Chem. Res. Toxicol. 1988,1, 379-384
379
A Chemical Comparison of Methanesulfonyl Fluoride with Organofluorophosphorus Ester Anticholinesterase Compounds Arthur W. Snow* and William R. Barger Chemistry Division, Naval Research Laboratory, Washington, D.C. 20375-5000 Received June 20, 1988
The chemical reactivity and physical properties of methanesulfonyl fluoride (MSF) were quantitatively measured for direct comparison with those of isopropyl methylphosphonofluoridate (GB, sarin) and diisopropyl phosphorofluoridate (DFP). The chemistry involved reaction rate constant measurements of MSF hydrolysis and for reactions with phenolic, amine, oxime, hydroxamic acid, phenyl N-hydroxycarbamate, and hydroxylamine compounds and cupric imidazole and bipyridyl complexes. Analogous chemistry with GB and DFP occurred only for the nucleophilic reactions. MSF was substantially less reactive toward interactions involving hydrogen bonding, cupric complex coordination, and quaternary amine formation. Physical characterization included water solubility, vapor pressure, heat of vaporization, solubility parameter, hydrogen bond basicity, W, Et,and NMR measurements. MSF is a much weaker hydrogen bond acceptor than the phosphorus esters. The hydrogen bond basicity P-scale measurement provides a key insight to the chemical reactivity divergence of MSF and DFP or GB. Its unexpectedly low value is interpreted that MSF does not possess a site with sufficient basicity to be protonated by strong acid, to enter into hydrogen bonding, or for coordination with the cupric ion.
Introduction Acetylcholinesterase inhibitors, such as isopropyl methylphosphonofloridate (GB, sarin) and diisopropyl phosphorofluoridate (DFP), have applications as chemical warfare agents and insecticides (1). These compounds function by phosphorylation of a serine residue in the enzyme active site (2). Likewise, methanesulfonyl fluoride (MSF) is an acetylcholinesterase inhibitor that functions by sulfonylation of the esteratic site (3-5). However, the inhibition mechanism and reaction characteristics are different from those of GB and DFP as indicated by a comparative study of MSF and dimethylphosphoryl fluoride (6). Our interest in MSF originated from an erroneous report that it is a nontoxic simulant for g-agents (7). In the process of using it in such a manner, we observed a divergance in chemical reactivity between that of MSF and GB or DFP. A search of the literture for nucleophilic and hydrolytic reactivity of MSF found very few examples and no kinetic measurements. This indicated a need for a more complete and quantitative physical and chemical characterization of MSF for comparison with GB and DFP, which was the objective of this work. The approach was to subject MSF to a series of chemical reactions and physical measurements which could be compared with those in the literature for GB and DFP. An interesting finding is that the extremely weak basicity of MSF could account for the sulfonyl fluoride-phosphoryl fluoride reactivity divergence which may relate to the acetylcholinesterase inhibition and reactivation differences. Experimental Section All reagents and solvents were of reagent grade quality, purchased commercially and used without further purification unless otherwise noted. Triethylamine and pyridine were freshly distilled before use. The following reagents were recrystallized from the associated solvent before use: imidazole/ chloroform, catechol/ toluene, resorcinol/toluene, pyrogallol/toluene, and hydroquinone/carbon tetrachloride and sublimed. Spectroscopic data
were obtained from the following associated instruments: UVjPerkin-Elmer Lambda 5, IR/Perkin-Elmer Model 1430, and 'H and '9NMFt/Varian EM390 (90 MHz) with TMS and CFC13 references, respectively. Gas chromatography measurements were obtained by using a Varian GC Model 3700 equipped with a Hewlett-Packard Model 3390A integrator. Melting points are uncorrected. Elemental analysea were performed by Schwartzkopf Microanalytical Laboratory. Benzohydroxamic acid was synthesized by reaction of hydroxylamine with methyl benzoate (8). Yield 53%; mp 122-125 "C (from benzene); BH (DMSO-de) 7.45 (3 H, m), 7.75 (2 H, m). Phenyl N-hydroxycarbamate was synthesized by reaction of phenyl chlorocarbonate with hydroxylamine (9). Yield 57%; mp 104-105 "C (from benzene); 6~ (DMSO-de) 7.25 (5 H, m), 9.05 (1 H, s, D20 exchangeable), 10.25 (1H, s, D20 exchangeable). 1-Ethylpyridinium-&aldoxime bromide was synthesized by reaction of ethyl bromide with pyridin-4-aldoxime in ethanol at 65 "C for 18 h (10). Yield 90%; 8H (DMSO-d6)1.52 (3 H, t, Me), 4.70 (2 H, q,CH,CH,), 8.26 (2 H, d), 8.47 (1 H, s), 9.17 (2 H, d), 12.8 (1 H, s, OH, D20 exchangeable). 0-(Phenylcarbamyl) benzohydroxamate, C6H6CONHOCONHCsH5, was prepared by reaction of 0.48 g of benzohydroxamic acid and 0.25 mL of methanesulfonyl fluoride in 50 mL of 0.20 M NaHC03 buffer. Yield 0.28 g, 63%; mp 238-240 OC (from 5 7 CHC13/EtOAc) [lit. 234 "C (11)]. Anal. Calcd for Cl4Hl2N2O3: C, 65.6; H, 4.7; N, 10.9. Found: C, 65.4; H, 4.7; N, 10.9. 6~ (DMSO-de) 7.5 (10 H, m), 10.3 (1H, s), 12.3 (1 H, 8 ) . Phenyl N- [(methylsulfonyl)oxy]carbamate, C&,OCONHOS02CH3, was prepared by reaction of 0.54 g of phenyl Nhydroxycarbamate and 0.25 mL of methanesulfonyl fluoride in 50 mL of 0.20 M NaHC03 buffer. The precipitated product was collected after acidification to pH 3.0 with concentrated HC1. Yield 0.51 g, 63%; mp 97-98 "c (from CHC13/CeHd. Anal. Cdcd for CBHBN05S:C, 41.6; H, 3.9; N, 6.1; S, 13.9. Found C, 41.6; H, 4.1; N, 6.3; S, 14.0. 6~ (DMSO-de) 3.34 (3 H, 9, CH3), 7.3 (5 H, m), 12.4 (1 H, s, NH, D20 exchangeable). The MSF reaction rates were studied by measuring the MSF concentration with gas chromatography. MSF is highly toxic and should be handled with gas-tight syringes, in an efficient hood with a concentrated alcoholic hydroxide bath available to decontaminant glassware. The pH was monitored with a pH meter and regulated by using a 0.2 M phosphoric-acetieboric acid plus additions of 1.0 M NaOH buffer for the MSF pH-dependent
This article not subject to U.S. Copyright. Published 1988 by the American Chemical Society
Snow and Barger
380 Chem. Res. Toxicol., Vol. 1, No. 6, 1988 hydrolysis study and a 0.2 M sodium bicarbonate buffer for the MSF coreactant studies. The MSF hydrolysis reaction was conducted at 23 "C in 50-mL solutions of pH control buffer prepared from the following quantities in mL of 0.2 M phosphoric-acetic-boric acidll.0 M NaOH components with the corresponding pH: 50.0/0, 1.8; 46.0/4.0, 2.05; 43.017.0, 2.50; 40.0110.0, 4.1; 37.0113.0, 5.1; 34.0116.0, 6.45; 32.0118.0, 7.15; 31.3118.7, 8.15; 30.3119.7, 8.75; 29.51205, 9.1; 29.4120.6, 9.25; 29.0121.0, 9.6; 28.0122.0, 10.1; of MSF 27.0123.0, 10.9. A precise volume of 0.050 mL (0.70 "01) was added with a gas-tight microsyringe to the stirred buffer solution. Samples for MSF concentration measurements were withdrawn as 1.0-mL volumes by gas-tight syringe at intervals ranging from 30 to 1200 s depending on the speed of the reaction. At pH >10.9 the reaction was too fast for this sampling procedure. The pH was monitored during the reaction and observed not to vary by more than 0.5 unit. The 1.0-mL reaction samples were immediately transferred to 3.5-mL septum-capped vials containing 0.200 mL of CH2C12which contained a precise concentration (0.025 mLl25.00 mL of CH2C12)of mesitylene or toluene as an internal GC reference. For the fast hydrolysis reactions (at pH >8) a drop of concentrated phosphoric acid was premixed with a 0.200-mL CH2C12solution to immediately quench the hydrolysis reaction. The mixture was rapidly stirred for 60 s to extract the unreacted MSF. After phase separation approximately 0.1 mL of the CHZClz phase was transferred by gas-tight syringe to a 1-mL septumcapped vial for GC analysis. For GC analysis, 20% DC550 silicone in. stainless on Gas-Chrom P-A/W DMCS 45/60 in a 4 ft X 'I4 steel column was used. The linear temperature program consisted of a 40 O C - 1 min initial time followed by a 4 OC/min temperature increase to 160 "C and a 1-min hold time at 160 OC. The CH2C12, MSF, toluene, and mesitylene peaks were resolved with the following respective retention times: 6.0,11.5, 16.5, and 25 min. Peak areas were obtained by electronic integration. Using MSF, toluene, and mesitylene solutions of known concentration, an MSF CH2C12/H20partition coefficient of 11.8 was determined, and the MSF peak mea of the reaction mixture samples was converted to a molar concentration by calculation. For the MSF-coreactant study, the reaction was conducted in a 50-mL solution consisting of 40 mL of a 0.20 M sodium bicarbonate buffer (pH = 8.2), a 10-mL aqueous solution of the coreactant (with pH adjusted to 8.2), and 0.050-mL microsyringe addition of the MSF. A coreactant:MSF molar ratio of 5:l was used, and the MSF concentration as a function of reaction time was measured by a procedure analogous to that for the hydrolysis study. The coreactanta included phenol, hydroquinone, resorcinol, catechol, pyrogallol, triethylamine, pyridine, imidazole, copper sulfate, cupric imidazole, and bipyridyl complexes (prepared by stoichiometric addition of imidazole or bippidyl to copper sulfate solutions), hydroxylamine, phenyl N-hydroxycarbamate, benzohydroxamic acid, and ethylpyridinium-4-aldoximebromide. Solubility of MSF in water and cyclohexane was measured from the UV absorbance of saturated solutions as extrapolated from Beer's law plots of analytical solutions. Vapor pressure-temperature measurements on MSF over a 20-80 "C temperature range were made by using an apparatus designed for small samples (12). Hydrogen bond acceptor strengths and complex formation constants were measured by using 'gF NMR with p-fluorophenol and p-fluoroanisole as the reference compounds (13). The sensitivity of the instrument limited the concentration of these reference compounds to 0.02 M instead of the 0.01 M as prescribed (13).
Results
MSF Chemical Reactivity.
Hydrolysis of G B a n d
DFP is catalyzed by both hydroxyl a n d hydronium ions (14).Spontaneous hydrolysis by water is also measurable. T h e hydrolysis reaction a n d rate laws may be written as follows with MSF in place of GB.
CH3S02F + H2O d [MSF] --= dt k,[MSF]
H+ or OH-
CH3S03H + HF
+ ~ o H [ O H - ] [ M S F ]+ ~ H [ H + ] [ M S F ]
I
I
,
0
1
I
I
I
2
3
4
Time
I 0
,
, 2
,
, 4
.
i 5
6
7
5
9
( r e c ) a 163
,
,
6
, 8
I
, 10
12
14
PH
Figure 1. Kinetics of MSF hydrolysis at 23 "C: (a) first-order dependence on MSF concentration and (b) pH dependence of observed MSF hydrolysis rate constant compared with that of DFP (data from reference 14 a t 25 "C). Table I. Spontaneous and Hydroxyl Ion Catalyzed Hydrolysis Rate Constants for MSF. GB. and DFP compd k,, s-* koa, L mo1-ls-l MSF 1 x 104 21 GB 3 x 1048' 26' DFP 2 x 1odc 0.86d
Reference 16, note 1. Reference 14, Table 2. Reference 16, p 1375. dReference 16, Table 7. W h e n t h e pH is maintained constant, a n integrated rate law with an observed first-order rate constant is obtained. ln [MSF] = (k, k ' o ~= k h ) t = kobsdt (1) kobad = k, k o ~ [ O H - l4- ~ H [ H + ] (2) T h e MSF hydrolysis rate at constant p H was followed by gas chromatographic measurements of MSF concentration. T h e reaction was first order with respect t o MSF concentration at different constant p H values ranging from 1.8 t o 10.9 (Figure l a ) a n d was too fast for t h e sampling procedure above p H 10.9. A profile of the pH dependence of the observed rate constant is presented in Figure l b with comparable literature data for D F P ( 1 4 ) . Noteworthy features are the following: MSF is more readily hydrolyzed t h a n D F P at neutral a n d alkaline p H , hydroxyl ion catalysis commences at pH 8-9, and hydronium ion catalysis of MSF does not occur within t h e pH 1.8 limit of this study. A plot of kobsd vs [OH-] is linear, yielding values for t h e hydroxyl ion catalyzed rate constant, koH,and t h e spontaneous hydrolysis rate constant, k,, from t h e slope a n d intercept as indicated by eq 2 with t h e last t e r m negligible. These values are presented in Table I with corresponding values for G B and DFP. T h e M S F and G B koH values are remarkably close. T h e reaction of G B with phenolic compounds has been studied with t h e conclusion t h a t phenols react in their
Comparison of MSF with Organofluorophosphorus Esters Table 11. Phenolic Reactivity toward MSF and GB MSF MSF GB phenolic compd PK how, kobd/kphenol k2/k2tphenol) 0.00033 none (control) 1.0 1.0' 9.78 0.00034 phenol 1.3 2.2' hydroquinone 10.12 0.00044 resorcinol 1.3 1.P 9.62 0.00044 1.6 11' catechol 9.60 0.00054 1.6 27b 9.12 0.00055 pyrogallol
'Calculated from Table I of reference 17. Calculated from Tables IV and VI1 of reference 18. ionized form and that positioning of a second phenolic hydroxyl ortho to the phenolate nucleophilic center enhances the reaction rate by way of a postulated hydrogen-bonded transition state (17). 0
II
Chem. Res. Toxicol., Vol. 1, No.6,1988 381
-
'Reference 20, p 2127. bReference 20, Table 11. Table IV. Cupric Complex Catalyzed Hydrolysis of MSF and DFP MSF MSF DFP coreactant kow, s-l kow/kmnml rate/rate(control) none (control) 0.00030 1.0 1.00 0.00032 1.1 4.0' cuso4 imidazole 0.00033 1.1 3.4' CuIml 0.00037 1.2 13' 0.00036 1.2 33' CuIm2 CuIml 0.00037 1.2 33' CuBipy, 0.00035 1.1 40b CuBipy, 0.00034 1.1
