Rotational transitions in excited vibrational - cchem.berkeley.edu
Journal of Molecular Spectroscopy 208, 219–223 (2001) doi:10.1006/jmsp.2001.8378, available online at http://www.idealibrary.com on
Rotational Transitions in Excited Vibrational States of D2 O Ernest A. Michael,1 Christy J. Keoshian, Serena K. Anderson, and Richard J. Saykally2 Department of Chemistry, University of California, Berkeley, California 94720-1460 Received March 15, 2001; in revised form May 8, 2001; published online July 31, 2001
Rotational transitions of D2 O with J < 12 in excited vibrational states (and the ground state) were measured with submegahertz precision in a pulsed discharge supersonic slit expansion of a D2 O/NeHe mixture while scanning the range from 1473.69 to 1685.91 GHz using the Berkeley terahertz laser sideband spectrometer. Assignments were made using all available previous IR results, which include the (000), (010), (020), (001), (100), (111), (130), (031), (012), (210), and (121) vibrational states with C 2001 Academic Press levels up to 7500 cm−1 . ° Key Words: terahertz; submillimeter; hot D2 O spectrum; rotational transitions; excited vibration.
Terahertz spectra of a supersonic expansion of a pulsed slit discharge plasma containing first-run neon–helium mix bubbled through D2 O were measured. The source is described elsewhere (6) for similar conditions, except that this scanning was performed with half the current. Over 400 discharge-generated lines, mostly unassigned, were observed in the continuously scanned range from 1473.69 to 1685.91 GHz with signal-tonoise ratios between 3 and 3200. Assignments were made with the assumption that the spectra originated from rotational transitions among vibrationally excited states of D2 O: The energy level data available from previous IR results, containing the (000), (010), (020), (001), (100) (1, 4) and the (130), (031), (012), (210), (111), (121) states (2, 3), were used to compute b-type pure rotational transitions as well as b- and a-type rotational–vibrational transitions in the FIR. The accuracy of these calculated energy levels was reported to be ca. 0.001 cm−1 (2, 3) and, depending on the individual level, in the range from 0.0003 to 0.02 cm−1 (4). One hundred twenty-eight pure rotational transitions with the selection rules (1J, 1τ ) = (0, ±2), (±1, 0), or (±1, ±2) (where 1τ := 1K −1 − 1K +1 ) were calculated for the scanned range, of which 81 coincided with observed transitions to within four times the given uncertainty. The rest were not observed or were outside the error of the predictions (see Table 1). The reported assignments for rotational transitions in the ground state and the first bending state show that the accuracy of the FTIR measurements is actually better than reported in (4), so that one has confidence in the assignments of higher excited states. Transitions from the (000) to the (201) vibrational state have been measured earlier in the IR with a resolution of 0.07 cm−1 (5). Rotational transitions within the (201) state determined from 1 Present address: 1. Physics Institute, University of Cologne, Z¨ ulpicher Str. 77, 50937 K¨oln, Germany. E-mail: [email protected]. 2 To whom correspondence should be addressed. E-mail: saykally@uclink4. berkeley.edu.
those data are not precise enough to be compared with our observed dense spectrum; i.e., the list of candidates would be too long. To characterize the physical conditions in the expanding plasma, rotational lines were assigned in the (010), (020), (100), (001), (210), and (111) states to fit a (nonlocal thermal equilibrium) rotational temperature of 120 K (±30%), leaving out tabulated rotational intensities. Furthermore, total populations were estimated for each of the assigned states, from which a vibrational temperature of 2500 K was estimated. The (000) state was not included in this analysis because the intensities are strongly distorted by the thermalized off-pulse background gas. It can be seen that (210) is more populated than (012), and (130) more than (031), showing that symmetric modes are preferentially excited in this discharge expansion. This is consistent with the observation that rovibrational transitions from the (130) to the (031) state occur in absorption and not in emission (see Table 2). The assumption that the relaxation times for the symmetric and the asymmetric stretching modes are similar suggests a mechanism for production of the high vibrational excitation. We propose that high levels, perhaps up to the dissociation energy, are populated “from the top down” in the expansion by dissociative recombination of hydronium ion clusters which are the dominant ions in these supersonic plasmas: D3 O+ (D2 O)n + e− → D2 O∗ + D + (D2 O)n . This mechanism mainly populates the bending and symmetric stretch vibrations due to the difference in the bond angle between D3 O+ and D2 O. On the other hand, no direct IR data are reported in literature for the (110), (030), (011), (200), (120), (030), (101), (040),
219
0022-2852/01 $35.00 C 2001 by Academic Press Copyright ° All rights of reproduction in any form reserved.
220
MICHAEL ET AL.
TABLE 1 List of All Calculated Pure Rotational Transitions for Vibrational States in the Scanned Range
Note. All candidates within a range of four times the uncertainty in the calculated levels are included or marked n.o. (not observed). The FIR laser used for observation is given for reference.
C 2001 by Academic Press Copyright °
ROTATIONAL TRANSITIONS OF D2 O
TABLE 1—Continued
C 2001 by Academic Press Copyright °
221
222
MICHAEL ET AL.
TABLE 1—Continued
C 2001 by Academic Press Copyright °
223
ROTATIONAL TRANSITIONS OF D2 O
TABLE 2 Assigned Rovibrational Transitions
(011), (021), (002), and (050) states. From this large gap in the data an equal number of calculated lines can be anticipated for the scanned frequency range. Therefore, many more of the transitions reported here will probably be assigned to even higher vibrational states in the future.
REFERENCES 1. 2.
3.
ACKNOWLEDGMENT
4. 5.
This work was supported by the Experimental Physical Chemistry Division of the NSF.
6.
R. A. Toth, J. Mol. Spectrosc. 162, 41–54 (1993). A. D. Bykov, O. V. Naumenko, L. N. Sinitsa, B. P. Winnewisser, P. S. Ormsby, and K. N. Rao, J. Mol. Spectrosc. 166, 169–175 (1994). P. S. Ormsby, K. Narahari Rao, M. Winnewisser, B. P. Winnewisser, O. V. Naumenko, A. D. Bykov, and L. N. Sinitsa, J. Mol. Spectrosc. 158, 109–130 (1993). R. A. Toth, J. Mol. Spectrosc. 195, 98–122 (1999). Y. Cohen, I. Bar, and S. Rosenwaks, J. Mol. Spectrosc. 180, 298–304 (1996). E. A. Michael, D. R. Wagner, C. J. Keoshian, S. K. Anderson, and R. J. Saykally, Chem. Phys. Lett. 338, 277–284 (2001).
C 2001 by Academic Press Copyright °
Rotational Transitions in Excited Vibrational States of D2 O Ernest A. Michael,1 Christy J. Keoshian, Serena K. Anderson, and Richard J. Saykally2 Department of Chemistry, University of California, Berkeley, California 94720-1460 Received March 15, 2001; in revised form May 8, 2001; published online July 31, 2001
Rotational transitions of D2 O with J < 12 in excited vibrational states (and the ground state) were measured with submegahertz precision in a pulsed discharge supersonic slit expansion of a D2 O/NeHe mixture while scanning the range from 1473.69 to 1685.91 GHz using the Berkeley terahertz laser sideband spectrometer. Assignments were made using all available previous IR results, which include the (000), (010), (020), (001), (100), (111), (130), (031), (012), (210), and (121) vibrational states with C 2001 Academic Press levels up to 7500 cm−1 . ° Key Words: terahertz; submillimeter; hot D2 O spectrum; rotational transitions; excited vibration.
Terahertz spectra of a supersonic expansion of a pulsed slit discharge plasma containing first-run neon–helium mix bubbled through D2 O were measured. The source is described elsewhere (6) for similar conditions, except that this scanning was performed with half the current. Over 400 discharge-generated lines, mostly unassigned, were observed in the continuously scanned range from 1473.69 to 1685.91 GHz with signal-tonoise ratios between 3 and 3200. Assignments were made with the assumption that the spectra originated from rotational transitions among vibrationally excited states of D2 O: The energy level data available from previous IR results, containing the (000), (010), (020), (001), (100) (1, 4) and the (130), (031), (012), (210), (111), (121) states (2, 3), were used to compute b-type pure rotational transitions as well as b- and a-type rotational–vibrational transitions in the FIR. The accuracy of these calculated energy levels was reported to be ca. 0.001 cm−1 (2, 3) and, depending on the individual level, in the range from 0.0003 to 0.02 cm−1 (4). One hundred twenty-eight pure rotational transitions with the selection rules (1J, 1τ ) = (0, ±2), (±1, 0), or (±1, ±2) (where 1τ := 1K −1 − 1K +1 ) were calculated for the scanned range, of which 81 coincided with observed transitions to within four times the given uncertainty. The rest were not observed or were outside the error of the predictions (see Table 1). The reported assignments for rotational transitions in the ground state and the first bending state show that the accuracy of the FTIR measurements is actually better than reported in (4), so that one has confidence in the assignments of higher excited states. Transitions from the (000) to the (201) vibrational state have been measured earlier in the IR with a resolution of 0.07 cm−1 (5). Rotational transitions within the (201) state determined from 1 Present address: 1. Physics Institute, University of Cologne, Z¨ ulpicher Str. 77, 50937 K¨oln, Germany. E-mail: [email protected]. 2 To whom correspondence should be addressed. E-mail: saykally@uclink4. berkeley.edu.
those data are not precise enough to be compared with our observed dense spectrum; i.e., the list of candidates would be too long. To characterize the physical conditions in the expanding plasma, rotational lines were assigned in the (010), (020), (100), (001), (210), and (111) states to fit a (nonlocal thermal equilibrium) rotational temperature of 120 K (±30%), leaving out tabulated rotational intensities. Furthermore, total populations were estimated for each of the assigned states, from which a vibrational temperature of 2500 K was estimated. The (000) state was not included in this analysis because the intensities are strongly distorted by the thermalized off-pulse background gas. It can be seen that (210) is more populated than (012), and (130) more than (031), showing that symmetric modes are preferentially excited in this discharge expansion. This is consistent with the observation that rovibrational transitions from the (130) to the (031) state occur in absorption and not in emission (see Table 2). The assumption that the relaxation times for the symmetric and the asymmetric stretching modes are similar suggests a mechanism for production of the high vibrational excitation. We propose that high levels, perhaps up to the dissociation energy, are populated “from the top down” in the expansion by dissociative recombination of hydronium ion clusters which are the dominant ions in these supersonic plasmas: D3 O+ (D2 O)n + e− → D2 O∗ + D + (D2 O)n . This mechanism mainly populates the bending and symmetric stretch vibrations due to the difference in the bond angle between D3 O+ and D2 O. On the other hand, no direct IR data are reported in literature for the (110), (030), (011), (200), (120), (030), (101), (040),
219
0022-2852/01 $35.00 C 2001 by Academic Press Copyright ° All rights of reproduction in any form reserved.
220
MICHAEL ET AL.
TABLE 1 List of All Calculated Pure Rotational Transitions for Vibrational States in the Scanned Range
Note. All candidates within a range of four times the uncertainty in the calculated levels are included or marked n.o. (not observed). The FIR laser used for observation is given for reference.
C 2001 by Academic Press Copyright °
ROTATIONAL TRANSITIONS OF D2 O
TABLE 1—Continued
C 2001 by Academic Press Copyright °
221
222
MICHAEL ET AL.
TABLE 1—Continued
C 2001 by Academic Press Copyright °
223
ROTATIONAL TRANSITIONS OF D2 O
TABLE 2 Assigned Rovibrational Transitions
(011), (021), (002), and (050) states. From this large gap in the data an equal number of calculated lines can be anticipated for the scanned frequency range. Therefore, many more of the transitions reported here will probably be assigned to even higher vibrational states in the future.
REFERENCES 1. 2.
3.
ACKNOWLEDGMENT
4. 5.
This work was supported by the Experimental Physical Chemistry Division of the NSF.
6.
R. A. Toth, J. Mol. Spectrosc. 162, 41–54 (1993). A. D. Bykov, O. V. Naumenko, L. N. Sinitsa, B. P. Winnewisser, P. S. Ormsby, and K. N. Rao, J. Mol. Spectrosc. 166, 169–175 (1994). P. S. Ormsby, K. Narahari Rao, M. Winnewisser, B. P. Winnewisser, O. V. Naumenko, A. D. Bykov, and L. N. Sinitsa, J. Mol. Spectrosc. 158, 109–130 (1993). R. A. Toth, J. Mol. Spectrosc. 195, 98–122 (1999). Y. Cohen, I. Bar, and S. Rosenwaks, J. Mol. Spectrosc. 180, 298–304 (1996). E. A. Michael, D. R. Wagner, C. J. Keoshian, S. K. Anderson, and R. J. Saykally, Chem. Phys. Lett. 338, 277–284 (2001).
C 2001 by Academic Press Copyright °
