Appl. Phys. B 63, 51-56 (1996)

Applied

Physics B and , ù , eOptics r, © Springer-Verlag 1996

Pulsed infrared selective dissociation of CF3Br temperature, fluence and wavelength J. I. del Barrio 1, R. Fernändez Cézar 1, E. Martin 1, F. M. G. Tablas 1, W.

as

a function of pressure,

Fuss 2

i Departamento de Quimica, Universidad Autónoma de Madrid, Canto Blanco, E 28049 Madrid, Spain Fax: + 34-1/397-4187, E-mail: PTABLAS@vml. sdi. uam. es) 2 Max-Planck-Institut für Quantenoptik, D-8046 Garching, München, Germany. Received: 2 August 1995/Accepted: 15 January 1996

Abstraet. The pulsed InfraRed (IR) photodissociation of pure CF3Br and a 1:3 mixture with H2 has been studied as a function of pressure, temperature, fluence and wavelength. The results have been obtained by mass spectrometry. Maximum enrichment for carbon 13 is obtained when irradiating with the 1046.9 c m - 1 (9P20) line, but the selectivity is stronger at longer wavelengths. The enrichment factor tends towards 1 when pressure increases with the pure substance, hut it is less affected in the mixture. It seems that hydrogen has a buffer effect. The selectivity shows maxima for all lines with the fluence and increases when the temperature is decreased from 22 to - 85°C. Some enrichment is observed for the bromine 79 isotope when irradiating with 1046.9 c m - 1 (9P20).

increase in selectivity when the temperature is lowered to - 85oC. Of all the variables studied, temperature and wavelength are clearly mainly responsible for the selection process, more important than fluence and pressure. To avoid a very common uncertainty, we start by defining clearly the steps by which selectivity is measured. The selectivity S is obtained following a similar procedure as has been used by Fuss [4] and Lyman [5]. The enrichment factor fi is calculated from the mass peak intensities in the remaining CF3Br; I149 corresponding to 13CF3V9Br and I14 s to 12CF379Br. fl -=

(I149/1148)afterirr (I149/I14S)before

PACS: 82.40

The CF3X and other halogenated methane compounds have been amply studied because they show absorption bands that overlap with the lines of the CO2 laser. Several researchers [1] have studied different aspects of the Multiple-Photon Dissociation (MPD) process. Avatkov et al. [2] have studied CF~Br under eo!lisional conditions. With mixtures of trifluorobromomethane and nitrogen, they observed that the enrichment factor increases with pressure up to 300 Torr, then decreases. A similar behavior is observed for the selectivity. Recently, Parthasarathy et al. [3] have attempted the carbon isotope separation in mixtures of CF3Br and C12. We have analysed the effect of hydrogen pressure on the carbon selectivity in CF3Br and compared these results with the photodissociation of the pure substance. Out experiment does not use sharp focussing in the static cell hut rather a constant fluence along the 1 m long waveguide. Although out experimental conditions were rather different from the previous studies, we obtained similar results. Contrary to CF3I which does not show selectivity at room temperature, CF3Br does, but we also see an

--

expl - (1 - S - l ) 13uni ~

(1)

irr

fu being the dissociation probability per pulse for isotope i, n the number of pulses and S the selectivity. As can be inferred from the definition of/~, a decrease in its value means an increase in the setective dissociation of 13C-bearing molecules. The remaining fraction of the most abundant isotope after n pulses is defined as: L = cù/co,

(2)

with cù the isotope concentration after n pulses and Co the initial one. From (1), the selectivity can be expressed as S = 1 + In/~/lnfù.

(3)

From the C2F 6 generated, we can estimate the product enrichment factor e as ~. after irr 0~ - - ~ b e f o r e i r r '

(4)

where 21121 + 112o ~" -- 21119 + I12o"

(5)

When S » 1 , as in our case, from (1) one obtains

13u'n = - lnfi/(1 - S -1) ~ - in/3.

(6)

52 Expressing ~ as a function of ~3u 1 - e x p ( - 13un) (7)

1 - e x p ( - S-113un) ' we finally obtain for S » 1 -ln/~ S~c~--

(8)

1-/~

The dissociation products are analysed with a SXP Elite mass spectrometer from Fisons capable of detecting 1 ppm. The mass spectrometer is kept at 10 .9 Torr. It is also possible to follow the amount of substance dissociated with a continuous CO2 laser from Edinburgh Instruments, but CF3Br did not show any strong absorption in the wavelengths emitted by this laser. CF3Br was supplied by Merck and hydrogen by SEO. Both were used without further purification.

1 Experimental 2 Results and discussion The TEA CO2 laser is a modified Lambda Physics E M G 200 that can be tuned. Figure 1 shows the experimental setup. The CO2 laser is filled with N2: CO2: He at 60: 80:320 mbar. The operating mode is TEMoo. An aperture of variable diameter is introduced in the cavity and allows to select the energy per shot. The maximum energy per pulse is 2 0 0 m j for the 9P18 line. The laser pulse consists of a 80 ns spike followed by a low-intensity microsecond tail. The frequency of the laser was 32 shots per minute. The laser beam is brought to the stainless steel waveguide (1 m long, 1.8 mm diameter) with KBr windows by a collection of mirrors and is focussed onto its entrance by a KC1 lens of 120 cm focal length. A metallic support bears the waveguide and allows to cool it to different bath temperatures. A slush bath of decane/heptane/liquid nitrogen allows to reach - 40°C and another one of hexane/liquid nitrogen reaches around - 85°C. Gentec sensors 200 are placed before and after the gas cell to measure the input and output energies. The ratio of these measurements reaches a value of 60% in the worst cases. The spot-size diameter at the entrance of the waveguide is approximately 0.45 mm. The waveguide is maintained at 10 .6 mm Hg. The windows and valves for the gas supply are mounted as close as possible to the ends to decrease the non-irradiated volume. The relation of both volumes is approximately 1 to 1. A MKS Baratron which has an output of 1 mV per mTorr is used at the entrance of the waveguide to control the pressure until 1 Torr. For higher pressures another Baratrou is used.

When not mentioned otherwise, the results were obtained at a sample pressure of 0.5 Torr, room temperature, 35 mJ per laser pulse and 960 pulses per sample.

2.1 Wavelength dependence 2.1.1 Pure CF3Br. For different wavenumbers of exciting light, from 1053.9 (9P12) to 1035.5 c m - 1 (9P32),/~, S and c~ have been studied. ~ shows a minimum around 1046.9 c m - 1 (9P20), as can be seen from Fig. 3, S is obtained using (8), and c~increases with longer wavelengths. This behavior is shown in Figs. 4 and 5, respectively. A similar trend has been observed by other researchers [2]. No selectivity is detected when the gas is irradiated with 9P34 line for 60 min (1920 pulses). The reason may be due to the small absorption coefficient that the molecule shows at this wavenumber, as can be seen from Fig. 2. Therefore, we obtained maximum S when irradiating with the 9P32 line (1035.5 cm-l), although the amount of product formed is very small. The reason for the 13C selectivity observed may be due to the near resonance of the vl mode of I»CF3Br at 1058 cm 1. There is a difference between the absorption and dissociation maxima of 23 cm ~ towards the red, as usual. As the absorption spectrum shows, the absorption coefficient of the vl mode of t3CF3Br is considerably larger than that of the 3 v3 mode of 12CF3Br. So we can justify the selective dissociation of ~3C-bearing molecules. 2.1.2 Mixture 1"3 ofCF3Br: H 2. Figure 3 shows a similar trend of/? vs wavelength in the mixture compared with the pure substance.

KCI

/

lens

TEA

I----,~

I~-

I I

~Da

I ow

k----\

I

D4

i I

veguide

\~) ~ _ OD2

I I I I

Y

D1 Fig. 1. Experimental setup (TEA pulsed COz laser; CW continuous-wave CO2 probe laser; Dl, D2, D3,/)4- IR detectors; V vacuum and gas supply; MS mass spectrometer)

2.2 Pressure dependence 2.2.1 Pure substance. We have studied the effect of pressure when exciting with the 1046.9 cm 1 line (9P20). Starting with 0.5 Torr, we observe that, as pressure increases,/3 moves slowly towards 1 (Fig. 4). For pressures larger than 3 Torr, fl becomes practically 1 indicating a total loss of selectivity. This fact can be explained by the effect of collisions distributing the absorbed energy among all molecules. It is also seen in the behavior of S and c~that are plotted in Figs 5 and 6, respectively. 2.2.2 Mixture. Figure 4 shows considerably different B values for the same number of CF3Br molecules in the

53 2.0-

II

'°9'311

I'

Will,0,0°0 cF3~r10cm

1.8-

1.6-

v1?' I

1034.18

1.4

1.2

| l

1117.81

/ '"°.d

A 1.0-

/

II nlllIL,

0.8-

i

/

1031.59 11032.40

l llii

I

II

J]l

|

l,o,~o~~,~% "/ I ~'1102;,

108,=7

0 . 6 "

0.4-

0 . 2 -

0

415o 44'4o d3o

41'oo doo 4o'9o 40'80

d2o

10'70 cm -1

10'60 10'50 10'40 1030

10'20 10'10 10'00

Fig. 2. Absorption spectrum of CF3Br at different pressures

-

_o

0

© ©

rn

©

©

•

o.1

o.1

0 , 0 1

,

1030

,

,

I

,

1035

,

,

I

. . . .

1040

q

. . . .

1045

~

1050

,

I

,

1055

0.01

,

1060

O

0.1

1

O

10

SamplePressure(Torr) [ • Pure CF3Br o CF3BdH2I

Wavenumber( cm ~) I o Pure CF3Br • MixtureI Fig. 3. Enrichment factor/~ vs wavenumber for the pure substance and the hydrogen mixture at room temperature. Sample pressure = 0.5 Torr, fluence = 1.4 J/cm 2

Fig. 4. /~ vs pressure for the pure substance and the mixture when irradiated with 1046.9 cm- 1 (9P20). Conditions are the same as in Fig. 3

waveguide, for example, 0.5 T o r r of pure substance and 2 T o r r of the mixture. This indicates that H2 impedes the occurrence of the following abstraction reaction that would decrease the enrichment [6]:

The pressure dependence of the enrichment factor has also been studied in the range of wavelengths indicated above (Fig. 7). It can be seen that there are no significant differences in /3 when pressure increases from 0.5 to 1.0 Torr, p r o b a b l y due to the relatively few collisions in both cases. We have only observed small selectivity at 1046.9 c m - 1 (9P20) for the bromine isotopes with pressures of 0.5 and

13CF3 + 12CF»Br ~ 13CFBr + 12CF3.

(9)

As a consequence, the selectivity and Πare also changed by the presence of h y d r o g e n (Figs. 5, 6).

54 100 0

10

O O

0.1

1

10

Sample Pressure ( Torr ) [ • Pure CF3Br

•CF3Br/H2]

Fig. 5. Selectivity vs pressure. Exciting energy 1046.9 cm -1 and standard conditions 100

O D

=~

- 85°C. Figure 8a shows that, for all wavelengths,/~ decreases when the temperature is lowered, but more markedly for the wavenumbers 1046.9 and 1049 cm-~. A corresponding behavior is observed for the selectivity and ct (Figs. 8b, c). It is known that lowering the temperature concentrates the molecules in the lower vibro-rotational levels, favoring the absorption from these. It supports our assumption that the v~ mode of 13C-containing molecules is responsible for the main absorption and the improved selectivity. Another explanation for the observed temperature dependence is that the abstraction reaction (9) will be less probable at lower temperature Il.a]. 2.3.2 Mixmre. For the same wavenumber interval and at 22, 0 and - 4 0 ° C , we have measured the enrichment factor. The results obtained are represented in Fig. 9. The B values decrease when lowering the temperature; the same as in the case of the pure substance. The changes are also larger for the same wavelengths. The effect is not as

10

I

1

0.1

Sample Pressure ( Torr ) / " Pure

•

10 0.1

• MixtureJ

Fig. 6. c~vs pressure. Exciting energy = 1046.9 cm- 1 0.01 1030

1035

1040

1045

1050

1055

1060

Wavenumber (cm ~) (a)

1000

o.1 •

t

1

I,

#

0.01 .... ~ 1030 1035

* ~ _ = 22~'C ~' T = -85°C I

100

.» õ

'

.

.

1040

.

.

1045

t

.

.

.

.

1050

. . . .

1055

1060

03

Wavenumber ( cm -~ )

O

10 O

• P = 0.500 Torr • P = 1.000 Torr I Fluence = 1.9 J/cm2

1030

Fig. 7. fl vs wavenumber at 0.5 and 1.0 Torf of the H2:CF3Br mixture

1035

1040

1045

1050

1055

1060

Wavenumber (cm -~ )

]" T=22°C

(b)

1.0 Torr, obtaining enrichment factors of 1.08 and 1.11 respectively. These values were calculated here by using

• T = -85°C ]

1000 1

(I150/ I148)afterirr B= (I150/I148)beforeirr"

100 <

This implies that we are dissociating selectively the molecules that contain the isotope 79. No selectivity was observed with the other lines used.

10

. . . .

1030

1035

1040

i

,

1045

,

,

i

1050

,

,

,

i

,

1055

,

,

1060

Wavenumber (cm -~ )

2.3 Temperature dependence


2.3.1 Pure substance. For wavenumbers ranging from 1053.9 to 1035.5 cm 1,/~, S and c~were obtained at 22 and

Fig. 8a-c Enrichment factor (a), selectivity (b) and ~ (c) at two temperatures vs wavenumber for pure CF3Br

I " T = 22 oc

• T = -85 ° c I

55

pronounced as in the pure substance because in the mixture the temperature has only been lowered to - 40°C; this is a temperature decrease of 45°C less than in the pure substance.

A

1 •

•

•

•

•

c~

0.1

0.01 1030

tt 6

¢ä m

6

•

''

. . . . . . . 1035 1040

li

' . . . . . . . . . . . 1045 1050 1055

1060

Wavenumber (cm 1)

0.1

[ • 1.01 J / c m -2 • 1.4 J / c m 2 • 1.9 J / c m -2]

Fig. 11. Mixture enrichment factor for different fluences vs wavenumber 0.01

•

,

1030

,

i

,

,

,

1035

~ ,

,

,

1040

I

. . . .

q

1045

,

1050

,

r

1055

1060

Wavenumber ( cm -1 ) [.T=22°C

•T=O°C

2.4 Fluence dependence

mT=-40°C]

Fig. 9. Mixture enrichment factor for three temperatures vs wavenumber

A

Q

©

•

o

I

•

O

0.1

,

0.01

1030

I

,

1035

,

I

,

,

,

1040

I

•

,

,

1045

I

. . . .

I

1050

1055

1060

Wavenurnber (cm -~ )

[

• 0.94 J / c m 2 • 1,4 J / c m -~ • 1.7 J / c m -2 o 1920 pulses (1.4 J / c m -2)

(a) 1000

100

•

2.4.1 Pure substance. For the wavenumbers that generate a substantial amount of product, we observe a clear dependence of the enrichment factor with fluence when it is changed from 0.94 to 1.7 J/cm 2 (Fig. 10a). The larger changes correspond to the region of maximum isotope separation. Figure 10b and c show the effect of fluence on selectivity and c~, respectively. For both there is no systematic trend in the values seen. 2.4.2 Mixture. The effect of fluence on the mixture follows a similar trend as in the pure substance (Fig. 11). The larger changes also correspond to the region where the B values are minima, around 1046.9 cm ~. Although we have not been able to estimate selectivity values for different fluences, we believe that, as in the case of the pure substance, the selectivity does not change systematically with fluence in the above-considered region.

j

.>

õ

-$ cO

3 Conclusion

10

1

'

'

1030

,

I

. . . .

1035

!

.

.

.

1040

.

.

.

I

1045

,

I

1050

,

1055

1060

Wavenumber (cm 1)

(b)

I • 0.94 J / c m 2 • 1.4 J / c m 2 • 1.7 J/cm-21

1000 0

100 am Il

10

1

,

1030

,

r

,

1035

,

,

I

,

1040

,

,

1045

,

,

I

,

1050

,

I

,

1055

, ~

1060

Wavenumber ( cm -1 )

(c)

I•

0.94 J / c m -2 • 1.4 J / c m -z • 1.7 J / c m -21

Fig. 10a-e /?(a), selectivity (b) and c~(e)for pure CF3Br for different fluences vs wavenumber.

In order to achieve isotope separation, the most relevant parameter is the exciting wavelength. Under our conditions, the enrichment factor is maintained until partial pressures around 4 Torr where it approaches 1, and the selectivity tends to disappear because of collisional energy transfer. It seems obvious from our results that hydrogen acts as a good buffer affecting enrichment. Lowering the temperature seems to be advisable to improve isotope separation. Within the range of fluence studied, we have observed significant changes in the enrichment factor. The depletion of bromine 79 in the remaining CF3Br when irradiating with the 9P20 line (1046.9 cm 1) could be explained through the absorption of the 3v3 overtone. This overtone falls at 1048.2 c m - ~ for CF» 79Br while for CF»a~Br appears at 1042.5 cm -1, as can be seen in Fig. 2, which shows the absorption spectrum of CF»Br at different pressures. Comparing the differences between the absorption of each molecule and the line that shows selective

56 d i s s o c i a t i o n (2 c m - 1 to the red for CF3 79Br a n d 5 c m to the blue for CF3 8~Br), we can c o n c l u d e t h a t the enrichm e n t factor we o b t a i n is due to the preferential a b s o r p t i o n of the V9Br-containing molecules. All the o t h e r i r r a d i a t i n g lines o v e r l a p with the a b s o r p t i o n b a n d s of b o t h types of molecules m a k i n g the selection process very difficult. W e have centered o u r discussion o n the e n r i c h m e n t factor i n s t e a d of the selectivity because we consider it m o r e meaningful physically.

Acknowtedgements. We thank the Max Planck Society and especially Professor Karl L. Kompa and Dr. Werner Fuß for the equipment that has allowed to carry out part of this work. We also thank the CAYCIT for financial support under contract PB90-0198. References 1. s. Bittenson, P.L. Houston: J. Chem. Phys. 67, 4819 (1977) M. Gauthier, P.A. Hackett, M. Drouin, R. Pilon, C. Willis: Cdn. J. Chem. 66, 2227 (1978)

2.

3. 4. 5. 6.

M. Drouin, M. Gauthier, R. Pilon, P.A. Hackett, C. Willis: Chem. Phys. Lett. 60, 16 (1978) R.V. Ambartzumian, V.S. Letokhov, G.An. Makarov, A.A. Puretzky: Tech. Dig. Xth Int'l Quantum Electronics Conf., Atlanta, GA (1978) Paper N-8 A.S. Sudbo, P.A. Schulz, E.R. Grant, Y.R. Shen, Y.T. Lee: J. Chem. Phys. 70, 912 (1979) K. Sugita, P. Ma, Y. Ishikawa, S. Arai: Appl. Phys. B 52, 266 (1991) V. Parthasarathy, S.K. Sarkar, K.K. Pushpa, K.A. Rao, K.V.S. Rama Rao, J.P. Mittal: Appl. Phys. B 56, 101 (1993) M. Hattori, Y. Ishikawa, K. Mizuta, S. Arai, S. Sugimoto, Y. Shimizu, S. Kawanishi, N. Suzuki: Appl. Phys. B 55, 413 (1992) A. Baldacci, A. Passerini, S. Ghersetti: J. Molec. Spectrosc. 91, 103 (1982) O.N. Avatkov, V.B. Laptev, E.A. Ryabov, N.P. Furzikov: Sov. J. Quantum Electron. 15, 375 (1985) V. Parthasarathy, S.K. Sarkar, N.V. Iyer, K.V.S. Rama Rao, J.P. Mittal: Appl. Phys. B 56, 321 (1993) W. Fuss: Chem. Phys. 36, 135 (1979) J.L. Lymann, S.D. Rockwood, S.M. Freund: J. Chem. Phys. 67, 4545 (1977) M. Gauthier, P.A. Hackett, C. Willis: Chem. Phys. 45, 39 (1980)