutions contained >95% acetone. A 2.0 ml aliquot of each standard solution was mixed with 2.0 ml of 0.7 mm Na 2 SO 3 aqueous solution. This amount of sulfite was more than twice that of formaldehyde in the highest concentration standard. The volume of the mixture was reduced to 1 ml under a stream of ultrapure nitrogen. This step took about 1 h. Into this solution were added 15 L of 0.5 M NaOH and subsequently 3 L of 1 M H 2 O 2 (an amount approximately twice that of sulfite added in the first step). The mixture was allowed to react for 20 min before 30 L of 0.5 M HCl was added to acidify the solution. Finally, 1 ml of 10 mm DNPH solution prepared in 2.0 M of HCl was added. The reaction vial was sealed by Teflon tape and covered by aluminum foil. The mixture was allowed to react at room temperature for 1 h [11]. The solution was then extracted with three portions of 500 L chloroform. The chloroform extracts were combined and evaporated to dryness under a mild stream of nitrogen. The residue was reconstituted with 500 L acetonitrile. The DNPH derivatives were analyzed by injecting 20 L of the sample in acetonitrile into an HPLC system (HP1100, Agilent, Santa Clara, CA, USA) equipped with a photodiode array detector. The column for separation was a reversed-phase column (Hypersil ODS, 4.6 mm 150 mm 5 m, Alltech, Deerfield, IL, USA). The mobile phase consisted of 60% acetonitrile and 40% water. The isocratic flow rate was 1.0 ml min 1 for 15 min. Absorbance at 360 nm was monitored for quantification Analysis of samples An analytical-grade acetone (Labscan, Stillorgan, Ireland) and a commercial nail polish remover were used to demonstrate the method. An aliquot of 4.0 ml of each sample was mixed with 2 ml of 0.7 mm Na 2 SO 3. The mixture was then treated the same way as the standards, with the amounts of the various reagents (i.e., NaOH, H 2 O 2, HCl) doubled considering that the volume of acetone samples were double that of the standards. Finally the DNPH derivative of formaldehyde was reconstituted to 100 L acetonitrile and analyzed using HPLC. 3. Results and discussion 3.1. Optimization of reaction conditions and reagent amounts Formaldehyde readily reacts with sulfite. The rapid nature of this reaction is the basis for using sulfite to scavenge formaldehyde in the industrial production of formaldehyde [12]. No kinetic experiment was conducted. A reaction time of 20 min was adopted for convenience. Comparison of calibration results for formaldehyde and HMS (described in Section 3.2) confirmed that 20 min was sufficient for quantitative conversion of formaldehyde to HMS. The stability of HMS is strongly dependent on ph. When HMS was mixed with DNPH under neutral or acidic conditions, no formaldehyde DNP hydrazone was detected, suggesting negligible dissociation of HMS back to formaldehyde. HMS quantitatively decomposes to formaldehyde and sulfite under basic ph conditions and in the presence of an adequate amount of sulfite oxidizer (e.g., H 2 O 2 ) [7]. Both conditions (i.e., alkaline ph and the presence of H 2 O 2 ) were needed to achieve complete release of formaldehyde from HMS. Different molar ratios of NaOH to HMS (0:1, 2:1, 5:1, and 20:1) were tried. No significant difference was observed between the ratios of 2:1 and 20:1. Consequently, a ratio of 2:1 was adopted. When only H 2 O 2 was deployed while keeping the ph neutral, the resulting formaldehyde DNP hydrazone response was only 13%. When left alone sulfite undergoes oxidation due to the pres-

3 136 analytica chimica acta 604 (2007) Fig. 1 Analytical scheme for the determination of trace amounts of formaldehyde in acetone. ence of dissolved oxygen [9], but the addition of a stronger oxidant such as H 2 O 2 allows complete destruction of sulfite for the next analytical step to proceed without lengthy delay. Experiments to optimize the reagent ratio of H 2 O 2 :HMS were also carried out by comparing the destruction efficiency at four H 2 O 2 :HMS molar ratios (0:1, 2:1, 5:1, and 20:1). An amount of the H 2 O 2 oxidizer in one time excess of the amount of sulfite was found sufficient to destroy sulfite rapidly. The extraction step in which formaldehyde DNP hydrazone was separated from other reagents in the aqueous solution was found to significantly improve the chromatogram quality through two means. First, the extraction step in effect stopped further formation of undesired products from the reaction of H 2 O 2 with DNPH, which otherwise could accumulate during the period prior to HPLC analysis and lead to deteriorated HPLC chromatograms. Second, inorganic salts, such as sulfate, was found to elevate the baseline noise with UV detection. The extraction step using CH 3 Cl therefore, serves to avoid interference of inorganic salt. Despite the multiple steps of chemical manipulation, formaldehyde levels in blank samples were fairly low. Fig. 2a shows a chromatogram of a blank sample of distilled water, which was processed in the same way as the standards and the samples were. Formaldehyde and the other two common carbonyls, acetaldehyde and acetone, were present in the procedure blank sample. Their presence and detection were anticipated due to their ubiquitous presence in the lab air and in reagents (e.g., water and DNPH). A comparison with chromatograms of real samples (Fig. 2b and c) indicates the levels of formaldehyde were much lower than those found in the samples. The reaction intermediate, HMS, albeit a compound known to be present in ambient aerosols, its presence is at trace levels. As a result, dust particles in the lab are not expected to contribute to the formaldehyde level detected in the blank samples. Unlike formaldehyde, HMS does not have other known contamination sources in typical university lab environments Calibration curves and method detection limit The calibration curves were established by spiking known amounts of formaldehyde in acetone solvent. The spiked volumes of formaldehyde were within 5% of the volume of acetone to ensure that the matrix of the calibration standards closely resembled that of the samples. The calibration curves were plotted as the peak area versus the concentration of formaldehyde in the injection solutions. An example is shown in Fig. 3a. The slopes generated in three different sets of calibration experiments were within 6% agreement. One can see from the analytical scheme that the part of the scheme after obtaining HMS solutions is applicable for determination of HMS. Due to our interest in HMS in atmospheric

4 analytica chimica acta 604 (2007) Fig. 3 Calibration curves for (a) formaldehyde in acetone and (b) HMS. Fig. 2 Example HPLC chromatograms of (a) a blank sample, (b) an analytical-grade acetone sample, and (c) a nail polish remover sample. samples, we also used this analytical scheme to determine HMS. For the determination of HMS, HMS standards in sodium salt form were prepared in water. Subsequently, HMS was decomposed using the same procedure as for formaldehyde and quantified in the form of formaldehyde DNP hydrazone. Fig. 3b shows a calibration curve for HMS. The slope values in the calibration curves for formaldehyde and HMS in mm (Fig. 3) are within 5% agreement. Such a good agreement in effect demonstrates good recovery during the step in which formaldehyde was isolated from the bulk acetone solvent. The method detection limit (MDL) was based on the precision of replicate injections of an analyte according to the definition of U.S. Federal Code of Regulations [13]. Seven replicate laboratory fortified blanks were spiked with formaldehyde, which resulted in a concentration of mg L 1. The MDL was calculated to be 3.14 times the standard deviation of the seven replicate analyses (Student s t-value is 3.14 at the 99% confidence level and 6 degrees of freedom). The MDL at the 99% confidence level was found to be mg L Application of the method to acetone-based samples Demonstration of the method was made using two acetonebased samples, an analytical-grade acetone and a nail polish remover. Table 1 lists the concentrations of formaldehyde in

5 138 analytica chimica acta 604 (2007) Table 1 Concentrations of formaldehyde in two acetone-based samples Sample No. of determination, N Concentration (mg L 1 ) Analytical grade acetone ± Nail polish remover ± the two samples determined using our method. Formaldehyde was found to be present at a level of mg L 1 (or %) in the acetone reagent, confirming the level of formaldehyde to be less than 0.002% as indicated in the product specification. The formaldehyde in the nail polish remover was quantified to be mg L 1. Replicate analyses demonstrated precision better than 3%, indicating good reproducibility of the method. The HPLC chromatograms of these two real samples (Fig. 2b and c) clearly show that the majority of acetone had been removed using the current method and acetone did not create interference in the detection of the trace amounts of formaldehyde. In addition to formaldehyde, acetaldehyde was also detected in the two acetone-based samples at levels significantly higher than that in the blank sample (Fig. 2). The carbonyl DNP hydrazone peaks in the sample chromatograms (peaks 2, 3, and 4 in Fig. 2) are slightly broaden in comparison with those in the blank chromatogram as a result of much higher concentrations of these derivatives in the samples. Two points are worth noting. First, this result indicated that trace levels of acetaldehyde were also present in acetone solvents. Second, the same analytical scheme may also be applicable to the determination of trace levels of acetaldehyde in acetone-based samples; however, whether the chemical transformation steps were quantitative for acetaldehyde needs further experiments to find out. 4. Conclusions We have developed a sensitive technique for the determination of sub-ppm levels of formaldehyde in acetone-based samples. Sulfite was used to isolate the trace amounts of formaldehyde from the bulk acetone through forming a stable and non-volatile product (i.e., HMS). The formed HMS was then decomposed back to formaldehyde under basic conditions in the presence of H 2 O 2 and quantified as formaldehyde DNP hydrazone using HPLC. This method is capable of achieving a detection limit of mg L 1, and a batch of samples can be processed within 4 h. This method can also be used to quantify HMS in atmospheric samples. Our method offers an analytical tool to monitor formaldehyde levels in acetone-based industrial and consumer products. The feasibility of this method was demonstrated using two real samples, an analytical grade acetone solvent and a commercial nail polish remover. Acknowledgements This work was supported by the Research Grants Council of Hong Kong, China (621806). We thank Ping Hei Lam for assistance during development of the analytical procedures and Dr. Eric Chun Hong Wan for helpful discussion. references [1] WHO, Environmental Health Criteria 207, Acetone, [2] USEPA, Significant New Alternatives Policy (SNAP), Final Rule Summary, [3] J.A. Swenberg, W.D. Kerns, R.I. Mitchell, Cancer Res. 40 (1980) [4] WHO, WHO Regional Publications, European Series, No. 91, Copenhagen, [5] USEPA, Compendium of methods for the determination of toxic organic compounds in ambient air. Method TO-11A; Center for Environmental Research Information, Office of Research and Development, Cincinnati, [6] S.S.H. Ho, J.Z. Yu, Environ. Sci. Technol. 38 (2004) 862. [7] S. Dong, P.K. Dasgupta, Environ. Sci. Technol. 21 (1987) 581. [8] T. Nash, Biochem. J. 55 (1953) 416. [9] L.M. De Carvalho, G. Schwedt, Fresen. J. Anal. Chem. 368 (2000) 208. [10] Y. Li, M. Zhao, Food Control 17 (2006) 975. [11] H. Tuß, V. Neitzert, W. Seiler, R. Neeb, Fresen. J. Anal. Chem. 312 (1982) 613. [12] J.G.M. Winkelman, M. Ottens, A.A.C.M. Beenackers, Chem. Eng. Sci. 55 (2000) [13] J.A. Glaser, D.L. Foerst, Environ. Sci. Technol. 15 (1981) 1426.