Laser-ablated carbon plasmas: emission spectroscopy and thin film growth

Laser-ablated carbon plasmas: emission spectroscopy and thin film growth

411[/Af65 ELSEVIER Surface and Coatings Technology 73 (1995) 170 176 Laser-ablated carbon plasmas: emission spectroscopy and thin film growth Rajesh...

603KB Sizes 2 Downloads 64 Views

411[/Af65 ELSEVIER

Surface and Coatings Technology 73 (1995) 170 176

Laser-ablated carbon plasmas: emission spectroscopy and thin film growth Rajesh K. Dwivedi, Raj K.

Thareja *

Department of Physics and Centre for Laser Technology, Indian Institute of Technology, Kanpur 208 016, India Received 25 April 1994; accepted in final form 14 October 1994

Abstract

The effect of the ambient gas pressure on thin films of carbon deposited by laser ablation using third harmonics at 355 nm of an Nd:YAG laser is reported. Scanning electron microscopy, X-ray diffraction and Raman spectroscopy have been used to characterize the deposited films. The nucleation density and morphology of the films and microcrystals are strongly affected by the pressure of the ambient gas. It is found that there is an optimum pressure at which nucleation is more pronounced. The optical emission spectra of C2, recorded as a function of the laser energy and background gas pressures, is used to estimate the vibrational temperature of the species. The vibrational temperature of the molecular C2 species is estimated using Swan bands. The intensity of the Swan bands increased as the pressure of the background gas (argon) was increased. The vibrational temperature depends on the laser energy, and on the choice of the ambient gas and its pressure. The correlation of the vibrational temperature with characteristics of the deposited film and with plasma plume parameters is discussed. Keywords: Plasma; Thin film deposition; Vibrational temperature; C60; Diamond-like carbon

1. I n t r o d u c t i o n

Pulsed laser deposition methods are emerging as among the leading techniques for depositing thin films of varying composition [1]. Recently, because of its several merits - - including the simplicity of implementation and operation, deposition of materials with high melting points, and easy control of film microstructure - - the laser-ablated plasma technique has been utilized in microelectronics fabrications [2], the production of cluster species for basic research [3], for diamond-like carbon films [4-10], for superconducting thin film deposition [ 11,12], and for hydrodynamic studies [ 13 ]. The structural properties of the thin films prepared by the pulsed laser deposition (PLD) technique strongly depend on the laser and processing parameters, especially on the properties of the laser-induced plasma. The plasma controls the growth, structure and properties of the thin film produced. However, while the behavior of the laser-ablated particles in the ablated plume affects the quality of the deposited films, the characteristic behavior of the particles being deposited is not yet fully understood. * Corresponding author. 0257-8972/95/$09.50 © 1995 Elsevier Science S.A. All rights reserved SSDI 0 2 5 7 - 8 9 7 2 ( 9 4 ) 0 2 3 8 0 - 8

Deposition by pulsed laser ablation essentially involves coupling of the laser energy to the target material, removal of material from the target, transfer of target material as vapor and/or plasma to the substrate via the gas phase, and the growth of the thin film on the substrate. At laser fluences above the evaporation threshold, vaporization is found to be a dominant mechanism, while plasma formation plays a dominant role in governing the material transfer from the target to the substrate at higher laser fluences. The optical emission characteristics of the plasma plume can yield important information on the process and understanding of the ablation-etched surface morphology [14]. Koren and Yeh have characterized the emission from the laser-induced polymide, Mylar, polymethyl methacrylate (PMMA) and graphite plasma created using laser radiation at wavelengths of 193, 248 and 351 nm [ 14]. Recently, pulsed laser deposition has been used to fabricate smooth fullerene films with continuously modified structure and properties [15]. The deposition of high purity TiN films on a cold silicon substrate from a multipulse excimer laser-irradiated titanium target in a low pressure ambient N z gas has also been reported [ 16]. It has been observed that the quality and homogeneity

R. I~ Dwivedi, 1~ K. Thareja/Surface and Coatings Technology 73 (1995) 170-176

of the film are affected by the energy of the particles being deposited and, hence, depend on the temperature. Therefore, to optimize the film characteristics, it is necessary to estimate the temperature of the plasma species which dominate in the film. The temperature of the target surface in turn gives information with respect to the ablation mechanism. It has been established that PLD of high T~ superconductors and cubic BN films with short wavelengths (UV) and short pulse widths (nanosecond range) yields superior quality films [ 11 ]. This is probably because of the strong absorption of shorter wavelength laser photons by photo-fragments in the laser-produced plasma. The absorption by the fragments heats the plume to higher temperatures, where further fragmentation into small fragments, including atoms and ions, occurs. As a result of this, fewer and smaller particulates in the film are expected. Thin foil experiments to study the laser wavelength dependence of the mass ablation and heat flux inhibition in laser-produced plasmas have shown that the mass ablation rate increases at short wavelengths [ 17]. In general, the properties of carbon films prepared by pulsed laser ablation range from soft and graphitic to hard and diamond like, depending on the deposition parameters, such as the laser power density and wavelength, as well as the background gas conditions. Several types of laser have been used in the deposition of diamond-like carbon films, including KrF (248 nm) and XeC1 (308 rim) [7-9,18], Nd:YAG (1064 nm) [-6], and CO2 CW lasers [19], with the laser intensities ranging from 108 to 1011 W cm -2. Depending on the intensity of the incident laser radiation, it is possible to obtain higher ionic states or molecular states [20]. In diamond-like carbon films, the variation in diamond-like character results from the different temperatures of the various carbon species in the laser plasma being deposited on the substrate. Optimum films are obtained only around a critical threshold laser intensity of about 108 W cm -2. At this power level, the evaporated carbon contains only neutral species. It is observed that C2 is a dominant constituent of diamond-like carbon; hence, the studies of C2 emission are helpful in optimizing the parameters for diamond-like carbon film deposition and to correlate carbon clusters with the plasma dynamics. Recently, an extensive study on the fragmentation of C6o in a microwave discharge sustained in 1 Torr of argon has shown that C2 is a major species which plays a critical role in the deposition of diamond-like films [21]. Since the report of Kroto et al. [22] on the existence of stable carbon clusters of 60 atoms in the mass spectrum of laser-ablated graphite, there has been renewed interest in laser-ablated carbon plasmas. Recently, there has been a vast amount of research on carbon-based films. Hard carbon films with a unique

171

combination of a diamond-like character have attracted considerable attention among carbon films for a wide variety of applications [23]. Prasad et al. [24] reported the presence of C6o, C7o and C84 fullerences in the carbon soot produced from a laser-produced carbon plasma in vacuum with 1.06 ~tm irradiance. The relative abundance of clusters depends on parameters such as the laser wavelength, laser intensity, background gas, etc. The present study is aimed at understanding the role of ambient gases and to correlate the morphology of the deposited film with the plasma plume parameters. A systematic investigation of optical emission studies of molecular carbon, using third harmonics of an Nd : YAG laser at different laser energies and argon gas pressures, is reported. The microstructure and the surface morphology of carbon films deposited on silicon substrates in the presence of argon gas is discussed.

2. Experimental set-up

The experimental set-up and data acquisition arrangement used in the present study are similar to those described elsewhere [-20,25]. An Nd:YAG (DCR-4G) laser and its harmonics (delivering up to 1 J in 8 ns at its fundamental at 10pps) was used to produce the carbon plasma. The graphite rod was continuously rotated and translated so that each laser pulse fell on a fresh graphite surface. The plasma emission was imaged onto a monochromator (Jobin Yvon, HRS-2). The vacuum chamber was evacuated to a pressure of less than 10 -3 Tort and argon gas was flushed three times before filling it in the chamber at the desired pressures. For thin film deposition, the laser beam was line focused on the graphite target enclosed in a vacuum chamber, using a cylindrical lens of focal length 25 cm. The ablated carbon was deposited on a silicon substrate positioned about 1 cm away and parallel to the target surface at different pressures of argon gas in the range 10- 2-100 Torr. Characterization of the carbon films deposited at different argon pressures was carried out by scanning electron microscopy (SEM), X-ray diffraction (XRD) and Raman spectroscopy.

3. Results and discussion

At higher laser irradiances, such as laser energies of 600 mJ and above, the plasma emission was dominated by atomic and/or ionic species from C V to C 1 [26], which recombine through atomic carbon to molecular carbon far away from the target surface. However, at low irradiances, molecular emission - - particularly C2 - - was found to dominate using laser radiation at a wavelength of 355 nm. At argon pressures where the optical emission showed intense C2 Swan band emission,

R. K. Dwivedi, R. K. Thareja/Surface and Coatings Technology 73 (1995) 170-176

172

the carbon films were deposited on silicon substrates. The silicon substrates were examined with the deposited film using SEM, XRD and Raman spectroscopy. Emission spectra of the C2 d-a, Av = - 1 Swan band sequence [-27] were taken in the presence of argon gas pressures at various laser energies. Fig. 1 shows the spectra of the C2 d-a, Av = --1 Swan band sequence for 10 -1 Torr of argon gas at laser energies of 30, 60 and 90 mJ using an N d : Y A G laser with a wavelength of 355nm. Swan band heads of (v', v'), i.e. (0-1) at 563.5 nm, (1-2) at 558.5 nm, (2-3) at 554.0 nm, (3-4) at 550.1 nm, and (4-5) at 547.0 nm, were also observed.

The intensity of the C2 band heads was found to increase with the laser energy. To see the effect of argon gas on the C2 molecular bands, a vibrationally resolved emission spectrum of the Swan band sequence A v = - 1 was recorded at a laser energy of 40 mJ with various argon gas pressures ranging f r o m 10 - 3 t o 100 Torr. Fig. 2 shows the A v = - 1 Swan band sequence at argon gas pressures of 10 -1, 1 and 10 Torr. The intensity of the C2 band heads increased as the pressure increased from 10 - 3 to 100 Torr. The intensity of the laser for observing these bands, being not very large, implies that the C2 bands are probably formed from atomic carbon recombination. At this

p(c

(c)

ur)

I,

"E

(b)

!

J

eIIJ C

a) (a)

s4s.o

sss.o

Wavetength

ses.o

(nm)

Fig. 1. C2 d - a Av = - 1 Swan band sequence at argon gas pressures of 10 - 1 T o r r and laser energies of(a) 30, (b) 60 and (c) 90mJ.

s~s.o sss.o ses.o Wavelength (nm) Fig. 2. C 2 d - a Av = -- 1 Swan band sequence at argon gas pressures of(a) 10 -1, (b) 1, and (c) 100 Torr.

R. K. Dwivedi, R. K. Thareja/Surface and Coatings Technology 73 (1995) 170-176

power level, the evaporated carbon contained a very small percentage of ionic species. Because our vibrational spectrum is rotationally unresolved, we have used the band head intensities [28,29] to estimate the vibrational temperature of the C2 bands. The relative population of the upper vibrational level, as derived from the measured intensities, was plotted against the vibrational quantum number. The slope of the curve gives the vibrational temperature. The F r a n k Condon factors used for our analysis are those of Spindler [30]. Fig. 3 shows the relative population of the upper vibrational level as obtained from the measured intensities, against the vibrational quantum number for argon gas pressure from 10 -3 to 100 Torr for laser radiation with a wavelength of 355 nm and laser energy of 40 m J; the slope of the curve gives the vibrational temperature. The vibrational temperature was also calculated at different laser energies for laser radiation with a wavelength of 355 nm at various argon gas pressures. Fig. 4 shows the variation of the vibrational temperature with the laser energy at an argon pressure of 10-~ Torr. We found that the vibrational temperature is optimum at a given energy. It first increases with laser energy and then decreases. Thin film deposition at the optimum laser energy can play an important role in deciding the quality of the film. The vibrational temperature was also calculated for other laser wavelengths. Fig. 5 shows the variation of the vibrational temperature at various argon gas pressures for a wavelength of 1.06 gm at a laser energy

173

12000 A = 355 nm P = 10-1Torr Ar

~ 11ooo

10000

~~ 9000 O -: x2_ > aoo0

7000 20

I

I

I

40

t

I

I

60 Energy (mI)

80

100

Fig. 4. Vibrational temperature as a function of laser energy at argon gas pressures of 10 -1 Torr.

15000 n •~ o

13000

0.355 p m 0.532 pm 1.060 p m

--1

1100C El.

rl

E [3 tO

900C

.I3

,---<

700~

-

5000

o

I

10-3

0 II

10-z

o

I

I

]

10-1

10o

101

102

103

Pressure (Torr)

z

Fig. 5. Vibrational temperature as a function of argon gas pressures for laser wavelength of energy 1.06 gm at energy 32 mJ, for 0.532 gm at 45 mJ and for 0.355 gm at 45 mJ.

~7 >

z

~

"

102

10q Torr 1 Torr 10 Torr 100

I

0

I

I

1.0

I

I

I

I

2.0 3.0 V ' - (V'= O)

I

Torr

]

4.0

Torr I

5.0

Fig. 3. Relative population of the upper vibrational level of the C 2 A v = - 1 Swan band sequence, vs. the vibrational quantum number for argon gas pressures from 10- 3 to 100 Tort at a laser energy of 40 mJ.

of 32 mJ, for 0.532 gm at 45 mJ and for 0.355 gm at 45 mJ. The vibrational temperature is found to decrease with the argon gas pressure for a particular wavelength. The presence of ambient gas helped to cool the molecular species and increased the recombination rate. We found the vibrational temperature to be larger for laser irradiation with a wavelength of 355 nm. Presumably, this is because of the changes in the optical penetration depth and photoionization of evaporated target material, as well as other secondary processes

R. K. Dwivedi, R. K. Thareja/Surface and Coatings Technology 73 (1995) 170 176

174

during ablation. This results in efficient coupling of the optical energy to the target, the enhancement of the production of atomic species, and an increase in the kinetic energy of the ablated species for deposition. It was reported by Koren et al. [ 11] that UV laserdeposited thin films results in lower resistivities. Xiang and Chang 1-18] have also shown that the quality of carbon films deposited using UV lasers is superior to that of those prepared by laser ablation with longer wavelengths, with regard to their mechanical hardness and optical properties. The wavelength dependence of the photoablation of carbon in the presence of helium gas has also indirectly suggested the usefulness of UV laser wavelengths for thin film deposition [-31]. It follows from our observation that the intensity of the C2 Swan band increases with decreasing laser wavelength. Since C2 is a critical species for diamond-like carbon, the deposition at UV laser wavelengths may result in good quality films. Fig. 6 shows the emission spectrum of the C2 d-a Swan band sequence with 10 Torr of helium and argon gas at a laser energy of 45 mJ. It is found that the plume emission can be enhanced significantly by an appropriate choice of ambient gas; however, the enhancement is more pronounced in the case of argon gas. The difference could be the result of the effect of the mass of the ambient gas atoms. Gases of heavy atoms exert a greater opposing force on the plume expansion than do gases of light atoms. As the result, the plume maintains a different density during its movement and, therefore, has a different emission intensity. Fig. 7 shows the variation of the vibrational temperature with the ambient gas pressures. We observed the vibrational temperatures to be larger in the case of argon, except at low pressures. This difference could be the result of the effect of confinement of the gases. At low pressures, the opposing force exerted by the background gas is very small; however, at higher background

(o)

t-

(b) ~

o

m

m

m

r

545.0

I

555.0

565.0 545.0 555.0 Wovetength (nm)

565.0

Fig. 6. C 2 d-a Swan band sequence at 10 Torr of (a) argon and (b) helium gas with laser wavelength of 355 nm and laser energy of 45 mJ.

15000 o o

Argon Helium

I ........ , 100 101

I 10 l

13000 L

rt

E 11000

-6 t-

Q

{3

O

rt :>

9000

7000 10-3

I 10-z

I 10-1

Pressure

103

(Torr)

Fig. 7. Variation of vibrational temperature as a function of ambient gas pressures for laser wavelength of 355 nm at a laser energy of 45 mJ.

gas pressures, the restriction of plasma expansion could be brought about by the surrounding gas. This will result in the transfer of part of the plume expansion energy to the background gas atoms as thermal energy. The rate of removal of thermal energy by the gas atoms varies inversely as its mass; hence, the heat removal by argon gas atoms will be less than that for helium. In turn, it results in a higher plume temperature and, hence, in a larger emission intensity. The deposited films have been characterized by SEM, XRD and Raman spectroscopy [32]. Fig. 8 shows SEM images of the carbon films deposited on silicon substrates with a wavelength of 355 nm and laser energy of 90 mJ at various argon gas pressures. The figure shows the impact of the argon gas pressure on the film morphology. The surface morphology of the films changes dramatically with increasing argon gas pressure, with densely packed spherical features dominating the film. Large spherical protrusions of different diameter are observed on the silicon surface. Recently, the adsorption of C6o molecules on silicon surfaces has been studied by various investigators [33-35], using scanning tunnelling microscopy (STM). It has been shown that the C6o molecules are highly spherical and that a considerable amount of charge transfer from the substrate to the C6o is responsible for the formation of monolayer films of C6o on various surfaces. We found that the nucleation of clusters on the substrate is optimum at an argon gas pressure of 1 Torr. It can be seen from Fig. 5 that the vibrational temperature is greater at an argon gas pressure of 1 Torr. This higher temperature at 1 Torr may lead to fragmentation

R. K. Dwivedi, R. K. Thareja/ Surface and Coatings Technology 73 (1995) 170-176 77% 2L2~r%" -2 2.~..". . • .z,:'~

-

.

-

.

,~ . - ' - -

-

.

175 . ..... ,~

.

.

ITA~'355

(a)

(b)

(c)

(d)

Fig. 8. SEM images of laser- ablated carbon films at argon gas pressures of (a) 10-1, (b) 1 and (c) 10 Torr, and (d) single microcrystalat 1 Torr. and ionization of larger clusters, resulting in dense film dominated by small clusters. The micro-Raman spectra of the deposited films showed two well-defined characteristic peaks at 1350cm -1 (D line) and a broad peak at 1580cm -1 (G line). These peaks have been attributed in the literature to polycrystalline graphite or amorphous carbon with graphitic bonding [-36]. The D line is a common feature of all disordered graphitic carbon, whose intensity relative to the G peak, ( I ( D ) / I ( G ) ) has been shown to vary inversely with the size of the graphitic crystallite [-37]. We found that, for the film deposited at 1 Torr, the relative intensity was smaller than that for the films deposited at pressures less than or greater than 1 Torr. This confirms the presence of larger crystallite sizes formed at 1 Tort as compared with at other pressures. Also, the broad G peak shifts to 1550 cm -1 for the film deposited at 1 Torr, which is characteristic of diamondlike carbon. The peak width of the Raman lines, indicating crystalline perfection of the probed sample region, is also found to vary with the variation of the argon gas pressures. The width of the Raman lines for the film deposited at 1 Torr of argon gas is the least, and the spectrum broadens as the pressure is increased or decreased around this value. This agrees well with the SEM images, because the crystallites become more dominant at a pressure of 1 Torr. The XRD pattern of the film deposited at 1 Torr gave peaks at 43.5 °, 76 °, 91.5 ° and 120 °, respectively indicating the presence of (111), (220), (311) and (400) crystal-

line planes of cubic diamond. This further confirms the formation of diamond-like films at 1 Tort of argon.

4. Conclusions C2 emission spectra were recorded at various argon gas pressures and various incident laser energies. The presence of argon gas increased the intensity of all the band heads. The C2 vibrational temperature was calculated in the presence of argon gas. It was found that the vibrational temperature decreases with increasing pressure. The surface morphology of the deposited films was found to depend strongly on the argon gas pressure. We found that there is an optimum pressure, where the cluster formation is more pronounced.

Acknowledgments This work was partially supported by Council of Scientific and Industrial Research, New Delhi, India. The authors wish to thank A. Trivedi and A. Sircar for their help during the experiments.

References [1] R.K. Singh, A.K. Singh, C.B. Lee and J. Narayan, J. 67 (1990) 3448.

Appl. Phys.,

176

R. K. Dwivedi, R. K. Thareja/Surface and Coatings Technology 73 (1995) 170-176

[2] J.R. Wachter, in L.J. Radziemski and D.A. Cremers (eds.), Laser-induced Plasmas and Applications, Marcel Dekker, New York, 1989, Chap. 6. [3] E.A. Rohlfing, D.M. Cox and A. Kaldor, J. Chem. Phys., 81 (1984) 3322. [4] J. Krishanaswamy, A. Rengen, J. Narayan, K. Vedan and C.J. McHargue, Appl. Phys. Lett., 54 (1989) 2455. [5] X. Chen, J. Majumdar and A. Purohit, Appl. Phys. A, 52 (1991) 328. [6] F. Davanloo, E.M. Juengermann, D.R. Jander, T.J. Lee and C.B. Collins, J. Appl. Phys., 67 (1990) 2081. [7] D.L Pappas, K.L. Saenger, J.J. Cuomo and R.W. Dreyfus, J. Appl. Phys., 72 (1992) 3966. [8] D.L Pappas, K.L. Saenger, J. Bruley, W. Krabow, J.J. Cuomo, T. Gu and R.W. Collins, J. Appl. Phys., 71 (1992) 5675. [9] T. Sato, S. Furuno and S. Iguchi, Jpn. J. Appl. Phys., 26 (1993) L1487. [10] S.S. Wagal, E.M. Juengerman and C.B. Collins, Appl. Phys. Lett., 53 (1988) 187. [11] G. Koren, A. Gupta, R.-J. Basemann, M.I. Lutwyche and R.B. Laibowitz, Appl. Phys. Lett., 55 (1989) 2450. [12] C. Champeaux, P. Maachet, J. Aubreton, J.-P. Mercurio and A. Catherinot, Appl. Surf. Sci., 69 (1993) 335. [13] E.W. Kreutz, A. Voss, M. Alunovic, J. Funken and H. Sung, Surf. Coat. Technol., 59 (1993) 26. [14] G. Koren and J.T.C. Yeh, J. Appl. Phys., 56 (1984) 2120. [15] M. Yoshimoto, T. Arakane, T. Asakawa, K. Horiguchi, K. Hirai and H. Koinuma, Jpn. J. Appl. Phys., 32 (1993) L1081. [16] I.N. Mihailescu, N. Chitica, L.C. Nistor, M. Popescu, V.S. Teodorescu, I. Ursu, A. Andrei, A. Barborica, A. Luches, M. Luisa De Giorgi, A. Perrone, B. Dubreuil and J. Hermann, J. Appl. Phys., 74 (1993) 5781. [17] R. Fabro, E. Fabre, F. Amiranoff, C. Garban-Labaune, J. Virmont, M. Weinfeld and C.E. Max, Phys. Rev. A, 26 (1982) 2289.

[18] F. Xiang and R.P.H. Chang, Mater. Res. Soc. Syrup. Proc., 270 (1992) 451. [19] S. Fujimori, T. Kasai and T. Inamura, Thin Solid Films, 92 (1982) 71. [20] Abhilasha, P.S.R. Prasad and R.K. Thareja, Phys. Rev. E, 48 (1993) 2929. [21] D.M. Gruen, S. Liu, A.R. Krauss and X. Pan, J. AppL Phys., 75 (1994) 1758. [22] H. Kroto, H.R. Heath, S.C. O'Brien, R.F. Curl and R.E. Smalley, Nature, 162 (1985) 318. [23] H. Tsai and B.B. Bogy, J. Vac. Sci. Technol. A, 5 (1987) 3286. [24] P.S.R. Prasad, Abhilasha, and R.K. Thareja, Phys. Status Solidi A, 139 (1993) K1. [25] R. Tambay, R. Singh and R.K. Thareja, J. Appl. Phys., 72 (1992) 1197. [26] R.K. Thareja and Abhilasha, J. Chem. Phys., 100 (1994) 4019. [27] K.P. Huber and G. Hertzberg, Molecular Spectra and Molecular Structure IV, Constant of Diatomic Molecules, Van Reinhold, New York, 1979. [28] L.L. Danylewych and R.W. Nicholls, Proc. R. Soc. London, Set. A, 339 (1974) 197. [29] Abhilasha and R.K. Thareja, Phys. Lett. A, 184 (1993) 99. [30] R.J. Spindler, J. Quant. Spectrosc. Radiat. Transfer, 5 (1965) 165. [31] Abhilasha, R.K. Dwivedi and R.K. Thareja, J. Appl. Phys., 75 (1994). [32] R.K. Thareja et al., unpublished results. [33] T. Hashizume, X. Wang, Y. Nishina, H. Shinohara, Y. Saito, Y. Kuk and T. Sakurai, Jpn. J. Appl. Phys., 31 (1992) L880. [34] X. Wang, T. Hashizume, H. Shinohara, Y. Saito, Y. Nishina and T. Sakurai, Jpn. J. Appl. Phys., 31 (1992) L983. [35] Y.Z. Li, M. Chander, J.C. Patrin and J.H. Weaver, Phys. Rev. B, 45 (1992) 13837. [36] D.S. Knight and W.B. White, J. Mater. Res., 4 (1989) 385. [37] F. Tuinstra and J.L. Koenig, J. Chem. Phys., 53 (1970) 1126.