2879
Anal. Chem. 1984, 56,2879-2882 ciety: Washington, DC, 1982; ACS Symposium Series, No. 199. Keifer, J. R.; Novlcky, M.; Akhter, M. S.: Chughtal, A. R.; Smith, D. M. Proc. SPIE-Int. SOC.Dpt. Eng. 1981, 289. Ellis, E. C.; Novakov, T. Sci. TotalEnviron. 1982, 23, 227-238. Uhdeova, J.; Rezl, V. Anal. Chem. 1981, 53, 164-167. Mackenzle, R. C. J . Therm. Anal. 1978, 73,387-391. Grosjean, D. Environ. Scl. Techno!. 1982, 16, 254-262. Hldy, G. M., et ai., Eds. "The Character and Origins of Smog Aerosols": Wlley: New York, 1980; Vol. 9. Wolff, G. T., Klimisch, R. L., Eds. "Particulate Carbon: Atmospheric Life Cycle"; Plenum: New York, 1962; p 19. Graedel, T. E. "Chemical Compounds In the Atmosphere"; Academic: New York, 1978. Benner, W. H. Anal. Chem. 1984, 56, 2871-2875. Gundel, L. A,; Novakov, T. Atmos. Environ. 1984, 18, 273-278. Dod, R. L. Lawrence Berkeley Laboratory, unpublished data, 1984. Wendiandt, W. W. "Thermal Methods of Analysis"; Why: New York, 1974; p 11.
(15) Buzas, I., Ed. "Thermal Analysis"; Proceedings of the International Conference on Thermal Analysis; Akademlai Kiado: Budapest, 1975; Voi. 3, p 1027. (16) Chang, S. G.; Brodzinsky, R.; Gundel, L. A,; Novakov, T. I n "Particulate Carbon: Atmospheric Life Cycle": Wolff, G. T., Kllmisch, R . L., Eds.; Plenum: New York, 1982.
RECEIVED for review May 24, 1984. Accepted July 27, 1984. This work was supported by the Director, Office of Energy Research, Office of Health and Environmental Research, Physical and Technological Research Division of the U S . Department of Energy under contract DE-AC03-76SF00098 and by the National Science Foundation under Contract ATM 82-10343.
Quantitative Analysis of Gas-Phase Formaldehyde Molecular Species at Equilibrium with Formalin Solution Avram Gold, David F. Utterback,*' and David S. Millingtod
Department of Environmental Sciences and Engineering, School of Public Health, University of North Carolina, Chapel Hill, North Carolina 27514
A method has been developed for quantltatlve analysis of vapor-phase formaldehyde condensates wlth water and methanol. Trimethylsllyl derlvatlves of the formaldehyde condensates were formed by gas-phase reactlon wlth N , O bls(tr1methylsilyl)trlfluoroacetamlde(BSTFA) which allowed their separatlon and quantltatlon. Methylal, methylene glycol, and a series of poly(oxymethylene) glycol monomethyl ethers accounted for the total formaldehyde content detected In the vapors at equlllbrlum with a formalln solution. Structures of the formaldehyde species were confirmed by caplllary gas chromatography/chemlcal loniratlon mass spectrometry (GWCIMS) using ammonia as the reactant gas.
Both oligomeric series, the poly(oxymethy1ene)glycols (I) and the poly(oxymethy1ene) glycol monomethyl ethers (II), are thermally unstable and decompose upon separation by gas chromatography (3). Trimethylsilylation of the oligomers in solution by BSTFA allows separation, identification, and quantitation through combined analysis of the respective Me3Si ethers by capillary GC/CIMS and GC with flame ionization detection (2). Indications of the presence of methylene glycol and low molecular weight oligomeric forms of formaldehyde in the gas phase appear in the literature. Hall and Piret (4) determined an expression for the equilibrium constant, Kp, over the temperature range of 40-160 "C for the ternary gas-phase system of water, formaldehyde, and methylene glycol: log K , = ( 3 2 0 0 / T ) 9.8
+
Formaldehyde is a highly reactive compound that equilibrates between many molecular forms in aqueous solutions. The addition of methanol to formaldehyde solutions as a stabilizer further increases the number of molecular species in which formaldehyde may exist (1-3), while molecular or monomeric formaldehyde is present only at very low concentrations (1). Many of the formaldehyde condensates with water and methanol have significant vapor pressure which results in their evolution into the gas phase, yet the molecular state of formaldehyde as it exists in the gas phase has not been verified. Formation of formaldehyde condensates, Le., poly(oxymethylene) glycols and the glycol monomethyl ethers, in formalin solution is represented by the following equilibria: CH2O + H2O --* CHZ(0H)Z CH2(0H)z CH20 HO(CH2O)ZH
+
HO(CH,O),H
+
+ CHzO
HO(CHZO),+lH HO(CHzO),H + C H 3 0 H --c HO(CH20),CH3 CH2O 2CH3OH CH30CHZOCH3
+
+
(I) (11)
+
'Present address: Department o f Health Science, California State University, Fresno, Fresno, CA 93740. Present address: Division of Genetics and Metabolism, D u k e University Medical Center (Box 3028), Durham, N C 27710.
0003-2700/84/0356-2879$0 1.50/0
On the basis of vapor pressure data, Hall and Piret predicted that at 10 "C above its boiling point, methylene glycol would be 95% dissociated into water and formaldehyde. Hence, formaldehyde may exist in the hydrated form even at 106 "C. Their calculations assumed that poly(oxymethy1ene) glycols were not present in significant concentrations, although there was no experimental evidence to support this concept. By indirect methods, Iliceto (5) investigated the formation of the dimeric species, oxydimethylene glycol, from formaldehyde and methylene glycol and calculated the enthalpy of formation to be -11.6 kcal/mol in the gas phase. This equilibrium has been supported by the work of Bryant and Thompson (6). Using techniques similar to those of Iliceto, they predicted an exothermic reaction for the formation of both methylene glycol and oxydimethylene glycol in the gas phase and also calculated the Gibbs free energies and the equilibrium constants for the reactions. The thermodynamic data are consistent with those reported by Iliceto and by Hall and Piret. Sawicki and Sawicki (7) assert that the 2-4% water present in paraformaldehyde causes repolymerization of formaldehyde generated in a gas stream that has passed over a paraformaldehyde permeation tube. Geisling et al. (8)also found that the decomposition of paraformaldehyde involves emission of water along with formic acid and methyl formate, and Walker 0 1984 American Chemical Society
2880
ANALYTICAL CHEMISTRY, VOL. 56, NO. 14, DECEMBER 1984
Table I. Formaldehyde Condensation Products with Water and Methanol Found in Vapors (70 mL) at Equilibrium with a Formalin Solution (mol) analysis I (30 MLof BSTFA/DMF) derivatization reaction phase compound liquid CH,OCH,OCH, 2.30 X lo-' CH,OCH,O(Me,Si) 1.37 X (Me,Si)OCH,O(Me,Si) 4.56 x lo-' CH,(OCH,),O(Me,Si) 1.83 x lo-' CH,(OCH,),O(Me,Si) 8.44 x total formaldehyde by gas-phase derivatization method total formaldehyde equivalents by chromotropic acid method (same concentration measured in two analyses)
gas 5.43 X 2.17 X N.D. N.D. N.D.
(91, citing similar results, adds that the presence of these products in the gas phase of formaldehyde promotes rapid polymerization at temperatures below 100 "C. Schnizer et al. (IO)reported that the yield of monomeric formaldehyde from acid-catalyzed depolymerization of trioxane was limited t o 89% due t o paraformaldehyde formation caused by the presence of water. In all these investigations, solid paraformaldehyde, the polymeric form of formaldehyde, was observed in the gas-phase emissions from a formaldehyde source. Previous work, therefore, suggests that short and intermediate length oligomers of formaldehyde must be present and involved in the gas-phase equilibrium formation of the long chain polymers from monomer. This paper describes a method by which these gas-phase formaldehyde condensates have been detected and quantitated in t h e headspace above a formalin solution.
EXPERIMENTAL SECTION Analysis was performed on a Varian Model 3700 capillary gas chromatograph (Varian Associates, Palo Alto, CA) equipped with a flame ionization detector. Separation was accomplished with a 30-m DB-5 (methyl phenyl silicone) fused silica capillary column (J & W Scientific, Rancho Cordova, CA). Helium carrier gas flow rate was 1.5 mL/min and a Grob-type split/splitless injector was operated a t a 1O:l split ratio. The injector was heated to 250 "C and the detector to 300 "C. After a 6-min isothermal period a t 30 "C, the column temperature was programmed to rise a t 6 "C/min to a limit of 230 "C. Detector output was recorded by a Hewlett-Packard Model m3380 integrator. Previously reported retention times (2) for oligomers on the same chromatographic system were used for identification of the silylated oligomers. Headspace derivatization was performed in vials that were evacuated to less than 1mmHg by an oil vacuum pump with an in-line liquid nitrogen vapor trap. A 5:l mixture of BSTFA (Supelco, redistilled 147-153 "C) and dimethylformamide (Supelco, derivatization grade) was introduced by syringe through the serum stopper into the reaction vial (70 mL). Headspace above a 37% reagent grade formalin solution (Fisher Scientific) at 24 "C in the 2-L headspace generator (Figure 1) was then bled into the reaction vial and the pressure equalized with atmospheric pressure. Two analyses were performed (Table I). The first utilized 30 pL of the BSTFA/DMF solution and the second used 40 pL. Quantitative analysis was performed on both the liquid and gas phases in the reaction vials by GC-FID. A derivatized gas-phase aliquot (100 pL) was taken for analysis directly from the reaction vial by a gastight syringe. The liquid phase from a separate derivatization reaction was transferred by pipet to a vial equipped with a screw-top Septum cap and a Teflon-lined septum from which an aliquot (1 pL) was taken for analysis. Structures were confirmed by analysis of a liquid-phase aliquot with capillary GC/CIMS using ammonia as reactant gas (2). The GC/MS system consisted of a Hewlett-Packard Model HP5710A gas chromatograph with a 30-m DB-5 fused silica capillary column
formaldehyde content 5.66 X l o T 6 3.54 X 4.56 X lo-' 3.65 X IO-' 2.53 X lo-' 9.64
X
analysis 11 (40 pL of BSTFA/DMF) derivatization reaction phase liquid
gas 6.26 X 1.73 X N.D. N.D. N.D.
4.70 X lo-' 2.22 X l o e 6 1.08 X lo-' 2.46 X l o - ' N.D.
lo-'
formaldehyde content 6.73 X 3.95 X 1.08 X lo-' 4.91 X lo-' N.D. 1.13 x 10-5
1.08 x 10-5
/,2L
\
Flask
Y
Figure 1. Schematic representation of headspace generator used In gas-phase derivatlzatlon technique.
interfaced with a VG-Micromass Model 7070F mass spectrometer and VG 2035 F/B data system (VG Analytical, Manchester, England). The ionizing voltage was 70 eV (tungsten filament, 2WpA trap current), and the scan cycle time over the mass range 20-300 was 1.3 s. The source temperature was 200 "C, the injector was 250 "C, and temperature program was the same as described above. Quantitative analysis of the total formaldehyde equivalents in the formalin headspace was performed by bubbling vapor (70 mL) from the headspace generator through an impinger containing distilled water (20 mL) (11). Suction was provided by a Sippin Model SP-1 pump a t a rate of 80 mL/min. An aliquot was transferred to a volumetric flask (25 mL) and diluted with distilled water to obtain a concentration in the linear range of the calibration curve. Standard solutions were generated by dilution of 37% formalin solution (1 mL) containing 10-15% methanol (Fisher Scientific) in a volumetric flask (500 mL) with distilled water. Aliquots (1 mL) of this stock solution were then diluted to generate solutions of 3.2, 1.6, and 0.4 pg/mL formaldehyde equivalents. Aliquots (4 mL) of the sample, standards, and a blank of distilled water were then concurrently derivatized by addition of 0.1% chromotropic acid (1mL) and concentrated sulfutic acid (6 mL). Absorbance of the solutions was measured a t 580 nm by a Coleman 124 UV-vis spectrophotometer.
RESULTS AND DISCUSSION Gas-phase trimethylsilylation of vapors in equilibrium with a formalin solution yields derivatives having sufficient vapor pressure to a p p e e in the vapor phase as well as in the excess liquid BSTFA/DMF contained in the reaction vial. Quantitative measurement, therefore, requires analysis of both gas
ANALYTICAL CHEMISTRY, VOL.
Vapor
%I
I1
56,NO. 14, DECEMBER 1984
CH,OCH,OCH,-
2881
- NH;
~
M 145
'i
100-
152
CH,OCH,O-TMS- -NH:
TMS-OCH,O-TMS--NH;
8
16
t i m , minutor
i' 'Ip I /I
Liquid
,t'
I
I
I
CH,O(CH,O)iTMS- -NH:
Figure 3. Chemical ionization mass spectra with ammonia as the reactant gas obtained from the liquid phase of the gas-phase derivatization of formalin headspace by BSTFA. Boldface labels refer to peak identification in Figure 2.
8
16
24
timo,minutor
Flgure 2. Chromatograms of the vapor phase (top) and the liquid phase (bottom) in the reaction vial after gas-phase derivatitatlon of formalin headspace with BSTFA, indicating the presence of methylene glycol (G,), methylal (M) and the first three oligomers of poly(oxymethylene) glycol monomethyl ether Me,Si derivatives (E,-3).
and liquid phases in the reaction vial as well as a determination of the volume of liquid in the vial. Distribution of the reaction vial contents between the liquid and gas phases after headspace derivatization was approximated by injection of serial amounts of the 5:l BSTFA/DMF solution into evacuated reaction vials followed by introduction of room air. Twenty microliters of the 5:l BSTFA/DMF
solution totally vaporized and was assumed to equal the volume of derivatization mixture in the gas phase. The volume of the liquid phase in the reaction vial after derivatization was then calculated (total injected minus 20 pL), and the total amount of derivatized formaldehyde oligomers in the formalin headspace could be determined. Since oligomers of formaldehyde in the equilibrium mixture may not be isolated and purified, the FID response factors for Me3Si derivatives of the formaldehyde condensates were estimated by BSTFA derivatization of chemically similar compounds (2). These derivatives indicate that FID response to Me3% derivatives of poly(oxymethy1ene) glycols and the glycol monomethyl ethers is a function of the T M S groups present (2, 3, 12). The response factor of the FID was determined to 8.16 X 10l2 area units/mol Me3Si. Quantitative analysis of methylal, a thermally stable compound present in formalin solution and headspace, was readily accomplished by determination of the molar response factor of reagent grade methylal (Fisher Scientific) dissolved in methanol. Detector response for methylal was determined to be 1.77 X 1013area units/mol. Gas chromatograms from the derivatization of formalin headspace by BSTFA are shown in Figure 2. Amounts of the formaldehyde condensation products as TMS ethers and methylal calculated with FID molar response factors are shown in Table I. Formaldehyde equivalents are calculated and compared with the total formaldehyde concentration in the headspace determined by the chromotropic acid method. The 30-pL injection provided a twofold excess of BSTFA for de-
2882
Anal. Chem. 1984, 56, 2882-2884
rivatization of all formaldehyde oligomers yet the 40-pL injection resulted in a greater yield. Competing reactions, including spontaneous decomposition of BSTFA, may have marginally limited its availability for derivatization of the glycols and monomethyl ethers with the 30-pL injection. Within the accuracy of measurement, all the formaldehyde equivalents in the vapor phase in equilibrium with formalin solution were in the form of methylal, methylene glycol, and the oligomers of poly(oxymethy1ene) glycol monomethyl ethers containing one, two, and three formaldehyde units. Mass spectra for the Me3Si derivatives of the formaldehyde condensates and methylal from the formalin headspace derivatization liquid phase appear in Figure 3 and are identical with the fragmentation patterns of the Me3Si derivatives obtained for formaldehyde condensates in formalin solution (2).
CONCLUSIONS Total formaldehyde content of the formalin headspace measured by a gas-phase trimethylsilylation procedure closely approximated the total formaldehyde content of the formalin solution headspace dete;mined by chromotropic acid analysis. The vapor-phase species in equilibrium with a formalin solution were shown to be methylal, methylene glycol, and three oligomeric poly(oxymethy1ene) glycol monomethyl ethers.
Registry No. CH20,50-00-0;H20,7732-18-5;CH,OH, 67-56-1; CH,(OH)2,463-57-0;HOCHZOCH,, 4461-52-3; HO(CH20)2CH,, 19942-08-6;HO(CH20),CH3,87728-58-3;methylal, 109-87-5. LITERATURE CITED (1) Walker, J. F. "Formaldehyde", 3rd ed.; Reinhold: New York, 1964; p 96. (2) Utterback, D. F.; Millington, D. S.;Gold, A. Anal Chem. 1984, 5 6 , 470-473. (3) Dankelman, W.; Daeman, J. M. H. Anal. Chem. 1978, 4 8 , 401-404. (4) Hail, M. N.; Piret, E. L. I n d . Eng. Chem. 1949, 4 1 , 1277-1286. (5) Iliceto, A. Gazz. Chim. Ita/. 1954, 84, 536-552. (6) Bryant, W. M. D.: Thompson, J. B. J . Po/ym. Scl., Part A 1971, 9 , 2523-2540. (7) Sawicki, E.; Sawlcki, C. R. "Aldehydes-Photometric Analysis": Academic Press: New York, 1978, p 194. (8) Geisling, K. L.; Miksch, R. R.; Rappaport, S. M. Anal. Chem. 1982, 5 4 , 140-142. (9) Walker, J. F. "Formaldehyde", 3rd ed.; Reinhold: New York, 1964; p 44. (10) Schnizer, A. W.; Fisher, G J.; McLean, F. J . Am Chem. SOC.1953, 7 5 , 4347-4348. (11) "Manual of Analytical Methods", 2nd ed.; National Institute of Occupational Safety and Health: Cincinnati, 1977; Vol. 1, p 127. (12) Sevcik, J. "Detectors In Gas Chromatography": Stulik, K., Transl.; Elsevier: New York, 1976, p 94.
RECEIVED for review May 31,1984. Accepted September 4, 1984. This work was supported by University of North Carolina Biomedical Research Support Grant 2S07 RR05450-21.
Simple Spectrophotometric Method for Determination of Carbonyl and Sulfonyl Chlorides Maciej Siewidski, Marianna Kuropatwa, and Apolinary Szewczuk* Biochemical Laboratory, Institute of Immunology a n d Experimental Therapy, Polish Academy of Sciences, 53-114 Wrocbw, Czerska 12, Poland
The method presenled Is based on the reactlon of carbonyl and sulfonyl chlorides wlth sodium azide In water-acetone solutlon at room temperature for 10 mln; the excess of azide Is determlned spectrophotometrlcallyafter converslon to a red complex wlth Fe3+. Reactlve anhydride can also be determined by thls method. The applicatlon of the new method was demonstrated for monitorlng conversion of sulfonyl chlorldes to sulfonyl fluorldes and for controlling the acylatlon of amino compounds wlth phthaloyl-L-glutamlc anhydrlde.
method is based on the reaction described by Curtius and Haas (6). Excess sodium azide can be easily determined spectrophotometrically, after its conversion to a red ferriazide complex. The last one was used for determination of inorganic azide (7) and for determination of organic acid azides after their alkaline hydrolysis (8,'9). So the new procedure for determination of acid chlorides proceeds according to the following two reactions:
R-S02C1 (R-COC1)
+ NaN3
-
R-S02N3 (R-CON,) A variety of methods for determination of carbonyl and sulfonyl chlorides have been described. Among them differential acid titrations, differential chloride titrations, and reductive titrations are the most precise (1). Spectrophotometric methods for determination of carbonyl chlorides are based on reaction with hydroxylamine, giving hydroxamic acid converted to red complexes with Fe3+ions (2), or on reaction with 2-(nitropheny1)hydrazine giving hydrazides, which in alkaline solution yield intensive colors (3). Carbonyl chlorides were also determined by high-performance liquid chromatography after their conversion to methyl esters (4). During our studies on covalent inhibitors of penicillin amidase from Escherichia coli (5) we developed a new, precise, and fast method for determination of sulfonyl chlorides. This
+ NaCl
N3- + Fe3+ + (FeNJ2+ (red ferrazide complex)
(I) (11)
EXPERIMENTAL SECTION Materials. Reagents synthesized in our laboratory are hexanesulfonyl chloride (10) redistillated at 114 "C/lO mmHg, phenylacetyl chloride (1I), phenylmethanesulfonyl chloride (12) recrystallized from benzene-cyclohexane, para-substituted phenylmethanesulfonyl chlorides (10,12) and phthaloyl-L-glutamic anhydride (13). Commercial reagents acetyl chloride, butyryl chloride, butyric anhydride, 54dimethylamino)-l-naphthalenesulfonyl chloride, methynesulfonyl chloride, propionic anhydride, p-toluenesulfonyl chloride were purchased from Fluka AG. Chloroacetamide was obtained from Schering VEB Adlershof, phenylmethanesulfonyl fluoride was from Sigma Chemical Co., and sodium azide (over 99%) was from Serva. Other reagents were from POCh, Gliwice.
0003-2700/84/0356-2882$01.50/0 0 1964 American Chemical Society