ARTICLE pubs.acs.org/JPCA
Infrared Emission Following Photolysis of Methylisothiocyanate and Methylthiocyanate Elisabeth A. Wade* and Jennifer L. Pore Department of Chemistry and Physics, Mills College, 5000 MacArthur Blvd., Oakland, California, 94613, United States
David L. Osborn* Combustion Research Facility, Mail Stop 9055, Sandia National Laboratories, Livermore, California, 94551-0969, United States ABSTRACT: Methylisothiocyanate (CH3NCS) was photolyzed at 193 and 248 nm, and the resulting time-resolved infrared emission was observed. Similar experiments were performed on methylthiocyanate (CH3SCN) photolyzed at 193 nm. Previous work suggested that these isomers undergo excited-state isomerization prior to dissociation, but other experiments have contradicted this claim. In the infrared emission experiments, we observed the same products from both starting materials, supporting the theory of excited-state isomerization prior to dissociation. Methylisothiocyanate is the active ingredient in a widely used pesticide and has been observed to form highly toxic methyl isocyanate (CH3NCO) under environmental conditions. The mechanism for this formation has been unclear, but must involve some oxygen-containing species. At 248 nm, methylisothiocyanate was photolyzed alone and with three atmospheric oxidizers: O2, NO, and NO2. No chemical reaction was observed with O2, whereas secondary reactions were observed with NO and NO2. When methylisothiocyanate was photolyzed with NO2, methyl isocyanate (CH3NCO) was observed, suggesting a likely environmental mechanism for methyl isocyanate formation.
’ INTRODUCTION Methylisothiocyanate (CH3NCS, MITC) is the primary decomposition product of metam sodium (CH3NHCS2Na), the second most commonly used fumigant in the United States.1 A fumigant is a broad-spectrum pesticide used to enhance the yield and quality of agricultural produce. When metam sodium is applied to fields, up to 60% of the MITC escapes into the atmosphere,2 although the air concentration varies by application method.3 Alvarez and Moore found that, at 308 nm, MITC photolyzes to produce methyl isocyanide (CH3NC) and sulfur atoms with a quantum yield of 0.98 ( 0.24.4 Based on a quantum yield of 1 and the absorbance of MITC at actinic wavelengths (λ g 295 nm), they estimated MITC’s environmental lifetime as 41 h.4 The final reaction products of MITC photolysis are of interest environmentally because CH3NC is not photochemically active in the troposphere, but as a carbene, it should react with other atmospheric species. Geddes et al. observed the reaction products of MITC photolysis in smog chambers5 and detected methyl isocyanate (CH3NCO), which is acutely toxic to humans.6 However, Alvarez and Moore found that, when MITC was photolyzed at 308 nm in pure oxygen, no CH3NCO was observed.4 In this study, we photolyzed MITC at 248 nm in the presence of several atmospheric oxidizers, namely, O2, NO, and NO2, to determine whether these reactions could result in CH3NCO production. Although 248 nm is not an actinic wavelength, it excites the same electronic transition in MITC as 308-nm radiation, albeit with stronger absorbance by MITC. Although we did not have access to a 308-nm laser, there is a significant disadvantage in 308-nm r 2011 American Chemical Society
excitation for the present purposes. Specifically, excitation at 248 nm is near the absorbance minimum of NO2, so that any observed reactions will be dominated by MITC photolysis rather than reactions with O(3P) atoms from NO2 photolysis. MITC has an isomer, methylthiocyanate (CH3SCN, MTC), that is also photochemically active. Possible dissociation channels for MITC and MTC, including that observed by Alvarez and Moore,4 are shown in Figure 1. Several groups have compared the photoproducts of MITC and MTC photolysis. The earliest study, by D’Amario et al., used vacuum ultraviolet (120 220 nm) photodissociation to produce electronically excited NCS(A) and measured the threshold for production of electronically excited NCS(A).7 Tokue et al. used low-energy electron impact to improve the threshold measurement for NCS(A) production, at 7.5 ( 1.0 eV for MITC and 6.8 ( 0.5 eV for MTC.8 Most recently, Northrup and Sears used UV photodissociation at 193 and 248 nm with laser-induced fluorescence (LIF) detection to measure the internal excitation of NCS(X).9 They observed both NCS and CN produced by a single-photon process for both MITC and MTC photolysis. The structure of MITC, CH3NCS, means that a CN fragment cannot be produced by simple fission of a single bond. Northrup and Sears suggested that both MITC and MTC isomerize on the excited state before dissociation. However, later work by Hall and Wu attempted to detect CN following photodissociation of Received: January 2, 2011 Revised: March 24, 2011 Published: May 09, 2011 5319
dx.doi.org/10.1021/jp2000305 | J. Phys. Chem. A 2011, 115, 5319–5323
The Journal of Physical Chemistry A
ARTICLE
Figure 1. Schematic of reaction channels for MITC and MTC. Thermodynamic data were taken from refs 1216.
CH3CH2SCN and HSCN; they were unable to observe any CN.10 They suggested that the CN detected by Northrup and Sears was actually due to photolysis of a thiocyanate contaminant rather than photolysis of the isothiocyanate. As reported herein, by comparing state-resolved infrared emission from MITC and MTC photolysis products following excitation at 193 nm, we found that CN fragments are produced from both MITC and MTC, supporting the excited-state isomerization mechanism proposed by Northrup and Sears.
’ EXPERIMENTAL SECTION The experimental system used here has been described elsewhere.11 Molecular emission from a room-temperature flow cell containing multipass collection optics was modulated by a timeresolved Fourier transform spectrometer, operating in step-scan mode. Calibrated mass flow controllers delivered MITC or MTC seeded in helium buffer gas to the flow cell, for a total flow rate of ∼20 standard cm3/min (sccm). Helium was bubbled through neat MTC or MITC. The total pressure in the bubbler was maintained at ∼570 Torr. The vapor pressure of MITC at room temperature is 14 Torr and that of MTC is 10 Torr, giving mole fractions in the mixture of ∼2.5% and ∼1.7%, respectively. When atmospheric oxidizers (O2, NO, or NO2) were added, the pure gas was delivered to the cell by a separate mass flow controller at 060 sccm. In all cases, a ∼100 sccm flow of helium purged the photolysis laser windows. The total pressure in the cell was typically 350 mtorr. An unfocused ArF laser was used as the 193-nm photodissociation source (30 Hz, 20 mJ cm2 pulse1), and an unfocused KrF laser was used as the 248-nm photodissociation source (30 Hz, 55 mJ cm2 pulse1). We attempted experiments using 355-nm excitation, where absorption by both photolytes is very weak; however, signals were not observed in either case. Following the photodissociation pulse, infrared emission was collected perpendicular to the photodissociation laser beam by a pair of 10-cm-diameter silver-coated f/1 spherical mirrors, separated by 20 cm. These multipass collection optics focus the emission to the small area required for fast infrared detectors. The emission passed through a Bruker IFS 66v/S Step-Scan Fourier transform spectrometer, operated in step-scan mode. The modulated emission was then detected using a liquidnitrogen-cooled InSb detector, which is sensitive to emission above 1850 cm1. In some spectra, bandpass filters were used to limit the detected infrared emission to particular regions of the
Figure 2. Relative infrared emission immediately after photolysis of MITC (solid line) and MTC (dotted line) at 193 nm. The resolution of the infrared spectra is 64 cm1. Spectra in both parts A and B are normalized, so that the largest emission peaks are equal in intensity. However, because the mid-IR emission from MTC is very weak, the normalized signal for part B is relatively noisy.
spectra. A 16-bit, 200 kS/s transient digitizer collecting one interferogram every 5 μs was used throughout.
’ RESULTS AND DISCUSSION A. Photolysis of MITC and MTC at 193 nm. Relative infrared emission following 193-nm photolysis of MITC and MTC is shown in Figure 2. Figure 2A shows the emission from the InSb cutoff at 1850 to 12000 cm1, whereas Figure 2B focuses on the mid-IR emission between 1850 and 4000 cm1. For both MITC and MTC, the same three product channels can be seen, although with varying relative intensities:
CH3 SCN or CH3 NCS f CH3 þ NCSðXÞ CH3 SCN or CH3 NCS f CH3 S þ CNðAÞ CH3 SCN or CH3 NCS f CH3 NC þ S The band origins of the observed infrared emission are summarized in Table 1. We observed electronically excited CN(A) through its A (2Π) f X (2Σ) emission.16 Power studies performed in our laboratory indicate that the observed emission is due to single-photon processes, which is consistent with the results of Northrup and Sears.9 Although the relative emission signals of the photoproducts are quite different for the two precursor molecules, it is generally not possible to directly relate infrared emission ratios to product branching ratios, because only electronically or vibrationally excited products are detected. For MTC, almost all of the observed signal is due to electronic emission from CN(A). However, because the Einstein A coefficient for electronic emission is much larger than the Einstein A coefficient for infrared emission from NCS and 5320
dx.doi.org/10.1021/jp2000305 |J. Phys. Chem. A 2011, 115, 5319–5323
The Journal of Physical Chemistry A
ARTICLE
Table 1. Band Origins of Observed Speciesa chemical species CH3NC
a
band origin (cm1)
band description
2166
CN stretch
2966
CH stretch
CH3NCO
2263
NCO stretch
CO2
2349
antisymmetric stretch
HNO
2756
HN stretch
HCNO
2193
CNO stretch
NCS
1942
CN stretch
NO
1904
NO stretch
Taken from ref 12.
Figure 3. Emission following MITC (solid line) and MTC (dotted line) photolysis, with the vibrational bands for CN A(2Π)X(2Σ) assigned based on ref 16. The resolution of the infrared spectra is 64 cm1, and the spectra were measured 3 μs after photolysis.
CH3CN, more signal from CN(A) does not imply that CN is the main photoproduct. Small amounts of vibrationally excited NCS and CH3NC are also observed, as is shown more clearly in Figure 2B, where the signal is renormalized. For MITC, the major source of emission is still electronic emission from CN(A), but the emission from vibrationally excited NCS and CH3NC has a much stronger signal. This observation is also in agreement with the work of Northrup and Sears, who found that the ratio of CN to NCS was much higher for thiocyanates (where CN can be produced by simple bond fission) than for isothiocyanates (where CN can be produced only following isomerization).9 The CN radical observed following MITC photolysis also exhibits significantly more vibrational excitation than does that produced by photolysis of MTC.16 Comparison spectra are shown in Figure 3. Whereas MTC photolysis primarily produces CN(A) with zero or one quantum of vibrational excitation, MITC photolysis produces significant amounts of CN(A) with up to four quanta of vibrational excitation. This result cannot be explained simply by available energy, because, as Figure 1 shows, the products of MTC photolysis have more available energy than those produced by MITC photolysis. It is likely that the greater vibrational excitation observed in CN(A) following MITC photolysis is produced by the excited-state isomerization, suggesting that reversing the NCS moeity results in significant vibrational excitation in the CN(A) radical. These observations are not consistent with Hall and Wu’s10 theory of CN production to photolysis of an MTC impurity in MITC samples. If the CN that we detected following MITC photolysis were due to an MTC impurity in our sample, we would expect that the CN(A) would have the same vibrational distribution as that observed from a pure MTC sample.
Figure 4. Emission following MITC photolysis at 248 nm: (A) 20 sccm of 2.5% MITC in He, (B) 20 sccm of 2.5% MITC in He and 40 sccm of pure oxygen gas. All spectra were collected with 8 cm1 resolution.
We also observed infrared emission following MITC photolysis at 248 nm. At this wavelength, as shown in Figure 1, the CN(A) channel is no longer energetically available through a one-photon process, and we no longer observed that emission. We did still observe the other two energetically accessible channels, as shown in Figure 4A. However, we could not compare MITC and MTC photolysis, because MTC, which has a much smaller absorption cross section than MITC (7.6 1020 versus 3.6 1018 cm2),4 did not produce measurable emission at 248 nm. B. Photolysis of MITC with Atmospheric Oxidizers at 248 nm. MITC was photolyzed at 248 nm with three atmospheric oxidizers: O2, NO, and NO2. Although 248 nm is not an actinic wavelength, excitation at 248 nm excites the same electronic transition as actinic wavelengths do, so we would expect the observed reaction products to be comparable to those observed environmentally.4 Moreover, 248 nm is near an absorbance minimum for NO2,17 so that when MITC is photolyzed, one should primarily observe reactions of NO2 with MITC photolysis products with minimal interference from reactions caused by O(3P) atoms arising from NO2 photolysis. However, at 248 nm, the CH3 þ NCS channel is energetically available (see Figure 1), which is not the case at actinic wavelengths. Figure 4 compares the photolysis of MITC with and without molecular oxygen present. The flow rate of oxygen was allowed to vary from 0 to 60 sccm, but the flow rate did not change the observed emission. Two energetically available channels can be observed CH3 NCS f CH3 þ NCS CH3 NCS f CH3 NC þ S The presence of oxygen gas does not change the observed emission dramatically. No CH3NCO or other reactive products were observed. This result is consistent with the steady-state infrared absorbance measurements of Alvarez and Moore.4 We did observe a small peak appearing at long times at ∼2100 cm1, which we tentatively identified as HSO2, although the band remained very small and broad, making identification difficult. This result might suggest the presence of SO2, due to the reaction of oxygen with the liberated sulfur atom, but all SO2 vibrational bands are below 1850 cm1 and therefore could not be detected in this experiment. Figure 5 shows the emission observed when MITC was photolyzed in the presence of NO (Figure 5A) or NO2 (Figure 5B). The band origins of observed species are summarized in Table 1. 5321
dx.doi.org/10.1021/jp2000305 |J. Phys. Chem. A 2011, 115, 5319–5323
The Journal of Physical Chemistry A
ARTICLE
CH3NC cannot be produced directly from MTC unless there is isomerization prior to dissociation. At the energies available in this experiment (see Figure 1), CN cannot be produced directly from MITC unless there is prior isomerization. These observations supports Northrup and Sears' proposal that MITC and MTC share an electronically excited state and that they can isomerize through that excited state.9 MITC was also photolyzed at 248 nm both alone and in the presence of O2, NO, or NO2. When photolyzed alone or with oxygen gas, two energetically available channels were observed: CH3 NCS f CH3 þ NCS Figure 5. Emission following MITC photolysis at 248 nm: (A) 20 sccm of 2.5% MITC in He and 60 sccm of NO, (B) 20 sccm of 2.5% MITC in He and 30 sccm of NO2. All spectra were collected with 8 cm1 resolution. When MITC was photolyzed in the presence of NO2, CH3NCO was observed as a product. The band origins of observed species are summarized in Table 1.
When NO and NO2 were present, there was clear evidence of post-photolysis chemical reactions, and increasing the flow rate of NO or NO2 increased the emission from these secondary products. NO appears to react with the methyl group of MITC (or possibly its photoproducts), because the hydrogen-containing products HNO and HCNO were detected. Whereas NO might be important for environmental removal of MITC, it does not appear to lead to the formation of CH3NCO. NO2, on the other hand, appears to release an oxygen atom and does form CH3NCO, as shown in Figure 5B. We also observed a significant NO peak, suggesting that the reaction is CH3 NC þ NO2 f CH3 NCO þ NO To verify that the observed chemistry was MITC photochemistry, rather than NO2 photochemistry, we also photolyzed NO2 alone at 248 nm. Although we did observe a very small amount of NO forming promptly and then disappearing, the NO signal was greatly enhanced when MITC was present. The NO signal also had a very different time profile, forming at much later times and lasting longer when MITC was present, as expected if NO is produced in a chemical reaction rather than by photolysis. This is the first observation of a potential pathway to form CH3NCO from MITC, although more work at environmentally realistic wavelengths and conditions should be done to confirm this possibility. If reaction of MITC photolysis products with NO2 is the main source of CH3NCO arising from use of metam sodium, this reaction should be most important near polluted urban environments in the mornings. Under these conditions, the photolytic conversion of NO2 to NO þ O is not yet complete, but photons are available to drive the photochemistry of MITC.
’ CONCLUSIONS MITC and MTC were photolyzed at 193 nm, and the resulting infrared emissions were observed. Products consistent with three reaction channels were observed for both starting materials: CH3 NCS or CH3 SCN f CH3 þ NCSðXÞ CH3 NCS or CH3 SCN f CH3 S þ CNðAÞ CH3 NCS or CH3 SCN f CH3 NC þ S
CH3 NCS f CH3 NC þ S No apparent reaction was observed with oxygen. When MITC was photolyzed with NO, secondary chemical reactions were observed that suggest that NO reacts with the methyl group of MITC or its photoproducts. When MITC was photolyzed with NO2, CH3NCO was observed. CH3NCO is a highly toxic chemical, and its production under environmental conditions would be of great concern. CH3NCO has been observed in smog chambers containing MITC.5 We suggest that O-atom abstraction from NO2 by the CH3NC photoproduct is a valid mechanism for formation of CH3NCO that might play a role in the environment.
’ AUTHOR INFORMATION Corresponding Author
*E-mail:
[email protected] (E.A.W.),
[email protected] (D.L.O.).
’ ACKNOWLEDGMENT This research was supported by the Division of Chemical Sciences, Geosciences, and Biosciences, the Office of Basic Energy Sciences, of the U.S. Department of Energy. Sandia is a multiprogram laboratory operated by Sandia Corporation, a Lockheed Martin Company, for the United States Department of Energy’s National Nuclear Security Administration under Contract DE-AC04-94AL85000. We also thank Howard Johnsen for technical assistance. J.L.P. thanks the Faculty Development Program at Mills College for support. ’ REFERENCES (1) Sullivan, D. A.; Holdsworth, M. T.; Hlinka, D. J. Atmos. Environ. 2004, 38, 2457. (2) Leistra, M.; Crum, S. J. H. Water, Air, Soil Pollut. 1990, 50, 109. (3) Woodrow, J. E.; Seiber, J. N.; LeNoir, J. S.; Krieger, R. I. J. Agric. Food Chem. 2008, 56, 7373. (4) Alvarez, R. A.; Moore, C. B. Science 1994, 263, 205. (5) Geddes, J. D.; Miller, G. C.; Taylor, G. E. Environ. Sci. Technol. 1995, 29, 2590. (6) Mehta, P. S.; Mehta, A. S.; Mehta, S. J.; Makhijani, A. B. J. Am. Med. Assoc. 1990, 264, 2781. (7) D’Amario, P.; Di Stefano, G.; Lenzi, M.; Mele, A. J. Chem. Soc., Faraday Trans. 1972, 68, 940. (8) Tokue, I; Kobayashi, K.; Honda, T.; Ito, Y. J. Phys. Chem. 1990, 94, 3485. (9) Northrup, F. J.; Sears, T. J. J. Chem. Phys. 1990, 93, 2337. (10) Hall, G. E.; Wu, M. J. Phys. Chem. 1993, 97, 10911. (11) Clegg, S. M.; Parsons, B. F.; Klippenstein, S. J.; Osborn, D. L. J. Chem. Phys. 2003, 119, 7222. 5322
dx.doi.org/10.1021/jp2000305 |J. Phys. Chem. A 2011, 115, 5319–5323
The Journal of Physical Chemistry A
ARTICLE
(12) Jacox, M. E. Vibrational and Electronic Energy Levels of Polyatomic Transient Molecules. In NIST Chemistry WebBook, NIST Standard Reference Database Number 69, Linstrom, P.J., Mallard, W.G., Eds.; National Institute of Standards and Technology: Gaithersburg, MD, 2009; http://webbook.nist.gov (accessed Feb 15, 2010). (13) Bradforth, S. E.; Kim, E. H.; Arnold, D. W.; Neumark, D. M. J. Chem. Phys. 1993, 98, 800. (14) Zhang, J. Photodissociation of free radicals. In Modern Trends in Chemical Reaction Dynamics; Yand, X., Liu, K., Eds.; World Scientific Publishing: Singapore, 2004; p 475. (15) Ruscic, B.; Berkowitz, J. J. Chem. Phys. 1994, 101, 7975. (16) Weinberg, J. M.; Fishburne, E. S.; Rao, K. N. J. Mol. Spectrosc. 1967, 22, 406. (17) Johnston, H. S.; Graham, R. Can. J. Chem. 1974, 52, 1415.
5323
dx.doi.org/10.1021/jp2000305 |J. Phys. Chem. A 2011, 115, 5319–5323
