- Email: [email protected]
The observation of hot bands in the 288 nm system of the FeCl2 radical
Journal of Molecular Spectroscopy 240 (2006) 265–268 www.elsevier.com/locate/jms
Note
The observation of hot bands in the 288 nm system of the FeCl2 radical Philip J. Hodges a, John M. Brown a
a,*
, Stephen H. Ashworth
b
Physical and Theoretical Chemistry Laboratory, Department of Chemistry, University of Oxford, South Parks Road, Oxford OX1 3QZ, UK b The School of Chemical Sciences and Pharmacy, University of East Anglia, Norwich NR4 7TJ, UK Received 8 September 2006 Available online 18 October 2006
Abstract The 288 nm band system of FeCl2 has been recorded with a sample produced in a warmed, free-jet expansion at moderately high resolution (with a linewidth of 0.28 cm1). Under these conditions, several hot bands are observed involving excitation of the symmetric and antisymmetric stretching vibrations. The wavenumbers determined as a result for FeCl2 in its ground 5Dg,4 state are m001 ¼ 352:34ð12Þ cm1 and m003 ¼ 504:8ð2Þ cm1 . No hot, sequence bands in the bending vibration m002 were observed. The most likely explanation is that the wavenumber for m2 is essentially the same in both the electronic states involved (88 cm1). Additional strong hot bands are observed that are unrelated to the previously assigned electronic transitions; they appear to emanate from a low-lying electronic state of FeCl2. Ó 2006 Elsevier Inc. All rights reserved. Keywords: Electronic spectroscopy; Transition metal compounds; Iron dichloride; Hot bands; Vibrational intervals
The electronic transition of FeCl2 at 288 nm has been assigned as X = 5–4, emanating from the lowest spin component of the 5Dg ground state [1]. In two recent publications, we reported vibrational parameters for FeCl2 in its ground electronic state [2] and in the excited state [1]. The molecule was formed in a sample at low vibrational temperatures (<100 K) in a free-jet expansion, and no hot bands were observed in the excitation spectrum [1]. The least-squares fits of the ground-state vibrational bands observed in dispersed emission measurements [2] were supplemented by measurements of some hot band features in an early study of the spectrum recorded at higher vibrational temperature by Ashworth and Brown [3]. Since the calibration of this spectrum (by means of the optogalvanic effect) was slightly less reliable than expected, we have made a new recording of the spectrum and obtained more accurate vibrational parameters. The experimental setup is the same as that described in the previous publications [1,2], except that we have reintroduced the ‘‘light shield’’ [4] that was designed to prevent blackbody radiation from the heated nozzle swamping *
Corresponding author. Fax: +44 1865 275410. E-mail address: [email protected] (J.M. Brown).
0022-2852/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.jms.2006.10.006
the photomultiplier tube. The light shield consists of a stainless steel disc mounted beyond the alumina nozzle. It contains a small hole of diameter 4 mm, through which the molecular beam emerges. In previous experiments with the light shield [4,5], the sample was formed at an increased rotational temperature, possibly due to backscatter from the molecular beam striking the edge of the small exit hole. In the present experiments, effective warming was observed when a 25 mm spacer was inserted before the light shield, such that the 4 mm hole in the shield was 40 mm from the nozzle exit hole. A portion of the warm vibrationally resolved excitation spectrum of FeCl2 is shown in Fig. 1, and at higher resolution (with a linewidth of 0.28 cm1) in Fig. 2. The spectrum was calibrated against the accurately determined positions of the R-branch heads of bands recorded at rotational resolution [1]. As usual, the dye laser scanning system was found to be highly linear in wavelength. The vacuum wavenumbers of the origins of assigned vibronic bands observed in excitation [1] and dispersed emission [2], and the hot bands assigned from the present experiments (102 , 301 , 101 , 110 301 , 111 , 102 301 , 121 , 101 310 , 131 ) are given in Table 1. For vibronic bands that have not been recorded at rotational resolution, the position of the band origin has
266
P.J. Hodges et al. / Journal of Molecular Spectroscopy 240 (2006) 265–268
120 110 †† 110
v00 110 310 110 310
*
35100
34900
*
34700
310 34500
34300
34100
Wavenumber / cm-1
Fig. 1. A portion of the vibrationally resolved spectrum of the FeCl2 band system at 288 nm, recorded under ‘‘warm’’ conditions. The bands marked with asterisks have been tentatively assigned to transitions between the second lowest spin components in both electronic states (X = 4–3). The bands marked with daggers are thought to correspond to a strong, separate transition from a low-lying electronic state of FeCl2.
110
110 310
102 310 122
34450
34400
34350
34300
Wavenumber / cm-1
Fig. 2. A portion of the ‘‘warm’’ vibrationally resolved spectrum of the FeCl2 band system at 288 nm, recorded under ‘‘warm’’ conditions with a laser linewidth of 0.28 cm1. The chlorine isotope splittings are clearly visible in these bands.
been calculated from the measured position of the Rbranch head in the vibrationally resolved spectrum, using rotational constants predicted from Ref. [1]. The wavenumbers for the bands observed in dispersed emission are quoted to the nearest integer and are estimated to be accurate to ±10 cm1. The table also includes residuals from a least-squares fit to vibrational parameters for the totally symmetric stretching vibration, m1. The results of the fit are given in Table 2. Table 3 lists known vibrational wavenumbers for a number of transition metal dichlorides. This table first appeared in our recent publication concerning the CoCl2 radical [6]. It has been amended to include the results from the present work. The observation of two progressions in the symmetric stretching vibration in the excitation spectrum has been explained [1] in terms of allowed transitions to an electronic state of gerade symmetry, and vibronically induced tran-
sitions (through the antisymmetric stretching vibration) to a nearby state of ungerade symmetry. The resultant value of excited-state m3 was tentatively declared to be 429.5 cm1. The assignment was said to be speculative because the differences between the values of m1 and m3 in the ground and excited states appeared to be inconsistent with a valence force field model. In the warm spectrum reported here, a progression in excited state m1 based on the 301 hot band is observed; the value determined for m3 of 504.8(2) cm1 is in good agreement with that determined by dispersed fluorescence spectroscopy [2] but considerably more accurate and precise. The observation of this complementary progression built on one quantum of m3 in the ground state reinforces our assignments of the vibronically induced progression referred to as Progression B in Ref. [1]. Additional support comes from the observation of 101 310 and 102 310 hot bands.
P.J. Hodges et al. / Journal of Molecular Spectroscopy 240 (2006) 265–268 Table 1 Vacuum wavenumbers for the origins of assigned vibronic bands in the 288 nm system of 56Fe35Cl2 v01
v03
v001
v003
Wavenumber/cm1
0 0 0 0 0 0 0 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 3 3 4
0 0 0 0 1 1 1 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0
0 1 2 0 0 1 2 0 1 2 0 0 0 0 0 0 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 0 0 1 0
0 0 0 1 0 0 1 0 0 0 1 0 0 2 4 6 8 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
34696.69 34344.96 33993.48 34191.89c 35126.21c 34774.49c 34424.12c 34897.20 34545.57 34194.04 34393.35c 35321.84b 35092.37 34050 33043 32023 31007 34590.48c 34740.68c 34367 34050 33704 33352 33029 32665 32337 32013 31658 31331 31007 30678 30346 30010 29682 29361 29042 28720 28398 28075 27761 27443 27161 26815 35514.53c 35282.75b 34931.29c 35468.58b
o ca 0.1b 0.5c 0.4c
0.0b 0.7c 0.2c
Table 2 Vibrational parameters for Vibrational Parameter
0.1b
The results of the fit are quoted in Table 2. a Observed calculated wavenumber (cm1). b These wavenumbers have been determined from analysis of rotationally resolved spectra. Their Doppler-limited accuracy is ±0.1 cm1. c These wavenumbers have been determined from the laser excitation spectrum to an accuracy of 0.2 cm1. They are given correspondingly greater weight in the fit. d Transition given zero weight in the fit (e.g. overlapping lines).
Even in the warm excitation spectrum, no bands have been observed that involve excitation of the m2 bending mode; this is particularly surprising since m002 88 cm1 [7]. It strongly suggests that the bending wavenumber is
Fe35Cl2 288 nm excited state
a
354.16(12) 0.909(23) 504.8(2)
205.34(18) 2.490(55) 429.5(2)
All values are in cm1. a The figures in parentheses give 1 standard deviation of the leastsquares fit, in units of the last quoted decimal place.
Table 3 Selected vibrational wavenumbers for the first row transition metal dichlorides Molecule
Band system (nm)
m001
m01
56
288 327a 460b 650c
352.34(12) 358.1(17) 360.24(10) 369.365(2)
200.36(18) 504.8(2) 429.5(2) 195.7(12) 356.380(1) 520.4(10)d 530.9546(4) 335.881(9) 522.185(4) 482.2(10)
Fe35Cl2 Co35Cl2 58 Ni35Cl2 63 Cu35Cl2 59
0.2b
56
Ground state
x1 x11 m3
0.1b
0.4 22 8d 10 2 22 1 12 25d 6 14 22 24 20 12 8 10 13 11 6 0 0 5 23 14d
267
m003
m03
All values are in cm1. a Ref. [6]. b Refs. [5,9]. c Refs. [10–12]. d One half of the 2m3 interval.
almost identical in both electronic states. In this respect, this system of FeCl2 differs markedly from the blue system of NiCl2 at 460 nm [5,8] in which many sequence bands in the bending vibration were observed. Again, this represents a considerable deviation from the expectations of a simple valence force field model. All the bands identified so far in this spectrum of FeCl2 arise from the lowest spin component (X = 4) of the ground 5Dg state. Two new bands have been observed in the warm spectrum that are tentatively assigned to transitions between the second lowest spin components in both states (X = 4–3); they are indicated in Fig. 1 with asterisks. The m00 band of this system is observed at 34626.43 cm1, and the 110 band at 34832.61 cm1. Although nothing is revealed about the absolute magnitude of the spin-orbit splitting in either electronic state, this assignment suggests that the splitting is about 70 cm1 larger in the ground state. The value of m01 is 206.2 cm1, 5.6 cm1 larger than in the lowest, X = 5, component. In addition, two new, intense features are present in the warm spectrum that are barely detectable under cold conditions; they are even more intense than the m00 band (see Fig. 1 where they are marked with daggers). The wavenumbers of the strongest R-branch heads in these two bands are 34970.49 and 34986.69 cm1. At higher wavenumber, there is another pair of new bands at 35161.32 and 35181.11 cm1 that are separated by approximately m1 from the first pair. None of these bands involve excitation from the zero-point level of the ground state. A possible explanation is that they correspond to a strong, separate electronic transition from a low-lying electronic state of FeCl2 that is populated under the warmer conditions.
268
P.J. Hodges et al. / Journal of Molecular Spectroscopy 240 (2006) 265–268
Acknowledgment We thank the Engineering and Physical Sciences Research Council for support of P.J.H. References [1] P.J. Hodges, J.M. Brown, S.H. Ashworth, J. Mol. Spectrosc. 237 (2006) 205–217. [2] P.J. Hodges, S.H. Ashworth, I.R. Beattie, J.M. Brown, Chem. Phys. Lett. 422 (2006) 160–164. [3] S.H. Ashworth, J.M. Brown, unpublished work, 1990. [4] S.H. Ashworth, D. Phil. Thesis, University of Oxford, 1991.
[5] F.J. Grieman, S.H. Ashworth, J.M. Brown, I.R. Beattie, J. Chem. Phys. 92 (1990) 6365–6375. [6] P.J. Hodges, J.M. Brown, T.D. Varberg, J. Chem. Phys. 124 (2006) 204302, 1–10. [7] K.R. Thompson, K.D. Carlson, J. Chem. Phys. 49 (1968) 4379–4384. [8] S.H. Ashworth, F.J. Grieman, J.M. Brown, Chem. Phys. Lett. 175 (1990) 660–666. [9] S.H. Ashworth, F.J. Grieman, J.M. Brown, J. Chem. Phys. 104 (1996) 48–63. [10] I.R. Beattie, J.M. Brown, P. Crozet, A.J. Ross, A. Yiannopoulou, Inorg. Chem. 36 (1997) 3207–3208. [11] A.J. Ross, private communication. [12] M.P. Barnes, D. Phil. Thesis, University of Oxford, 1995.
Note
The observation of hot bands in the 288 nm system of the FeCl2 radical Philip J. Hodges a, John M. Brown a
a,*
, Stephen H. Ashworth
b
Physical and Theoretical Chemistry Laboratory, Department of Chemistry, University of Oxford, South Parks Road, Oxford OX1 3QZ, UK b The School of Chemical Sciences and Pharmacy, University of East Anglia, Norwich NR4 7TJ, UK Received 8 September 2006 Available online 18 October 2006
Abstract The 288 nm band system of FeCl2 has been recorded with a sample produced in a warmed, free-jet expansion at moderately high resolution (with a linewidth of 0.28 cm1). Under these conditions, several hot bands are observed involving excitation of the symmetric and antisymmetric stretching vibrations. The wavenumbers determined as a result for FeCl2 in its ground 5Dg,4 state are m001 ¼ 352:34ð12Þ cm1 and m003 ¼ 504:8ð2Þ cm1 . No hot, sequence bands in the bending vibration m002 were observed. The most likely explanation is that the wavenumber for m2 is essentially the same in both the electronic states involved (88 cm1). Additional strong hot bands are observed that are unrelated to the previously assigned electronic transitions; they appear to emanate from a low-lying electronic state of FeCl2. Ó 2006 Elsevier Inc. All rights reserved. Keywords: Electronic spectroscopy; Transition metal compounds; Iron dichloride; Hot bands; Vibrational intervals
The electronic transition of FeCl2 at 288 nm has been assigned as X = 5–4, emanating from the lowest spin component of the 5Dg ground state [1]. In two recent publications, we reported vibrational parameters for FeCl2 in its ground electronic state [2] and in the excited state [1]. The molecule was formed in a sample at low vibrational temperatures (<100 K) in a free-jet expansion, and no hot bands were observed in the excitation spectrum [1]. The least-squares fits of the ground-state vibrational bands observed in dispersed emission measurements [2] were supplemented by measurements of some hot band features in an early study of the spectrum recorded at higher vibrational temperature by Ashworth and Brown [3]. Since the calibration of this spectrum (by means of the optogalvanic effect) was slightly less reliable than expected, we have made a new recording of the spectrum and obtained more accurate vibrational parameters. The experimental setup is the same as that described in the previous publications [1,2], except that we have reintroduced the ‘‘light shield’’ [4] that was designed to prevent blackbody radiation from the heated nozzle swamping *
Corresponding author. Fax: +44 1865 275410. E-mail address: [email protected] (J.M. Brown).
0022-2852/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.jms.2006.10.006
the photomultiplier tube. The light shield consists of a stainless steel disc mounted beyond the alumina nozzle. It contains a small hole of diameter 4 mm, through which the molecular beam emerges. In previous experiments with the light shield [4,5], the sample was formed at an increased rotational temperature, possibly due to backscatter from the molecular beam striking the edge of the small exit hole. In the present experiments, effective warming was observed when a 25 mm spacer was inserted before the light shield, such that the 4 mm hole in the shield was 40 mm from the nozzle exit hole. A portion of the warm vibrationally resolved excitation spectrum of FeCl2 is shown in Fig. 1, and at higher resolution (with a linewidth of 0.28 cm1) in Fig. 2. The spectrum was calibrated against the accurately determined positions of the R-branch heads of bands recorded at rotational resolution [1]. As usual, the dye laser scanning system was found to be highly linear in wavelength. The vacuum wavenumbers of the origins of assigned vibronic bands observed in excitation [1] and dispersed emission [2], and the hot bands assigned from the present experiments (102 , 301 , 101 , 110 301 , 111 , 102 301 , 121 , 101 310 , 131 ) are given in Table 1. For vibronic bands that have not been recorded at rotational resolution, the position of the band origin has
266
P.J. Hodges et al. / Journal of Molecular Spectroscopy 240 (2006) 265–268
120 110 †† 110
v00 110 310 110 310
*
35100
34900
*
34700
310 34500
34300
34100
Wavenumber / cm-1
Fig. 1. A portion of the vibrationally resolved spectrum of the FeCl2 band system at 288 nm, recorded under ‘‘warm’’ conditions. The bands marked with asterisks have been tentatively assigned to transitions between the second lowest spin components in both electronic states (X = 4–3). The bands marked with daggers are thought to correspond to a strong, separate transition from a low-lying electronic state of FeCl2.
110
110 310
102 310 122
34450
34400
34350
34300
Wavenumber / cm-1
Fig. 2. A portion of the ‘‘warm’’ vibrationally resolved spectrum of the FeCl2 band system at 288 nm, recorded under ‘‘warm’’ conditions with a laser linewidth of 0.28 cm1. The chlorine isotope splittings are clearly visible in these bands.
been calculated from the measured position of the Rbranch head in the vibrationally resolved spectrum, using rotational constants predicted from Ref. [1]. The wavenumbers for the bands observed in dispersed emission are quoted to the nearest integer and are estimated to be accurate to ±10 cm1. The table also includes residuals from a least-squares fit to vibrational parameters for the totally symmetric stretching vibration, m1. The results of the fit are given in Table 2. Table 3 lists known vibrational wavenumbers for a number of transition metal dichlorides. This table first appeared in our recent publication concerning the CoCl2 radical [6]. It has been amended to include the results from the present work. The observation of two progressions in the symmetric stretching vibration in the excitation spectrum has been explained [1] in terms of allowed transitions to an electronic state of gerade symmetry, and vibronically induced tran-
sitions (through the antisymmetric stretching vibration) to a nearby state of ungerade symmetry. The resultant value of excited-state m3 was tentatively declared to be 429.5 cm1. The assignment was said to be speculative because the differences between the values of m1 and m3 in the ground and excited states appeared to be inconsistent with a valence force field model. In the warm spectrum reported here, a progression in excited state m1 based on the 301 hot band is observed; the value determined for m3 of 504.8(2) cm1 is in good agreement with that determined by dispersed fluorescence spectroscopy [2] but considerably more accurate and precise. The observation of this complementary progression built on one quantum of m3 in the ground state reinforces our assignments of the vibronically induced progression referred to as Progression B in Ref. [1]. Additional support comes from the observation of 101 310 and 102 310 hot bands.
P.J. Hodges et al. / Journal of Molecular Spectroscopy 240 (2006) 265–268 Table 1 Vacuum wavenumbers for the origins of assigned vibronic bands in the 288 nm system of 56Fe35Cl2 v01
v03
v001
v003
Wavenumber/cm1
0 0 0 0 0 0 0 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 3 3 4
0 0 0 0 1 1 1 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0
0 1 2 0 0 1 2 0 1 2 0 0 0 0 0 0 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 0 0 1 0
0 0 0 1 0 0 1 0 0 0 1 0 0 2 4 6 8 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
34696.69 34344.96 33993.48 34191.89c 35126.21c 34774.49c 34424.12c 34897.20 34545.57 34194.04 34393.35c 35321.84b 35092.37 34050 33043 32023 31007 34590.48c 34740.68c 34367 34050 33704 33352 33029 32665 32337 32013 31658 31331 31007 30678 30346 30010 29682 29361 29042 28720 28398 28075 27761 27443 27161 26815 35514.53c 35282.75b 34931.29c 35468.58b
o ca 0.1b 0.5c 0.4c
0.0b 0.7c 0.2c
Table 2 Vibrational parameters for Vibrational Parameter
0.1b
The results of the fit are quoted in Table 2. a Observed calculated wavenumber (cm1). b These wavenumbers have been determined from analysis of rotationally resolved spectra. Their Doppler-limited accuracy is ±0.1 cm1. c These wavenumbers have been determined from the laser excitation spectrum to an accuracy of 0.2 cm1. They are given correspondingly greater weight in the fit. d Transition given zero weight in the fit (e.g. overlapping lines).
Even in the warm excitation spectrum, no bands have been observed that involve excitation of the m2 bending mode; this is particularly surprising since m002 88 cm1 [7]. It strongly suggests that the bending wavenumber is
Fe35Cl2 288 nm excited state
a
354.16(12) 0.909(23) 504.8(2)
205.34(18) 2.490(55) 429.5(2)
All values are in cm1. a The figures in parentheses give 1 standard deviation of the leastsquares fit, in units of the last quoted decimal place.
Table 3 Selected vibrational wavenumbers for the first row transition metal dichlorides Molecule
Band system (nm)
m001
m01
56
288 327a 460b 650c
352.34(12) 358.1(17) 360.24(10) 369.365(2)
200.36(18) 504.8(2) 429.5(2) 195.7(12) 356.380(1) 520.4(10)d 530.9546(4) 335.881(9) 522.185(4) 482.2(10)
Fe35Cl2 Co35Cl2 58 Ni35Cl2 63 Cu35Cl2 59
0.2b
56
Ground state
x1 x11 m3
0.1b
0.4 22 8d 10 2 22 1 12 25d 6 14 22 24 20 12 8 10 13 11 6 0 0 5 23 14d
267
m003
m03
All values are in cm1. a Ref. [6]. b Refs. [5,9]. c Refs. [10–12]. d One half of the 2m3 interval.
almost identical in both electronic states. In this respect, this system of FeCl2 differs markedly from the blue system of NiCl2 at 460 nm [5,8] in which many sequence bands in the bending vibration were observed. Again, this represents a considerable deviation from the expectations of a simple valence force field model. All the bands identified so far in this spectrum of FeCl2 arise from the lowest spin component (X = 4) of the ground 5Dg state. Two new bands have been observed in the warm spectrum that are tentatively assigned to transitions between the second lowest spin components in both states (X = 4–3); they are indicated in Fig. 1 with asterisks. The m00 band of this system is observed at 34626.43 cm1, and the 110 band at 34832.61 cm1. Although nothing is revealed about the absolute magnitude of the spin-orbit splitting in either electronic state, this assignment suggests that the splitting is about 70 cm1 larger in the ground state. The value of m01 is 206.2 cm1, 5.6 cm1 larger than in the lowest, X = 5, component. In addition, two new, intense features are present in the warm spectrum that are barely detectable under cold conditions; they are even more intense than the m00 band (see Fig. 1 where they are marked with daggers). The wavenumbers of the strongest R-branch heads in these two bands are 34970.49 and 34986.69 cm1. At higher wavenumber, there is another pair of new bands at 35161.32 and 35181.11 cm1 that are separated by approximately m1 from the first pair. None of these bands involve excitation from the zero-point level of the ground state. A possible explanation is that they correspond to a strong, separate electronic transition from a low-lying electronic state of FeCl2 that is populated under the warmer conditions.
268
P.J. Hodges et al. / Journal of Molecular Spectroscopy 240 (2006) 265–268
Acknowledgment We thank the Engineering and Physical Sciences Research Council for support of P.J.H. References [1] P.J. Hodges, J.M. Brown, S.H. Ashworth, J. Mol. Spectrosc. 237 (2006) 205–217. [2] P.J. Hodges, S.H. Ashworth, I.R. Beattie, J.M. Brown, Chem. Phys. Lett. 422 (2006) 160–164. [3] S.H. Ashworth, J.M. Brown, unpublished work, 1990. [4] S.H. Ashworth, D. Phil. Thesis, University of Oxford, 1991.
[5] F.J. Grieman, S.H. Ashworth, J.M. Brown, I.R. Beattie, J. Chem. Phys. 92 (1990) 6365–6375. [6] P.J. Hodges, J.M. Brown, T.D. Varberg, J. Chem. Phys. 124 (2006) 204302, 1–10. [7] K.R. Thompson, K.D. Carlson, J. Chem. Phys. 49 (1968) 4379–4384. [8] S.H. Ashworth, F.J. Grieman, J.M. Brown, Chem. Phys. Lett. 175 (1990) 660–666. [9] S.H. Ashworth, F.J. Grieman, J.M. Brown, J. Chem. Phys. 104 (1996) 48–63. [10] I.R. Beattie, J.M. Brown, P. Crozet, A.J. Ross, A. Yiannopoulou, Inorg. Chem. 36 (1997) 3207–3208. [11] A.J. Ross, private communication. [12] M.P. Barnes, D. Phil. Thesis, University of Oxford, 1995.