J. Phys. Chem. 1983, 87, 2185-2190
0
1.0
2.0
3.0
4.0
5.0
log ( 7 ) Figure 8. The guest emisslon llne width (fwhm) normalized by I? vs. the interactlon volume n. From bottom to top, these curves are for guest mole fractions p = 1.0, 0.5, 0.2, and 0.1, respectively.
might be expected to be important in BCN. The 365 sites contained in the interaction volume would yield an interaction radius, R,, of 182 for a one-dimensional interaction. Using @ = 7.4 cm-’, T~ = 20 ms for BCN, one obtains form eq 8b a value for direct exchange of yd = 0.1. Exchange interactions are usually of shorter range, i.e., yd >> 1, with yd = 5 being typical for many aromatic systems. For superexchange, one can estimate AE in this system as the energy separation between absorption and emission maxima. Using AE = 76 cm-’, one finds B = 66 cm-l for the I-D superexchange mechanism. This value of 6 is about an order of magnitude larger than the value
2185
determined for DBN. Considering three-dimensional interactions with n = 365 sites, one obtains Ro = 4.5. For direct exchange in three dimensions, using 0 = 7.4 cm-’ as an upper limit for the isotropic exchange interaction, one obtains yd < 7 using eq 8b. For superexchange with AE = 76 cm-l one obtains an isotropic p = 0.6 cm-l from eq 8c. For dipoledipole coupling in three dimensions one obtains from eq 8a a value for the nearest-neighbor transfer time 7 = 2.6 p s . The results of this analysis of the phosphorescence profile provide further support to the conclusion that the energy transfer is three dimensional for excitations on the low-energyside of the absorption profile. A t higher excitation energies where nearest-neighbor transfer is important, the energy transfer may be one dimensional. The fit of the emission profile to eq 13 gives the interaction volume for the interaction with the largest range. An interaction with a smaller range, even with a stronger coupling within that range, would be unimportant in determining the emission profile.
Acknowledgment. The authors thank Prof. P. Prasad for our initial collaboration on this work and for supplying us with the sample. M.A.E. thanks Drs. A. Blumen, S. Fischer, and E. Knapp of the theoretical group of the Technical University of Munich for many helpful discussions. M.A.E. thanks the Alexander von Humbolt foundation for a Senior Humbolt Award spent in Munich (Feb-Aug 1982). The financial support of the Office of Naval Research is greatly appreciated. Registry No. BCN, 53220-82-9.
Klnetlcs of Volatilization In Rapid Heating Mass Spectroscopy: Activation Energies for Some Hydrogen-Bonding Specles Brian Wesley Wllllams, A. Peter Irsa, Hagal Zmora,+ and Robert J. Beuhler’ hpertment of Chemisby. Brookhaven National Laboratwy, Upton, New York 11973 (Received: September 2 1, 1982)
Activation energies and characteristictemperatures for the volatilizationof several hydrogen-bondingcompounds with mild to rapid heating mass spectroscopy techniques are reported. A homologous series of N-acetyl(L-alanine), methyl esters, di- through pentaalanine, show a regular increase in activation energy of 7-9 kcal/mol per alanine residue as studied by single-ion-impacttandem mass spectroscopy. The rapid heating-chemicalionization mass spectroscopy of glycerol, meso- and pentaerythritol, D(+)-mannose,D(+)-glucose, sucrose, and maltose with ammonia and methane reagent gases gives activation energies for the production of protonated or cationized molecular ions which resemble experimental heats of vaporization for the polyalcohols and amount to -4.4 kcal/mol per CHOH group in these compounds. Trends in these data, along with the production of fragment ions in the case of the disaccharides, are briefly discussed.
Introduction Recently a great deal of research in mass spectroscopy has been concerned with the development of techniques for the analysis of thermally labile, nonvolatile species. Methods such as laser desorption and “in beam” electron impact or chemical ionization have all be applied to compounds such as carbohydrates or peptides resulting in spectra indicative of molecular weight or structure.’ Despite this practical success, however, the volatilization process itself remains unclear. Currently, heating an unstable, nonvolatile compound is viewed as causing a kinetic competition between decomposition and volatilOn leave from Soreq Nuclear Research Center, Yavneh, Israel. 0022-365418312087-2185$01.50/0
ization, where decomposition is favored since the activation energy for breakage of intramolecular bonds in these compounds is less than that required for breakage of intermolecular bonds. From this viewpoint, any technique enhancing volatilization must cause a change in the relative rates favoring the production of unfragmented molecular species. Two factors have so far been found to be effective: the nature of the surface from which the sample is desorbed or volatilized, and the rate at which the sample is heated.2 An inert surface tends to decrease surface(1)G. D. Daves, Jr., Acc. Chem. Res., 12,359 (1979),and references cited therein. (2)R. J. Beuhler, E. Flanigan, L. J. Greene, and L. Friedman, J.Am. Chem. SOC.,96, 3990 (1974).
0 1983 American Chemical Society
2186
The Journal of Physical Chemistry, Voi. 87,No. 12, 1983
molecule bonding and minimizes complementary pyrolysis or catalysis leading to fragmentation. The effect of rapid heating is somewhat more subtle; again, although the activation energy for decomposition may be less than that required for volatilization, an Arrhenius plot of the processes would show an intersection at some value of 1/T. At temperatures above this point, volatilization is favored over decomposition. Heating the sample rapidly minimizes the time in which decomposition is favored over volatilization. The present study is an investigation of the kinetics of the volatilization process for some thermally unstable species undergoing mild to rapid heating. We report here results involving two classes of model compounds investigated by two separate techniques: single-ion-impact tandem mass spectroscopy applied to a series of Nacetyl(L-alanine),, methyl esters, and the rapid heatingchemical ionization mass spectroscopy of a group of polyalcohols and mono- and disaccharides. These individual techniques provide similar information regarding the temperature dependence of the rate of molecular or quasi-molecular ion production, allowing the activation energy for sample volatilization to be found from an Arrhenius plot. The activation energies observed increase regularly with molecular weight in both experiments, and quantitatively resemble heats of vaporization obtained from experiment. Since intermolecular hydrogen bonding is a dominant interaction reflected in the heat of vaporization, this suggests that activation energies for volatilization can be quantitatively related to the degree of hydrogen bonding.
Experimental Section Sample Preparation. The N-acetyldi-L-alanyl methyl ester and the N-acetyltri-L-alanyl methyl esters were products of Fox Chemical Co. (No. 4318 and No. 4315). The N-acetyltetra- and N-acetylpenta-L-alanyl methyl esters were synthesized from the respective peptides [(Lala)(, Miles Laboratories No. 4 AL-3; (L-ala)&,Miles Laboratories No. 5 AL-51 in the following manner: The peptide (0.5 mg) was dissolved in 0.2 mL of acetic acid and 0.2 mL of acetic anhydride was added while stirring. After the solution was stirred for 18 h at room temperature the mixture was evaporated to dryness. Analysis of the reaction mixture with an amino acid analyzer prior to esterification indicated that the acylation reaction had proceeded to the extent of 97-9890. The acetylated peptides were esterified by dissolving in 0.2 mL of methanol saturated with anhydrous HC1. After 18 h at room temperature, the solution was evaporated to dryness, redissolved in methanol or CF3CH20H,and again evaporated to dryness. The esters were dried in vacuo over KOH. Analysis by thin-layer chromatography indicated that the derivatives were at least 95 90 homogeneous, based on the technique of serial dilution of the sample. Equal sensitivity to the thin layer chromatographic detection reagents for both major and minor compounds was assumed. Tandem Mass Spectroscopy. A tandem mass spectrometer system, previously described,2was used to generate low-velocity CH3NH3+ ions which ionized Nacetyl(L-alanine),,methyl esters via a gentle proton transfer process. Ionization takes place in a heated, Teflon-lined collision chamber into which the esters have been evaporated from a copper sample inlet probe covered with 0.002-in. Teflon foil. Fragments and protonated-parent (MH') ions formed in this collision chamber are focused into a quadrupole mass spectrometer, mass analyzed, and detected. Samples are dispersed on the probe by application of 5-hL aliquots of the esters in solution, 0.01-1.0
Williams et 81.
,
I
d
'
I
1
I
\
2'2
2'3
2'4
2'5
2'6 2!7 218 I / T ( O K ) x IO3
2'9
30
3I
'
Flgure 1. Determination of activation energies for volatilization of N-acetyk-alanyl methyl esters. The ion intensity of the quasi-parent ions X absolute Tis plotted against reciprocal absolute T. The peptides were dispersed on a Teflon probe and ionized with CH3NH3+in a Teflon collislon chamber. Sample probe and collision chamber surfaces were kept at the same temperature and the rate of heating was 1-5 Klmin. The amounts of samples of the derivatized peptides were 43 nmol of Ac(Ala),OMe, 70 nmol of Ac(Ala),OMe, 56 nmol of Ac(Ala),OMe, and 23 nmol of Ac(Ala)50Me. The activation energies calculated from the slopes of the lines are given for each peptide.
mg/mL in water or methanol, which undergo drying under a stream of helium. Aliquots of the same solutions were subjected to acid hydrolysis and amino acid analysis to determine the amount of ester delivered to the probe. The probe was heated at rates of 0.5 to 10 K/min from 20 to 225 "C via radiation from the collision chamber or direct heating. Chamber temperatures were determined from thermocouples placed in contact with the Teflon surfaces. Rapid Heating-Chemical Ionization Mass Spectroscopy. The experimental arrangement used is identical with that described by Udseth and Friedman.3 Samples were introduced into the ion source by placing 0.5-1-pL aliquots of aqueous solutions (1-5 nmol/pL) on a rhenium ribbon (0.030 in. X 0.0012 in. X 0.200 in.) mounted on an insertion probe and dried by brief exposure to a heat lamp. The rhenium filament was heated inside the ion source by a square top current pulse with variable amperage and duration up to 6 A and 10 s, with pulse rise time < 100 ps. Generally currents of 3-5 A and pulse durations of 100 ms were used, with the sample fully volatilized in a period of a few milliseconds. Heating rates under these conditions were up to 4000 K/s. Both methane and ammonia were used as chemical ionization reagents. Filament temperatures were determined by computer-controlled measurements of the resistivity of the rhenium filament, using a voltage to frequency converter circuit, simultaneously with mass analysis by a computer-controlled Extranuclear Laboratories q ~ a d r u p o l e . ~The analyzed ions were detected with postacceleration and secondary electron detection, also previously described.6 Sugars, glycerol, and meso- and pentaerythritol were all obtained from commercial sources and used without further purification.
Results and Discussion Activation Energies for Volatilization of N-Acetyl(LAlanine),, Methyl Esters by Tandem Mass Spectroscopy. The activation energies and temperatures required for volatilization in this homologous series are presented in Figure 1. With the rate of appearance of the protonated-parent ion having been chosen as a kinetic parameter related to the activation energy, the data represent the (3) H. R. Udseth and L. Friedman, Anal. Chem., 63,29 (1981). (4) T.P.Radus, H. R. Udaeth, and L. Friedman, J.Phys. Chem., 83,
2969 (1979). ( 6 ) R. J. Beuhler and L. Friedman, Int. J.Mass. Spectron. Ion Phys., 23,81 (1977).
The Journal of Physical Chemistty, Vol. 87, No. 12, 1983 2107
Kinetics of Volatilization In Rapid Heatlng Mass Spectroscopy
-
'-1i7
t
i
40
SUCROSE
CI SPECTRUM
0
i
m
0 10 (D
m
T
E 0
D
i
W
I
r 400
200
300 MASS ( a m u )
Figure 3. Chemical ionizatlon (NHd mass spectrum of sucrose. The spectrum shown is the superposltlon of four partial spectra. -
0
0
1
2
3
4
5
6
NUMBER OF A L A N I N E R E S I D U E S , n
Figwe 2. Actlvatkm energy for volatlllzatlon of Nacetykalanyl methyl esters as a function of number of alanine residues per molecule. The data are taken from Figure 1.
natural logarithm of MH+ ion intensity ( I )times the absolute temperature ( T ) plotted against the reciprocal absolute temperature (l/T). The factor of T makes the product I T proportional to the rate of appearance by providing a minor correction for the temperature dependence of the rate of effusion of neutral molecules from the collision chamber. Here, heating rates are much slower than in the subsequent rapid heating experiments, and the collision chamber used as well as the slower time scale of the experiment allows much more effusion of neutral molecules than the chemical ionization chamber used in later experiments. As might be expected, the dialanine derivative evaporates at the lowest temperature (50 "C) and with the least activation energy (12 kcal/mol). The pentaalanine derivative requires 180 "C, the highest temperature, and an activation energy of 39 kcal/mol. When these data are plotted against the number of alanine residues as in Figure 2, a regular increase in activation energy of 7-9 kcal/mol per alanine residue results. This value of 7-9 kcal/mol per alanine residue compares favorably with a value of 5-10 kcal/mol ascribed to an amide-nitrogen-H-oxygen bond between neutral peptide molecules.6 Before this comparison of energies can be used as evidence for the influence of hydrogen bonding on volatility, however, other factors possibly affecting the energetics of volatilization need to be considered. These would most likely arise from possible ionic interactions in the condensed phase, and the possible interaction of the esters with occluded water. The esters investigated here should be less likely to form zwitterions than single amino acids, and the protonated molecular ions formed through proton transfer correspond to species expected from the evaporation of neutral an-
hydrous molecules, suggesting that exposure to the high vacuum of the mass spectrometer has removed any water from the sample. Since it appears that the methyl esters are indeed anhydrous neutral molecules in the condensed phase, the experimental value of 7-9 kcal/mol represents essentially hydrogen bonding. With this conclusion, experimental values of the activation energies indeed correlate with the number of alanine residues, a necessary prerequisite for any model of the volatilization process based on hydrogen bonding. A model calculation for the evaporation of these alanine peptides based on a rate constant for multiple critical oscillators has been compared to the present experimental data in a separate publication.' The temperature required for the evaporation of the N-acetyldi-L-alanyl methyl ester and N-acetylpenta-Lalanyl methyl ester illustrates a second aspect of the energetics of the volatilization process which deserves comment: i.e., the molecular excitation associated with the temperature at which the molecule is volatilized. The higher the temperature, the greater the probability of molecules having excess internal energy. These excited neutral gaseous molecules would then fragment more extensively after proton transfer and indeed this was observed for some mass spectra taken on the underivatized tetra- and pentaalanine peptides.* Activation Energies and Characteristics Temperatures for Volatilization of Sugars and Polyalcohols by Rapid Heating-Chemical Ionization Mass Spectroscopy. The volatilization of the compounds glycerol, meso- and pentaerythritol, D(+)-mannose, D(+)-glucose, sucrose, and maltose were investigated via rapid heating-chemical ionization mass spectroscopy. A typical chemical ionization (CI)mass spectrum of sucrose is given in Figure 3,with ammonia used as the reagent gas. The spectrum shown is the superposition of four partial spectra, each covering a mass range of 80 mass units. Each partial spectrum was obtained by integration of about 10 mass scans of 4-ms duration each. The lower mass end of the spectrum was ~
(6) G. C. Pimentel and A. L. McClellan, Annu. Rev. Phys. Chem., 22,
347 (1971).
(7)B. W. Williams and R. J. Beuhler, Chem. Phys. Lett., 94, 126 (1983). (8)R. J. Beuhler, unpublished data.
2188
The Journal of Physical Chemistry, Vol. 87, No. 12, 1983
Williams et al.
TABLE I: Activation Energies and Characteristic Temperatures for Selected Ion Peaks compd (mol wt)
CHOH subunits
glycerol ( 9 2 )
3
meso-erythritol ( 1 2 2 )
4
pentaerythritol(l36) D ( +)-mannose (180)
4 5
D(t)-glUCOSe (180)
5
sucrose (342)
8
maltose (342)
8
E,, kcalimol 17.9 * 2.0 17.7 * 2.2 22.4 i 2.6 22.9 * 2.6 21.5 i 1.5 30.4 i 3.2 32.1 i 2.2 23.3 i 2.6 24.4 i 3.6 24.1 t 2.5 25.0 i 2.0 39.8 t 4.5 22.0 i 3.6 57.7 i 7.7 38.9 i 4 . 3 37.0 i 5.1 35.4 t 4.6 40.3 i 5.4 41.0 i 2.4 42.7 I 3.1
mle (CI gas) 9 3 (CH,) 110 (NH,) 1 2 3 (NH,) 139 (CH,) 137 (NH,) 181 (CH,) 198 (NH,) 181 (CH,) 198 (NH,) 215 (NH,) 1 6 3 (NH,) 180 (NH,) 198 (NH,) 360 (NH,) 1 6 3 (NH,) 180 (NH,) 198 (NH,) 343 (NH,) 360 (NH,)
calibrated with an Xe electron impact mass spectrum, while the higher end was calibrated with tryptophan (mass 205). One observes the parent molecule with or without addition of ammonium ion (NH4+),as well as fragments with (H,O), or (CH,O), molecular subunits removed. Sucrose and the other sugars were also analyzed under E1 conditions, but no useable spectrum could be obtained when the probe filament was removed from the direct electron beam. However, when sucrose was examined under conditions where the sample heating filament was placed directly in the electron beam, mass spectra were obtained similar to those seen under CI conditions. These results for sucrose are in agreement with those reported earlier by Anderson et aL9 Activation energies for the production of parent and fragment ions were studied by following their rate of production as a function of the reciprocal absolute temperature of the rhenium sample heating filament. Typical Arrhenius plots for sucrose are shown in Figure 4. The results for the various sugars and alcohols are tabulated in Table I. Given are the compound and its molecular weight, the number of CHOH subunits in the compound, the masslcharge ratio of the ion observed with the CI gas used, the activation energy (derived from the slope), and a characteristic volatilization temperature, Tl12,taken as the temperature at which the observed counting rate is half of its maximum value. The purpose of reporting TlI2 values is twofold. First, this value serves as a figure of merit of the reproducibility of the experimental heating curves observed. If a given curve were to have a T l j z greatly different than those previously observed for a particular ionic species, this would be indicative of some problem in either filament heating or temperature determination. Second, in the case of such species as maltose, where numerous fragment peaks are observed, comparison of Til, values among the peaks helps to correlate the data and suggest conclusions about the volatilization process. Errors reported for the activation energy indicate the statistical standard deviation of the experimental data, where each measurement was repeated at least five times. The errors reported for the characteristic Tl12volatilization temperature also reflect the statistical standard deviation of the experimental data and do not exceed 10% relative error throughout the experiment. However, the possible
-
(9) W. R. Anderson, Jr., W. Frick, and G. D. Daves, Jr., J.Am. Chem. SOC.,100 1974 (1978).
t'
TI,,,
405 t 395i 415i 425 f 440i 350 t 360 i 3701 380 i 360t 405 t 385 i 425t 375 i 390 i 410+ 430i 405t 410 i 405i I
I
I
I
K 20 30 20 10 10 5 5 5 20 10 5 5 25 10 5 5 20 10 10 10 I
SUCROSE
I o3
c
i
m / e 163 0 m/e
10
2.3
2.4
360
2.5
2.6
2.7
2.0
I / T ( ~ - 1 x 103)
Flgure 4. Determination of activation energies for volatilization of rapidly heated sucrose, obtained with NH, CI. The small arrows represent the points at which T,,* was determined.
uncertainty in the absolute temperature measurement may be as large as &20%due to uneven heating of the rhenium filament. In all cases, however, since the activation energies are derived from relative temperature differences, these values are not sensitive to large possible systematic errors in the measurement of absolute temperature, and retain the uncertainties reported. Several trends can be noted in the data in Table I. The most basic trend is toward greater activation energy with increasing molecular weight. For those species studied with both ammonia and methane as chemical ionization reagents, little difference is noted in the activation energy observed, although the mass of the predominantly observed ion varies. A t the source pressures of the current experiment, ammonia generally gives what is attributed to a parent plus ammonium ion peak (M + NH4+)while methane usually gives a protonated parent (MH+). Also,
Kinetics of Volatilization in Rapid Heating Mass Spectroscopy
the use of ammonia often gives more intense parent or cationized parent for similar amounts of sample compound. These differences can easily be understood on the basis of relative proton affinities and hydrogen-bonding capabilities. If one envisions a schematic ionization process of proton transfer followed by addition of a molecule of unionized reagent gas, the use of ammonia results in a protonated species with greater internal relaxation, favoring unfragmented products and possible subsequent addition. Ammonia will also be favored over methane in any subsequent addition reactions because of its unpaired electrons. The observed similarity in activation energies, however, does not allow any conclusions about the mechanism of ionic production. The evaporation of neutral species with ionization in the gas phase is certainly consistent with this result, but a mechanism involving protonation or other interaction with the surface cannot be ruled out. Indeed, earlier results involving methane and polystyrene suggest such surface intera~tion.~ What can be noted is that, in any possible surface interaction involving protonation, the possible exothermicity of methane over ammonia causes no effect on the observed activation energy of the volatilization process. The trend of the data for maltose suggests a conclusion in this case regarding the fragmentation process producing the observed species. The similarity of activation energies and characteristic volatilization temperatures suggests that the rate-limiting step in the production of both parent and fragment species is the same. While surface fragmentations are probable it would appear from the overall behavior that these fragments would require smaller activation energies than the parent or quasi-parent species for volatilization, and therefore probably do not contribute substantially to the observed species and energies. A mechanism involving fragmentation in the gas phase after volatilization of the parent would be favored. Again, the specific issue of surface vs. gas-phase ionization cannot be settled, because the activation energy measurement only pertains to the rate-limiting step which may occur before or after ionization of an unfragmented parent. In contrast to the other data in Table I, the data reported for sucrose show quite different activation energies for different m / e ions. It would appear that the 360 and 180 peaks are produced through a gas-phase fragmentation mechanism because the E, values are similar, approximately 39 kcal/mol. However, the 163 and 198 peaks have values of E, which are significantly different, which could result from a dominant surface fragmentation process. Two values of E, for the m / e 198 peak have been reported, because either one was equally probable for similar conditions of reagent gas and heating rate. The reason for this behavior is not known though it might reflect several possibilities: an alternative surface fragmentation mechanism, the production of a dimer species which subsequently fragments, or the effect of strong intramolecular hydrogen bonding in the sucrose crystal.1° The mass spectra were examined for dimer species but no evidence for their production was found. The rapid heating mass spectroscopy of sucrose will require further investigation before any more definitive conclusions can be reached. The availability of experimental measurements of the heat of vaporization for the polyalcohols used here, as well as the correlation noted earlier between the number of alanine residues and the observed activation energies, suggests that a similar correlation be attempted with the ~~
~
(10) G. M. Brown and H. A. Levy, Science, 141, 921 (1963); Acta Crystallogr., Sect. B , 29, 790 (1973).
The Journal of Physical Chemistry, Vol. 87, No. 12, 1983 2189
50 45.
c I
9 1
401 35
5
!251 E 301
20
"v
1
IO
- - 5
1
2
3 4 5 6 7 N U M B E R O F CHOH GROUPS
8
Flgure 5. Activation energy for vdatilizatkm of polyalcoholsand sugars as a function of the number of CHOH groups per molecule. The data are taken from Table I, where the open circles represent experimental activation energies and the open squares represent previously determined heats of vaporization for polyalcohols.
data in Table I. Figure 5 represents experimental heats of vaporization, represented as open squares, and experimental activation energies, represented as open circles, plotted against the number of CHOH subunits in the polyalcohols and sugars investigated. The line represents a least-squares linear fit to the experimental heat of vaporization data (methanol, 9.0 kcal/mol; 1,2-ethanediol, 14.0 kcal/mol; glycerol, 18.2 kcal/mol; meso-erythritol, 22.3 kcal/mol).l' A correlation similar to that noted for the alanine results, where a regular increase of -4.4 kcal/mol per CHOH subunit results. While activation energies and enthalpies in general are not strictly equivalent, in this case the comparison is meaningful because the activation energy for the reverse process (condensationon the surface) would be expected to be small, and the activation energy for volatilization primarily reflects the energy required to remove a species into the gas phase. The observed value of -4.4 kcal/mol per CHOH resembles the value of -4.0 kcal/mol per hydroxyl group noted for hydrogen bonding in liquid water,12and suggests that, as in the case of the alanine residues, the activation energy for volatilization is primarily determined by the degree of hydrogen bonding in the molecule to be volatilized. The correlation between heats of vaporization, rather than sublimation, and activation energy may also imply that a regular crystal structure either is not formed on the surface or is lost upon rapid heating prior to volatilization. This difference is most striking in the case of pentaerythritol, which has reported experimental heats of sub(11) Heats of vaporization for methanol, 1,2-ethanediol,and glycerol were taken from 'Handbook of Chemistry and Physics", 53 ed, R. C. Weast, Ed., Chemical Rubber Co., Cleveland, Ohio, 1972; for mesoerythritol from S . Seki and K. Suzuki, Bull. Chem. Soc. Jpn., 26, 63 (1953). (12) W. A. P. Luck, 'Hydrogen Bonds in Liquid Water" in 'The Hydrogen Bond", Vol. 111, P. Schuster, G. Zundel, and C. Sandorfy, Ed., North Holland Publishing Co., Amsterdam, 1976, p 1369.
J. Phys. Chem. 1983, 87, 2190-2193
2190
limation of 31.4 and 34.4 kcal/mol13 and an experimental activation energy of only 21.5 f 1.5 kcal/mol. This effect might conceivably arise due to the larger surface/volume ratios expected for the absorbed samples as opposed to a bulk crystal, although no systematic effect of sample size upon activation energy was noted in the present work. (13) S. Seki and K. Suzuki, Bull. Chem. SOC.Jpn., 26,63 (1953); R. S. Bradley and S. Cotson, J . Chem. SOC.,1684 (1953).
Acknowledgment. This research was carried out at Brookhaven National Laboratory under contract with the U.S. Department of Energy and supported by its Office of Basic Energy Sciences. Registry No. AcOCH3, 79-20-9; A~(Ala)~oMe, 30802-26-7; Ac(Ala),OMe,26910-17-8;Ac(Ala),OMe,30802-29-0;Ac(Ala)50Me, 85083-58-5;glycerol, 56-81-5; meso-erythritol,149-32-6; pentaerythritol, 115-77-5; D-mannose, 3458-28-4;D-glucose, 50-99-7; sucrose, 57-50-1;maltose, 69-79-4.
A Fourier Transform Infrared Study of the Kinetics and Mechanism for the Reaction HO -t CH,OOH H. Nikl", P. D. Maker, C. M. Savage, and L. P. Breltenbach Research and Engineering Staff, Ford Motor Company, Dearborn, Michigan 48 12 1 (Received: October 14, 1982)
The analysis of products in the photolysis of mixtures containing C2H60N0,NO, and CH300H in ppm concentrations in 700 torr of O2-N2diluent by FT IR is consistent with a value of (kl,/klb) = 0.77 (Et20%)for the competitive H-abstraction channels HO + CH300H CHzOOH + H20 (la) and CH300 + H20 (lb). An absolute value of (kla + k l b ) = 1.0 X lo-'' cm3molecule-' s-' was derived from the decay rates of CH300H relative to those of C2H4 and CH3CHO.
-
-
Introduction Methyl hydroperoxide (CH,OOH) is a potentially important atmospheric constituent formed in the photooxidation of CHI and other organic compounds, i.e., CH300 + HOO CH300H + Oz.' The assessment of its subsequent fate is, therefore, of considerable interest. However, no quantitative kinetic or mechanistic studies of the gas-phase bimolecular reactions of CH300H have been made to date, primarily because of numerous experimental difficulties associated with sample preparation, gas handling, identification, and quantification of the CH,00H.2 In the present long-path FT IR work, these problems have been largely overcome and product studies have been made in the HO-radical initiated oxidation of CH300H. The major mechanistic features to be addressed are the relative importance of the two possible H-abstraction channels, in one instance from the CH, group and in the other from the HO group, i.e., reactions l a and Ib,
-
and the subsequent reactions of the ensuing radical CH,OOH, e.g., reaction 2, the relevant reactions of CH300 CHzOOH CHzO + HO (2) being reasonably well understood as discussed later. For comparative purposes, product studies were also carried out for the C1-atom initiated oxidation of CH,OOH to examine the relative importance of the competitive H-abstraction reactions 3a and 3b analogous to reactions
-
CH200H f
HCI
(3a) (3b)
l a and lb. According to an empirical kinetic correlation among related C1-atom reactions? reaction 3a is expected to be fast, -1 X 10-lo cm3 molecule-' s-', and dominant over reaction 3b. Interestingly, in our earlier semiquantitative study of the C1-CH300H reaction in the presence of NOz, a substantial yield of CH300NOz formed via CH,OO + NOz CH300NOZwas observed.2 There are two possible sources for the CH,OO radical in this system, i.e., reaction 3b and, alternatively, reaction 3a followed by reactions 2 and lb. The results obtained in the present study are consistent with the latter mechanism.
-
Experimental Section General features of the long-path FT IR method used were described p r e v i ~ u s l y . ~A~3-m ~ long Pyrex IR cell (180-m path length, multipassed, KBr windows, and internal mirrors) equipped with UV fluorescent lamps (GE F4OBLB) served as the photochemical reactor. Reference absorbance spectra for the reactants and products were generated and calibrated by using a 50-cm long quartz cell (1-m path length, double passed, KBr windows, and external mirrors), directly monitoring partial pressures of pure samples in the range of 0.1-1 torr. The interferometer was operated with either a KBr beam splitter-CuGe detector or a CaFz beam splittel-InSb detector. Spectra with peak-to-peak signal-to-noise ratios exceeding 1OO:l throughout the 600-4000-cm-' region at 1/ 16 cm-' theo-
~
(1) See, for example, H. Levy, Planet, Space Sci., 20, 919 (1972); P. J. Crutzen, Annu. Reu. Earth Planet. Sci., 7,443 (1979); 3. A. Logan, M. J. Prather, S. C. Wofsy, and M. B. McElroy, J. Geophys. Res., 86,7210 (1981), and references therein. (2) H. Niki, P. D. Maker, C. M. Savage, and L. P. Breitenbach, Chem. Phys. Lett., 55, 289 (1978). 0022-3654183J2087-2 190$01.5010
(3) J. V. Michael, D. F. Nava, W. A. Payne, and L. J. Stief, Chem. Phys. Lett., 77, 110 (1981). (4) P. D. Maker, H. Niki, C. M. Savage, and L. P. Breitenbach, ACS Symp. Ser., 94, 161 (1979). ( 5 ) H. Niki. P. D. Maker, C. M. Savaae. and L. P. Breitenbach, J.Mol. Struct. 59, 1 (1980).
0 1983 American Chemical Society
