摘要:

Kinetic Study of Electronically Excited Carbon Atoms C(2'SJ BY DAVID HUSAIN AND PETER E. NORRIS t Department of Physical Chemistry, University of Cambridge, Lensfield Road, Cambridge CB2 IEP Received 29th November, 1978 A kinetic study of the collisional behaviour of the electronically excited optically metastable carbon atom, C[2p2('So)], 2.684 eV above the Z P ' ( ~ P ~ ) ground state, is presented. C(2lSO) was generated by the repetitive, pulsed irradiation of CC14 in a flow system and monitored photoelectrically in ab- sorption by resonance line attenuation at I. = 247.9 nm [3s('P?) +-2pz('S0)]. The experimental system, incorporating pre-trigger photomultiplier gating, signal averaging and computerised analysis of the atomic decays, yields considerably improved rate data for the 'So state compared with those reported hitherto from the results of " single-shot " measurements.Collisional rate data for all the states of atomic carbon and also silicon arising from the npz configuration, namely, C(Z3P,, 2'D2, 2'S0) and Si(33PJ, 31D2, 3'S0), are considered in general terms within the context of symmetry argu- ments on the nature of the potential surfaces involved using the weak spin-orbit coupling approxima- tion. The data for C + Ha, O2 and C2H, are discussed in some detail. Correlation diagrams con- necting the states of C -t H2 and CH + H tlia the state manifold of CH2, and of C + O2 and CO + 0 via the state manifold of CO, are presented. An approximate potential energy diagram for CH2-CH2 in CZu symmetry, based on electron occupancy arguments, is also presented and employed to discuss the rate data for C(23PJ, 2'D2, 2'S0) + C2H4./c\ The collisional behaviour of group IV atoms in specific electronic states, in par- ticular, states arising from the overall np2 configuration (3PJ, ID2, '&), constitutes one of the major development areas which permit a general consideration of the relationship between atomic reactivity and electronic structure.' As these excited states are characterised by electric dipole-forbidden emission to lower states they are amenable to direct study by time-resolved atomic absorption spectroscopy.' The general vehicle for considering reactivity has been to employ symmetry arguments on the nature of the potential surfaces involved using the weak spin-orbit coupling approximation for light atom-molecule collisions 2*3 and (J, Q) coupling for heavy atom-molecule collisions.' Notwithstanding the limitations of the correlation dia- gram approach, particularly the fact that such diagrams are not potential energy diagrams and do not indicate the presence or absence of energy barriers, especially if the infinitely separated reactants and product species are not correlated via states of the least symmetrical complex, symmetry arguments for the light atom-molecule collisions are more useful than those for heavy atom-molecule collisions.' This arises because, in ( J , Q) coupling, the spin degeneracy is lost, the number of potential surfaces correlating with a particular pair of colliding species becomes larger as does the number of surface crossings. Thus, the collisional behaviour of the light atoms, carbon and silicon, are more amenable to considerations of symmetry arguments and constitute the broad framework of this paper.It is only recently that the detailed absolute rate data derived from direct monitor- ing of the atomic states have been reported for silicon, principally by the present t Present address : I.C.I. Petrochemicals, P.O. Box 90, Wilton, Middlesborough.274 ELECTRONICALLY EXCITED C ATOMS authors. Thus, we have described absolute rate constants for the collisional removal of Si(33PJ),495 si(31D2)697 (0.781 eV)' and si(31S0)9i10 (1.909 eV)' by a wide range of gases using time resolved attenuation of atomic resonance radiation following pulsed initiation. Furthermore, we have shown by monitoring of the individual spin-orbit states in the appropriate kinetic experiments [Si(33P0-,,2): J = 0,O; J = 1,77.15 cm-'; J = 2,233.3 1 cm-']' that the rate data are consistent with the maintenance of a Boltz- mann equilibrium between the spin-orbit levels during the kinetic decays.Davis et aZ.11*12 have reported rate data for the collisional removal of Si(33PJ) by F2, NO, 0, and N 2 0 derived from resonance line absorption measurements on a flow discharge system. Rate data for the collisional behaviour of the analogous states of atomic carbon, derived from atomic absorption methods, have been reported in the litera- ture for some years now. The experimental methods include flash photolysis coupled with plate photometry in the vacuum ~ltraviolet,'~ photoelectric monitoring of reso- nance line absorption following pulsed irradiation,1620 and resonance line absorption following plasmolysis in a flow system" and pulsed radiolysis.22 Thus rate data have been reported or may be derived for the collisional removal of C(23PJ),1s'6*21 C(21D2)13917-19 (1.263 eV)' and C(21S0)13920-22 (2.684 eV)' by a large number of gases.In the various measurements on C(23PJ), it has been assumed that the spin-orbit levels maintain a Boltzmann equilibrium throughout the kinetic decays in view of the low electronic energies to be transferred on collision ( J = 0, 0; J = 1, 16 cm-'; J = 2, 43 cm-').* Previous kinetic investigations on C(2'S0) show these all to be subject to severe experimental limitations. Braun et aZ.13 showed the low photolytic yield of C(2'S0) derived from C302 to be dependent on a power of the flash intensity greater than unity.These authors reported one limit for the quenching rate constant of this atomic state, namely, by H2 from plate photometry of the atomic line at a fixed time delay. Although Meaburn and Perner 21 monitored C(2'S0) by resonance absorption following pulsed radiolysis, they did not, in fact, translate their limited number of observed half-lives for the atomic state into absolute rate constants. Such a procedure was only adopted by later workers in order to achieve estimates of the appropriate rate constant^.'^ The more detailed resonance line absorption measurements of Husain and KirschI9 still only resulted in lower limits for the rate of quenching of C(2lSO) by gases added to the photochemical precursor, C302.This arose on ac- count of the high flash energies and the vacuum ultraviolet photolysis (A > 105 nm) necessary to achieve particle densities of C(2'SO) sufficiently large for time-resolved resonance line attenuation measurements in the " single shot " mode,20 and the sub- sequent complexity to the photochemistry.20 In this paper, we describe an experi- mental study of C(2'S0) which considerably improves upon this earlier work. The method involves the relatively low energy repetitive pulsed photolysis on a flow system, coupled with resonance line absorption using pretrigger photomultiplier gating, signal averaging and computerised analysis of the results. The rate data for C(2lSO) are considered with those for C(23PJ) and C(21D2) using the symmetry arguments de- scribed above, and are also compared with the rate data for Si(33PJ, 3'D2, 3'S0).In limited cases, it has been possible to correlate reactants and products through states of the triatomic collision complex. EXPERIMENTAL The experimental arrangement employed a system similar to that described hitherto for the kinetic study of Si(3ID2).' C(21S0) was generated by the repetitive pulsed irradiation ( E = 125 J, 0.2 Hz) of a CC1, + He mixture ([He]: [CCI,] !z 5 x lo4: 1) in a coaxial lampD. HUSAIN AND P . E . NORRIS 215 and vessel assembly,' with a common wall of high purity quartz (Spectrosil, 1 > c 165 nm), that constituted part of a flow system kinetically equivalent to a static system.The transient atom was monitored by resonance absorption at L = 247.9 nm [3s('P;) +2pz ('So),* gA = 1.9 x lo9 s - ' ] ~ ~ by means of a " pre-trigger " gated photomultiplier (E.M.I. 9816 QB), the operation and circuits for which have been described p r e v i o ~ s l y , ~ ~ mounted on the exit slit of a grating monochromator (Czerny-Turner mount, McPherson Corporation, U.S.A.). The resonance source comprised a microwave-powered atomic emission flow lamp (pco = 1.3 N m-', ptotal with He = 133 N m-', incident power = 60-80 W) using a power generator (E.M.I. type T 1001) incorporating a high tension filter to reduce output instability and mains ripple. This, combined with the use of an optical system incorporating the combination of short focus Spectrosil lenses, yielded unattenuated signals, I,, of high intensity and with good signal-to-noise ratio, and permitted kinetic measurements at degrees of resonance absorption < z5 %.The resulting photoelectric signals at A = 247.9 nm were amplified without di~tortion,'~ captured, digitised and stored in a fast-response transient recorder (Data Laboratories DL 920) used in the " A/B " mode4 in order to record It, and I,, the attenuated and unattenuated signals, on different time bases. The signals were then transferred to a signal averager (Data Laboratories DL 4000) whose contents, generally representing the result of 32 individual experiments, were then transferred onto paper tape (Data Dynamics punch 11 33) in ASCII code for direct input into the University's IBM 370 computer.All data were subjected to the numerical smoothing proceeding of Savitsky and Golay.26 All materials were prepared essentially as described in previous publications [He, Kr (for the photoflash lamp), CCI4, H2, 0 2 , NO, N20, C2H4],4-5*8*11*27.28 RESULTS AND DISCUSSION Fig. l(a) gives an example of the computerised output of the digitised light inten- sity at il = 247.9 nm indicating the decay of resonance absorption by C(2'S0) follow- ing the pulsed irradiation of CC14. Fig. l(b) and (c) show the effect on the lifetime of C(2'S0) by the addition of 0,. First-order kinetic plots for the decay of C(2lSo) are shown in computerised form in fig. 2 for the data given in fig. 1. Such plots, involving low degrees of resonance light absorption, are scattered, but represent the first clear kinetic decays of the photochemically generated C(2'SO) atoms as opposed to using the tail of an oscilloscope trace.20 In some cases, the initial yield of C(2'SO) is affected by the addition of added quenching gases, particularly at higher pressures.The method does not readily permit the standard empirical determination of a modi- fied Beer-Lambert law,29 principally on account of the variation of the yield of C(2'SO) with flash intensity. We therefore employ the Beer-Lambert law itself [It, = I, exp(-&cZ) where the symbols have their usual meaning] to translate the ex- tent of resonance absorption into relative atomic concentrations. Hence the slopes of the plots given in fig. 2 are given by -k', the overall first-order decay coefficients for C(2'S0) in a given experiment.The first-order decay coefficients are given in the standard form, k' = K + kQ[Q], where k , is the absolute second-order quenching rate constant for the gas, Q. Fig. 3 and 4 show the variation of k' with the concentration of the gases CCl,, 02, NO and C,H4. The remaining gases investigated here, namely, He, H2 and N20, only yielded upper limits. The slopes of the plots in fig. 3 and 4 yield the appropriate absolute second order rate constants. These are listed in table 1 together with the upper limits derived from this investigation, previous rate data for C(2'S0) and those for C(23P,) and C(2'D2). This table provides a summary of rate data for atomic carbon in specific electronic states. Table 2 gives the analogous summary represent- ing known rate data for specific states of atomic silicon.The general reactivities of atomic carbon and silicon in the three low lying states have been discussed in previous papers, referenced in table 1 , using the weak spin-orbit276 ELECTRONICALLY EXCITED c ATOMS ..............................................................................,...............-... .. 0-40'- . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ................................... ................................. .................................. ............................................. ......................................... ....................................... .................................... . . . . . . . . . . . . . . . . . . . .. .................... ................. ............ . , ........... . . . . . . . . . . . ......... ........... ....... .......... ........ ....... ....... . . . . . . ......... ....... ....... .... . . . . . . , . ..... I . ...... ..... ..... ..... ...... ..... , . * 0.38:+ .. ... l a 1 ;I ..... 1 ..................... 1 ..................... I ..................... 1 ....................... 1 2 3 4 5 0.6 1.0 1.4 1.8 time / m s FIG. 1.-Digitised time-variation of the transmitted light intensity at 1 = 247.9 nrn [3s('P,0)+ 2pz ('So)] indicating the decay of resonance absorption by C(2lS0) in the presence of O2 following the pulsed irradiation of CCI,. [CCl,] = 1.3 x 1013 molecules ~ m - ~ , . [ H e ] = 7.2 x IOl7 atoms C M - ~ ; E = 125 J; repetition rate = 0.2 Hz; gate duration = 120 ps.[O2](rnolecules ~ m - ~ ) : (a) 0.0; (6) 4.3 x (c) 5.2 x lot3. No. of experiments for averaging: ( a ) 16; (b) 32; (c) 32.D. HUSAIN A N D P. E . NORRIS - 3.0 - 4 .o la I I I I I I I . .. I.. . .... I.. ...... ... ........... . . 8 ..... , . . . . . . . . . . . ... , .... ...... ...... - .......... ........ .... .... ... ... # ... ..... .... ....... ...... .... * ....... ....... ........ * ............ ...................... ............... L .............. .......... ..... I . . . I ..... I ..... I ...... , ..., .... , . .I ... . . I -5.0+ * , i .......... I...-.- ................ 1 ..................... 1 ..................... 1 ..........---------.I . o ..I . I 1.0 1.5 2.0 2.5 I . . ........ ......... . . ....... ........... .. . ...... I . . 4 I .. -3.5- :* I . - 4 . 0 - 0.6 277 time / ms FIG. 2.-Pseudo first-order plots for the decay of C(2'SO) obtained by monitoring the absorption of resonance radiation at IE = 247.9 nm [3s('P?) t-2pz('So)] following the pulsed irradiation of CCI4 in the presence of 0,. [CCl,] = 1.3 x 1013 molecules ~ m - ~ , [He] = 7.2 x 10'' atoms ~ m - ~ ; E = 125 J ; repetition rate = 0.2 Hz; gate duration = 120 ps. [O,] (molecules cmP3): ( a ) 0.0; (b) 4.3 x (c) 5.2 x IOl3. No. of experiments for averaging: (a) 16; (b) 32; (c) 32.278 ELECTRONICALLY EXCITED c ATOMS 2.0 t - I a I I I 0 1 2 3 L 5 6 /molecules cm-3 lo-% 0 2 A aj 1 6131 c CIS/ mole cu ~ e s c m-3 FIG. 3.-Plots of pseudo first-order rate coefficients ( k ' ) for the decay of C(2'S0) in the presence of (a) CC14 and (b) 02.a 0 1.0 2.0 3.0 16'3[N0,C2H,] /molecules cm-3 FIG. 4.-Plots of pseudo first-order rate coefficients (k') for the decay of C(2'SO) in the presence of NO and C2H4: A, NO; 0, C2H4.D. HUSAIN AND P . E . NORRIS 279 TABLE RATE DATA FOR THE COLLISIONAL REMOVAL OF C(z3PJ, YO2, 21S0) BY VARIOUS (2nd-order rate constants, cm3 molecule-' s-' ; 3rd-order rate constants, cm6 molecule-2 GASES a s-'; T = 300 K) gas c(23pJ) C(21D2)(1 .263 eV) C(21S,)(2.684 eV) He Ne Ar Kr Xe H2 NO < 3 x 10-1B(18) < 10-16 b 4 1 0 - 1 5 ( 1 8 ) < 2 x 10-'6(20) 1.1 i 0.4 x (18) 9.4 -i: 1.6 x (18) 1.1 0.3 x 10-lo (18) 4.15 Y lo-" (13) 6.9 &- 1.2 x (M = He)(16) c 2.6 & 0.3 x 10-l0 (17) < 5 x b 7.1 f 2.5 x < 4 x 10-14 (20) < 7 x (21) < 5 x (13) x 2 x 1 0 - 1 4 (22) (M = He) (14, l 5 ) c 3.1 -+ 1.5 x (M == Ar) (15) c 4.1 :k 1.2 Y 10-"(17) < 3 X (20) x 2.5 x lo-'? (13) 2.6 f 0.3 :i 10-l' (16) 2.6 x lo-" (19) 9.9 5 1.8 x 3.5 f 1.5 x lo-" (14, 15) < 5 x 10-12(13) x 5.0 x lo-" (22) X 3.3 x 10-11 (13) 2.5 x lo-'? (21) 4.8 i 0.8 x (16) 4.7 -& 1.3 x lo-" (19) 4.8 f 0.5 x b 7.3 f 2.2 x (14, 15) 9.2 ..: (13) 1.1 x (13) 6.3 f 2.7 x (M -= He) (15) 1.6 1 0 .6 x 10-" (19) < 6 s (20) 6 3.5 x 10-'8 (22) < 10-15 (16) <10-14 (14: 15) 1.3 * 0.3 x lo-" (16) 1.4 0.5 A 10-'O(19) < 5 x lO-'*h 2.5 f 1.6 x 3.7 5 1.7 x 10-11 (19) ,< 1.0 x 10-lS(22) (14, 15) (16) z 1.7 x 10-"(19) < 3.6 x (14. 15) < 2.5 x 10-l5 (15) 2.1 0.5 \: (19) <lo-" (20) < 5 x 10-15 (13) 3.2 x lo-" (13) x 3.0 x (22) < 6 x l0-l7 (21) -C 6 x lO-l7(21) 3.7 'i 10-'0(19) 9.0 f 1.6 x lo-" h -C 6 x lO-"(21) 1.8 f 0.2 >.: 10-'"(16) 1 x 1 0 - 1 0 (20) 2.7 0.5 x b o Figures in brackets denote appropriate reference numbers; b this work; c 3rd-order.TABLE RATE DATA (k'cm3 molecule-' s-I, 300 K ) FOR THE COLLISIONAL REMOVAL OF Si(33PJ), Si(3'DJ and Si(3'S0) BY VARIOUS GASES a (2nd-order rate constants, cm3 molecule-' s- ' ; 3rd-order rate constants, cm6 s-'; T = 300K) ~ ~ -~~~ Si(33P~) Si(3'D2)(0.781 eV) Si(31S,)(1.909 eV) gas G 1.3 >: 10-15 (9) He < I O - I 5 (6) Kr < 4 :: 1 0 . - ' 5 (7) < 4 Y 10-15 (7) Xe i 6 x 10- l 5 (7) < 6 x 10-lK (7) H Z NZ 0 2 F* c 1 2 co NO 10-33(M : He) (5) b 8.1 / IO-" (6) <lo-" (10) 4 x 10p3?(M = He) (5) < 5 x 1 0 - 1 2 (7) <lo-" (10) 2.7 & 0.3 x 10-lu(4) 9.8 f 4.9 x lo-" (11) 1.2 f 0.6 x 10-l0 (12) (600 K) 3.3 & 0.3 x 10-lo (5) 1.1 f 0.1 x 10-10 (5) 7.1 x (7) 1.2 f 0.05 x 10-'(lO) 2.0 * 1.0 x I0-'1 (11) c 7.6 f 3.8 x (11) c 1.1 f 0.1 x lo-" (5) 1.7 x (7) 1.7 f 0.3 x lO-ll(l0) 2.3 Y lo-" (6) 1.5 f 0.2 x lo-" (9) 6.1 x lo-" (7) 7.3 * 0.1 x 10-11 (10) < 3 x 10-33(M = He) (5) b 1.1 x 10-11 (7) <lo-" (10) coz CH, < 10-14 (5) 1.3 x I O - ' O (7) 9.4 & 1.2 x 10-11 (10) CF, 2.4 f 0.3 x 10-12 (5) 4.2 10-"(7) 4.3 i 0.8 x (10) C2H2 4.9 * 0.3 x 10-lo (5) 2.0 >.10-10 (7) 1.1 5 0.1 x lO-'"lO) GH4 2.2 f 0.2 '< 10-10 (5) 3.7 :< 10-'0(7) 2.5 f 0.3 x 10-lo(lO) SiCI* 7.2 f 1.2 < lo-" (4) N,O 1.9 f 0.2 x lo-" (4) 1.7 x (7) 4.3 f 0.4 x lo-" (9) 8.2 f 4.1 ': lo-" ( I l ) c 2.9 >.: 10-"(6) 9.1 5 1.4 x lo-" (9) a Figures in brackets denote appropriate reference numbers; b Slow, third-order kinetics presumed; c 350 K.280 ELECTRONICALLY EXCITED c ATOMS coupling approximation.Space clearly prevents a similar treatment here, even for each reactant studied in this investigation. We restrict our discussion to the reaction of atomic carbon with the molecules H,, 0, and C2H, as, in the cases of the first two molecules, we may correlate states of reactants and products through the manifold of states respectively for CH, and C02, and for the latter, we may construct a crude P\ potential energy diagram in CZv symmetry for the species CH,-CH,. Fig. 5 shows the diagram connecting the states of C + H, and CH + H, corre- lated through the state manifold of CH2 and employing standard thermochemical data.30-32 The states of CH2 merit detailed discussion; however, we restrict our considerations here to the low lying C'A, state whose energy above the Z3B1 ground state is a matter of dispute. We employ the value reported by Danon et aZ.33 derived from laser-induced fluorescence measurements of CH,(a"A,) (0.273 eV) rather than that reported from ab iriitio calculations (0.48 eV).34 Apart from these two sources, we essentially follow Herzberg3' for the states of CH,. The energy of CH(a4ZC-) is taken as 0.742 0.008 eV foflowing the laser photoelectron spectrum of CH- given by Kasdan et al.35 The third-order removal of C(23PJ) by H2 (table 1) is clearly in accord with the correlation diagram, reaction proceeding on the 3A" surface and corre- lating through CH,(Z3B,).Rapid removal of C(2'D2) by H, (table 1) is in accord with symmetry-allowed, exothermic chemical reaction to yield CH(X211) + H( 12S), proceeding through the states CH2(C'A,) and CH($'B,) via 'A' and IA" surfaces, respectively. There are no correlations leading exothermically ria symmetry-allowed routes either to states of CH2 or of CH 4- H (fig. 5) for C(2'SO) + H, and the slow rate reported for the atomic removal process (table 1) is in accord with this. We have previously presented a correlation diagram connecting the states of Si + H, and SiH + H through those of SiH,' following the calculation of the state manifold of SiH, reported by W i r ~ a r n . ~ ~ The rate data for the removal of Si(33PJ, 3lO2, 3'SJ by H2 are similarly seen (table 2) to be in accord with the correlation diagram. The correlation diagram connecting the states of C + 0, and CO + 0 through those of C02 must employ calculated states of the latter.Whilst there is clearly an extensive electronic spectroscopy of the CO, r n ~ l e c u l e , ~ ~ * ~ ' detailed energies and electronic configurations of the full manifold of the low lying states have not been derived from experimental measurements as many of the excited states are bent. Hence, spectra involving transitions from the linear ground state to these bent states have yielded broad envelopes for the vibronic transitions in accord with the Franck- Condon principle. There are various calculations of electronic states for the C02 molecule.We are concerned here primarily with a complete state manifold, correctly ordered, rather than with high accuracy in absolute energy. For this purpose, we employ the results of the calculations of England et aZ.,38-40 who have reported the energies of the first eight excited states. The dramatic change in the ordering and energies of the states of CO, with geometry can be seen in the diagram given by England et Fig. 6 shows the diagram connecting the states of C + 02, CO + 0 and CO, up to the highest state calculated by England et aZ.38-40 Above this, states are correlated for reactants and products at infinite separation. Whilst there are extensive spectroscopic data on various high lying linear states of C02,31,37 fig. 6 employs correlations between infinitely separated reactants and products above CO,(B'A,), as a correlation diagram must employ a complete state manifold.Fig. 6 is self-explanatory in accounting for the rapid removal of C(23PJ) by 0, (table 1) where there are exothermically favourable correla- tions through states of CO, to products. Furthermore, C(23PJ) + 0, correlate with CO(XIZ+) + 0(2102) through the energetically most favourably lying states of C02 correlating linear and bent states of CO,.D . HUSAIN AND P . E . NORRIS 28 1 CHz(X 381 ) FIG. 5.-Correlation diagram connecting the states of C + Hz and CH + H via those of CH2.252 ELECTRONICALLY EXCITED c ATOMS FIG. 6.-Correlation diagram connecting the states of C + O2 and CO + 0 via those of COz. (X'Z;, alA1). This may be considered to support the observations of Thrush and coworkers 41 who concluded from infrared chemiluminescence measurements on C + 0, in a flow system that the overall route to CO(X'C+) + O(2lD2) AH = -4.02 eV)8930942 was preferred to the more exothermic route to CO(X'C+) + O(z3PJ) (AH = -6.00 eV).8930*42 Unfortunately, the states C(2'D2) + O2 and C(2lSO) + 0, lie above the energy below which correlations through states of CO, can be made.There are clearly symmetry-allowed pathways to infinitely separated products which are exothermic for these pairs of reactants (fig. 6) and the relatively rapid removal of both C(21D2) and C(2lSO) by 0, (table 1) is in accord with this. The collisional behaviour of C(23PJ, 2'D2, 2'S0) with ethylene is considered in terms of a semi-empirical potential energy diagram for CH,-CH, constructed in C,, symmetry.This particular symmetry envisages the course of reaction as the collision of the carbon atom with C2H4 along the perpendicular bisector of the C-C axis (the z-axis). The collisional interaction is approximately measured by the extent of the resulting P, electron occupancy. (We are indebted to Dr. A. B. Callear for suggesting this approach.) This reaction geometry is consistent with the experiments of Wolfgang et aZ.43 who showed that the collision of "C with C2H4 led to centre- labelled allene. The atomic wave functions for C(23P), C(2'D2) and C(2'S0) were constructed in the L, ML, S, Ms basis following Condon and S h ~ r t l e y ~ ~ using the standard Clebsch-Gordon coefficient^.^^ For C2, symmetry, linear combinations of these functions must be taken so as to reproduce the character table under the actionD. HUSAIN A N D P .E . NORRIS 283 of the symmetry operators of the CZv point group. This procedure is relatively standard and is not reproduced here. Fig. 7 shows the resulting potential energy diagram with the orbitals, together with the appropriate P, electron occupancies in parentheses. Before embarking on a discussion of the collisional behaviour of atomic carbon with ethylene using this diagram, some general comments on the diagram are appropriate. 9c H,C - CH, FIG. 7.-Schematic presentation of the potential energy diagram for CH2-CH2 in C2- symmetry. (a) W 3 P J + C2H4(-f1Ag), (6) C(2lD2) + C2H4(-f1Ag), (c) C(~'SO) + C2H4@'Ag)* In general, the initial interaction between the n-bonding system and the P, orbital is assumed to depend on the electron occupancy of the latter.If this interaction is extended to smaller values of z (fig. 7), the ordering of the resulting minima is such that these are in accord with a Walsh diagram constructed on the basis of the united atom approximation for CH2-C-CH2 (uiz. C02). The relative positions of the three lowest minima in fig. 7 ('Al, 3B1 + 3B2, 'B1 + 'Bz) immediately follow from the P, electron occupancy concept. The absolute positions of the potential minima of the 'A1(2/3) (C'D + C2H4) and [3B1(l/2) + 3B2(l/2)] (C3P + CzH4) states are taken from the CNDO calculations of Field.46 The small minima ascribed to the surfaces labelled lA2(0) and 3A2(0) with zero P, electron occupancy, are arbitrary but consistent within the united atom approximation using C02 as an analogy.The minimum in the zero P, occupancy 'A,(O) orbital is presumed to result from repulsion with the 'A1(1/3) orbital which arises from the states labelled (c) in fig. 7. Given the attractive interactions as presented in fig. 7, the exact positions of the surface crossings do not significantly affect discussion of the collisional behaviour of atomic carbon. Within the context of fig. 7, the collisional behaviour of C + C2H4 becomes, for the most part, self-explanatory. The rapid removal of C(2'SJ by C2H4 (table 1) is clearly in accord with the potential energy diagram. Mixing of the ' A , surfaces arising from states (c) and (b) will be highly effective in the event of geometric de- formation and the subsequent loss of CZv symmetry.Alternatively, if C2v symmetry284 ELECTRONICALLY EXCITED c ATOMS is maintained, the non-adiabatic transition between the ‘ A , surfaces is clearly facilitated by surface crossing at favourable energies. Clearly, for C(21D2) + C2H4, there are a number of favourable removal channels for the excited atom (fig. 7), particularly the one leading to a linear product. The rapid removal rate for C(2’D2) (table 1) is fully consistent with the potential energy diagram. The rate data for C(z3PJ) + C2H4 (table 1) reported by Wolf et d . * l are clear!y at variance with the potential energy diagram (fig. 7) and, indeed, the rate is extremely slow by comparison with rates for a large body of atomic reactions with this unsaturated molecule.’ For example, the reaction between Si(33PJ) + C,H4 proceeds at a rate essentially that approaching the collision number (table 2).In our opinion, the preliminary plasmolysis experiments of Wolf et ~ 1 . ~ ’ require more detailed investigation. A detailed investigation in the vacuum ultraviolet of the reaction rate between C(23PJ) + C2H4 in the time-resolved mode would further, in our opinion, be justified. We thank the S.R.C. and the Pye Unicam Company for a CASE research student- ship awarded to one of us (P.E.N.) during the tenure of which this work was carried out. We also thank the S.R.C. for an equipment grant. D. Husain, Ber. Bunseriges. phys. Chem., 1977, 81, 168. R. J. Donovan and D. Husain, Chem. Rev., 1970, 70, 489.D. Husain and P. E. Norris, J.C.S. Faraday IZ, 1978, 74, 93. D. Husain and P. E. Norris, J.C.S. Faraday ZI, 1978, 74, 106. D. Husain and P. E. Norris, Chem. Phys. Letters, 1978, 53, 474. D. Husain and P. E. Norris, J.C.S. Faraday ZI, 1978, 74, 1483. C. E. Moore, ed., Nat. Bur. Stand. Circular 467, Atotnic Energy Lecels (U.S. Government Printing Office, Washington D.C., 1958), vol. 1-111. D. Husain and P. E. Norris, Chem. Phys. Letters, 1977, 51, 206. P. M. Swearingen, S. J. Davis and T. M. Niemczyk, Chem. Phys. Letters, 1978, 55, 274. R. A. Armstrong and S. J. Davis, Chem. Phys. Letters, 1978, 57,446. I3 W. Braun, A. M. Bass, D. D. Davis and J. D. Simmons, Proc. Roy. SOC. A , 1969,312,412. l4 D. Husain and L. J. Kirsch, Chem. Phys. Letters, 1971, 8, 543.D. Husain and L. J. Kirsch, Trans. Faraday SOC., 1971, 67, 2025. l6 D. Husain and A. N. Young, J.C.S. Faraday II, 1975, 71, 525. D. Husain and L. J. Kirsch, Cheni. Phys. Letters, 1971, 9, 412. D. Husain and L. J. Kirsch, Trans Faraday SOC., 1971, 67, 2886. l9 D. Husain and L. J. Kirsch, Trans. Faraday SOC., 1971, 67, 3166. ’O D. Husain and L. J. Kirsch, J. Photochem., 1973/74, 2, 297. ’’ G. M. Meaburn and D. Perner, Nature, 1968, 212, 1042. 23 C. H. Corliss and W. R. Bozmann, Experimental Transition Probabilities for Spectral Lines of Seuenty Elements, Nat. Bur. Stand. Monograph no. 53, (U.S. Government Printing Office, Washington D.C., 1962). ’ K. E. Shuler, J. Chem. Phys., 1953, 21, 624. lo D. Husain and P. E. Norris, J.C.S. Faraday II, 1978, 74, 335.F. F. Martinotti, M. J. Welch and A. P. Wolf, Chem. Comm., 1968, 15. 24 D. Husain and P. E. Norris, J.C.S. Faraday IZ, 1977, 73, 415. 25 W. H. Wing and T. M. Sanders Jr, Rev. Sci. Znstr., 1967, 38, 1341. 26 A. Savitsky and J. E. Golay, Analyt. Chem., 1964, 36, 1627. ’’ I. S. Fletcher and D. Husain, J.C.S. Faraday IZ, 1978, 74, 203. ’* I. S. Fletcher and D. Husain, J. Phys. Chem., 1976, 80, 1837. 2 9 R. J. Donovan, D. Husain and L. J. Kirsch, Trans. Faraday SOC., 1970, 66, 2551. 30 B. Rosen, Spectroscopic Data Relative to Diatomic Molecules (Pergamon, Oxford, 1970). 31 G. Herzberg, Electronic Spectra of Polyatomic Molecules (Van Nostrand, New York, 1966). 32 V. I. Vedeneyev, L. V. Gurvich, V. N. Kondratiev, V. A. Medvedev and Ye. L. Frankevich, 33 J. Danon, S. V. Filseth, D. Feldmann, H. Zacharias, C. H. Dugan and K. E. Welge, Chem. 34 L. B. Harding and W. A. Goddard 111, J. Chem. Phys., 1977,67, 1777. 35 A. Kasdan, E. Herbst and W. C. Lineberger, Chem. Phys. Letters, 1975,31, 78. Bond Energies, Ionisation Potentials and Electron Afinities (Nauka, Moscow, 1970). Phys., 1978, 29, 345.D . HUSAIN AND P . E . NORRIS 285 36 B. Wirsam, Cheni. Phys. Letters, 1972, 14, 214. 37 M. J. Hubin-Franskin and J. E. Collin, J . Electron Spectr., 1975, 7, 139. 38 W. B. England, B, J . Rosenberg, P. J. Fortune and A. C. Wahl, J . Cheni. Phys., 1976, 65, 684. 39 W. B. England, W. C. Ermler and A. C. Wahl, J . Chem. Phys., 1977, 66, 2336. 40 W. B. England, D. Yeager and A. C. Wahl, J . Chetri. Phys., 1977, 66, 2344. 41 E. A. Ogryzlo, J. P. Reilly and B. A. Thrush, Chem. Phys. Letters, 1973, 23, 37. 42 A. G. Caydon, Dissocintion Energies and Spectra of Diatoniic Afolecrrles (Chapman and Hall, 43 M. Marshall, C. Mackay and R. Wolfgang, J . Atner. Cheni. SOC., 1964, 86, 4741. 44 E. U. Condon and G. H. Shortley, The Theory of Atotriic Spectra (Cambridge University Press, 4s P. W. Atkins, Molecular Qiiaiitirtii Mechanics (Clarendon, Oxford, 1970), parts I and 11, p. 184. 46 D. E. Field, personal communication in paper by D. Husain and L. J. Kirsch, Trans. Faraduy London, 1968). London, 1963). Soc., 1971, 67, 3166.

ISSN:0301-7249    

DOI:10.1039/DC9796700273

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

Reactions of O(2'0,) and O(2"P,) with Halogenomethanes BY MICHAEL C. ADDISON, ROBERT J. DONOVAN AND JOHN GARRAWAY Department of Chemistry, University of Edinburgh, West Mains Road, Edinburgh EH9 355 Received 20th December, 1978 Product branching ratios for the reaction of 0(2'D2) with the halogenomethanes CF3CI, CF3Br, CF31 and CFzHCl are presented. The dominant channel is shown to be abstraction yielding a halogen oxide. This contrasts with the behaviour observed with hydrocarbons, where insertion into C-H bonds dominates. Quenching of 0(2*D2) to the ground state is also observed with the halo- genomethanes and accounts for "N 30 % of the total removal cross-section. Reaction of 0(2'D2) with CF2HCl leads to the formation of CIO (55 %) and to the elimination of HCl (40 %).The latter process is accompanied by the formation of CF2 and OG13PJ). The reactions of 0(2'D2) are compared with those for O(23PJ), where these are known, and the absolute rate for reaction of O(23PJ) with CFJ is determined as (1.1 & 0.3) x lo-" cm3 molecule-' s-l at 300 K. The results are discussed in terms of the main topological features on the potential surfaces involved. Reactions of O(2l D2) with hydrocarbons have been studied e~tensively.l-~ The reaction cross-sections are large and the main reaction channel involves insertion into C-H bonds. Insertion has been shown to proceed indiscriminately and the total reaction cross-section found to be proportional to the number of C-H bonds in the m~lecule.~ A number of other reaction channels have also been recognised and may be summarised as follows, 0(2ID2) + R H - + ROHI ?-+ ROH (65 %) (1) (2) (3) (4) + R + OH (20-30 %) -+ R'O + H2( <lo %) -+ RH + 0(23~,)(<3 %).It is clear that quenching is negligible and that a direct abstraction reaction, leading to OH formation, plays an appreciable role. By comparison the reactions of 0(2lD2) with halogen-containing molecules have been little studied, although it is known that the reaction cross-sections are again The formation of halogen oxide products has been observed and lower limits for branching into this channel In the present work we have made a detailed study of the branching ratios into different reaction channels for a number of halogen-containing molecules. The domi- nant channel is shown to be abstraction of a halogen atom.Quenching to the ground state is also an important process. We also present data for the reaction of O(23PJ) with CFJ and compare these, together with data for the other halogenomethanes, with those for the analogous re- actions involving O(2lD2).M. C. ADDISON, R . J . DONOVAN AND J . GARRAWAY 287 EXPERIMENTAL Three separate experimental arrangements were employed for this work, all of them based on the flash photolysis technique. (i) FLASH SPECTROSCOPY A conventional arrangement, suitable for photographing transient spectra in the visible and ultraviolet regions, was used to obtain kinetic data on the halogen oxides and CF2. Spectra were dispersed on a Hilger-Watts medium quartz spectrograph and recorded on Kodak Panchro-Royal film. A more detailed description of this technique and the data processing has been given in ref.(6) and (7). (ii) TIME-RESOLVED PHOTOMETRY I N THE VACUUM ULTRAVIOLET This apparatus employed a conventional flash photolysis unit coupled to a vacuum ultra- violet monochromator and fast photometric recording system. It was used to monitor the formation and decay of O(23P.,) (via the resonance lines at A = 130 nm), following quenching of O(2lD2) by the halogenomethanes, and also to obtain absolute rate data for reaction of OQ3P,) with CF31. for work on S(33PJ); however, for the present work an EMR542 solar blind photomultiplier was used. The use of this photomultiplier eliminated the effect of scattered light from the flash lamp and allowed kinetic measurements to commence during the flash.A flow system was used for the atomic lamp and the best results were obtained when very low (<O.l %) oxygen/helium ratios were passed through the microwave discharge. An extensive series of experiments was carried out to establish that this new arrangement gave a linear photo- metric response with stable molecules such as 03. Curves of growth for O(23PJ) were then determined by photolysing O3 under optically thin conditions [in the presence of excess Nz to quench O(2lD2) to O(23PJ)] over a range of pressures (fig. 1). As a final check the rate of the reaction between O(z3PJ) and NO2 was determined * as k == (1.1 C 0.3) x lo-" cm3 molecule-' s-', in excellent agreement with the accepted result obtained by resonance fluorescence' [k -- (9.12 The experimental arrangement was similar to one described previously 0.44) x 10-l' cm3 molecule -'s-' at 295 K]. 0.4 O s 5 1 0.3 0.2 0.1 0 1 2 3 6 7 0 ( 23PJ ) / a r b .units FIG. 1.-Curve of growth for O(z3PJ) using the three resonance lines at 130.2, 130.5, 130.6 nm. The O(23PJ) concentration was taken to be proportional to that of 03, which was varied over the range 0.4-6.1 N m-'. * In these experiments O(23PJ) was formed by photolysis of NOz ( x 1 %) in the visible and near U.V. regions.In all of these experiments the output from the photomultiplier was fed to a fast analogue- to-digital converter (Datalab DL905) and data were processed in the standard way. (iii) The yield of OH from reaction of 0(2'D2) with CF2HCI was determined using an arrange- ment similar to that described by Morley and Smith." The intense OH emission produced by a microwave discharge through a flowing mixture of water vapour in argon carrier gas was focused through the reaction vessel and onto the slit of a McKee-Pederson (MP1018B) monochromator which selected lo the Q13 line at 308.15 nm.A chlorine gas filter surrounded the reaction vessel and reduced scattered light from the flash to a negligible level. The output from the photomultiplier was fed to a transient recorder (Datalab DL905) and data were pro- cessed as in section (ii) above. For all experiments 0(2'D2) was produced by the ultraviolet photolysis of 03(1 = 200- 300 nm) and, where required, O(23P.,) was formed by adding an excess of N2, to quench O(2lD2) to O(23P,). The experimental conditions used with the three different techniques varied significantly and will be described in the appropriate section dealing with results.TIME-RESOLVED PHOTOMETRY I N THE NEAR-ULTRAVIOLET RESULTS ABSOLUTE CONCENTRAT IONS OF o(2'02) PRODUCED BY THE FLASH Photolysis of 0, in the ultraviolet (200-300 nm) is known to produce almost exclusively O(2'0,) and thus the absolute yield of this atomic state can be determined by observing the amount of O3 removed by the flash. In pure 0, (or 0, + SF, and O3 + He mixtures) O(2l0,) reacts rapidly with a second O3 molecule, and under our conditions the amount of 0, removed immediately after the flash ?ill in fact be twice the amount photolysed in the primary photochemical step. However, by adding excess CO, to the ozone, the effect of the secondary reaction can be eliminated, as O(2'0,) is quenched to the ground state.We therefore carried out experiments to determine the amount of O3 removed after the flash (30 ,us) both in the presence and absence of CO,. The depletion in the presence of CO, was found to be 12 & 1 % (Po, = 26.6 N m-,) and in the absence of CO, 24 & 2 %, this gives a yield of O(2'0,) of 8 x loL4 atoms cm-3 per flash. These results confirm that the yield of O(23PJ) in the ultraviolet photolysis of 0, is negligible (< 10 %) and that O(2'0,) is removed en- tirely by reaction with 03, physical quenching being unimportant. The decay of 0, at times > 30 p s was observed to be very slow, as expected from the known slow rates for reactions involving O(23PJ) and O2(dAg) with 0,.REACTION OF O(2'0,) W I T H CF3C1 When O3 is photolysed in the presence of excess CF3Cl (PCFICI = 2.7 kN m-? a strong spectrum of C10 is observed via the (A2n t- X 2 n ) system, and its rate of formation closely follows the integrated form of the flash. No C10 is observed when CO, or N, is added to quench O(2'0,) and it is clear from these, as well as earlier experiment^,^*' that C10 results from a fast reaction between O(2l0,) and CF3C1. There are, however, three possible mechanisms for C10 formation. The first and most obvious is the direct formation of C10 in a primary abstraction step O(2'0,) + CF3C1+ CIO + CF3. ( 5 ) A second possibility is insertion into the C-C1 bond followed by fragmentation to yield C10 0(2lD,) + CF3Cl -+ CF30Cll -+ CF3 + C10.(6)M . C . ADDISON, R . J . DONOVAN AND J . GARRAWAY 289 The third possibility is that C1 atoms are produced by a displacement reaction, fol- lowed by the fast reaction of C1 with 03, i.e., O(2'0,) + CF3C1 -+ CF30 + C1 (7) (8) C1+ 0,- C10 + 0,. Under the conditions of our experiment it would be difficult to distinguish between these three mechanisms simply by observing the rate of formation of C10 as they are all very rapid. We can, however, use a chemical method to distinguish between the first two and the third mechanisms. By adding a small amount of ethane to mixtures of 0, and CF,Cl (PCZHs = 67 N mW2), any C1 atoms formed in the primary step can be removed before reacting with 0,; the yield of C10 will then be reduced by an amount which depends on the yield of free chlorine atoms in the primary step.As the pressure of C2H6 used is very much lower than that of CF,CI it will not interfere by reacting with O(2l0,) (>95 74 of the excited oxygen atoms react with CF,Cl). Our results show that the yield of CIO is only slightly reduced by addition of C2H6 and C1 atom formation accounts for <20 % of the total O(2'0,) removal by CF,Cl. Quantitative yields of C10 were determined oia the (5, 0) band of the A211 +- X211 system" [see ref. (7) for a detailed discussion], and when these are compared with the amount of O(2'0,) produced by the flash we find that 65 % of the excited oxygen gives rise to C10 formation in a primary step. Measuring the yield of OQ3PJ) by time resolved photometry in the vacuum ultra- violet proved more difficult than first envisaged.Absorption by CF3Cl reduced the intensity of the oxygen resonance line reaching the photomultiplier (the 130.6 nm line was used), as expected; however, we also observed a change in the sensitivity with which OQ3PJ) could be detected : as the extent of absorption by CF3Cl increased, the sensitivity for detecting O(23PJ) decreased. Thus in order to measure the yield of OQ3PJ) quantitatively, new curves of growth were determined over a range of condi- tions under which the oxygen resonance lines were attenuated by an absorbing gas such as CF,CI. For the present work relative O(z3PJ) concentrations were read directly from the appropriate curve of growth and the yield of O(23PJ) formed by the quenching of 0(2102) by CF3C1 was determined by comparing the concentrations produced in the absence and presence of excess N2 [the N, quenches a large and calculable fraction of the 0(2'D2) directly to the ground state].By this means the branching ratio for O(23PJ) formation was determined as 30 2:; %. Typical conditions in these ex- periments were Po, = 0.5 N m-2, PCFBCl = 4.0 N m-2, with a flash energy of 180 J. The branching ratios for all the channels determined in this work are summarised in table 1. We also include a recent estimate 12a for the elimination channel 12b O(2'0,) + CF3Cl--+ CF20 + FCl (9) which is seen to have a small branching ratio. REACTION OF 0(2'D2) W I T H CF3Br AND CF31 Reaction of 0(2'D2) with CF,Br results in the rapid formation of BrO; however the decay is also rapid and this makes the absolute determination of the BrO yield extremely diffic~lt.~ Nevertheless we can obtain a useful lower limit for the yield of BrO and using the extinction coefficient given by Clyne et al.13 for the (4,O) band of the A211-XZll system we find a branching ratio for BrO formation of >25 %. It should be noted that reaction between Br and O3 is much slower than the corre-290 REACTIONS OF O(2'0,) AND 0(33P,) WITH HALOGENOMETHANES sponding reaction for C1 atoms, and that we can therefore distinguish between BrO formed in a primary step and that formed by secondary reaction of Br with 03.Experiments to determine the yield of I 0 from reaction of O(2'0,) with CF31 are considerably more difficult than the corresponding experiments with CF3Cl and CF3Br. A strong spectrum of I 0 is observed; however, some photolysis of CF31 occurs, (it absorbs in the same region as 0,) and it is known that iodine atoms react rapidly with 0, to yield 10.Preliminary results from our laboratory show that spin-orbit excited TABLE BRANCHING RATIOS FOR PRODUCT CHANNELS IN THE REMOVAL OF 0(2'D2) BY HALOGENOMETHANES ~~ ~ ~ reactant products/% quenching halogen halogen other to 0(23~.,) oxide atom products CF3Cl 30 2 % 65 i 10% <20 % FCl( z 10 %) CF2HCl 28 ? ::%* 55 10% <lo% OH(5%) CF3Br >25 % ~~~ * Yield of OQ3PJ) based on CF, formation (i.e., uia the dissociative excitation channel yielding CF2 + HCl + 0) is 45 f 10 %. iodine atoms, I(S2Pli2), react more rapidly with 0, than ground state I(52P3/2) atoms; the rate constant for the ground state iodine atom reaction has been determined14 as k = 0.8 x Thus I 0 can be formed by more than one reaction and detailed experiments are required to distinguish between the various possibilities.Our present results indicate that the yield of I 0 from reaction of O(2'0,) with CF31 is substantial and we hope to report a quantitative measure for the branching ratio into this channel at the Discussion. cm3 molecule-' s-l. REACTION OF O(2'0,) WITH CF2HCl Strong transient spectra of C10 and CF2 were observed following the photolysis of 0, or N20 in the presence of CF2HCl (Po, = 13.3 N m-2; PCFzHC1 = 2.7 kN m-,). Both spectra were completely suppressed by addition of excess N2, showing that they resulted from reaction of O(2'0,) with CF,HCl.Photolysis of CF,HC1 in the far- ultraviolet is known to produce CF,; however, this region is not transmitted by our equipment and CF, was not observed when CF,HCl (PCFaHC, = 2.7 kN m-2) alone was flashed in the reaction vessel. The yield of C10 was measured as described for CF,Cl and found to be 55 & 10 % of the initial 0(2'D2) yield. Addition of small amounts of ethane to the system had no significant effect on the C10 yield, showing that C1 atom formation is of little importance (< 10 %) in the removal of O(2'0,) by CF2HC1. The yield of CF, was determined using the known extinction coefficient for the v; = 6 band (249 nm) of the A'B, +- X'A, system, given by Tyerman." The branching ratio into this channel was determined as 45 zt 10 %. The branching ratio for O(2,P,) formation was measured using the method de- scribed above for CF3C1 and found to be 28 Z i2 %, which suggests that both CF2 and O(23PJ) must be formed in the same process. The formation of OH radicals was not observed using plate photometry, however we would expect OH to react rapidly with CF,HCl under the conditions employed. Using the more sensitive technique of time resolved spectrophotometry at 308 nmM .C . ADDISON, R . J . DONOVAN AND J . GARRAWAY 29 1 (PCFIHCI = 2.0 kN m-2, POa = 40 N m-2) formation of OH was detected but in very low yield. A careful calibration of the system was achieved using the reactions of O(2'0,) with H20 and CH,. Assuming that H20 gives two OH radicals for each O(2'DJ atom reacting, the yield of OH from CH, was found to be 80 %, in good agreement with previous work.I6 The yield of OH from CF2HCI, based on the same method, was found to be only 5 %.REACTION OF O(2,PJ) WITH CF31 The kinetics of O(2,PJ) removal by CF,I were investigated using time-resolved spectrophotometry at 130 nm (the slit width used was 800 pn and thus the three atomic lines at 130.2, 130.5 and 130.6 nm were transmitted by the monochromator). By photolysing 03(P0, = 1.33 N m-2) in the presence of excess N2(PN2 = 800 N m-2), suitable concentrations of O(23PJ) could be generated ( ~ 3 % photolysis of 0, oc- curred). The decay of the ground state oxygen atom under these conditions was found to be very slow, as expected. Addition of small partial pressures of CF31 (0.13-0.6 N m-2) resulted in a marked increase in the rate of decay and by measuring the pseudo first-order rate coefficients for removal of O(2,PJ) over a range of CF,I pressures (data are shown in fig.2) the second-order rate constant was determined as, ko(23p,) + CFsI = (1.1 & 0.3) x cm3 molecule-' s-'. I I I I I I I 0 0.1 0.2 0.3 0.4 0.5 0.6 PCF3, / N m-2 FIG. 2.-Plot of the first-order rate coefficients for removal of O(23PJ) against partial pressure of CF31. (Pas = 1.3 N m-'; PNa = 800 N m-'). A small correction'' was made for departure from Beer-Lambert behaviour and the slope of fig. 2 should be multiplied by 1.3 ( y = 0.76 based on the data in fig. 1) to obtain the rate constant given above. Some photolysis of CF31 will inevitably occur under the conditions used; however, the percentage photolysis will be much less than that for 0, (ie., <3%), due to the lower extinction coefficient for CF,I, and should have no effect on the kinetics of the oxygen atom decay.As a check, further experiments were carried out over a range of flash energies (180-320 J). No significant difference in the decay rate for O(2,PJ)could be detected and we conclude that radical-radical reactions do not influence the observed kinetics and that photolysis of CF31 is unimportant. Some slow regeneration of O(z3PJ) will occur via the reaction of 02(a1Ag) with 03, but this is entirely negligible on the time scale used here. DISCUSSION REACTION OF o(2'02) WITH CF3C1, CF,Br A N D CF31 A major channel in the reactions of 0(2'D2) with halogenomethanes (excepting attack on C-F bonds)," is clearly the formation of a halogen oxide molecule. We shall concentrate our discussion on the reaction with CF3Cl, as the data for this molecule are most complete, but we expect the same general points to apply for CF3Br and CF31.Formation of C10 from CF3Cl can in principle occur by two mechanisms, the more direct being abstraction of a chlorine atom. The second possible mechanism involves insertion of 0(2ID2) into the C-Cl bond, to form a vibrationally excited hypochlorite CF30Cl '+, followed by fragmentation. CF30C1 is a stable molecular species and its thermal and photochemical reactions have been examined. The results suggest that the favoured primary dissociation channel is formation of CF30 and C1 (thermochemically this is the most favourable dissociation process).Thus if insertion of 0(2'D2) into C-Cl bonds was important, we would expect a high yield of Cl and not C10, contrary to observations. Our results therefore suggest that C10 formation occurs by a direct abstraction mechanism. Similar behaviour has been reported previously for reactions of singlet methylene (CH,), which is isoelectronic with O(2lD2), with ha loge no me thane^.^^-^^ Thus, while both singlet methylene and 0(2'D2) undergo fast insertion reactions into C-H bonds, the main reaction channel with halogenomethanes involves direct a b s t r a c t i ~ n . ' ~ - ~ ~ The above behaviour can be understood when we consider the strong interaction that will occur between the vacant p-orbital of 0(2'D2) (or CH,) and the lone pairs on the halogen atom.Thus the potential surface contains an attractive basin which surrounds the halogen atom and facilitates attack at this point in the molecule. A further attractive region must exist on the potential surface, corresponding to insertion of O(2'0,) into the C-C1 bond (the minimum corresponding to the ground state configuration for CF30Cl); however, it appears that this region is less accessible, possibly due to inertial effects; both Cl and CF3 are relatively heavy and need to move a substantial distance for insertion to occur (contrast this with the situation for C-H insertion where the much lighter H atom can move rapidly to accommodate the insertion process). Our data also provide information on another aspect of the singlet potential surface discussed above.Thus the singlet surface must be sufficiently attractive to be crossed by one or more triplet surfaces correlating with O(23PJ) + CF3C1 and non- adiabatic transitions at these crossings must be favourable, as evidenced by the rela- tively high branching ratio for O(23PJ) formation. For O(2'0,) interacting with CFJ the singlet surface may pass below the asymp- tote for O(23PJ) + CF31 (fig. 3) and could therefore influence the dynamics of the reaction between O(23PJ) with CF,I (see below). Stable compounds with the structure RIO can be prepared (e.g., iodosobenzene, C6H510) showing that the singlet surface has a very deep minimum in the region occupied by the lone pair electrons of iodine. * Removal of 0(2'D2) is much slower by CF, groups4*' and appears to proceed entirely by quench- ing.I8M .C . ADDISON, R . J . DONOVAN AND J . GARRAWAY 293 0(210,) + R I O ( ~ ~ P J -L t RI /- - - - R + I O FIG. 3.-Section through the proposed potential surfaces for O(Z3PJ) and 0(2'D2) interacting with an iodide. The lowest singlet surface is shown by the continuous line and the triplet surfaces by dashed lines. R I O REACTION OF 0(2'D2) WITH CF2HCI Lin23 has studied the photolysis of 0, in the presence of a number of hydrogen containing halogenomethanes, including CF,HCI, and observed stimulated emission from vibrationally excited hydrogen halide molecules formed in these systems. He proposed that this resulted from the insertion of 0(2'D2) into C-H bonds followed by the elimination of a vibrationally excited hydrogen halide molecule from the hot intermediate, e.g., 0(2'D2) + CF2HCl -+ CF,ClOHI --+ CFClO + HFI.(10) With CF2HCl, only HF emission was observed, although the formation of HCl is more exothermic. The present results clearly show that HF elimination cannot account for more than 10-20 %* of the total reaction cross-section and that elimina- tion of ground state HC1 is a more important process. It seems unlikely that chemical laser emission would result from a minor reaction channel and an alternative explanation for Lin's result is that excited HF is produced by secondary radical reactions. In a separate series of studies, Lin24 has suggested that the reaction, O(23PJ) + CF,H+ CFO + HF (AH = -433 kJ mo1-I) can give rise to H F laser emission. Our results show that both O(z3PJ) and CF,H are major products of the interaction of O(2'0,) with CF2HCl and we therefore sug- gest that reaction (11) could account for Lin's observations in the O3 + CF,HCl photochemical laser system.,, The dominant channel in the interaction of 0(2'D2) with CF2HC1 is clearly that leading to the formation of ClO(55 %) and, as the branching ratio is similar to that for CF,CI, we infer that the mechanism is the same.The second most important channel involves dissociative excitation viz., 0(2'D2) + CF,HCl+ CF, + HC1 + O(23PJ). (12) This channel is thermoneutral within the bounds of current thermodynamic data (AH = 17 -& 19 kJ mol-') and it is surprising that it competes so effectively with the other highly exothermic channels. However, the observed rapid formation of CF, and O(2,PJ) cannot be accounted for by any other process.We have shown that C1 atom formation is unimportant ( e l 0 % of the total cross-section) which rules out reactions such as O(2'0,) + CF2HC1+ CF2 + OH + C1. (13) * This is an upper limit based on the error bounds for the products which are directly observed.This is further confirmed by the very low yield of OH(5 %) observed. to be formed by the disproportionation reaction CF, is known 2CF,H+ CF, + CF,H2 (14) however, this could only account at most for 10 of the CF, observed as the dominant removal channel for two CF,H radicals is dimerisation. We, therefore, conclude that dissociative excitation [reaction (12)] accounts for z 40 % of the total cross-section. It is interesting to note that both the thermal and infrared multiphoton d i s s ~ c i a t i o n ~ ~ of CF,HCI lead to the formation of CF, and HCI.The other surprising feature is that OQ3P,) escapes from the force field of CF,, as CF,O is a very strongly bound molecule. This can, however, be understood when it is realised that CF2(X1A1) and O(23P,) do not correlate directly with the ground state of CF,O, but with an excited triplet state which may not allow efficient combination. From the bond additivity relationships suggested by Cvetanovic et al.3 and by Davidson et al.,5 we would expect OH formation to account for z 30 % of the total cross-section. This is clearly not the case and it appears that the distribution in the product channels does not follow the simple additivity relationship suggested for the total removal rates. The low yield of OH is in fact similar to the situation previously encountered with singlet methylene reactions, where it was found that attack at C-H bonds was reduced to a very low level when a chlorine atom is present on the same, or adjacent, carbon at om.26 The branching ratio for OH formation ( 5 "/o) is surprisingly low. REACTION OF O(23P,) WITH CF,Br AND CF31 Reaction of O(23PJ) with CF,Br, to yield BrO, is strongly endothermic (AH = +65 & 5 kJ mol-') and negligibly slow at 300 K. However, the reaction has been studied at elevated temperatures (800- 1200 K) and Arrhenius parameters determined 27 as A = (1.5 -& 0.5) x lo-" cm3 molecule s-' and E, = 57 4 kJ mo1-'. The activation energy for reaction is thus close to the endothermicity and the pre-expo- nential factor ( A ) is low when compared with reactions involving O(2'0,).We shall return to the latter point after discussing the corresponding reaction with CF,T. The reaction of OQ3PJ) with CF31 has been studied in some detail by Gorry et aL2* using the molecular beam technique and has been shown to involve the formation of a weakly bound collision complex. The product scattering (10) changes from a mainly backward, to a near isotropic distribution as the kinetic energy of the incident OQ3PJ) is increased. It was suggested" that at low collision energies the lifetime of the complex is shorter than its rotational period (as it is probably formed in low impact parameter collisions with low angular momentum).At higher collision energies the rotational period is reduced (higher angular momentum) and this leads to an increase in the forward scattering. The total cross-section for reaction was not determined in the molecular beam work but a thermally averaged (300 K) cross-section can be obtained from the present data as CJ 21 2 A'. It is clear that the reaction must be close to thermoneutral and our results provide an upper limit for the activation energy of E, < 6 kJ mol-I. When this is combined with the bond strength of CF31,29 D(CF,-I) = 221 +- 5 kJ mol-', we obtain a lower limit for the bond strength of I 0 as, D,(IO) 3 210 kJ mol-', which is consistent with the value given earlier by Radlein et al.30 We might expect the Arrhenius pre-exponential factor for reaction of O(23PJ) with CF31 to be similar to that for the analogous reaction with CF,Br, and the fact that the rate constant (at 300 K) for O(23P,) + CF,I is close to the pre-exponential factor forM .C . ADDISON, R . J . DONOVAN AND J . GARRAWAY 295 OQ3P,) + CF,Br, suggests that this is probably the case. These values are sur- prisingly low when compared with the analogous reactions for 0(2'D2) (where any kinematic constraints should be the same), but appear to be characteristic of reactions involving O(23P,) with halogen, or halogen-containing molecules. As these reactions involve attractive potential surfaces and a bound collision complex we would normally expect a substantial reaction cross-section or large pre-exponential factor. It has been suggested that the low values observed result from a very restrictive reaction geometry and that a near collinear collision is required before reaction can This was rationalised in terms of the molecular orbital structure for the collision inter- mediate which favours a linear 0-X-Y structure for lowest energy on the triplet potential surface.However, the above discussion on the quenching of O(2lD2) by halogenomethanes leads us to suggest an alternative explanation. We have seen that crossings between triplet and singlet surfaces must occur and that for iodoso com- pounds one of these may be close to the dissociation asymptote for O(23P,) +- RI (fig. 3). Thus the low reaction cross-section could result from a " low " triplet- singlet transition probability, while the scattering dynamics would be determined by the potential minimum in the singlet surface. CONCLUSIONS Reactions of O(2lD2) with halogenomethanes proceed with a large total cross- section, the dominant channel being abstraction to yield a halogen oxide.The singlet potential surface, on which these reactions occur, is strongly attractive and is crossed by lower lying triplet surfaces correlating to O(23PJ). This provides an efficient mechanism by which 0(21D2) is quenched to the ground state. The reactions of 0(2'D2) closely parallel those of singlet methylene. Reactions of O(z3PJ) with halogenomethanes have relatively low total cross-sections (and Arrhenius pre-exponential factors) and may involve a triplet-singlet surface crossing. We thank Drs H. Gillespie and G.Black for help in initiating this work and I.C.I. Ltd for the gift of samples of CF2HCl. H. Yamazaki and R. J. Cvetanovic, J. Chem. Phys., 1964, 41, 3703. A. J. Colussi and R. J. Cvetanovic, J. Phys. Chem., 1975, 79, 1891. P. Michaud, G. Paraskevopoulos and R. J. Cvetanovic, J . Phys. Chenz., 1974, 78, 1457. I. S. Fletcher and D. Husain, J. Phys. Chern., 1976, 80, 1837. J. A. Davidson and H. I. Schiff, J. Chem. Phys., 1978, 69, 4277. H. M. Gillespie and R. J. Donovan, Chem. Phys. Letters, 1976, 37, 468. H. M. Gillespie, J. Garraway and R. J. Donovan, J . Photochetn., 1977, 7, 29. R. J. Donovan and D. J. Little, Chern. Phys. Letters, 1972, 13, 488. D. D. Davis, J. T. Herron and R. E. Huie, J. Chem. Phys., 1973, 58, 530. M. A. A. Clyne and J. A. Coxon, Proc. Roy. SOC. A , 1968, 303, 207. and J. Wolfrum, Nature, 1976, 262, 204. lo C. Morley and I. W. M. Smith, J.C.S. Faraday ZZ, 1972, 68, 1016. l2 (a) J. Wolfrum and K. Kaufmann, personal communication; (b) R. J. Donovan, K. Kaufmarm l 3 M. A. A. Clyne and H. W. Cruse, Trans. Faraday SOC., 1970, 66, 2214. l4 M. A. A. Clyne and H. W. Cruse, Trans. Faraday Soc., 1970, 66, 2227. l5 W. J. R. Tyerman, Trans. Faraday SOC., 1969, 65, 1188. l6 C-L Lin and W. B. DeMore, J . Phys. Chem., 1973, 77, 863. l7 R. J. Donovan and H. M. Gillespie, Reaction Kinetics (Specialist Periodical Report, Chemical Society, London, 1975), vol. 1, p. 14. R. G. Green and R. P. Wayne, J. Photochem., 1977, 6, 371. l9 D. W. Setser, R. Littrelhand J. C. Hassler, J. Amer. Chern. SOC., 1965, 87, 2062. " C. H. Bamford, J. E. Casson and R. P. Wayne, Proc. Roy. SOC. A, 1966,289,287. 21 C . H. Bamford, J. E. Casson and A. N. Hughes, Proc. Roy. SOC. A, 1968,306, 135.296 REACTIONS OF 0(210~) AND 0(23~,) WITH HALOGENOMETHANES 22 R. L. Johnson and D. W. Setser, J. Phys. Chem., 1967, 71, 4366. 23 M. C. Lin, J. Phys. Chem., 1972, 76, 1425. 24 M. C. Lin, Int. J. Chem. Kinetics, 1973, 5, 173. 25 J. C. Stephenson and D. S. King, J. Chem. Phys., 1978, 69, 1485. 26 C. H. Bamford and J. E. Casson, Proc. Roy. SOC. A, 1969, 312, 163. 27 T. C. Frankiewicz, F. W. Williams and R. G. Gann, J . Chem. Phys., 1974, 61, 402. 28 P. A. Gorry, C . V. Nowikow and R. Grice, Chem. Phys. Letters, 1978, 55, 19. 29 E. N. Okafo and E. Whittle, Int. J. Chem. Kinetics, 1975, 7 , 273. 30 D. St. A. G. Radlein, J. C . Whitehead and R. Grice, Nature, 1975, 253, 37. 31 D. D. Parrish and D. R. Herschbach, J. Amer. Chem. Soc., 1973, 95, 6133.

ISSN:0301-7249    

DOI:10.1039/DC9796700286

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

State-to-State Photocheniical Reaction Dynamics in Polyatomic Molecules BY KARL F. FREED AND MICHAEL D. MORSE James Franck Institute and Department of Chemistry, University of Chicago, Chicago, Illinois 60637, U.S.A. AND YEHUDA B. BAND Department of Chemistry, Ben Gurion University of the Negev, Beer Sheva, Israel Received 23rd November, 1978 The generalized Franck-Condon theory of the dissociation of linear triatomic molecules is pre- sented, including proper descriptions of the bending and rotational motions on the bound and unbound electronic potential energy surfaces. The role of angular momentum conservation is made explicit through analytical expressions for rotational and orbital angular momentum distributions of the photofragments. The theory is also applied to the calculation of photofragment angular distributions and preliminary results are described.Molecular dissociation processes play a vital role in the description of chemical reactions, in atmospheric processes, and in astrophysical and biological problems. Despite the central nature of dissociation phenomena, they have only rarely been studied as isolated primary events in polyatomic molecules. In polyatomic molecules the molecular fragments may contain internal vibrational, rotational and electronic excitation which could, in principle, be employed to perform useful work by provid- ing population inversions which yield lasers 'i2 or to generate vibrationally excited species that can serve as the reactants for subsequent chemical reactions, since vibra- tionally excited dissociation fragments can overcome activation barriers to permit reactions that otherwise might not be accessible.In certain cases the photofragments may be formed primarily in particular excited vibronic states so that these photo- fragments can be utilized for further state-to-state kinetic ~tudies.~ Furthermore, photodissociation or predissociation of polyatomic molecules holds considerable promise as a means of separating isotopes. Studies of photodissociation and pre- dissociation in polyatomic molecules are also important as a means of testing theories of vibrational energy randomization and unimolecular reaction dynamics. For example, the molecule can be photoexcited to individual nearly isoenergetic predissoci- ating levels to determine whether the fragment energy distributions are solely dependent on total energy (complete randomization) or whether memory of the initially excited level persists (lack of complete randomization).Considerable theoretical difficulties abound in the description of photodissociation and predissociation. In the limit of weak coupling between the bound and dissocia- tive states the golden rule of time dependent perturbation theory is sufficiently accurate for the calculation of the rate of production, rf,, of a final state, f, from an initial state, i, Tfi = l(fl Vli)12/ti. (1298 PHOTOCHEMICAL REACTION DYNAMICS Here V is the interaction operator coupling the bound state to the continuum, and the continuum wavefunction is normalized to 2n times a delta function in energy.In the Born-Oppenheimer approximation, the wavefunctions for the initial and final states separate into a product of an electronic factor and a vibrational-rotational factor. Often the coupling V depends only weakly on vibrational or rotational coordinates, so this dependence may be ignored (if necessary it may be treated dire~tly).~ Inte- gration over electronic coordinates gives an average coupling strength and the ex- pression where xf and xi are now the wavefunctions for the nuclear motion alone. Thus the problem of photodissociation or predissociation is reduced to this multidimensional Franck-Condon overlap integral, involving a bound initial state and an unbound final state. All that is required is the determination of the wavefunctions and the evalua- tion of the integral.FR ANK-C 0 ND 0 N AM PLI T UD E As a simple example, consider a linear triatomic molecule. The initial, bound state wavefunction involves two stretching vibrations (no1 necessarily harmonic), a bending vibration and overall rotation. The final, dissociated state wavefunction involves a relative interfragment translation, vibration and rotation of the diatomic fragment, and orbital motion of the fragments about one another. Thus, the natural coordinates to use in the description of the bound initial state are quite diffeient from those used in the final dissociative state. These features imply [see eqn ( 5 ) below] that eqn (2) is, in general, a multidimensionaI, nonseparable bound-continuum integral. This difficulty never arises in the (one-dimensional) description of the photodissociation of diatomic molecules.A drastic approximation has been used to avoid this problem in the " quasidiatomic " model, wherein the initial state of the molecule is taken to consist of uncoupled bond oscillators, which then have normal modes identical to those of the final ~ t a t e . ~ - ' ~ This enables the evaluation of the integral in eqn (2) as products of one-dimensional integrals. In later work the correct normal modes are used, but a collinear approximation is retained, thereby constraining all three atoms to lie on a non-rotating Such an approach results in a lack of information about the interesting effects of angular momentum conservation on the photodissociation. In the work presented here, we explicitly include the effects of bending and rotational degrees of freedom to calculate expected rotational and angular distribu- tions in the photofragments as well as the effects on the fragment vibrational and translational distributions that are imposed by constraints of angular momentum conservation.We start with the simplest physically reasonable basis functions for the nuclear wavefunctions. For example, we assume harmonic oscillators and rigid rotors, yielding for the initial state basis functions Here Q, and Q2 are the normal (or local, if necessary) modes for the stretching vibra- tion and are linear combinations of the bond displacements, and wn,(Ql) and wn,(Q2) are harmonic oscillator wavefunctions. wb(6) is the wavefunction for the doubly degenerate bending mode with v quanta and vibrational angular momentum k.DGK(cc, p, 7) is the rigid-rotor wavefunction with angles cc, j? giving the orientation of the molecular axis and y the azimuthal angle of the plane of the instantaneously bentK. F. FREED, M. D. MORSE A N D Y. B . BAND 299 molecule measured about this axis (see fig. 1). Vibration-rotation interaction or vibrational anharmonicity imply that xi is a linear combination of terms of the form of eqn (3), but the integrals (2) may be evaluated term-by-term, so a single function (3) may be utilized without loss of generality. FIG. 1 .-Coordinates used in the theory. Similarly, the wavefunctions for the final dissociative state are written in terms of product functions with diatomic rotational angular momentum j and its z-projection, m, orbital angular momentum of the atom about the diatom, I, and its z-projection, p.These are coupled to give fixed values of total angular momentum J' and its z-projec- tion, M', Bond 2-3 breaks in the photodissociation process. xf = 2 { 2 (J'M'llj~m>Yi,d&F, + s F > Y j m ( P , a)l~n(Q;>~"Ejl(Q;>- (4) nlj pm In expression (4) Q; is the vibrational coordinate of the fragment diatomic, and Q; is the distance between the atom and the centre of mass of the diatomic molecule. The sum over n, j and I is required because the Schrodinger equation for motion on the unbound potential surface is nonseparable, and the coefficients v/E{(Q;) must be determined by a close-coupled solution of this Schrodinger equation. This is a well- studied problem in scattering theory, requiring an enormous computation when large numbers of final states are allowed by energy conservation.Because scattering theor- ists are well acquainted with the solution of the problems of this kind, we focus on the feature of the problem which is new, the calculation of the Franck-Condon integrals. For this purpose it is only necessary to consider individual terms in eqn (4) one at a time. The scattering calculation is necessary to provide the relative amplitudes, This half-collisional scattering process describes the vibrational and rotational relaxation on the unbound electronic surface. The primary difficulties in evaluating such Franck-Condon factors are the large number of degrees of freedom which occur in polyatomic systems and the different coordinate systems that are appropriate to the initial and final states.For the ex- ample of a linear triatomic molecule it is necessary to evaluate the integral between basis functions, Q); for these individual terms. where a rotation through the angles (a, p, y ) takes (OSF, $SF) to (0, 0), as indicated on fig. 1, and 6 = arctan [sinO/(cos0 - A ) ] , with A = [mlrl2/(ml + m2>l/[r23 + m1r12/(m1 + m2)I. In the harmonic approximation (Q,, Q2) are linear functions of (Q;, Qi).300 PHOTOCHEMICAL REACTION DYNAMICS Shapiro 1 8 9 1 9 has provided a different method to alleviate some of these difficulties and correctly to incorporate the effects of inelastic scattering as the fragments separate. His method involves the solution of a large set of coupled differential equations (for both the initial and final states), with the same nuclear coordinate system on both potential surfaces and a fictitious channel included.For collinear processes this possesses some important advantages, but when bends and rotations are included, the sheer size of the problem precludes such a treatment. For these reasons we focus on the Franck-Condon factors themselves to gain insight into the dynamics of the photodissociation process and to provide direct analytic approximations or more accurate reductions of the full Franck-Condon amplitude ( 5 ) to simple one-dimen- sional integrals. After considerable angular momentum algebra, the integration over the coordin- ates (a, p, y ) may be performed. The resulting expression for the Franck-Condon amplitude may be approximately separated into a vibration-translation factor that is identical to that evaluated in the collinear theory multiplied by a factor that involves the bending and rotational degrees of freedom.20p21 The correction terms may be evaluated when necessary, but in applications to date it has been found that the single leading term is sufficient. Furthermore, since in the initially bound state the bending vibration constrains 6 to be small, a small angle approximation is valid for purposes of computing this Franck-Condon integral.With such approximations the rate of production of final basis state f [denoted by the nth term in the wavefunction (4)] from basis state i [denoted by eqn (3)] is Here we invoke a scalar coupling, such as that expected in a predissociation, where there is no preferred direction in space, (provided the method of preparation of the initial state is ignored). The numerical results are almost identical when the proper dipole approximation is made for the case of direct photodissociation, in cases where the distribution of diatomic angular momentum, j, is wide.In expression (6), FEI is the two-dimensional nonseparable overlap integral that arises in the collinear dissociation case ; its exact reduction to one-dimensional integ- rals and interpretation have been well-studied. The only remaining molecular parameter is p = (1 - A ) 2 / ~ 2 where A is a geometri- cal constant and depends on the reduced mass and frequency of the bending mode and where K is given by the expression RESULTS Let us momentarily ignore IFEl l 2 and examine the conditional rotational distribu- tions provided by P,(l,j) = l(fli)12/1FE112 as these are universal distributions for all molecules with the only molecular parameter being p .The conditional probabilityK . F . FREED, M . D. MORSE A N D Y . B . BAND 30 1 of producing orbital angular momentum 1 under these conditions is obtained as This conditional distribution is independent of the total angular momentum J, so as J increases the distribution in diatomic rotational quantum number, j, is expected to shift to higher values, with j z J . We may likewise obtain the conditional distributions of diatomic rotation j , PJ(j), but the sum has been analytically evaluated only for J = 0, giving A comparison of PJ(j) and PJ(I) [eqn (9) and (lo)] shows them to be identical when J = K = 0.Expressions (9) and (10) give normalized probabilities, provided sums 0ver.j and I are replaced by integrals over (-+, a). The case of J = 0 is instructive, since it enables the separation of those features due to bending vibrations alone from those due to overall molecular rotation. Ex- pression (10) can be used to obtain the average diatomic rotational energy, ( E j ) and its r.m.s. deviation, a(Ej) = [(EJ2) - ( E j ) 2 ] ) , as a(E,) = - - + v + 1 P 2 r2 For J = k = 0 (for which these formulae apply), ( E j ) is proportional to the diatomic rotational constant, B, and to the energy initially in the bending mode, given by For large values of v, (E,). Although these hm(v + 1).(A factor of cc, is embedded in - = ~ o(E,) is also proportional to these factors, since a(Ej) - - d2 provide the actual Franck-Condon distributions only if the dependence of [FE1I2 on j and I is ignored, they are qualitatively satisfying because of the proportionality of the average fragment rotational energy (for J = 0) to the initial bending energy. The factor of B/p in eqn (1 1) and (12) gives particular insight into the dissociation process. The amplitude of the zero point bending motion is specified by V ' ~ / K , so (1 - A)v'?/K measures the amplitude of this motion in terms of 8. When K is small (floppy bending modes), little energy is partitioned into rotations, provided the mole- cule is initially nonrotating. Similarly, for large K (stiff bending modes), substantial rotational excitation is predicted (still given the neglect of the dependence of FEI on j , J and I).This general result may be understood in terms of the Heisenberg uncertainty principle which requires AqAp 2 h for any generalized coordinate q and its conjugate momentum p . The available region in 8 space is specified by (1 - A)V'?/K, so the IC2 ) 1 p (1 - A)2 *302 PHOTOCHEMICAL REACTION DYNAMICS range in its conjugate angular momentum is roughly Arc/<2(1 - A ) . Thus for large ~ / ( 1 - A)the range in rotational quantum numbers j and I is large, and for small K / ( 1 - A ) this range is small. Rotation energy is proportional to the square of angular K 2 1 momentum, so we expect ( E j > oc - - - as found in eqn (1 1). Fig.2-4 present the distributions P J ( j ) for selected values of J and bending quantum numbers v, k. The abscissa is chosen to b e j - J , to emphasize the peaking of P J ( j ) aboutj = J , especially for large values of J . These plots ignore the variation of FEI with j and I, so the probability distribution P,(I) is identical to that of PJ = o(j), as discussed above. All curves correspond to p = 0.0151, which is appropriate for ICN. When photodissociation originates from excited bending states considerable struc- ture occurs in the conditional rotational distributions because of nodes in the bending wavefunction. This structure may be washed out by the variation of FEI with j and I, by the effects of averaging over J for the initial bound state, and by half-colli- sional rotational relaxation.The effects of the variation of FEl withj and I have been included for ICN photodissociation to provide distributions that are reasonably in accord with the available experimental data, assuming two excited electronic states are In addition to the above, we have constructed linear combinations of final dissoci- ated states [with wavefunctions of the form of eqn (4)] of various I to give an outgoing wave in the direction (8, #), and have thereby calculated the angular distributions of photofragments in specific internal quantum states resulting from direct photo- diss~ciation.~~ The appropriate final dissociated state wavefunction (again ignoring (1 - A)2 - p (Comparable values are found for many other molecules.) V=k=O V= k = l -20 -10 0 10 20 -20 -10 0 10 20 30 j - J j - J FIG.2.-Probabilities of producing fragment rotational state j from initial triatomic rotational state J for p = 0.0151 and bending quantum numbers v , k , ignoring the dependence of FEl on I, j and J, and ignoring final state interaction effects. Values of J a r e : ( a ) 30, (b) 20, ( c ) 10, ( d ) 0 ; (e) 31, (f) 21, (g) 1 1 , (h) 1.K . F . FREED, M . D . MORSE AND Y . B . BAND 303 -20 -10 0 10 20 -20 -10 0 10 20 j - J j - J FIG. 3.--Probabilities of producing fragment rotational statej from an initial quantum state specified by J , v and k , ignoring the dependence of F,, on I, j and J and final state interactions. Values of J are: (CI) 30, (b) 20, (c) 10, (d) 0, ( e ) 32, ( f ) 22, (g) 12, (h) 2. Frc;. 4.-Rotational distribution P J ( j ) for the dissociation of a molecule with an excited bending node ( v = 3), ignoring the dependence of FEI on I, j and J and final state interactions.Values of J are: (a> 31, ( h ) 21, (c) 11, (4 1, (el 33, (f) 23, (g) 13, (h) 3.304 PHOTOCHEMICAL REACTION DYNAMICS scattering effects on the unbound potential surface) is found to be where the appropriate generalizations for the presence of diatomic electronic angular momentum, Ad, and atomic angular momentum (projected along the atom-diatom vector), A,, have been made. 6l is the phase shift of the Ith partial wave on the un- bound surface. After considerable manipulation along the lines outlined above we obtain the rate of production of diatomic molecules in states specified by j and n at an angle 0 relative to the laboratory-fixed z-axis to be given by I Ad - Ai - k I' (Ad - Ai - k 1 - M" M" k + Ai - A, In this expression Ai, Ad and A, are the axis-projections of the electronic angular momentum of the initial bound triatomic, the diatomic fragment, and the atomic fragment, respectively.Parallel transitions are described by M" = 0, and perpendicu- lar by M" = & 1. Linearly polarized light with its electric vector in the z-axis corre- sponds to q = 0, and circularly polarized light propagating in the z-direction corre- sponds to q = 1. W(p-1 In ; I l l ) represents a Racah coefficient, (: - t :), etc. are 3-j symbols and !(I + k + A,)!(I - Aa)! - k - A,)!(I + A,)! If bends are ignored (2' = FEI) and we consider either the nonrotating ( j = J = 0) or high translational energy case, expression (14) reduces to Ifi cc 1 + pP2(c0sO), where p = 2 for q = M" = 0, p = -1 for q = k l , M" = 0 or q = 0, M" = & l , or p = + for q = & l , M" = & l .These are the results expected intuitively, or on the basis of the simpler diatomic case. Only for q = i-1, M" = &1, is there aP,(cosO) term, but this is expected to be removed by the presence of both M" = & 1 transitions in a perpendicular band. Terms with n > 2 will never contribute to eqn (14). Calculations in progress on ICN high-energy dissociations show that in general p(j, J ) depends primarily on I j - JI, except for j close to zero, with structure appear- ing in these cofficients when bending modes are excited in the initialK .F. FREED, M . D . MORSE AND Y . B. BAND 305 CONCLUSION This paper has shown how a Franck-Condon theory of polyatomic photodissocia- tion or predissociation may be constructed to provide detailed state-to-state transition probabilities. The expressions obtained show the nodal structure associated with Franck-Condon factors and emphasize the importance of angular momentum and energy conservation in polyatomic dissociations. In the future we can hope for a fruitful interplay between theory and experiments in this field, leading to new ideas that should greatly improve our understanding of state-to-state processes in general. K. F. F. is supported in part by N.S.F. Grant no. CHE77-24652. M. D. M. holds a Fannie and John Hertz Foundation Fellowship and Y.B. B. is grateful to the Barecha Fund for Science for grants. M. J. Berry, Chetii. Phys. Letters, 1974. 29, 329. A. P. Baranavski, R. G. Miller and J . R. McDonald, Chetn. Phys., 1978, 30, 119. P. Fink and C. F. Goodeve, Proc. Roy. Soc. A, 1937, 163, 592. D. Porret and C. F. Goodeve, Proc. Roy. Soc. A, 1938, 165, 31 ; Tram Faraday SOC., 1937, 33, 690. H. Friedman, R. B. Bernstein and H. E. Gunning, J . Cheni. Phys., 1957, 26, 528. K. E. Holdy, L. C. Klotz and K. R. Wilson, J . Chem. Phys., 1970, 52, 4588. F. E. Hendrich, K. R. Wilson and D. Rapp, J . Chcnz. Phys., 1971, 54, 3885. M. Shapiro and R. D. Levine, Chem. Phys. Letters, 1970, 5, 499. ' G. A. West and M. J. Berry, J . Cheni. Phys., 1974, 61, 4700. ' K. F. Freed and S. H. Lin, Chew. Phys., 1975, 11, 409. ' A. Gordus and R. B. Bernstein, J . Chetn. Phys., 1954, 22, 790; J . Cheni. Phys., 1959,30, 973. l2 R. G. Gilbert and I . G. Ross, Austral. J . Cheui., 1971, 24, 1541. l 3 H. Gebelein and J. Jortner, Theor. Chit??. Acta, 1972, 25, 143. l4 S. Mukamel and J. Jortner, J . Cheni. Phys., 1974, 60, 4760. l5 Y. B. Band and K. F. Freed, Chetu. Phys. Letters, 1974, 28, 328. l6 Y . B. Band and K. F. Freed, J . Cherii. Phj.~., 1975, 63, 3382. 0. Atabek, J. A. Beswick, R. Lefebvre, S. Mukamel and J. Jortner, J . Chenr. Phys., 1976, 65, 4035. l8 M. Shapiro, J . Chem. Phys., 1972, 56, 2582. l9 M. Shapiro, Israel J . Chem., 1973, 11, 691. 'O M. D. Morse, K. F. Freed and Y . B. Band, Chern. Phys. Letters, 1976, 44, 125. 21 M. D. Morse, K. F. Freed and Y. B. Band, J . Chem. Phys., in press. 22 M. D. Morse, K. F. Freed and Y. B. Band, J . Chew. Phys., in press. 23 M. D. Morse K. F. Freed and Y . B. Band, to be published.

ISSN:0301-7249    

DOI:10.1039/DC9796700297

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

Photofragmentation Dynamics and Reactive Collisions of Laser-excited Electronic States BY STEVEN L. BAUGHCUM,~ HUBERT HOFMANN,~ STEPHEN R. LEONE$ AND DAVID J. NESBITT Joint Institute for Laboratory Astrophysics, National Bureau of Standards and University of Colorado, and Department of Chemistry, University of Colorado, Boulder, Colorado 80309, U.S.A. Received 24th November, 1978 Tunable laser excitation followed by observation of infrared fluorescence provides a means of readily studying electronically excited photoproducts and their reactions. A number of specific examples involving the production of I*(52P1/2) and Br*(42P1 c 2 ) upon molecular photodissociation are considered. The quantum yield of I * production from HgIz is obtained as a function of wave- length. An inconclusive search was made for unobserved states in IBr which lead to I* product atoms and for states of BrCl which lead to Br*.Photodissociation of CHJz in an intense laser field is observed to undergo multiphoton dissociation. Quenching and reactive collisions of Br* and I* with halogens, I t , Br, and CI2, and interhalogens, IBr, ICI and BrCI, are investigated. Electronically adiabatic reactive channels are detected for the collisions I * + Brz -f IBr + Br* and I* + IBr + I2 -t Br*. Vibrationally excited HBr product molecules are observed in 10 ”/, of the quenching collisions of Br* with H2S. Tunable laser, infrared fluorescence techniques provide a powerful means of studying photodissociation and reaction dynamics of electronically excited states. A great amount of data is now available which characterizes the electronic states of atoms produced on photodissociation I and the rates of deactivation of these electroni- cally excited atoms.’ As yet, however, there are many unknown aspects of the photo- dissociation pathways and reactive kinetics of electronically excited atoms produced.With narrow band, tunable dye laser sources and detection of infrared emission from electronically excited atoms, it is possib!e to study many of these processes in more detail. In this work, the production of I*(52P,,2) and Br*(42P,:,) on molecular photodissociation and the reactions of these electronically excited atoms are investi- gated. The experimental techniques described here provide an excellent method to identify photodissociation products which can be detected via infrared emission.Kinetic information for the excited products is obtained directly from the time development of the infrared emission after the pulsed laser excitation. Information is obtained on the quantum yields of excited atom production as a function of photolysis wavelength, enabling more complete specification of the nature of the electronically excited states participating in the photofragmentation. In addition, detailed rate constants are obtained for specific reactive channels occurring in the deactivation of the I* and Br* atoms. t NRC-NBS Postdoctoral Fellow. 5 Present address: Battelle-Institut e.V., 6000 Frankfurt am Main 90, West Germany. $ Staff member, Quantum Physics Division, National Bureau of Standards, and Alfred P. Sloan Fellow.LASER-EXCITED ELECTRONIC STATES 307 EXPERIMENTAL A schematic diagram of the experimental arrangement is shown in fig 1.A pulsed, flashlamp-pumped tunable dye laser with frequency doubling capability produces 1-10 mJ of tunable radiation in the 265-365 nm and 440-720 nm ranges. The E 1 p s pulses are used to photodissociate a variety of molecules to study the generation of electronically excited I* and Br* and their reactions in a fluorescence sample cell. A flowing gas cell is used for reac- tive studies and a heated cell is used for low vapour pressure compounds. Beam splitters tunabl com pu t er /signal averag er mon ochromator , e flashlamp dye laser, f i l t e r t o vacuum handling system meter FIG. 1 .-Schematic diagram of experimental apparatus used for tunable laser photodissociation and reaction studies.Detection of both time and amplitude variation of excited species is by infrared fluorescence. See text for remaining details. separate portions of the laser beam to serve as a trigger for the data acquisition and as a wave- length monitor. The energy of the laser beam transmitted through the experimental cell is monitored by a calibrated thermopile. The output of the thermopile is read by a milli- voltmeter and continuously recorded on a strip chart recorder. The excited iodine and bromine atoms are detected via their infrared emission on the corresponding 2Pl,z -+ 2P3i2 transitions at 1.315 and 2.713 pm, respectively. Typically a liquid nitrogen cooled InSb infrared detector, 1.27 cm x 1.27 cm in area, is used, viewing through a narrow band inter- ference filter to select the specific atomic transitions.Other emissions from vibrationally excited molecules are detected through appropriate interference filters. The signals are amplified and recorded in a transient digitizer. The digital signals are stored and summed with a signal averaging computer and subsequently plotted with an x-y recorder. The result- ing signals, normalized for the number of laser pulses, pressure of photolysis gas and laser energy, carry complete time and relative concentration information for the excited species produced in the photolysis or by reactive events. Reproducibility of time and amplitude behaviour after signal averaging 100-3200 laser pulses is better than 5 %.The use of rela- tively low energy photolysis pulses ( z 10 mJ per pulse) provides reliable and reproducible results. Rigorous precautions are taken to provide the highest purity sample preparations using vacuum distillations of all materials when necessary and following good vacuum practices througho~t.~-~308 LASER EXCITED ELECTRONIC STATES RESULTS Detailed studies of photodissociation dynamics reveal a wealth of new information regarding the electronic states of the parent molecule as well as the fundamental dynamics of the photofragmentation.6 The quantum yield of excited I* atoms was obtained for the linear triatomic HgI, molecule as a function of dissociating wave- length in the range of the first long wavelength absorption band, from 265 to 350 nm.5 Data were taken at a constant cell temperature of 453 K and a constant sidearm tem- perature of 361 K.This provides a number density of HgI, molecules of 2.6 x l O I 4 ~ m - ~ . This low density was selected to ensure that only a small fraction of the laser light is absorbed in the cell over the entire wavelength region and so that a relatively slow deactivation decay of 10 p s could be readily extrapolated to obtain an accurate maximum amplitude at time equal to zero. A total of 19 measurements of the I* atom decay signal amplitude as a function of time were made at laser wavelengths ranging from 265 to 320 nm. The signal strength at each wavelength was obtained by an extrapolation of the observed exponential decay to zero time using a least-squares fit based on eight data points from each decay.The laser power was recorded continuously, and after normalization for the laser power, relative I* quantum yields are ~ b t a i n e d . ~ In an independent experiment, the absorption cross section of HgI, was measured. To obtain directly the absolute quantum yield for I* production would entail further elaborate measurements of detection efficiency, geometry, etc. Therefore a com- parison method was chosen to determine the absolute value of the quantum yield of I* production from HgI,. From the branching ratio work of Donohue and Wiesen- feld,7*8 it is known that the fractional yield of I* atoms from n-C3F71 is >0.99. The absorption cross section for n-C3F71 was also obtained in an independent measure- ment, and the relative I* yields from HgIz and n-C3F71 were determined from the infrared fluorescence signals in a set of measurements with constant excitation condi- tions and detection geometry.The ratio of I* yields obtained at 270 nm is 4Hg12/ $C3F71 = 1.0 & 0.12, where the quoted error is most affected by the extrapolation procedure to obtain the maximum fluorescence amplitudes at zero time and by the temperature fluctuations in the fluorescence and absorption cells, which affect the HgI, number density. The results of the photodissociation I* quantum yield for HgI, are shown in fig. 2. Both the total absorption cross-section, curve (a), and the fractional cross-section for I* production are shown, curve (b). The difference between (a) and (b) shows clearly that the cross-section in this first long wavelength absorption is composed of two distinct components.One component, with a maxi- mum at 270 nm, is associated with the formation of excited I*(52P1,2) atoms, while the other component [difference (a)-(b) in fig. 21, with a maximum around 310 nm, is attributed to the formation of ground state I(52P3/2) atoms. Numerous electroni- cally excited states of HgI, may participate in the absorption in the region 265-350 nm. However, the observed behaviour is best described by a simplified picture of two paral- lel repulsive states (fig. 3). More information could be obtained from measurements of the relative kinetic energy of the departing fragments, as well as a determination of the internal energy of the HgI fragment after the photodissociation.Photodissociation of the halogens (I,, Br,) and interhalogens (IBr, BrCl, etc.) has long been a subject of intensive investigation. A great amount of data is now available which characterizes the electronic states of the atoms produced on photo- dissociation,8-10 the quantum yields of predissociation and the nature of collisional predissociation and r e l e a ~ e . ~ . ~ Much has been learned from spectroscopic investiga- tions t h e m ~ e l v e s . ~ ~ - ~ ~ Some of the most extensive and elegant studies on photo- No I* signals were observable above 320 nm.s. L . BAUGHCUM, H . HOFMANN, s. R . LEONE AND D. J . NESBITT 309 i difference l o ) - / b l 1 - 1 < I I I I I I , h i I I I I I ] 265 275 285 295 305 315 325 335 345 wavelength I nm FIG.2.-Measured total absorption cross-section ( a ) of Hg12 and fractional components (b) leading to excited I(52f'1/2) and ground state [(a)-(b)] I(52p3/2) atoms upon photodissociation of Hg12 as a function of wavelength. Filled circles are experimental I* quantum yield points. The laser gap is a region where no I* yield data could be taken due to low dye power. A typical error bar for the quantum yield results is shown. ,I =270n A = 310 n R(Hg 1-11 FIG. 3.-Plot of potential energy as a function of internuclear separation for Hg12. Schematic of parallel repulsive curve mechanism, most likely responsible for the experimental photofragmentation results in Hg12. Shorter wavelengths access the excited I* repulsive curve. Longer wavelengths lead to ground state I in a separate absorption feature.310 LASER EXCITED ELECTRONIC STATES dissociation involve photofragmentation spectroscopy carried out in molecular b e a r n ~ .~ ~ * ' ~ ~ ' ~ Direct detection of electronically excited atomic fragments by their infrared fluorescence emission provides a sensitive indicator of their quantum yield as a function of dissociating ~ a v e l e n g t h . ~ , ~ ~ For example, in the photodissociation of interhalogen molecules such as IBr and BrCI, states which correlate to electronically excited (2P1/2) Br* and C1*, respectively, have been observed2 or determined spectro- scopically.12 Bound or repulsive states which might lead to the I* limit in IBr or the Br* limit in BrCl have not been observed. These states are expected to exist, but may be much weaker due to a displacement of the potential curves to larger inter- nuclear separation with respect to the ground state (see fig.4).12 In addition, these states would be buried in the strong line and continuum absorptions of other states, making it difficult to observe them in absorption. Thus, atomic absorption measure- ments after flash photolysis of 1Br have observed only large fractions of Br* and ground state 1.'' A search for states which might lead to an I* product upon photodissociation of IBr and to Br* from BrCl was made using the tunable laser, infrared fluorescence techniques described above. The laser was tuned from 440 to 460 nm, around the wavelengths required to reach the dissociation limits, for both IBr (450 nm) and BrCl (460 nm), and also well above the dissociation limit at 300 nm for IBr.No I* signal was observed from IBr at 300 nm. Over the 440-460 nm range in IBr, no I* emission was observed which could be directly attributed to photolysis of IBr, although weak I* signals from the dissociation of the small equilibrium fraction of I2 in IBr was seen. This was confirmed by a separate experiment on I2 alone. Based on the known absorption coefficients of I2 and 1Br,l8 and the equilibrium fraction of I2 in IBr,394 the ratio of the I* quantum yields of photodissociation from I2 and IBr can be given in terms of an upper bound. A similar experiment was carried out for BrCI, trying to detect the Br* product. In this case, excess C1, was added to shift the equilibrium to BrC1, thereby minimizing Br,, which is known to produce Br*.Again, no detectable Br* could be directly attributed to photodissociation of BrCI. A Iimit of the quantum yields of Br* produc- tion from BrCl and Br, can be given. This limit is $Brcl(Br* at 460 nm)/$,,,(Br* at 460 nm) 5 These experiments on IBr and BrCl at their dissociation limits indicate that (a) either the Franck-Condon factors for absorption to the unobserved states correlating to Br-I* and CI-Br* are very weak, or (b) the states themselves under- go curve crossing mechanisms which always lead to ground state atoms or dissociation to the Br* and Cl* limits which are ordinarily observed (fig. 4). The null experiment at 300 nm in IBr gives further support that the curve crossing mechanism may be im- portant, since it would be expected that some part of the repulsive Br-I* curves would be accessible above the dissociation limil.Several laser photolysis, infrared fluorescence experiments are currently being carried out on CHJ and CH212, using the 193,248 and 308 nm outputs of a high energy rare gas halide excimer laser. Although the primary goal of these experiments has been the detection and study of vibrationally excited free radical^,'^ in the course of this work a number of preliminary photofragmentation results on electronically excited species have been obtained. Under low power ( z lo5 W cm-2), the photo- dissociation of CH212 and CH31 with 248 nm laser light is observed to produce sub- stantial fractions of electronically excited I*. In the case of CH2T2, but not CH31, highly vibrationally excited radicals (CH,I) have also been observed." At 308 nm, I* is observed from CH212, but vibrationally excited CH,I radicals are not.Under these low power conditions, typical decay rates are observed: 2.2 x cm3 molecule-' s-l for I* deactivated by CHJ and 4 x cm3 molecules-' for I* This limit is $,Br(I* at 450 nm)/$r,(l* at 450 nm) 5s. L . BAUGHCUM, H. HOFMANN, s. R . LEONE AND D. J . NESBITT 311 3 . c 2 . c > a, 1 . 1 . c - 1 T laser range I nvestigai - - I 0.1 0 . 2 I 0.3 0.4 0.5 0 . 6 R/nm FIG. 4.-Approximate potential curves as a function of internuclear separation for IBr, showing the three lowest possible dissociation limits: I + Br, I f Br*, and I* + Br. The dotted curves are un- observed states in IBr, leading to the I* product.The inability to observe these states may be due to either ( a ) poor Franck-Condon factors, schematically indicated by the large displacement of the I* t Br states to larger internuclear separation than the vertical transition, or (b) by fortuitous curve crossings, indicated by the circled region in the figure, which produces Br* or ground state atomic products whenever the I* + Br states are accessed. deactivated by CH,I,. Under highly focused conditions (107-10s W cm-2) at 248 nm, the photolysis of CH,12 is accompanied by immediate visible light emission, most likely arising from a multiple photon dissociation process followed by electronic excitation of the products at 248 nm. Brief and intense infrared signals are observed around 3 pm.The intensity of these signals has a third order dependence on the laser power. The decay lifetimes of these signals are limited by the response time of the i.r. detector ( z 300 ns), possibly indicating that the emitters are short-lived electronic states. A sequential two photon absorption at 266 nrn in CHJ,, thought to produce CH, and two I atoms, has been previously reported in photofragmentation molecular beam312 LASER-EXCITED ELECTRONIC STATES experiments.20 In our experiments visible light is not observed with the focused 248 nm excitation in 1, alone. Therefore, it is likely that the visible light produced by photolysis at 248 nm is due to electronically excited CH, radicals formed by a three photon process, CH,I, -% CH,I + I hv_ CH, + I % CH; + I.Studies of CH212 in the range 125-200 nm have reported the production of electronically excited molecular 12.'l In our experiments with unfocused 193 nm light, an immediate visible emission is observed in the photolysis of CH212, but not in CH21, possibly from direct production of I, in excited states. However, we have also found that molecular 1, itself is excited directly by 193 nm light with subsequent visible fluorescence, making analysis of the photodecomposition process difficult. For both CHJ and CH,I,, no I* signals were observed following photolysis at 193 nm. l " ' 1 0 20 30 40 t i m e I ,us FIG. 5.-I* and Br* product signal intensity against time from the process I* + Brz+ IBr -t Br* as a function of time. The Br* signal has been plotted on a scale 16 times more sensitive than the I* signal.Both the The pressures used were 400 N m-' C3F71, 27 N m-2 Br, and 1330 N m-' argon. signals have been obtained from an average over 1600 laser pulses. In addition to spectral studies of photodecomposition into various excited elec- tronic states, it is possible to use various molecules as photolysis sources of excited atoms for further study of the reactive collisions of electronically excited species. Deactivation of I* and Br* atoms by both reaction 4 , 2 2 -23 and electronic-to-vibrational, rotational and translational energy transfer (E-V, R, T) 1 6 9 2 5 has been discussed. was the first to observe Br* production in the collision system I* + Br, --+ IBr + Br*, proving the presence of reactive, adiabatic channels in this deactiva- tion process.23 Hofmann and Leone" observed Br* production in the I* + IBr+ I, + Br* reaction as well.They estimated the fractions of the total deactivation rate constants responsible for the reactive channels in the above processes as 15 and 13 %, respectively (fig. 5). Wiesenfeld and Wolk '' have provided thorough measurements on I* + Br, which indicate that the fraction of Br* produced in the reaction is 72 x. These results are in better agreement with the fact that a population inversion and stimulated emission in Br* has also been reported with the chemical reaction I* +s. L. BAUGHCUM, H . HOFMANN, s. R . LEONE AND D . J . NESBITT 313 Br, --+ IBr + Br*.26 The discrepancy in the Br* yield measurements obtained by the laser infrared fluorescence results and those obtained using flash photolysis atomic resonance absorption spectroscopy techniques 24 has not been resolved.However, the infrared fluorescence measurements presently depend on the I* and Br* radiative lifetimes, values which are still in question2' and provide a margin for error.4 In addition, the signal-to-noise ratios in these particular infrared experiments were noted to be especially poor (fig. 5),4 and may preclude an accurate test of agreement between the two methods. For collisions of Br* and I* with certain molecules, the possibility exists for reactive quenching as well as competing E-V energy transfer channels. Preliminary results have been obtained on the reactive channel of Br* with H2S to produce HBr and HS, which competes with E-V transfer. The reaction of ground state Br with H,S is only 2 kcal mo1-I exothermic, so no vibrationally excited products are expected.Among the reactive channels of Br* + H2S there exists enough extra energy in Br* (10.5 kcal mol-') to populate either HBr or HS in the (v = 1) state. A fast flow reaction cell and the tunable dye laser have been used to investigate these processes. By tuning the laser to 580 nm, copious amounts of ground state Br atoms are produced by the photolysis of Br,, as evidenced by product HBr vibrational chemiluminescence signals from the Br + HI --+HBr(u = 1) + 1 reaction. As expected, no vibra- tionally excited products are observed by photolysis of Br, with 580 nm light in the presence of H,S. However, on tuning the dye laser to 480 nm, where the ratio of Br*/Br produced by the photolysis is nearly unity,' chemiluminescent HBr(u = 1) product is observed from the reaction, Br* + H,S+ HBr(u = 1) + HS.Proof that the signal is from HBr(v = 1) is obtained by blocking it entirely with a gas filter of HBr. A preliminary estimate based on the relative signal strengths and detectivity calibrations, transmission coefficients, and radiative lifetimes indicates that this reac- tive channel accounts for 10 % of the overall deactivation of Br*. The total quenching rate of Br* by H,S is measured to be (2.6 i. 0.1) x lo-'' cm3 molecule-'s-'. Work is now underway to investigate the extent of E-V transfer in the Br* + H,S system as well. In the examples given above, reactive encounters of electronically excited atoms are observed directly.There is also strong evidence that the rates of certain quenching processes are highly dependent on reactive type collisions, and the results can be ex- plained by a collision complex model for quenching. Specifically, this includes the quenching rates of 1*4 and Br*3 with the halogens, Br,, I,, CI,, and interhalogens ICl, IBr, BrCl, which have been measured. In order to measure the rates of quench- ing for the interhalogen gases, the parent molecules are mixed together in a wide range of mole fractions, producing three component mixtures, e.g. Br,, I, and 1Br. All three deactivation rate constants are then extracted from a three parameter least- squares fitting procedure. The signal-to-noise ratio is very high (fig. 6), and control over sample purity extremely good.The method produces highly consistent results and agrees well with several previously reported rate^.^.^ The data obtained are summarized in table 1. It may be seen from table 1 that the deactivation processes for I* are much more rapid than for Br*. Deactivation of I* in nonreactive collisions is typically one order of magnitude or more smaller than for deactivation by the halogens. In addition, there is an obvious trend in the Br* deactivation data. All molecules containing I atoms deactivate Br* very efficiently in comparison with BrCl and Cl,, and Br, is of intermediate value. The large differences in rates cannot be explained on the basis of a purely E-V energy transfer me~hanisrn.~*~ One possible mechanism is that the quenching of I* and Br* by halogens goes via formation of excited state trihalogen314 LASER EXCITED ELECTRONIC STATES 0 100 2 t i me I p s FIG.&--A typical Br* emission signal intensity as a function of time obtained in an equilibrium (BrZ + C1, + BrCl) mixture, PBrr = 111 N m-', PcLz = 111 N m-', PBrCl = 312 N m-2. Photo- lysis of Br, parent molecule at 480 nm produces the Br* observed. complexes followed by either formal reaction or simply breakup of the complex to give quenched starting partners. There is good qualitative agreement between the relative rates of Br* and I* quenching by the different halogens and the expected stability of the corresponding ground state trihalogen complexe~.~*~ Thus the order of stabilities, [Br I I] > [Br I Br] > [Br I CI] > [Br Br Br] > [Br Br Cl] > [Br C1 Cl], decreases with decreasing electropositive character in the halogen collision partner.This order of stabilities correlates well with the decreasing rates of deactivation on going from I,, IBr, ICl, Br,, BrCl to C1,. In the case of I*, not only does a similar trend exist, but reactive collisions are actually observed d i r e ~ t l y , ~ ~ ~ * confirming the validity of the mechanism. TABLE 1 .-TOTAL DEACTIVATION RATE CONSTANTS/CM3 molecule-' S-', FOR I* AND Br* (2P1,2) WITH HALOGENS AND INTERHALOGENS AT 293 K I2 Br2 c12 IBr ICl BrCl Br*3 1.86 & 0.37 4.7 f 0.4 2.2 3 1.4 1.0 f 0.14 9 4 2.9 1.4 I* 4 3.1 & 0.5 5.2 f- 0.3 1.7 i 0.2 6.6 & 0.3 2.3 + 0.2 2.7 & 0.2 x 10-12 x 1 0 - 1 3 x 10-14 x 10-12 x 1 0 - 1 3 x 1 0 - 1 4 x 1 0 - 1 1 x lo-" x lo-'* x lo-" x lo-" x lo-" DISCUSSION A variety of results in molecular photofragmentation and reactive collisions of electronically excited species have been presented.Tunable laser, infrared fluores- cence techniques provide a powerful means of extracting information about the quantum yield of excited species produced upon photodissociation and the detailed chemistry that ensues with electronically excited states. In many cases, only qualitative interpretations of the results can be made. Further theoretical work is obviously necessary to obtain deeper insights into the mechanisms and dynamics of electronic excited-state processes. It is evident from the experiments on the unobserved states in IBr and BrCl that there is still much work which can be done to understand the nature of electronically excited molecular states, even in these simple molecules.s. L .BAUGHCUM, H . HOFMANN, s. R . LEONE AND D. J . NESBITT 315 From the I* + Br, and IBr results, it appears that reactions of electronically excited atoms proceed in many cases via electronically adiabatic surfaces when possible. This leads to many interesting questions, for example, whether the exchange of Br* with Br, occurs more frequently than quenching. A new example is now available (Br* + H2S) in which the reaction of an electronically excited atom directly produces vibrational excitation in the HBr product. Finally, our understanding of the relative importance of reactive type collisions versus E-V transfer mechanisms has been substantially refined by the studies of Br* and I* collisions with halogens.It is ap- parent that a number of qualitatively different processes contribute to the quenching of electronically excited atomic and molecular states, depending on the particular excited state and energy and its interaction with the quencher. Further studies should provide much more detailed information on the potential surfaces involved. The authors gratefully acknowledge the support of the National Science Founda- tion and the Office of Naval Research, and wish to thank the support staff at the Joint Institute for Laboratory Astrophysics for their help. K. R. Wilson, in Excited State Chemistry, ed. J. N. Pitts, Jr (Gordon and Breach, New York, 1970), p. 33. R.J. Donovan and D. Husain, Chem. Rev., 1970, 70,489. H. Hofmann and S. R. Leone, Chenz. Phys. Letters, 1978, 54, 314. H. Hofmann and S. R. Leone, J . Chem. Phys., 1978, 69, 641. H. Hofmann and S. R. Leone, J . Chem. Phys., 1978, 69, 3819. J, P. Simons, in Gas Kinetics arid Energy Transfer, ed. P. G. Ashrnore and R. J . Donovan (The Chemical Society, London, 1977), vol. 2. T. Donohue and J. R. Wiesenfeld, Chem. Phys. Letters, 1975, 33, 176; J . Chem. Phys., 1975, 63, 31 30. A. B. Petersen and I. W. M. Smith, Chem. Phys., 1978, 30, 407. D. H. Burde, R. A. McFarlane and J. R. Wiesenfeld, Phys. Rev. A, 1974, 10, 1917. l o M. S. devries, N. J. A. van Veen and A. E. de Vries, Chem. Phys. Letters, 1978,56, 15. l 1 J. Tellinghuisen, J . Chem. Phys., 1973, 58, 2821. l2 M. S. Child and R. B. Bernstein, J . Chenz. Phys., 1973, 59, 5916. l3 R. S. Mulliken, J . Chem. Phys., 1971, 55, 288. l4 G. E. Busch, J. R. Cornelius, R. T. Mahoney, R. I. Morse, D. W. Schlosser and K. R. Wilson, l5 K. R. Wilson, in Excited State Chemistry, ed. J. N. Pitts, Jr (Gordon and Breach, New York, l6 S. R. Leone and F. J. Wodarczyk, J . Chem. Phys., 1974, 60, 314. l 7 R. J. Donovan and D. Husain, Trans. Faruduy SOC., 1968, 64, 2325. l9 S. L. Baughcum and S. R. Leone, Proc. SOC. Photo-Optical Instr. Eng., 1978, 158, 29. ’ O P. M. Kroger, P. C. Demou and S. J. Riley, J . Chem. Phys., 1976, 65, 1823. 21 P. J. Dyne and D. W. G. Style, J . Chem. Soc., 1952, 2122; D. W. G. Style and J . C. Ward, 22 K. Bergmann, S. R. Leone and C. B. Moore, J. Chem. Phys., 1975,63,4161. 23 P. L. Houston, Chem. Phys. Letters, 1977, 47, 137. 24 J. R. Wiesenfeld and G. L. Wok, J. Chem. Phys., 1978, 69, 1797, 1805. ” S. Lemont and G. W. Flynn, Ann. Rev. Phys. Chem., 1977, 28, 261. 26 D. J. Spencer arld C. Wittig, Tenth International Quantum Electronics Conference (1978). ” D. Husain, N. K. H. Slater and J. R. Wiesenfeld, Chem. Phys. Letters, 1977, 51, 201. ’’ R. J. Donovan, F. G. M. Hathorn and D. Husain, Trans. Faraday Sor., 1968, 64, 1228. Rev. Sci. Instr., 1970, 41, 1066. 1970). D. J. Seery and D. Britton, J . Phys. Chem., 1964, 68, 2263. J . Chem. SOC., 1952, 2125.

ISSN:0301-7249    

DOI:10.1039/DC9796700306

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

Studies of BrCl by Laser-induced Fluorescence Part 3.--Dynamics of Quantum Resolved Levels in the Excited B3n(O+) State BY MICHAEL A. A. CLYNE AND I. STUART MCDERMID Department of Chemistry, Queen Mary College, Mile End Road, London El 4NS Received 29th December, 1978 Laser-induced fluorescence has been used to determine the kinetics of decay of resolved ro-vibra- tional states of excited a1Br3sC1 B31T(O+) molecules. Fluorescence lifetimes have been measured as a function of J’(5 d J’ d 60), u’(4 d u’ d 6) and pressure of C12 (0.10 d p d 1.00 mTorr). The mean value of the collision-free lifetime for the stable levels of a1Br35C1 (B) was T,, = (40.2 & 1.8) ,us, for 6 2 u’ 2 3. Although electronic quenching of the B-state of BrCl was slow, rapid vibrational energy transfer within this state was found to occur in collisions with C12.This effect was manifested experimentally by a dependence upon u‘ of the second-order rate constant for collisional deactivation of BrCl (B). Predissociation of all rotational levels in the v‘ = 7 manifold was confirmed, and predissociation in u’ = 6 has been observed for the first time as a shortening of lifetime of levels with J’ d 42. The collision-free lifetimes of these levels varied from zo = 8.8 ,us for J‘ = 42, to to = 4.3 ,us for J‘ = 50, and showed a nearly linear dependence of 1 /to upon J’(J’ t- 1). Rotational levels up to at least J ’ = 70 in u‘ = 5 have been observed to be stable, indicating that the predissociation probably belongs to Herzberg’s case I(c). New close limits for the ground state dissociation energy, namely D$( a1Br35C1) = (17 934 26) cm-’, have been calculated, based on this observation.The B311(Of) - X’C+ transitions of several halogens and interhalogens currently are being considered as electronic-transition lasers, using optical or chemical pumping. The B-X transition of BrCl is promising in this respect, since the transition is strongly non-vertical; also, the lifetime of the B3n(O+) excited state is relatively long (40.2 & 1.8) ,us according to the present work. The Br, + OClO reaction gives intense BrCl (B-X) chemiluminescence; therefore, it may be possible to chemically pump a BrCl laser. Laser-induced fluorescence (LIF) studies of the collisional and non-collisional dynamics of (v’, J’) levels of the B state of BrCl provide important data for the design of a possible B-X laser.In addition to the interest in BrCl as a possible electronic-transition laser, BrCl is a simple prototype heteronuclear molecule for detailed dynamical studies using LIF. Particular interest attaches to collision-free predissociation phenomena, and to the relative importance of electronic quenching, collisional predissociation and vibrational transfer within the B state manifold. This paper describes new work dealing with these important questions. However, LIF studies of BrCl are not particularly easy, because of the low signal intensities involved. The Franck-Condon factors q u r , u j j for exciting B3n(O+) - X’C+ bands of BrC1, which originate from the ground-state vibrational level u” = 0, are very 10w.l How- ever, absorption in bands of the v’’ = l progression, although weak, at 298 K has sufficient intensity for several of these hot bands to be observed in fluorescence. Part t Present address : Jet Propulsion Laboratory, California Institute of Technology, Pasadena, California 91 103, U.S.A.M .A . A . CLYNE AND I . S . MCDERMID 317 1' describes the observation and analysis of the 3-1, 4-1, 5-1, 6-1 bands of BrCl (B-X), as well as several bands with u'' = 2, including the 7-2 band. The relevant values of qv!,uJ! are given in table 1 of ref. (1); the maximum intensity factor ( O y q u ~ , v ~ ~ ) , where Ouft is the Boltzmann vibrational population at 298 K, is 3.0 x for the 6-1 band. The first quantum-resolved determinations of the fluorescence lifetimes of ro- vibrational states of excited BrCl were described in part 2., The low efficiencies for electronic quenching of BrCl(B) by C1, and BrC1, which had been reported by Wright et u Z ., ~ were confirmed in our work.* A further important result2 was the observa- tion of rapid vibrational energy transfer within the B-state manifold of BrCl in collisions with C1,. Approximate rate constants for V-V transfer were reported ;2 although, because of the long radiative lifetime of BrCI(B) and the high pressures used, it was not possible in our earlier work2 to eliminate multiple collisions. The occur- rence of multiple collisions (over the pressure range p = 30-200 mTorr) in our pre- vious work2 also entailed a long and somewhat uncertain Stern-Volmer extrapolation of l / z against p .Thus it was possible only to estimate a value for the collision-free lifetime of the stable levels of BrCl(B), which was zo = 35 f b1 An even longer Stern-Volmer extrapolation was used by Wright et and their study did not employ quantum-resolved excitation of BrCl(B). Wright et al.3 reported zo = 18 ps, which agrees surprisingly well with the present value of (40.2 & 1.8 p s ) . Because of improved technique^,^ fluorescence lifetimes of BrCl(B) have been measured in the present work at total pressures as low as 0.15 mTorr. Therefore, it has been possible to measure zo under collision-free conditions, and in addition, to determine directly the rate constants for vibrational energy transfer within BrCl(B). Observations on predissociated states have led to new, close limits for the dissociation energy of ground-state BrCl.EXPERIMENTAL In our previous studies of the decay kinetics of BrCl(B),, at pressures in the ranges 50- 200 mTorr and 3-11 Torr, the analysis of the fluorescence decay curves was complicated by rapid energy transfer processes in collisions of BrCI(B) with C1,. In the present work, we have used a low pressure test chamber [fully described in ref. (4)] to measure fluorescence lifetimes in the pressure range 0.1-1.0 mTorr. Based on a hard-sphere collision model, there are about three radiative lifetimes of BrCl(B) (i.e.. 120 ps), between BrCl(B) -t C1, collisions at a total pressure of 1.0 mTorr. Thus, these experiments were essentially colli- sion-free and fluorescence from states formed by collisional relaxation was not significant.Molecular bromine and chlorine were mixed in the ratio 1 : 15 to produce bromine mono- chloride in equilibrium with starting materials, Br2 f Clz $ 2 BrCl. It was necessary to use an excess of Clz in order to minimise the concentration of free Br2, which fluoresces strongly. Themixture was allowed to equilibriate for at least 24 h, and the final composition was 15 % BrCl and 85 % C12, with Br, < 1%. Fluorescence from resolved quantum levels of excited B311(O+) states of *'Br3jC1 was excited using a narrow-band tunable dye laser. This laser produced pulses of z 7 ns dura- tion with an energy of 5-50 pJ at 10-15 Hz repetition rate. The laser band width was 1 pm, and ethanolic solutions of Rhodamine 6G and Rhodamine B were used to cover the wave- length range 570-600 nm.Details of the laser, and its application to recording the B-X excitation spectra of BrCl, have been published previously.' Lifetime measurements were made by detecting total fluorescence intensity beyond 605 nm (Wratten no. 29 long-pass filter) with a S20 photomultiplier tube (E.M.I. 9558QB, 44 nm diameter cathode). Individual decay curves following each laser shot were captured318 STUDIES OF BrCl BY LASER INDUCED FLUORESCENCE using a fast transient recorder (10 ns/channel Biomation 8100), and were averaged in the hard-wired averager section of a computer (Nicolet LAB 80). For this study, 5000 shots were normally averaged, except for one or two weak lines at the lowest pressures, for which 7500 shots were necessary.Averaged data were stored on floppy discs and reduced under software control in order to give fluorescence lifetimes, using a weighted fitting routine.' Details of the data acquisition and processing techniques have been de~cribed.~ RESULTS LASER EXCITATION SPECTRA OF BrCl(B-X) The laser excitation spectrum of a molecule is obtained by recording undispersed fluorescence intensity as a function of laser wavelength. Use of narrow band width, as in this study, results in a high-resolution (Doppler-limited) spectrum that can be analysed in order to identify a defined ro-vibrational state (u', J ' ) . If excitation occurs under collision-free conditions, the laser excitation spectrum unequivocally defines the emitting species and its quantum state.I n the present work, excitation spectra of the B-X transition of BrCl have been obtained at total pressures near 5 mTorr, containing z 400 pTorr of BrCl. These studies approximate more closely to collision-free conditions than our previous studies using 5 Torr pressure.' Due to technique improvements, the signal-to-noise ratio of the present BrCl spectra was superior to that of the earlier high pressure spectra. Also, resolution was noticeably better, mainly because of detectable pressure-broadening effects at 5 Torr. Fig. 1 shows part of the laser excitation spectrum of BrC1, near the heads of the 4-1 bands, from 584.6 to 585.3 nm. All four isotopic bands have been rotationally assigned (see fig. l), whereas in previous work' only the more intensc pair of bands (79Br35C1 and s1Br3sC1) could be identified.The isotopic abundances are approximately 3 :3 : 1 : 1 for the four species 79Br35C1, s1Br35Cl, 79Br37C1, s1Br37C1. The appearance of the 4-1 bands is particularly simple, because the P and R branches of all four species are totally overlapped (within the Doppler line width ~ 0 . 0 3 cm-l) over a wide range of J (J" < 25). Thus, P(J) lines are blended with R(J + 4) lines in all four 4-1 bands (fig. 1). Rotational states were assigned from combination differences, and from the calculated vibrational shifts between the 79Br35C1, 81Br35C1, 79Br37C1 and s1Br37C1 species. Coxon's data6 for the rotational constants were used to form combination differences for ground and excited states: R(J) - P(J) = 4Bt:(J -t +), R(J - 1) - P(J + 1) = 4B:I(J + t).For the 4-1 band, the rotational assignments were unequivocal. The PR branches of 79Br35C1 and 81Br35C1 could be followed to the 4-1 band heads. Except for the lowest few J-lines of the minor 79Br37C1 and s1Br37C1 species, these PR branches could also be assigned to the corresponding 4-1 band heads (fig. 1). The vibrational isotope shifts in the 4-1 bands, therefore, could be determined, with values as follows: gG79-81 - + (0.68 0.03) ern-'; and 6G35-37 = + (3.31 -+ 0.08) cm-'. These values were based on measurements of dG79-s1 for the Br3T1 species alone, and on measurements of 6G35-37 for both the *'BrCl and 79BrC1 species.B r C l R 79 35 P R P 81 35 M . A . A . CLYNE AND I . S . MCDERMID d 79 37 R /* 81 3 7 R \ i /f 4 17 1 16 15 15 1 14 1 13 14 13 I L l 11 I k 11 319 I I I i 584.8 584.6 I I I 585.2 5 8 5 .0 laser w a v e i e n g t h l n m FIG. 1.-Laser excitation spectrum near the band heads of the 4-1 bands of BrCl (B-X). assignment to all four isotope species : 79Br35C1, *lBr3’CI, 79Br37C1, a’Br37C1. this region, due to overlapping of P(J) + R(J t 4) lines in all bands. Note Spectrum is simple in For the minor (37C1) isotopic species, only the R lines are assigned in the figure. An approximate expression for 6Gi, which has been found to be valid within 50.10 cm-l for B-X bands of C12(12 3 v’ 3 7),7 can be given for BrCl bands with u” = 1, by eqn (I): 6G’ = CO’,(~’ + +)($ - 1) - CU;X:(V’ + $)2(p2 - 1) - A . (1) In eqn (I), o: = 222.68 and okx: = 2.884 cm-’ for 79B?35Cl;6 p 2 is the ratio of the corresponding reduced masses; and A is the vibrational isotope shift in the ground state levels u” = 1.For 6G79-81, A = 2.53 cm-’; and for 6G35-37, A = 12.71 cm-l. The calculated vibrational isotope shifts were 6G79-s’ = 0.83 cm-’ and 6G35-37 = 4.17 cm-l. These calculated values show some discrepancy with the observed magnitudes ; however, the ratios 6G’ (observed)/dG‘ (calculated) are similar for the two isotopic shifts, namely 0.82 for 6G79--81 and 0.79 for 6G35-37. Thus, the observed vibrational isotope shifts are self-consistent. It is not unexpected that the simple anharmonic- oscillator approximation [eqn (I)] gives slightly inaccurate values for 6Gi, in light of the known irregularity of the excited-state vibrational-energy levels in the B311(O+)320 Br CI 79 35 R P 81 35 R P 79 37 R P 81 37pR n 79 35 R P 81 35 $ STUDIES OF BrCl BY LASER-INDUCED FLUORESCENCE 31 29 27 25 I as er wavelen g t h I n m FIG.2.-Laser excitation of the 4-1 and 5-1 (high J ) bands of BrCI(B-X). In this wavelength region, PR doublets are beginning to separate in the 4-1 bands (compare fig. 1). Black spots show 79Br35C1 and 81Br35C1 lines of the 5-1 bands. manifold of BrC1.6 Coxon6 has shown that it is not possible to describe these energy levels by a simple polynomial function; in addition, the energies of the lowest two vibrational levels (u’ = 0, 1) are uncertaim6 Our previous rotational assignments for the 6-1 band [see fig. 2 of ref. (l)] have been checked, using the improved excitation spectra now obtained.The rotational assignments of the 79Br35C1 band were confirmed; however, that of the 81B135Cl band requires a minor modification. The R(J) and P(J) assignments in the 6-1 band of 81Br35C1 need to be exchanged. Thus, R(J) should be P(J - 4) and P(J) should be R(J + 3). For instance, R(31) should be P(27) and P(27) should be R(30). PREDISSOCIATION OF T H E L E V E L u’ = 7 OF BrCI(B) We first concluded that the entire u‘ = 7 level was predissociated from observa- tion, in fluorescence, of the excitation spectra of the overlapping 7-2 and 5-1 bands.l In these spectra, recorded at 5 Torr total pressure, the u’ = 7 band was found to beM . A . A . CLYNE A N D I . S . MCDERMID 321 less intense, by about a factor of 5, than expected from predictions based on the Franck-Condon factors and the thermal populations.In Part 22 of this series, the lifetimes of some levels in the 7-2 band were measured at 70 mTorr pressure, and were found to decrease from 0.9 p s for J’ = 7 down to 0.2 ps for J’ = 19. At much lower pressures in this work, the intensity of fluorescence from u’ = 7 was too low to enable systematic measurements to be made. Some measurements were made, however, and these confirm that the lifetimes of states with 21’ = 7 are < 1 ps. It was found, as for BrF,4 IF9 and Cl,,7 that when the excitation spectrum was recorded at low pressures (< 5 mTorr) to minimise collisions, all predissociated transitions were absent. For example, the excitation spectrum of the B3H(O+)- X’C’ transition of BrCl, when recorded at low pressure, showed no lines of the 7-2 band. Fig.2 is a part of this laser excitation spectrum from 585.6 to 586.2 nm, which shows intense 4-1 and 5-1 band lines, but no 7-2 band lines. The same spectral region recorded at 5 Torr pressure [see fig. 3 of ref. (l)] showed the 7-2 band with an intensity approaching that of the 5-1 band. This provides additional confirmation that the entire u’ = 7 level is predissociated. PREDISSOCIATION OF THE LEVELS u’ = 5 AND 6 OF BrCl(B) The higher rotational levels of (J’ 3 42) in u’ = 6 could not be observed in laser excitation spectra recorded at low pressures of BrCl. The chamber pressure was therefore increased to z 40 mTorr, whereupon transitions to these levels were easily observed; thus the laser wavelength could be tuned to the exact frequencies of these transitions.Once the laser had been tuned to the transition, the test pressure was reduced, and the lifetime was determined from fluorescence decay measurements. The onset of predissociation was sharp at J’ = 42. The lifetime of J’ = 41 was 40.6 p s , falling to 8.8 p s for J’ = 42. Thus, predissociation has been observed for the first time in the higher rotational levels of u’ = 6. The lifetimes of the stable and predissociated levels will be discussed below. This observation of predissociation in u’ = 6 led us to search the highest rotational state of u’ = 5 for a predissociation, in order that the dissociation energy and type of curve crossing could be accurately determined. Excitation spectra were carefully recorded at low pressure to eliminate the 7-2 band, and to facilitate observation of predissociation from the excitation spectrum.It was possible to assign all four isotopic P and R branches of the 4-1 band in this spectrum, and transitions involving high J values (J‘ = 59 to 67 in fig. 2) of the 5-1 band are clearly identified. For the higher J’ lines of the 5-1 band, additional overlapping at the 3-1 band head, commencing at 59 1.18 nm, precluded unambiguous assignments of transitions with J’ 3 71 in the state u’ = 5. However, we have shown that levels up to and including J’ = 70 in u’ = 5 are stable and this does allow the calculation of limits for the dissociation energy. Fig. 3 shows the relative total energies of the rotational lines in the states u’ = 5, 6 and 7. Fig.2 shows a typical example of these spectra. CURVE-CROSSING WITH THE B STATE; A N D GROUND-STATE DISSOCIATION ENERGY OF BrCl Although predissociation was not found in the state u’ = 5, its rotational levels up to and including J’ = 70 were stable. This result can be used to determine the maximum value of the internuclear distance, r, corresponding to the potential322 STUDIES OF BrCl BY LASER-INDUCED FLUORESCENCE 65 60 .......................... 15 - _ _ _ _ - _ - d ......................... 40 10 .......................... ......................... 65 35 30 50 17.9 FIG. 3.-Energies of (d, J’) states of 81Br35C1(B) in relation to predissociation. limits of predissociation energy, relative to the state (0,O) of 81Br35C1(X) as the energy zero. shown as full lines are observed to be stable.(- - - -), predissociated levels (z d 10 ,us); (. . . . .), Hatched area shows Levels predissociated levels (z d 1 ,us); (- - - - -), levels not observed due to overlapping of spectra. energy maximum over which BrCl(B) must pass in order to predissociate.1° The rotational constant B, and thus the internuclear distance, may be calculated from the difference in energy of the last stable (or first unstable) rotational levels in the con- secutive vibrational states u’ = 5 and 6. Assuming that the (u’, J ’ ) state (5, 71) is predissociated, a lower limit for B is provided, and thus an upper limit for r. The energy relations are as follows: AE = E (5,70) - E(6,41) = B(70 x 71 - 41 x 42) = 149.1 cm-l. Thus B 3 0.0459 cm-’ and r < 3.87 A. If the same calculation is made for increasingly higher J‘ values, the following results are obtained: for example, J’ = 75, r = 3.52 A; and J’ = 80, Y = 3.32 A.The minimum possible value for Y is the position of the right-hand limb of the potential curve at the predissociation energy, which in this case is E 3.0 A [see fig. 1 of ref. (l)]. If the potential energy maximum is located close to the right-hand limb of the B-state potential curve, then the type of crossing is Herzberg’s case I(c), which is caused by an avoided crossing with a repulsive state.lo The above results are compati- ble with such a curve-crossing, and in the interhalogens the repulsive curve is believedM . A . A . CLYNE A N D 1. S . MCDERMID 323 to be an O+ state. The calculations also suggest that predissociation in v' = 5 probably occurs in the energy range between J' = 80 and 90.An experiment at higher temperatures, to increase the Boltzmann populations of the corresponding ground- state rotational levels in order to populate J' > 80, would be interesting. The predissociation energies lead directly to an estimate for the dissociation energy D! of ground-state BrCl(X'C+). The effective B-values, which we have calculated, can be used to make a correction to the dissociation energy, in order to allow for the rotational energy barrier. An upper limit for D: can be found using ( 5 , 70) as the last known stable level. The effective B-value, 0.0481 cm-', can be used to make a correction to the energy at the known predissociation in the state (6, 42), namely: Dt(BrC1) = E(6, 42) - B(42 x 43).Thus the upper limit is Dt("Br3'CI) < 17 959.9 cm-'. A lower limit for D;(BrCI) can be found by calculating an effective B for an internuclear separation of 3 A (see above). As before, using B = h/8n2r2pc, this gives B = 0.0767 cm-' and leads to a lower limit Dt(s1Br35C1) 3 17 908.3 cm-l. The mean value deduced for Dt(s1Br35C1) is therefore (17 934 5 26) cm-I or (214.49 & 0.31) kJ mol-l. As has been pointed out in connection with the B311(O+) states of BrF4 and IF,8 measurements of predissociation energies give improved upper limits for Do0 when expriments are performed under collision-free conditions. In this way, our previous upper limit value' for Dt(BrC1) from observations of predissociation at higher pres- sures, Di(BrC1) < 18 035 cm-', has now been reduced to the range of values Do0 (81B135Cl) = (17 934 Earlier determinations of D:(BrCl) have been discussed previously; ' we note here that the new value of (17 934 & 26 cm-') should be more reliable than the thermo- chemical datum, D:(BrCl) = (18 010 + 100) cm-'.The agreement, however, is within the error limits, with a discrepancy of 0.9 kJ mol-1 between the values. 26) crn-'. LIFETIMES OF PREDISSOCIATED LEVELS I N u' = 6 The last stable level in v' = 6 was J' = 41 which had a collision-free lifetime, zo, of 40.6 ps. Above this level, zo fell monotonically from 8.8 ,us for J' = 42, down to 4.3 p s for J' = 50, which was the highest level measured. The logarithmic decay curves were used to identify the initial region of exponential decay, and the lifetime was calcu- lated from this section only.The plots were linear over, typically, the first 300 chan- nels, i.e., 15 ps, after which a low-intensity long-lived component was observed. This was probably due to a small extent of rovibrational relaxation into lower-energy, stable states, even at the low pressure of 10 mTorr at which these measurements were made. The data gave a good fit to an equation of the form : l/z0 = cc +- kJ'(J' + 1); and a plot of l/zo against rotational energy B'J(J' t 1) is shown in fig. 4. The inter- cept of the plot, a, was of small magnitude, and no obvious interpretation can be made of this. The gradient, k , however, is an indication of the strength of the rotationally- dependent predissociation and the value was k = 1.60 x lo4 s-'.The magnitude of k can be compared with those found for rotationally-dependent predissociations in u' = 8 of 79BrF and 81BrF,4 and in v' = 10 of IF:6 namely 79BrF, k = 4.0 x lo4 s-l; ''BrF, k = 3.2 x lo4 s-'; IF, k = 2.8 x lo4 s-l. LIFETIMES OF THE STABLE LEVELS OF BrCl(B) Fluorescence lifetimes were measured as a function of pressure for the stable (u', J ' ) states that are accessible in the B state manifold, i.e., the vibrational levels 6 3 u' 3 3.324 STUDIES OF BrCl BY LASER-INDUCED FLUORESCENCE P I 150 200 2 50 B’I‘( /‘+ 1 I / cm“ FIG. 4.-Predissociated levels in u‘ = 6 of 81Br35C1(B). Plot of 1 /zo against rotational energy is shown. The range of total pressures was 0.1-1.0 mTorr; of this pressure, z90 mol % was C1, and z 10 mol % was BrC1.Fig. 5(a) shows a typical fluorescence decay curve, based on the average of 5000 laser pulses. The data shown are for excitation of the (6, 14) state of 91Br35C1, using absorption in the R18 line at 0.51 mTorr total pressure containing 40 pTorr partial pressure of BrCl. B311(02) -XIC: fluorescence of C12 is very weak using excitation wavelengths near 580 nm,’ and thus did not interfere with the studies of BrCl(B-X) fluorescence. The lifetime data, obtained as a function of total concentration [MI, were analysed by the Stern-Volmer formulation, according to eqn (11) : l / Z = I/Zo + kM[M]. (11) Using eqn (II), values for the collision-free lifetime zo, and the rate constant k, for collisional depletion of the initially-excited state, may be obtained.Fig. 5(b) shows a typical set of results for the variation of z with pressure, plotted in the Stern-Volmer form. The data shown in fig. 5(b) are for initial excitation of the (6, 14) state, including the result of fig. 5(a). The interpretation of zo values as the radiative lifetimes of states BrCl(B) is straight- forward. However, several different processes may contribute to kM, the overall rate constant for collisional depletion of BrCl(B). k , may include terms due to electronic quenching, and collision-induced predissociation. We shall show below that, for collisions of BrCl (B, v’ 3) with Cl,, this effect is dominant for collisional depletion of excited BrCl, with kM = k,, where k, is the rate constant for V-V energy transfer. Table 1 summarises the data for zo and k, for the stable ro-vibrational levels that were studied.A range of resolved rotational states (60 3 J’ 3 6) were studied; these were the states that were accessible through light absorption by ground-state molecules within the Boltzmann rotational envelope at 298 K. No significant de- pendence of zo, nor of k,, upon rotational energy was observed in the states u’ = 4, 5 and 6 ; the data for v’ = 3 were not extensive. The variations in z, with u’ were within the standard errors of the determinations; Upward vibrational ladder climbing can also result in predissociation.M . A . A . CLYNE A N D I . S . MCDERMID 325 t l p s 1 0 0.5 1 t o t a l p r e s s u r e I mTorr FIG. 5.-Lifetimes of the stable levels of BrCl(B). Typical data are shown for fluorescence decay following excitation of the (6, 14) state of 81Br35C1(B).(a) Total pressure of 0.51 mTorr (40 pTorr of BrCl); 5000 laser pulses. Data shown are based on 2048 time channels with three operations of 3-point smoothing to facilitate reproduction of the traces. Top, intensity against time. Bottom, In Z against t . Note increased noise at lower end of In Z trace, due to statistics of counting limitations. (b) Stern-Volmer plot from 0.19 to 0.99 mTorr total pressure.326 STUDIES OF BrCl BY LASER- INDUCED FLUORESCENCE the overall mean value for zo, based on all 31 sets of Stern-Volmei plots for 6 3 vf 3 3 of "Br3'C1, was zo = (40.2 & 1.8) p s (lo). However, a significant trend was noted for kM to increase markedly, with an increase in vibrational energy.Thus, k , increased from (6.2 5 2.3) x lo-" cm3 molecule-' s-I for v f = 4, up to (2.1 * 0.4) x As noted before, the data for uf = 3, which state had to be pumped by the weak 3-1 band, were not extensive enought to define kM adequately for this state. A similar strong dependence of k , upon vf was reported in our earlier work' using higher pressures of BrC1. The values of k , reported previously (denoted k, in that work)' weie systematically lower than those now determined from experiments under single-collision conditions. It is probable that our previous measurements were affected by multiple collisions. Such collisions tend to vibrationally relax the initial uf state, and thus to reduce the rate of collision-induced predissociation via vibrational ladder-climbing.The value for zo reported previously' was zo = (35 t 6') pus, based on data for the states v f = 3 and 4. This value is in good agreement with the accurate value now reported, zo = (40.2 * 1.8) ps. However, it is noted that zo was underestimated in our previous work,2 probably due to a curved Stern-Volmer plot. This result is consistent with the occurrence of multiple collisions, which have been invoked above to explain the under-estimation of k , in our higher-pressure study.2 cm3 molecule-' s-l for v f = 6. TABLE 1 .-LIFETIMES OF STABLE LEVELS OF 81Br35C1 B3n(O+) (u) summary U' no. of runs range of J TOIPS kM/lO-ll cm3 molecule-' s-' J' 6 10 15 20 25 30 35 40 45 50 55 3 15-35 42.0 & 2.7 7 4-45 40.0 i 2.0 11 6-60 39.8 &- 2.0 10 10-41 40.0 i 0.7 (b) data for u' = 5" transition used no.of pressuresb R 5 6 P11 6 P16 5 R19 5 R24 6 R29 5 R34 6 R39 6 R44 5 R49 5 R54 5 - 6.2 & 2.3 9.6 & 2.8 20.7 -l 4.1 %IPS 42.7 38.9 37.6 37.7 39.6 37.9 39.0 39. I 41.3 42.2 42.6 mean value 39.8 & 2.0 a Full data for the states v' = 6,4, 3 are available as an Appendix from the authors. pressures (0.25-1 .OO mTorr total pressure) used to define a Stern-Volmer plot for one J' value. No. of .M . A . A . CLYNE A N D 1. S . MCDERMID 327 DTSCUSSION INTERPRETATION OF kM VALUES ; VIBRATIONAL ENERGY TRANSFER I N BrCl(B) It has been shown previously that electronic quenching of excited BrCl(B) by BrCl(X) or Cl,(X) has a low collisional efficiency. For mixtures of BrCl + Cl, similar to those used in the present w o ~ k, the rate constant (17,) for quenching of BrCl (B) was reported to be 3.4 x [ref.(3)] or 3.9 x [ref. (2)] cm3 molecule-' s-' at 298 K. Rapid vibrational energy transfer was found previously;2 the observed strong dependence of k , upon u' (table 1) indicates that the rapid collisional depletion of BrCI(B) found in the present work was controlled predominantly by upward vibrational energy transfer. Such vibrational ladder-climbing can cause initially-stable molecules in states 6 3 u' 2 3 to pass into predissociated states u' 3 7, which are unstable and fluoresce with negligi- ble quantum yields. A mechanism of this type is important not only in excited BrCl(B), but also in Cl,(B),* N here the probabilities of vibrational energy transfer are higher for Cl,(B) + Cl,(X) collisions,' than for BrCl(B) + Cl,(X) collisions.For BrCl(B), the binding energies AE below dissociation have been given in table 2 of ref. (2). Taking the state (7,O) of 79Br35C1(B) as the energy zero, the AE values for the levels u' = 6, 5, 4, 3 are 151, 317, 495 and 684 cm-', respectively. These energy values for BrCl(B) are similar in magnitude to those of the four or five levels of Cl,(B) that lie immediately below its first predissociated state (0' = 12 in this case). Thus, the Boltzmann factors for the ratio of upward- and downward- vibrational-energy transfer (ku,l/kc,-l) in BrCl(B) and CI,(B) are fairly similar in mag- nitude; at 298 K, the calculated ratio ku,l/ku,-l is about 0.5 for the sets of vibrational levels Values of AE/kT for BrCl(B) range from 0.73 (u' = 6) to 3.30 (u' = 3) at 298 K; thus, vibrational transfer with IAul > 1 is expected to have an appreciable probability.Values of k , for BrCl(B) collisional depletion may be analysed in the same manner as has been employed to obtain vibrational state-to-state rate constants for C12(B).8 Thus, k , for initial excitation of state u' is the summation over Au of all rate constants k,,*, for energy-transfer steps which can form an unstable state with u' 3 7. For example, if u' = 5, then upward V-V transfer with Au = +2, +3, +4 . . . can all deplete the concentration of unpredissociated (stable) excited BrCl : k , is evidently much less than the measured k, values. k M , 5 = k 5 , 2 + k 5 , 3 + k5,4 - - * Assuming, as for C12(B)," that ku,hu is not a strong function of u, we may write the following equations for the values of k , determined in this work for u' = 6, 5 and 4: k M , 6 = k u , l + ku,2 + ku,3 + kv.4 + * - * k M , 5 = k M , 4 = k",3 + ku,4 + * - * * kU.2 + ku,3 + ku,4 + - - * Using the data of table 1, the following results therefore may be deduced for collisions of BrCl(B, u') with Cl,: kU,' = kM,6 - kM,5 = (1.1 & 0.4) x lo-'' cm3 molecule-'s-'; ku,2 = kM,5 - kM,4 = (3.4 & 2.0) x lo-'' cm3 molecule-' s-'; ku.3 + ku,4 + ku,5 + .. . = kM,4 = (6.2 k 2.3) x The corresponding rate constants for downward vibrational transfer in collisions of BrCl (B) with C1, are about a factor of two larger than these data for upward transfer cm3 molecule-'^'-^.328 STUDIES OF BrCl BY LASER-INDUCED FLUORESCENCE (see above).Thus, the hard-sphere efficiencies for BrCl(B) + Cl,(X) collisions, in respect of vibrational transfer, are ~ 0 . 6 for Au = -1, and ~ 0 . 2 for AZJ = -2. The magnitudes are fairly similar to those reported* for Cl,(B) + Cl,(X) collisions, al- though smaller as might be expected. The above analysis to obtain kv,hv values is necessarily approximate, but is unaffec- ted by multiple collisions of BrCl(B), which were extremely improbable under the low- pressure conditions employed in the present experiments. Also, since k , was found to show no significant trend with J', it is valid to neglect the effects of rotational energy transfer in the case of BrCl(B) + CI,(X) collisions. In other cases, such as during collisions of BrF(B) + He,ll vibrational transfer is slow and rotational energy transfer plays a dominant role in the collisional kinetics of the excited interhalogen.'l RADIATIVE LIFETIME A N D DIPOLE MOMENT OF THE B-X TRANSITION OF BrCl As shown in table 1, there was no significant trend in the collision-free lifetime zo of BrCl(B), with u'. has been made for the B-X transition of BrC1, using the accurate mean lifetime, zo = (40.2 1.8) ps. The relevant analysis and discussion has been given elsewhere., The resulting mean value of was (0.96 & 0.10) x lo-' D2, without any significant variation with u'. This value may be compared with our previous value of (1.1 & 0.1) x lo-' D2, based on zo = (35 2 bl) ps. A similar change is required to the absorption coefficients given in part 2., However, the alterations in the quantities are small; and our previous suggestion of BrCl(B-X) as a possible laser for optical or chemical pumping,, is unaffected by the more accurate results now reported. Recalculation of the electric dipole moment We thank Steve Davis for helpful discussions and are grateful to the S.R.C. and the U.S. Air Force Office of Scientific Research (grant no. AFOSR-75-2843) for sup- port of this work. M. A. A. Clyne and I. S. McDermid, J.C.S. Faraday ZI, 1978,74,798. * M. A. A. Clyne and I. S. McDermid, J.C.S. Faraday ZZ, 1978,74, 807. J. J. Wright, W. S. Spates and S. J. Davis, J . Chem. Phys., 1977, 66, 1566. M. A. A. Clyne and 1. S. McDermid, J.C.S. Faruduy ZZ, 1978, 74, 1376. M. A. A. Clyne and M. C. Heaven, J.C.S. Faraday ZI, 1978, 74, 1992. J. A. Coxon, J. Mol. Spectr., 1974, 50, 142. ' M. A. A. Clyne and I. S. McDermid, J.C.S. Faruday IZ, 1978, 74, 1935. * M. A. A. Clyne and I. S. McDermid, J.C.S. Furnduy ZZ, 1979, 75, 131 3. M. A. A. Clyne and I. S. McDermid, J.C.S. Faruday I/, 1978, 74, 1644. lo G. Herzberg, Spectra of Diutornic Molecules (Van Nostrand, N.Y., 1953). l1 M. A. A. Clyne and 1. S. McDermid. J.C.S. Furodai, 11. 1978, 74. 644

ISSN:0301-7249    

DOI:10.1039/DC9796700316

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

Crossed Beam Studies of Chemiluminescent, Metastable Atomic Reactions Excitation Functions and Rotational Polarization in the Reactions of Xe(3P,,0) with Br, and CCl, BY CHARLES T. RETTNER AND JOHN P. SIMONS Chemistry Department, The University, Birmingham B 15 2TT Received 1 1 th December, 1978 The reactive scattering of Xe(3P,,0) by Br, and eel, has been studied under crossed beam condi- tions, using a supersonic, rotor-accelerated metastable atomic beam, a nozzle expansion cross-beam and a chemiluminescence monitoring system. Excitation functions for the production of XeBr [B(+), (C+)] and XeCl [B(+), C(+)] have been determined over the collision energy range, E, < 125 kJ mol-', as well as the energy dependence of the fluorescence polarization (and hence the product rotational polarization).The fluorescence spectra of the xenon halides have also been recorded under crossed beam conditions, at thermal energies. The results display a close similarity to the behaviour observed in the reactions of Cs with Br, and eel, and have been interpreted in terms of a transition towards a spectator stripping regime at high collision energies (Et 2 100 kJ mol-') in the reaction of Xe(3P2) with Br,, but a more complex interaction with Ccl4 where steric considerations should be considered. All of the polarization measurements indicate electronic transitions in the rare gas halides polarized parallel to the internuclear axis (AQ = 0). " One can have a great deal of fun, and indeed do some very interesting experiments, with levitated rotors .. ."I Some of the most interesting are those which have evolved from the pioneering molecular beam study of the reaction between Cs and CC14, reported by Bull and Moon2 at the last Faraday Discussion held in Birmingham, twenty-five years ago. Lately the use of rotors to accelerate molecular beams has been revived and greatly improved by the development of magnetically levitated rotors incorporating a carbon fibre composite (c.f.c.) haft.^-^^ These now permit the ready production of velocity selected pulsed molecular beams travelling at velocities <3 \ km s-1,394*7 with intensities z 10'' sr-I s-' . 4 * 7 Unlike seeded nozzle beams, with which they may usefully be ~ompared,~ rotor accelerated beams are generated from a single molecular species and are well suited to the production of supersonic, metastable atoms (or molecules) by directing the accelerated beam through an elec- tron b~mbarder.~-' Reactive or inelastic scattering of the metastable atoms is most conveniently studied by monitoring the emissions from luminescent products at the intersection with a second, reagent cross-beam. The technique has been applied initially to studies of the reaction of Xe(3P2,,) atoms with Br, and CCI, and the excitation functions, luminescence spectra and rotational polarizations of the reaction products XeBr* and XeCl* have been measured.The experiments offer a high level of state selection in the reagent channel through fine tuning of the rotor speed, the use of time-of-flight methods, the introduc- tion of an electron bombarder and the incorporation of a nozzle source in the cross- beam.Internal state selection in the product channel can be determined through analysis of the chemiluminescence spectrum and its polari~ation.~***~ The depend- ence of the internal energy disposal in the fluorescent products and their spatial330 Xe(3P2,0) + Br, OR CC14 REACTION polarizations are obtained directly and the excitation functions for the reactive chan- nels can be measured quickly and simply. The close analogy between the chemistry of the alkali metals and the (long range) chemistry of the metastable rare gas atoms is well documented." In some respects, the studies of the reaction + CCl, --+ XeCl[B(+) ; CCl, included in the present work, mirror those reported2 at the beginning of the " alkali age " l1 of molecular beams.EXPERIMENTAL CROSSED BEAM SYSTEM The essentials of the crossed beam system are shown in fig. 1 . A small fraction ( E lo-') of each rotor-accelerated molecular beam pulse is electronically excited during passage through the electron bombarder and chemiluminescence emitted at the cross-beam inter- - FIG. 1.-Schematic representation of the crossed beam system. A, rotor; B, axis of He/Ne laser; C, collimating aperture; D, E, electron bombarder grid and fiiament; F, ion deflector; G, light baffles; H, nozzle beamshroud; I, nozzle beam source; J, Wood's horn (or channeltron electron multiplier); K, lenses and photomultiplier; L, cryogenic shield; M, radiation shield; N, differentially pumped cross-beam vessel. section zone is monitored along the third perpendicular axis, using photon counting and multichannel scaling techniques.All three vacuum chambers were constructed from borosilicate glass, to avoid interaction with the r.f. fields which drive the rotors. In practice, the drive fields were not seriously perturbed by remote metal objects and it is not essential that the scattering and nozzle beam chambers be non-metallic. The vacuum requirements for studying chemiluminescent systems are modest in comparison with systems employing mass-spectrometric detection and the metastable reagents decay on collision at a surface. Although ultimate pressures x Torr* could be obtained in the system, in use the pressures in each chamber were x Torr (rotor chamber), 510-3 Torr (nozzle chamber) and Torr (scattering chamber).METASTABLE BEAM SOURCE The magnetically levitated, c.f.c. rotor (the construction of which has been described Torr) to generate a * 1 Torr = 133 Pa. el~ewhere)~~' was spun in a constant pressure of gas (typically xC. T. RETTNER AND J . P . SIMONS 33 1 pulsed molecular beam comprising x 10" atoms Maximum intensities were achieved using blade ended rotor shafts, but at the expense of a reduction in the maximum attainable rotor tip speed. In the early experiments blade ended rotors were preferred for speeds G1.4 km s-' (frequencies <3 kHz) while the higher speeds 5 2 km s-' (frequencies 54 kHz) were achieved with conical ended rotors. More recently, rotors have been con- structed from a stronger c.f.c. material developed at Harwell," which has raised the prac- tical limits to G3.5 and G5 kHz, respectively.The highest speed corresponds to a kinetic energy <2.5 A4 kJ mol-', where A4 is the relative molecular mass of the accelerated beam. After passing through a collimating aperture 3 cm in length and 3 mm in diameter, the rotor beam passed through an electron bombarder, 5.5 cm in length, operating at a constant emission current (typically z 4 mA) and accelerating voltage [20 V for excitation of Xe(3P2,0)]. Ion deflector plates and a series of light baffles were placed at the exit of the bombarder source. Only the metastable excited atoms survive long enough to reach the cross-beam interaction zone. When the channeltron was aligned along the axis of the metastable atomic beam, it could be used to measure the metastable flux as a function of the rotor speed, the gas pressure in the rotor vessel and the grid-cathode potential in the electron bombarder.Fig. 2 shows the relative Xe(3P2,0) fluxes arriving at the interaction zone as a function of the accelerating voltage, using (i) a channeltron electron multiplier (Mullard B419 BL/Ol) and (ii) the chemiluminescent reaction with a cross-beam of Br2 as the metastable atom detector. The slight displacement of the two curves can be understood if the efficiency of the channel- tron detection favours the more energetic 3P0 state while the reaction with Br2 favours the 3P2 state (see later discussion). The number of excited atoms per pulse passing through unit area remains sensibly independent of the rotor frequency, since the probability of excitation falls inversely with the velocity through the bombarder, while, to a fair approximation, the number of atoms in each pulse entering the bombarder is proportional to the mean molecular beam velocity.' Fig.3 shows the dependence observed experimentally; comparison of the incident and exit beam fluxes leads to an estimated excitation efficiency atom-' and an average metastable atomic flux z 3 x 10' cm-2 s-' at 3 kHz. I I T ? 10 20 30 LO 50 cathode-grid voltage (=electron impact energy) / V FIG. 2.-Relative flux of Xe(3P2,,,) as a function of the electron accelerating voltage for a rotor beam velocity of 1.6 km s-l. Open circles, detection by the channeltron electron multiplier; triangles, detection via chemiluminescent reaction with Brz.332 Xe(3P2,0) + Br, OR CCI,, REACTION e l I *.= 0 2 A00 * 1200 ' 1600 ' 2000 ' 2400 ' 2000 ' 3200, ' ' rotor frequency /Hz FIG.3.-Relative numbers of Xe(3P2,,,) atoms per pulse as a function of the rotor frequency. Dif- ferent symbols refer to different series of measurements. The pulsed beam of metastable atoms is superimposed on a steady thermal component, produced through excitation of the steady gas flow issuing from the rotor vessel. While the relative intensity of the thermal component is negligible before entry into the bombarder, the reduced pumping speed introduced by the series of light baffles at the exit, and their slower velocity, leads to a rise in the local pressure in the bombarder and a large amplifica- tion in the thermal component among the excited atoms.The two components are readily separated by the time resolved detection system, however, (see fig. 4 and later discussion) and the presence of the thermal beam, far from being a disadvantage, provides a very helpful reference signal against which separate experiments at different rotor speeds can be normal- ised. It also provides a convenient source for spectroscopic studies of crossed beam inter- actions at thermal energies under assured single collision conditions. m c 0 -tbase-line L thermal component ------ 0 266 time I2ps channel FIG. 4.-" Time of arrival " spectrum of XeBr* fluorescence showing the relative contributions from the pulsed rotor beam and the continuous thermal beam. Rotor frequency = 2300 Hz, correspond- ing to a most probable collision energy of z60 kJ mol-'.CROSS-BEAM SOURCE The cross-beam was generated by a nozzle expansion, using a system closely similar to that described by Cruse et aZ.13 It employed a nozzle aperture of 0.25 mm diameter, a skim- mer aperture of 1.5 mm and a nozzle-skimmer distance of 5 mm. The nozzle and skimmer assembly was surrounded by a copper box cooled by conduction to a cryogenic shield (see below and fig. 1) providing thereby a high pumping speed for condensibles. The nozzleC . T . RETTNER AND J . P . SIMONS 333 was held at a temperature of N" 420 K to prevent condensation and the emergent beam entered the scattering region via a collimating slit of dimension 6 mm x 2 mm. With a stagnation pressure z 50 Torr the stream temperature and velocity of the reagent gases were calculated to be 36 K and 350 m s-' (Br2) and 36 K and 425 m s- (CCl,); literature data for a com- parable source13 suggest a beam density x 10" cm-j at the collision zone.SCATTERING VESSEL AND OPTICAL SYSTEM The glass scattering vessel was equipped with a number of side ports through which the molecular beam sources and electrical feeds to a channeltron electron multiplier and to the electron bombarder could be passed. Cryogenic pumping was provided by blowing liquid nitrogen around a nickel-plated copper cylinder to which the nozzle bulkhead was attached. A concentric radiation shield reduced external heat transfer. Chemiluminescence emitted from the crossed beam interaction zone was collected along the third perpendicular axis by a fused (Spectrosil) silica lens suspended from the lid of the scattering vessel.The lens subtended a solid angle x0.3 sr, permitting a collection effi- ciency x 2%. Interference from scattered light was effectively eliminated by introducing a number of light baffles along the optical path and along the electron bombarder axis, and by placing a Wood's horn at the exit of the metastable beam in place of the channeltron (see fig. 1); all surfaces were painted matt black. In the measurement of excitation functions, light was passed through suitable optical filters and focused directly onto the cathode of a cooled, low noise photomultiplier (EMI, 9789B). Polarization ratios were determined by directing the fluorescence first through a sheet of polarizing film transmitting in the U.V.(which could be rotated through 360" about the optical axis), then through a suitable glass filter or a variable U.V. interference filter, and finally onto the cooled photomultiplier cathode. The variable interference filter could be set to transmit anywhere in the range (250-400) nm with a bandpass x 25 nm. Chemilumi- nescence spectra were recorded by focusing the light onto the entry slit of an f/4 scanning monochromator (Spex Minimate), equipped with a 1200 groove mm-' grating. Spectra were only recorded for reactions with the thermal beam of metastable xenon; this could easily be intensified (in the absence of a spinning rotor) by raising the pressure in the rotor chamber to 5 Torr, and/or increasing the electron bombarder emission current to give fluorescence signal count rates x 20 s-' with a bandpass of 5 nm.The dark count rate from the cooled photomultiplier was x 2 s-'. Three types of optical system were employed, depending on the experiment. DATA COLLECTION Tn all measurements which involved rotor-accelerated (as opposed to the continuous thermal) metastable atomic beams, the pulses from the electron or photomultiplier were monitored on a multichannel time-of-flight spectrum analyser (J. & P. Engineering). This incorporated a fast, 256 channel buffer store, which allowed sequential single photon counting into successive channels of widths variable between 1 and 5 , m After transfer into a per- manent store, the counting sequence was repeated following receipt of the next " start " pulse, derived from the chopping of a He/Ne laser beam by the rotor tip.A block diagram of the system employed in the measurement of polarization ratios is shown in fig. 5 and a typical time-of-arrival trace used in the determination of an excitation function is shown in fig. 4. Since the time-of-arrival traces resolved the spread in velocities of the rotor-accelerated species, the actual collision energy resolution could be enhanced by restricting integration of the signal count rate to a narrow, selected group of channels. Thus the achievable degree of reagent translational state selection could be greatly increased by focusing on a specified time of arrival. This persists in the chemiluminescence signal pulses, since the radiative lifetimes14 of the rare gas halide exciplexes are x2 orders of magnitude shorter than the t.0.f.channel widths. The chemiluminescence excited by the reaction of thermal atomic beams of Xe(3P2,0) was recorded by conventional photon counting techniques and the emission spectra were dis-334 Xe(J,,,) + Br, OR CCl, REACTION played on a chart recorder. In future experiments the emission spectra associated with pulsed rotor-accelerated beams may be derived by subtracting the thermal component from the nett spectrum recorded using the rotor beam system. It should not be difficult to monitor the onset of new chemiluminescent channels as the rotor frequency is increased. pre - @Zi- 6 polariser tl., strip - spectrum , filter ex ecade readout r 1 I -Lr-LJ- unit puke shaper 8l filter i FIG.5.-Schematic representation of the system used to measure fluorescence polarization ratios. RESULTS FLUORESCENCE SPECTRA Fig. 6 and 7 display the fluorescence spectra recorded during the reaction of a thermal beam of Xe(3P2,0) with Br, and CCl,. They closely resemble those observed by Setser and his c o - ~ o r k e r s ' ~ during reaction in a discharge flow system at low pressures. EXCITATION FUNCTIONS Relative fluorescence intensities were recorded for the reaction of Xe(3PP,,o) with Br, and CCl, at rotor frequencies in the range (800-3500) Hz using the t.0.f. detection system. Count rates were typically 2400 s-l for Br, and 250 s-l for CCl,. The fluorescence was passed through the broad band OX7 glass filter which was opaque in the visible, thereby excluding stray light, but transmitted between 240 and 400 nm, which covers most of the fluorescence emission spectrum for both reaction systems (see fig.6 and 7). This minimizes any errors that might arise following significant changes in the fluorescence spectral profile with increasing collision energy. (Experi- ments with narrower band-pass filters centred on the " primary " band in XeBr* at !z 280 nm and the " secondary " band at M 360 nm confirmed that the excitation func- tions determined at the two wavelengths were identical within experimental error). The relative cross-sections were determined by subtracting the thermal component from the total signal, scaling the integrated count rate to the relative metastable beam flux at each rotor speed (see fig. 4), correcting for the increase in the effective path length through the cross-beam in the c.m.frame with increasing rotor speed and, finally, correcting for any anisotropy in the fluorescence intensity distribution. The c.m.C. T . RETTNER AND J . P . SIMONS 335 f wavelength 1 nm FIG. 6.-Emission spectrum recorded during the crossed thermal beam interaction of Xe(3Pz,o) with Brz. Resolution, 5 nm. The shoulder at ~ 2 9 0 nm is attributed to emission from molecular bromine. ol,L-.L-. . , . . . . I ' 2LO 280 320 360 LOO 4.40 wavelength I nm FIG. 7.--Emission spectrum recorded during the crossed thermal beam interaction of Xe(3P2,0) with CCl,. Resolution, 5 nm. correction involved multiplication by the factor (ureI/umet), where ~ i ~ ~ , is the most prob- able relative velocity of the colliding beams and umet is the most probable velocity of the metastable beam, while the anisotropy correction involved scaling by the factor (1 - p / 3 ) where p is the measured degree of polarization16 [see next section and ref.(16)]. The relative cross-sections, related to an approximate absolute scale by a rough extrapolation to the values reported at thermal energies, 8b,15a*b are displayed in fig. 8. (The results for Xe + Br, supersede those reported earlier,' where the small aniso- tropy correction was not included). FLUORESCENCE POLARIZATION RATIOS The relative intensities of the thermal and pulsed components of the fluorescence signal were recorded initially, as a function of the polarizer angle. In all cases where there was a measurable anisotropy, the maximum signal corresponded to the polariza- tion direction aligned parallel to the most probable relative velocity vector of the colliding beams.The accumulated counts were summed in 25 channels centred around the most probable time of arrival and typically accumulation times <lo3 s were sufficient to achieve S/N ratios 3 2 5 : 1. The polarization ratios, p = (I1 - I l ) / ( I 1 + I l ) , where I1 and I 1 are the recorded intensities with the polarizer aligned parallel and perpendicular to the most probable relative velocity vector are], were measured over a range of rotor frequencies between 800 and 3600 Hz, in a series of over- lapping experiments. The polarization ratios for chemiluminescence emitted from Xe(3P2,,) + Br,, between 275 and 295 nm and between 340 and 365 nrn, and from Xe(3P,,,) + CCI,, monitored between 290 and 400 nm, are shown in fig.9. Each point represents the mean of several measurements.336 xe(3P2,,) + Br, OR Ccl4 REACTION energy /eY 0 0 25 0 50 0 75 1 00 1 2 5 la1 ! , I 1 I 0 ' 2 5 50 75 100 125 energy I kJ rnol-' FIG. 8.-Excitation functions for the reactions of Xe('P2,,) with (a) Br2 and ( 6 ) CC14. Absolute values are roughly scaled to cross-sections measured under thermal conditions by Setser and his co-workers.' energy / kJ mol-' FIG. 9.-Polarization of the fluorescence from XeBr(B, C) and XeCl(B, C) as a function of the collision energy in the crossed beam interaction of Xe(3P2,0) with Br2 and CC14. In the case of Brp the emission from XeBr(B) (solid circles) and XeBr( C) (crosses) was monitored separately.For CC14 (triangles) the chemiluminescence was monitored through a glass filter which transmitted most of the emission spectrum. The dashed horizontal line at p = 3 represents the limiting polarization.C . T . RETTNER AND J . P . SIMONS 337 DISCUSSION FLUORESCENCE SPECTRA The fluorescence spectra displayed in fig. 6 and 7 are associated with emission from XeBr* and XeCl*, respectively (together with a very minor contribution from Br,* in the former They are excited in the chemiluminescent reactions Xe(3P2,0) + Br2+ XeBr* + Br --+ Xe('So) + Br2* Xe('P,,,) + CC14-+ XeCl* + CC13 and are associated with electronic transitions from bound upper states of the rare gas halides to unbound or repulsive lower ~ t a t e s .~ ~ ' ~ ~ ' ~ The interaction of Xe('S,) + X(,P;,+) and of Xe+('P;,f) and X-('So) generates three low-lying [X($); A($, 3)] and three excited [B(+) ; C($) ; D($)] electronic states, respectively. Of the nine possible electronic transitions between them, four are polarized perpendicular to the inter- nuclear axis (AQ = -+1) and are expected to have very low dipole strength^,'^ while the other five, with parallel polarization (AQ = 0), are allowed. The intense bands at 280 (XeBr*) and 305 (XeCl*) nm, together with the oscillatory features lying to shorter wavelengths, have been assigned to the single transition B($) + X(3).I4 The broad maxima at FZ 360 (XeBr*) and FZ 430 (XeCl*) nm have been associated with the strongly allowed transition C(+) --+ A($), but ab initio calculation^'^ suggest that the broad continuum in XeCl* may also include contributions from the weaker B(+)+ A($) system.The bands located at 220 (XeBr*) and 235 (XeCl*) nm and associated with the D(+)--+ A'(+) tran~ition'~ could not be detected in the crossed thermal beam spectra, despite their high predicted dipole strengths l4 and their energetic accessibility. All the fluorescence polarization measurements obtained so far con- firm the proposed spectral assignments, since they indicate " parallel " transitions at both short and long wavelengths. The breadth and oscillatory nature of the emission spectra are consistent with high levels of vibrational excitation in the rare gas halides. On the basis of a band contour synthesis,8a Setser et aLSc estimate that two-thirds of the energy available in the reaction with CCl, at thermal collision energies appears as vibration in XeCl [I?($) or C(+)]; for the XeBr [I?($)] produced in the thermal reaction with Br2, we estimate (6) z 0.7-0.8.These vibrational energy disposals both fall short of the spectator stripping limits. EX CI T A TI ON F U NC TI ONS As expected for exothermic reactions proceeding without any threshold energy requirement, the cross-sections for reaction of Xe(3P2,0) with Br, and CCl, both fall steadily with increasing collision energy E, (see fig. 8). In the reaction with CCl,, the fall-off curve approximates a dependence of the form where am (z8 x cm2) and a ( z -10 kJ mol-') are parameters insensitive to changes in E,; the curve for the reaction with Br, falls off rather less steeply.Qualitatively, the behaviour with Br, parallels that observed in the reactions of alkali metals with halogens, where the reaction cross-sections have been discussed in terms of an electron-jump mechanism modified by an orbiting criterion l7-I9 [which may lead to a dependence of the form of expression (3)] , 2 0 y 2 1 or in terms of the velocity338 Xe(,P,,,) + Br, OR CCl, REACTION dependence of covalent-ionic curve-crossing probabilities 22 or a combination of both.20 The detailed comparisons should not be drawn too naively, however; the degeneracy of the metastable 3P2,0 and ionic 'Pa,+ states of the rare gas atoms contrasts with the much simpler neutral 2S+ and ionic 'So states of the alkali metals which greatly reduces the number of potentially reactive surfaces and imposes a symmetry conservation con- straint against reactive broadside collisions." Such constraints are absent in the highly degenerate metastable rare gas-halide systems, though Setser and his co-workers 8c have observed that more states are generated by the possible product channels than can be accommodated by the multiplicity of the metastable reagent channels; they suggest the possible involvement of intermediate ionic states derived from excited X2- ['lI(3, 9 1 ions as well as the ground X2- ['C($)] ions.For illustration, schematic minimum symmetry correlation and potential energy diagrams for the interaction of Xe(193P) and Br, are shown in fig. 10 and 11. They indicate (1) the possibility of populating XeBr(B) cia the direct reaction of Xe("*) over the lowest ionic potential surfaces, (2) that XeBr(C) may be populated following transfer onto the ionic surfaces correlating with the spin-orbit excited ion, Xe+('P+) but (3) that XeBr(D) is accessible only if the electron jump is delayed until the covalent surfaces intersect the more highly excited ionic surfaces correlating with Br,- ['II($)].On this basis, the low rate of reaction into XeBr(D) can be understood, while the con- tribution made by reaction of Xe(3P0) may be neglected on statistical grounds. [Assuming a relative 3P,: 3P2 population of 1 : 5 in the metastable atomic beam and a similar statistical weighting of 1 : 5 in the available reactive channels, only 4% of the observed products would derive from the reaction of Xe(3Po)].To summarize, there are good arguments for analysing the excitation functions of XeBr(B) and XeBr(C) in terms of a combined orbiting and multiple curve-crossing model similar to that developed by Gislason and Sachs.*O Similar considerations will apply to the production of XeCI(B and C ) from CC&. Ixe' + x, I FIG. 10.-Schematic minimum symmetry correlation diagram (for case c coupling) for the reaction Xe* + X2+ XeX* + X.C. T. RETTNER AND J . P. SIMONS 339 energy l e v ,5 10 FIG. 11 .-Schematic potential energy diagram for the reaction of Xe(3P2,0) with Br,. FLUORESCENCE AND ROTATIONAL POLARIZATION The use of the chemiluminescence monitoring technique lends itself to the study of rotational polarization in the fluorescent products of crossed beam interaction^,^.^^^^ though apart from the pioneering study by Jonah et al.few experimental studies have been reported.6 Assuming cylindrical symmetry about the reagent relative velocity vector, ure,, the corresponding distribution function for the product rotational angular momentum vector J', of the rare gas halide, can be expressed in terms of a Legendre polynomial e x p a n s i ~ n , ~ . ~ ~ where 0 is the polar angle measuring the angle between J' and urel. If this distribution is anisotropic, it will be reflected in a polarization of the fluorescence monitored along an axis perpendicular to the plane of the intersecting beams. In the classical limit (appropriate, in view of the high angular momenta of the colliding beams) the fluores- cence will tend to be polarised either parallel or perpendicular to the relative velocity vector, depending on the orientation of the transition dipole.Jonah et al.9 have shown that for " parallel '' (An = 0) and " perpendicular " (An = & 1 ) transitions the polar- zation ratios will be, respectively, - 3a, P" = 20a, - a2 ( 5 ) and340 Xe(3P2,0) + Br, OR CCl, REACTION Normalisation of the distribution function and use of the orthogonality relationship for Legendre polynomials, gives a, = 3 and (COS2U) = +($a, + 3) (7) and for transitions with AQ = 0 or & I 3 - p J - respectively. These expressions provide a simple relationship between the observed fluorescence polarisation, the orientation of the transition dipole vector and the product rotational anisotropy. The magnitude of the relative velocity vector is easily varied by altering the rotor speed, which allows the rotational anisotropy to be monitored over a wide range of collision energies.A simple calculation' shows that the angular spread of the relative velocity vectors is much too narrow to introduce any significant " blurring " TABLE VA VALUES OF THE ASYMMETRY PARAMETER, a2 reaction collision a2lao (cos2e) ref. energy/kJ mol-I 7 -0.14 & 0.14 -0.14 & 0.14" 24 -0.4 & 0.2 33 -0.9 f 0.3 48 -1.6 & 0.4 -1.5 0.4" 68 -2.0 & 0.3 89 -2.2 f 0.2 116 -2.3 & 0.2 0.31 0.31" 0.28 0.22 0.1 1 this work 0.10" 0.06 0.04 0.03 8 -0.6 & 0.06 0.32 71 -0.7 & 0.2 0.24 107 -0.9 & 0.14 0.21 Xe(3P2) + CC14 43 -0.5 f 0.2 0.27 this work K + Br2 Cs + Br2 Cs + CC14 Cs + CH31 K + HBr Cs + HBr Cs + HI CS + SF, thermal (510) - 0.6 9 , -0.5 9 , -0.3 9 , -0.2 9 , - 0.6 8 -2.14 & 0.30 9 -1.9 k 0.2 10 -2.2 & 0.2 0.25 0.27 0.29 0.31 Herschbach and co- 0.25 workers, ref.(24). 0.05 0.08 0.04 " Fluorescence monitored on the " secondary " C+ A band of XeBr* at ~ 3 3 6 nm. of the observed anisotropy, except at thermal collision energies. Table 1 lists values of the asymmetry parameter a2, determined over a range of collision energies, together with earlier results for the reaction of alkali metal atoms at thermal energies, reported by Herschbach and his co-workers 24 using the electric deflection technique. Three types of system can be distinguished:C. T. RETTNER AND J . P. SIMONS 34 1 (i) M + HX and M + CH3X, where the mass combination " heavy-light-heavy ", HLH, forces the product angular momentum J' to lie nearly parallel to the total reagent angular momentum vector J, and the high level of polarization reveals little information relating to the potential surface over which the reaction proceeds ; (ii) M + Xz, where the mass combination HHH is uniform, and the kinematics of the collision do not obscure the influence of the motion over the potential surface; (iii) M + CX,, where the mass combination is again uniform, but where the internal rotational angular momentum may be divided between both reaction products.Consider first the reaction between Xe(",) and Br, [type (ii)]. The crossed beam conditions are such that the total angular momentum of the colliding reagents J z L and in the limiting case of spectator stripping dynamics the product angular momentum J' must lie parallel to L, so that (cos20) -+ 0 and (a,/a,) -+ -2.5.This is exactly the behaviour observed for XeBr(B) [and XeBr(C)] at high collision energies, Et 100 kJ mol-l, but at the other extreme of thermal collision energies the polarization is very small and a, < -0.1. Under these conditions it must be possible for the re- pulsive energy released in the exit channel to cause precession of the product rotational angular momentum about L (or J ) over a wide range of angles, 0. This is consistent with the observed vibrational energy release at thermal energies being less than expected for a spectator mechanism. Similar behaviour is displayed in the reactions of Cs and K with Br224b (see table l), though the comparison may be clouded by their preference for collinear interaction.The polarization of the fluorescence of XeCl (B and C), produced in the reaction with CCl,, increases relatively slowly with the collision energy, but does not reach the limiting value p = ++ in the available energy range (see fig. 9). The contrasting behaviour compared with that observed with Br, can be ascribed to the disposal of both linear and rotational angular momentum in the CC13 fragment. Steric considera- tions suggest that in the latter case a larger proportion of the reactive trajectories will lead to momentum transfer between the fragments in the exit channel. This inter- pretation is reinforced when the relative behaviour of Xe(3P,) and Cs is considered.Table 1 shows a parallel reduction in the magnitude of the asymmetry parameter a, for the reactions with Br, and CCI,. There is also a marked difference in the velocity- angle distributions in the reactions of alkali metals with Br, and CCl,, where the scatter- ing pattern shifts from the forward hemisphere to a sideways d i r e c t i ~ n , * ~ * ~ ~ again consistent with an interaction between the product fragments. CONCLUSION The utility of rotor accelerated beams in studies of the molecular dynamics of the reactive scattering of metastable, electronically excited atoms has been demonstrated. The technique is particularly well suited to the determination of excitation functions and the dependence of product rotational polarization on the collision energy, and the current results reinforce the often-stated analogy between the chemistry of metastable rare gas and alkali metal atoms.While the initial studies have been restricted to chemiluminescent, reactive scattering, it should be possible to extend these to include laser induced fluorescence detection of electronically unexcited products, to study inelastic scattering processes and to capitalize on the time of flight method to record energy loss spectra and their angular distributions. Prof. P. B. Moon devised the rotor beam technique; Mr. S. Travers constructed the c.f.c. rotors; Mr. R. Dackus constructed the scattering, rotor and cross-beam glass vacuum chambers; Mr. J. D. A. Hughes of Harwell developed and supplied342 Xe(3P2,0) + Br, OR CC14 REACTION samples of ultra-high tensile strength c.f.c.; Dr.M. R. Levy and Mr. R. J. Hennessy contributed to different stages of the research. We are most grateful to all of them and to Prof. D. W. Setser for sending pre-publication manuscripts. Finally, we appreciate the financial support of the S.R.C., the Royal Society and the University of Birmingham. P. B. Moon, Rutherford Memorial Lecture, Proc. Roy. SOC. A , 1978, 360, 303. (a) P. B. Moon, M. P. Ralls and J. B. Saul, Bull. Inst. Phys., 1975, 25, 51 1; (b) M. P. Ralls, Ph.D. Thesis (Univ. of Birmingham, 1975); (c) A. J. Barker, P. B. Moon and M. P. Ralls, to be published. P. B. Moog, C. T. Rettner and J. P. Simons, J.C.S. Faraday I], 1978,74,630. M. R. Levy, C. T. Rettner and J. P. Simons, Chem.Phys. Letters, 1978, 54, 120. ( a ) C. T. Rettner and J. P. Simons, Chem. Phys. Letters, 1978, 59, 178; (6) R. C. Estler and R. N. Zare, J . Amer. Chem. SOC., 1978, 100, 1323. (a) K. Tamagake and D. W. Setser, J . Chem. Phys., 1977, 69, 4370; (6) D. W. Setser, T. D. Dreiling and J. H. Kolts, J . Photochem., 1978,9, 91 ; (c) D. W. Setser, personal communication. C. D. Jonah, R. N. Zare and Ch. Ottinger, J . Chern. Phys., 1972, 56, 271. lo See for example: (a) E. E. Muschlitz, Science, 1963,159,599; (b) M. F. Golde, Gas Kinetics and Energy Transfer, (Specialist Periodical Report the Chemical Society, London, 1977), vol. 2, p. 123. D. R. Herschbach, Faraday Disc. Chem. SOC., 1973, 55, 233. ’ T. H. Bull and P. B. Moon, Disc. Faraday SOC., 1954, 17, 54. ’ C. T. Rettner, Ph.D. Thesis (Univ. of Birmingham, 1978). l2 J. D. A. Hughes, H. Morley and E.E. Jackson, 1978, AERE report No. 8727. l3 H. W. Cruse, P. J. Dagdigian and R. N. Zare, Faraday Disc. Chem. SOC., 1973, 55, 277. l4 P. J. Hay and T. H. Dunning, Jr., J . Chem. Phys., 1978, 69, 2209. l5 ( a ) J. E. Velazco and D. W. Setser, J. Chem. Phys., 1975, 62, 1990; (b) J. E. Velazco, J. H. Kolts and D. W. Setser, J . Chem. Phys., 1976, 65, 3468; (c) J. H. Kolts, J . E. Velazco and D. W. Setser, J. Chem. Phys., to be published. l6 V. Aquilanti, P. Casavecchia and G. Grossi, J . Chem. Phys., 1978,66, 1499. J. Maya and P. Davidovits, J . Chem. Phys., 1973, 59, 3143. l8 R. Grice and D. R. Herschbach, Mol. Phys., 1974, 27, 159. l 9 R. W. Anderson and D. R. Herschbach, J . Chem. Phys., 1975, 62, 2666. 2 o E. A. Gislason and J. G. Sachs, J. Chem. Phys., 1975, 62, 2678. 21 J. E. Barker and M. E. Weston, Chem. Phys. Letters, 1973, 19, 238. 22 M. S. Child, Mol. Phys., 1969, 16, 313. 23 D. A. Case, G. M. McClelland and D. R. Herschbach, Mol. Phys., 1978, 35, 541. 24 ( a ) C. Maltz, N. D. Weinstein and D. R, Herschbach, Mul. Phys., 1972, 24, 133; (b) D. S. Y . Hsu and D. R. Herschbach, Faraday Disc. Chem. Soc., 1973, 55, 116; (c) D. S. Y . Hsu, N. D. Weinstein and D. R. Herschbach, Mol. Phys., 1975, 29, 257; (d) D. S. Y . Hsu, G. M. McClel- land and D. R. Herschbach, J. Chem. Phys., 1974, 61, 4927. 25 S. J. Riley, P. E. Siska and D. R. Herschbach, Earaday Disc. Chem. SOC., 1979, 67, 27. 26 (a) K. R. Wilson and D. R. Herschbach, J. Chem. Phys., 1968, 49, 2676; (b) J. C. Whitehead, D. R. Hardin and R. Grice, Mol. Phys., 1972, 23, 787.

ISSN:0301-7249    

DOI:10.1039/DC9796700329

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

GENERAL DISCUSSION Dr. M. S. Child (Oxford) said: It is interesting that Prof. Setser should report significant electronic energy transfer R* + X2 + R, + XI for the systems Kr/Cl,, Kr/Br, and Xe/I, because, if this process is a byproduct of the ionic/covalent harpoon mechanism, it implies that a significant fraction of the en- counters on the ionic surface fail to negotiate the corner which leads to the product channel. In the language of the paper of Child and Whaley, this indicates the exist- ence of a second trapped trajectory on the ionic surface even at the thermal energies of Prof. Setser's experiments. Dr. M. A. D. Fluendy and Mr. D. Sutton (Edinburgh) said: The results presented by Setser et al. in which they observed substantial production of electronically excited bromine molecules in collisions with rare gas metastables provide yet another contrast to the behaviour of the analogous alkali metal/halogen molecule systems.Prelimin- ary measurements of the excitation produced in K/Br, collisions using the time of flight crossed beam technique described earlier in this Discussion' are shown in fig. 1. 0 10 A E/eV FIG. 1.-K/Br2 Energy loss profile at a collision energy of 164 eV (centre of mass). The results are the average of 27 separate observations at angles between 0.45 and 1.78". The main peak at AE 21 0 is broadened, presumably by vibrational excitation of the Br,, but possible electronic transitions with AE > 1.4 eV account for (2% of the observed scattering. The second main feature is due to the K41 isotope which occurs with 6% abundance and provides a useful magnitude comparison.In these M. A. D. Fluendy, K. P. Lawley, J. McCall, C. Sholeen and D. Sutton, Faraday Disc. Chem. Soc., 1979, 67, 41.344 GENERAL DISCUSSION alkali metal/halogen systems, interactions between the ion pair potential surface and the surface leading asymptotically to the excited halogen molecule occur at extremely large distances and cannot therefore be an important route to these excited states. In contrast, as suggested, in the rare gas metastable analogue systems the same V(RG+, Br;)/ V(RG, Br;) interaction lies inside the initial electron transfer radius [ V(RG*, Br,)/V(RG+, Bri)]. Dr. W. M. Jackson (Washington, D.C.) said: We have combined the laser induced fluorescence technique with vacuum ultraviolet flash photolysis to study the dynamics of the C + NO reaction.This reaction is summarized in the following equation for ground state atoms : C[~P~(~P,)] + NO -+ CN(X2X, u", N") + O(3P); AH = 1.1 eV. The carbon atoms are produced by using a high pressure argon flashlamp, which It has been shown that the V.U.V. has a CaF, window to photodissociate the C302. photolysis ' of this molecule leads to the following photochemical reactions. -+ C(29,) + 2co --+ C(2'SO) + 2 c o . The relative output of the argon flash lamp is such that it decreases with decreasing wavelength. When the relative output of the flash lamp is folded into the absorption coefficient of C302 one predicts that most of the molecules are dissociated at 165 & 10 nm. Husain and Kirsch2 have shown that most of the carbon atoms produced in this wavelength region are in the (Z3PJ) electronic state.The rate constant for the reaction of carbon atoms in each of these is the same order of magnitude so that any reaction that is observed probably comes from the reaction of 3P atoms with NO. The rate constant2 for the reaction of 3P atoms with NO is 5 x lo-'' cm3 molecule-' s-l. At 0.150 Torr of NO this corresponds to a mean time between collisions of 5 ps. The mean time for non-reactive collisions at this pressure is 1 p s . In 2 ps, few of the CN radicals that are formed will have had time to undergo a relaxing colli- sion. The conclusion can therefore be drawn that any distribution observed under these conditions represents the nascent distribution of the radical.The spectra that are obtained under the above conditions show first that CN radi- cals are produced from the reaction of C atoms with NO, in agreement with the corre- lation arguments of Husain2 and that some of the exotherniicity of the reaction is used in producing vibrationally and rotationally excited radicals. Preliminary analysis of the data indicates that CN radicals are produced with a rotational tem- perature of z 1000 K. No analysis of the vibrational distribution has been made since only two vibrational levels are observed. A surprisal analysis of the observed distribution is calculated from the observed temperature which is then compared with a " prior " distribution. This prior distribution was calculated using the techniques described by Kinsey and LevineO3 The analysis gives a linear surprisal and it shows that the rotational distribution ob- served in these experiments is cooler than the predictions based upon the prior distri- bution.W. Braun, A. M. Bass, D. D. Davis and J. D. Simmons, Proc. Roy. SOC. A, 1969,312,417. D. Husain and L. J. Kirsch, Truns. Faruduy SOC., 1971,67, 3166. R. D. Levine and J. L. Kinsey, Atomic and Molecule Collision Theory: A Guide for the Ex- perimentalist, ed. R. B. Bernstein (Plenum Press, New York, 1978).GENERAL DISCUSSION 345 The prior distributions used in the present surprisal plots were computed using only the conservation of energy as a constraint. The linearity of the surprisal plot suggest that for the present results this constraint is adequate.It is therefore un- likely that the rotational states that are accessed are significantly constrained by con- servation of angular momentum. In conclusion, the present preliminary results show that the dynamics of A + BC reactions may be studied under bulb conditions using short duration flash lamps and LIF. Dr. G. Hancock (Oxford) said: The direct measurement of the quenching of C('S,) with 0, by Husain and Norris results in a rate constant for the process, 9.9 x lo-', cm3 molecule-' s-l, which is considerably larger than the previously reported estimate of 5 x Recently we have used the latter value in an interpretation of the kinetics of vacuum ultraviolet emission observed in the quenching of C,(a311n,) by 02,2 and this needs to be re-examined in the light of the new measurement.C, in the low lying a311, state was produced by the infrared multiple photon dis- sociation of 0.2 mTorr C2H3 CN diluted in 5 Torr Ar, and fig. 2 shows the rate of cm3 molecule-' s-l.l 2 c I * 0 0 50 100 150 200 02 pressure /mTorr FIG. 2.-First order decay rates of C,(a3Xu) in mixtures of 0.2 mTorr CtHJCN with varying amounts of added O2 at a total pressure of 5 Torr with added Ar. 0, laser excited fluorescence signal; 0 , vacuum U.V. emission. decay of the radical, measured by dye laser excited fluorescence, as a function of the pressure of added 0,. The slope of this plot gives a rate constant for the removal of C2(a31-In,) by O2 as 3.4 x lo-', cm3 molecule-' s-l. In view of the substantial exo- thermicity of several reaction schemes which can be written for C2 + 02, a search was made for vacuum U.V.emission accompanying the quenching by 02. Emission was observed in a wavelength region strongly suggesting that it originates from CO(A'I1). Furthermore, the decay of the vacuum U.V. emission followed first order kinetics with the same dependence upon O2 pressure as the removal rate of C,(a3rIu), again illu- strated in fig. 2. Clearly CO(A'll) is either being formed directly from the reaction C2(a3Hn,) + 02(x3Cg) -+ CO(ATI) + CO(x'C+) or by a secondary process involving the products of the C2 + 0, reaction, and taking place at a considerably faster rate. The reaction scheme C2(a311n,) + 02(x3z;) --f Co2(x1cS+) + C('S,) C(lS,) + 0 2 ( X 3 C g ) -+ c o ( ~ ' r I ) + 0(3~), (2) (3) G. M. Meaburn and D.Perner, Nature, 1966, 212, 1042. S. V. Filseth, G. Hancock, J. Fournier and K. Meier, Chern. Phys. Letters, 1979, 61, 288.346 GENERAL DISCUSSION which is exothermic for the production of CO(A'II), was rejected on the grounds that the previously reported value of k, was far too slow to account for the observed vacuum U.V. decay kinetics. Although the new value of k3 is faster than the rate constant for quenching of C,(a31-In,) by O,, reactions (2) and (3) would not imply a single expo- nential decay of the emission from CO(A'II), but would indicate that the signal should initially rise to a maximum value following the infrared laser pulse, and then fall, with a rate at long times equal to that of reaction (2). In all cases only a single ex- ponential decline in signal was experimentally observed ; the time resolution was sufficient to be able to detect any rise in signal following the infrared laser pulse com- patible with this kinetic scheme.On this evidence, the two-step process (2) and (3) can be rejected, and reaction (1) still appears to be the prime candidate for the produc- tion of vacuum U.V. emission in the C2 + 0, system. Dr. R. J. Donovan and Mr. M. C. Addison (Edinburgh) said: We had originally hoped to include work on the reactions of S(31D2) in our paper, but the results were not available in time for the manuscript deadline. Since then we have succeeded in making a number of direct studies of S(3'D2) using time-resolved atomic absorption photometry (1 = 167 nm) in the vacuum ultraviolet and have measured the absolute rate for its reaction with OCS as k = (1.2 & 0.3) x 10-l' cm3 molecule-' s-'.This is a particularly interesting reaction as the product S, is formed in its first electronically excited state, uiz: S(3'D2) + OCS --f S,(a'A,) + CO. We are currently making a detailed study of the state-to-state kinetics in this system and some other systems including reaction of S(3'D2) with O,, which produces SO(b'Z), and with N20, which produces SO(alA). Dr. J. E. Butler (Washington, D.C.) said: We have recently measured the product OH rotational distributions for u = 0, 1 formed in the reactions of electronically excited oxygen atoms, O('D), with H,, HCl, H2016 and H20I8. We have observed that the available energy in these exothermic reactions is not always statistically distributed.In the one case where a statistical distribution agrees with our observa- tions, O('D) + H2 --f OH(u = 0, N I 26; u = 1, N 5 IS), we believe the mechanism to be a distribution of " insertion " and " abstraction " events. Our observations are: (i) For H,, all observable rotational levels of OH(u = 0, N I 26; v = 1, N 15) can be fitted with an infinite temperature Boltzmann like distribution, with relative integrated rotational distribution of N(v = l)/N(v = 0) = 1.1 5 0.40. (ii) For HCl, the OH rotational distributions for u = 0, 1 can be charac- terized as the sum of two Boltzmann like distributions, one with T x 400 K for 0 I N 6, and with T x 4000 K for 7 5 N I 23 or 16 for v = 0 and 1 , respectively. The relative integrated rotational distributions gave N(v = l)/N(u = 0) = 0.3 & 0.15.(iii) For H2016, the OH rotational distributions observed for u = 0, 1 could be approxi- mated by the sum of two Boltzmann like distributions with characteristic tempera- tures of x400 K for 0 < N < 6, x2300 K for 7 I N I 18 or 15, respectively. The relative integrated rotational distributions gave N(v = l)/N(u = 0) = 0.35 & 0.10. (iv) For H2018 + 016('D), the 016H and 018H distributions were measured and could be approximated as above with characteristic temperatures of x 400 and x 3300 K for 016H and x 400 and x 1800 K for 0l8H. The integrated rotational distributions gave N(v = l)/N(u = 0) = 0.44 & 0.15 for 016H and 10.08 for 0"H. Trajectory calculations on the O('D) + H2 system by Sorbie and Murrell' and K.S. Sorbie and J. N. Murrell, Mol. Phys., 1976, 31, 905.GENERAL DISCUSSION 347 Whitlock et aZ.l indicate that one can factorise reactive encounters into two groups which can loosely be described as " insertion " into the H2 bond and " abstraction " of a H-atom by the attacking O(lD) atom. These two types of reactive events were predicted to give noticeably different rotational and vibrational product OH distribu- tions. " Insertion " events gave hot, even inverted rotational distributions and mono- tonically decreasing vibrational distributions (with vibrational energy), while " ab- straction " events gave rotationally cold distributions and vibrationally inverted or hot distributions. These trajectory results also agree with our observed OH distri- butions from O('D) + H2.Since our observed OH(u, J ) distribution for O(ID) + HC1, H20I6, H201' do not agree with the predictions of statistical theories, and since they all gave rotationally cooler distributions than that observed with H2 where " insertion " should be the easiest, we infer that the mechanism in these reactions is probably not the formation of a long lived complex in which energy is scrambled amongst available modes, but rather a distribution of" insertion " and " abstraction " events with the former more important in H2 and the latter in the HCl, H2016*18 systems. Dr. P. A. Gorry, Dr. C. V. Nowikow and Prof. R. Grice (Manchester) said: We wish to comment on the flash photolysis measurements of the O(3P) + CF31 reaction by Addison, Donovan and Garraway and their relation to a recent molecular beam study' of this reaction.The molecular beam data may be explained in terms of hard sphere scattering at small impact parameters combined with significant energy transfer to the CF, radical but do not indicate the existence of a long-lived collision complex. Thus the potential energy surface for the 0 + CF31 reaction may involve a shallow well or even be essentially level. The CF310 molecule3 is stable with respect to dispropor- tionation for temperatures ,<O "C and must correspond to a deep well E, 2 100 kJ mol-' on the potential energy surface. Such a well corresponding to the (presum- ably) singlet CF,IO molecule would give rise to a long-lived collision complex if it participated in the reaction dynamics.Accordingly we feel that the 0 + CF31 reaction dynamics may be explained by motion over a triplet potential energy surface and that it is not necessary to invoke transitions to the singlet surface. This would be in line with the situation which is believed to obtain4 for the reactions of 0 atoms with halogen molecules. The small total reaction cross section Q z 2 A' for the 0 + CF31 reaction indicated by the flash photolysis measurements would be attri- buted to reaction occurring only at small impact parameters together with an orienta- tion requirement for reaction. These conclusions concerning the O(3P) + CF31 reaction are not in conflict with the observation of some O(lD) quenching to O(3P) in the O('D) 4- CF3C1 reaction by Addison, Donovan and Garraway.Reaction of O(lD) with CF31 would yield a long-lived singlet CF310 collision complex persisting for 1 1 50 vibrational periods. Consequently the seam of intersection between the singlet and triplet CF310 surfaces may be traversed 2300 times in the O(lD) + CF31 reaction compared with a single time in an O(3P) + CF31 reaction proceeding via a direct mechanism. Accordingly, P. A. Whitlock, J. T. Muckerman and E. R. Fisher, Research Institute for Engineering Sciences Report (Wayne State University, 1976). P. A. Gorry, C. V. Nowikow and R. Grice, Chem. Phys. Letters, 1978, 55, 24; Mol. Phys., 1979, 38, in press. D. Naumann, L. Deneken and E. Renk, J. Fluorine Chem., 1975,5,509. D. D. Parrish and D. R. Herschbach, J. Amer. Chem. Soc., 1973,95,6133; D. St.A. G. Radlein, J. C. Whitehead and R. Grice, Mol. Phys., 1975, 29, 1813.348 GENERAL DISCUSSION the probability of a singlet ++ triplet transition may be higher in the O('D) + CF31 reaction than in the O(3P) + CF31 reaction by a factor 2300, provided that the transition probability is small for a single traversal of the seam. Dr. J . P . Simons (Birmingham) said: One of the simplest experimental methods of studying molecular photodissociation is to monitor the fluorescence of electronically excited fragments following predissociation from prepared states populated by monochromatic light absorption in the vacuum U.V. The initial wavefunction in the generalised Franck-Condon factors discussed by Freed et. al.' is then that of the photoexcited state rather than the ground state.For example, this technique has been used to determine the distribution over rotational (and vibrational) states in CN(B2C +) following the predissociation of HCN( CIA') from vibronic levels carrying 3 and 6 quanta in the bending mode.2 In this particular example, the distribution followed the simple mapping .Iparent --f jfragment, dictated by the requirement of angular momentum conservation, following the loss of the light H atom. In the case of CS2 however, the rotational energy distribution in CS(A'n) following predissociation from a linear Rydberg state at 130.4 nm,3 qualitatively conforms to the predictions of the full Franck-Condon mode1.l The distribution peaks at j M 16, well below the level j M 30, determined solely by angular momentum conservation (cf.the " bending " term in the generalised Franck-Condon factor)' but close to the level determined solely by the energy conservation requirement (cf. the " stretching " term in the generalised Franck-Condon factor) which, for a Boltzmann distribution in the parent molecule, would produce a maximum at j z 11. In contrast, the rotational distributions in CN(B2C +) produced following excitation of the cyanogen halides in the cc-continuum4 cannot be explained solely in terms of a Franck-Condon model. In these molecules, the topography of the potential energy surface over which the neglected " final state interactions " occur, is thought to be an important factor in determining the final energy d i ~ p o s a l . ~ The importance of adequately characterising the nature of the photoexcited state initially prepared by photon absorption cannot be overemphasised, particular in a Faraday Discussion devoted to the Kinetics of State Selected Species.A useful technique which helps in the spectroscopic assignment of the prepared state has been devised recently5 and termed polarised photofluorescence excitation spectroscopy. Examination of the degree, and particularly the sign, of the polarisation in the fluores- cence of electronically excited fragments, following direct or predissociation, can give valuable information related to the symmetry and the lifetime of the photoexcited parent m ~ l e c u l e . ~ * ~ Fig. 3 shows the excitation and polarisation spectra of the CN- (B --f X ) fluorescence produced through predissociation of BrCN in the B and C band systems.The high positive polarisation in the underlying continuum region indi- cates an upper state of either O+(linear) or A'(bent) symmetry, while the negative polarisation in the banded regions reflects the perpendicular orientation of the transi- tion m ~ r n e n t . ~ ~ ~ The B and C systems are assigned to the 311 and 'll components of the Rydberg transitions 27t -+ 3sa.' It is particularly striking that there is a dramatic K. F. Freed, M. Morse and Y. B. Band, Faraday Disc. Chem. Soc., 1979, 67, 297. M. N. R. Ashfold, A. M. Quinton and J. P. Simons, unpublished work. M. N. R. Ashfold and J. P. Simons, J.C.S. Faraday 11, 1978, 74,280. G. A. Chamberlain and J. P. Simons, J.C.S. Faraday ZZ, 1975,32,355.; M. T. Macpherson and J.P. Simons, Chem. Phys. Letters, 1977, 51, 261. M. T. Macpherson, J. P. Simons and R. N. Zare, Mol. Phys., to be published. ' M. N. R. Ashfold, M. T. Macpherson and J. P. Simons, Chem. Phys. Letters, 1978, 55, 84. ' A. S. Georgiou, M. T. Macpherson and J. P. Simons, unpublished work.GENERAL DISCUSSION 349 I f 0.06 0.04 0.02 0 -0.02 I . 142 144 146 148 150 152 A/nm FIG. 3.-Photofragment fluorescence excitation spectrum of CN(B) from BrCN and detailed polar- ization measurements in the region of the 'TI, 311 (2n +ma) band systems. Dashed curve shows variation in intensity of continuum photolysis source. change in the internal energy disposal in CN(B), on changing the exciting wavelength from 149.4 nm, which lies principally in the continuum region, to 147 nm,' which excites principally the 1JI,3s0 t X band origin.Prof. R. N. Dixon (Bristol) said: I would like to comment on the model for pre- dissociation discussed by Freed, Morse and Band and to suggest a need for a more detailed model. Mr. Noble and I have been studying the predissociation of HNO in its first 'Aff excited state using the technique of laser excited fluorescence at low pressure ( z 1 mTorr). Evidence for predissociation has also been obtained previously through breaking off in the structure of HNO chemiluminescence from the reaction H + NO + M --f HNO* + M,' and through line broadening3 This spectrum is also subject to extensive minor perturbations of the rotational structure. Table 1 shows the highest bound rotational levels of various vibronic states.TABLE HIGHEST BOUND ROTATIONAL LEVELS IN VIBRONIC STATES OF HNO JIA" vibronic highest bound level term vibronic states K' J' value/cm-' origin/cm- 000 13 ? 2 16 440* 13 154 010 10 ? 2 1 6 500* 14 575 020 4 12 16 469 15 956 100 3 14 16 452 16 009 100 4 11 16 485 16 009 101 0 0 16 971 16 971 030 0 0 17 310 17 310 * Ref. (1). M. N. R. Ashfold and J. P. Simons, J.C.S. Faraday ZZ, 1978, 74,280. M. J. Y. Clement and D. A. Ramsay, Canad. J . Phys., 1961,39,205. P. A. Freedman, Chern. Phys. Letters, 1976, 44, 605.350 GENERAL DISCUSSION Theoretical potential energy curves for HNO show a long range attraction in the correlation of the H + NO limit with the ground 2 'A' state of HNO, but a repulsive barrier between this limit and the A ' A " excited state.' The theory of Freed, Morse and Band would be directly applicable to the predissociation of the A state through the crossing of this barrier.However, there are a number of pieces of evidence to suggest that the predissociation of the levels in the table proceeds through a two-step mechanism involving high levels of the ground state: (i) The breaking-off limit is remarkably constant at about 16 488 cm-' even though for the 0,, vibronic level the K, quantum number is 13 and the total rotational energy is ~33300 cm-', whereas for the 100 level and K, = 3 the rotational energy content is only ~ 4 4 0 cm-'. The centrifugal barrier in the effective potential for the exit channels must therefore be small. This would be the case if the predissociation route involved the attractive ground state potential, but not for passage over the excited state barrier, which occurs at a conformation with a substantial a-axis rota- tional constant.(ii) We have found (table 1) that the O,, rotational levels of the 101 and 030 vibronic states are sharp, even though they lie well above the predissociation limit of the lower levels, whereas levels with higher J are predissociated. Furthermore, Freed has found that these levels have a width which is J-dependent.2 Thus the predissociation must be rotationally induced. (iii) By exciting laser induced fluorescence in magnetic fields up to 10 kG we have found that only the perturbed lines are highly sensitive to the field. From the analysis of these effects we conclude that the transition densities between the main and per- turbing levels have magnetic moments of the order of tenths of a Bohr magneton.The excited state alone should be diamagnetic, but the k and A" states together corre- late with a lA state of linear HNO. From intensity anomalies in perturbed levels close to the predissociation limit we conclude that the perturbations and the pre- dissociation are two manifestations of the same " internal conversion " mechanism. During the dissociation of HNO* the atoms will therefore move in the force field of the ground state, and the final distribution over quantum states of the products will not be simply related to the theory of Freed, Morse and Band. Presumably their conclusions concerning the conversion of reactant bending vibrational momentum into product rotational angular momentum should still hold, but the vibrational distribution will be greatly affected by the extra step.We hope to be able to measure this distribution in the NO formed. Internal conversion is known to be an important process in many polyatomic molecules and can provide a general mechanism for circumventing barriers to the dissociation of excited states. We may therefore expect our conclusions concerning the mechanisms of predissociation of HNO to hold for many other molecules. We hope that Prof. Freed will extend his model to include this important process. Prof. K. F. Freed (Chicago) said: We are glad to see the excellent experimental data on photodissociation processes which Dr. Simons and Prof. Dixon have pre- sented. We hope that experiments of this type will help to test and broaden the theories as well as to provide us with detailed information concerning the structure of repulsive potential energy surfaces.The theory described in our paper and previous 0nes~9~ is directly applicable to A. W. Salotto and L. Burnell, Chem. Phys. Letters, 1969, 3, 80. P. A. Freedman, Chem. Phys. Letters, 1976, 44, 605. Y . B. Band and K. F. Freed, J . Chem. Phys., 1975, 63, 3382. K. F. Freed and Y . B. Band, Excited States, 1978, 3, 109.GENERAL DISCUSSION 351 predissociation when the individual rotational states are broadened, but not over- lapping such that states with identical good quantum numbers (like total angular momentum) are nonoverlapping. In this case the initial state, li), in eqn (1) is the predissociating level, and all else remains the same.Dr. Simons' beautiful polariza- tion experiments indicate the more complicated situation in BrCN photodissociation to CN(B2C+) in the B and C band systems where there are contributions from direct and predissociation occurring simultaneously. (Perhaps even the predissociating levels are also overlapping.) This situation corresponds to the general problem discussed by Fanol which can also be applied to photodissociation as mentioned previously by US.^^^ However, this procedure involves the solution of a configura- tion interaction matrix involving mixing of the predissociating levels with the con- tinuum of directly dissociating states. This, of course, represents a very difficult calculation, and it would be preferable to have experiments which involve weak pre- dissociation through readily assignable quantum states, li).The three-dimensional theory in our paper discusses how vibrational and rotational relaxation on the repulsive surface contribute to the state-to-state photodissociation probabilities. Thus, the wavefunction (4) of our paper has a summation over all possible fragment quantum states to describe this half-collision scattering event. The calculations have not yet included this feature. Evidence from collinear calcu- l a t i o n ~ ~ ~ ~ shows how this half-collisional process can yield fragment vibrational excitation, and similar effects are to be anticipated in the full three-dimensional case. In face, ab initio calculations on excited surfaces4-' for HCN, H20, H02 and HCO all display conical intersections between surfaces and this is fairly general when the fragments are radicals with low lying excited states.In the H20 case' the surface leads to large torques on the departing OH fragment, producing high rotational excitation' in the half collision. More effort is required on this important problem. In fact, close coupled half collision scattering calculations would require the use of our Franck-Condon theory if optimal (and different) coordinate systems are utilized for the bound and dissociative surfaces. We should also emphasize that the theory predicts rotational distributions, as in our eqn (9), which can be fit approximately to Boltzmann distributions without any model of rotational energy randomization. These distributions (and ones for higher initial J ) have the exponential dependence on ( j + +)2[and ( J - j)2] due to the gaussian nature of the bending wave-function, having no thermal origins: p in our eqn (8) and (9) is dynamically determined.Dr. W. M. Jackson (Washington, D.C.) said: We have measured the rotational distribution of the CN(X2C) state radical produced by the predissociation of three different vibrational levels of (C'll,) state of C2N2. This information represents the most detailed experimental data currently available on photodissociation. It is therefore a challenge to theory. The data show how a molecule in a given electronic vibrational state predissociates into fragments with a fixed amount of translational energy. A surprisal analysis of the data indicates that excited molecules with 8400 and 10 500 cm-' of excess available U.Fano, Phys. Rev., 1961, 124, 1866. Y. B. Band and K. F. Freed, J. Chem. Phys., 1975,63, 3382. K. F. Freed and Y . B. Band, Excited Stares, 1978, 3, 109. G. J. Vazquez and J. F. Gouyet, Chem. Phys. Letters, 1978,57, 385 and in press. F. Flouquet and J. A. Horsley, J. Chem. Phys., 1974, 60, 3767. S. R. Langhoff and R. L. Jaffe, J . Chetn. Phys., in press. S. Iwata, Chem. Phys. Letters, in press. * T. Carrington, J. Chem. Phys., 1964, 41, 2012.352 GENERAL DISCUSSION energy have the same distribution with respect to the fraction of the available energy that ends up in rotation. At 6300 cm-' there is a difference in the observed distribu- tion. Currently, the experi- mentalists have been unable to measure the ro-vibronic distribution of the CN(A2n) state fragments that are produced and undetected.Theory should be able to do this. I hope that theorists will accept these experimental challenges. Dr. M. S. Child (Oxford University) said: In fig. 4 of his paper Prof. Leone attributes the inaccessibility of states of IBr which correlate with electronically excited Theory should be able to explain these differences. 18 500 18 000 17 500 17 000 c 5 2 16 500 16 000 500 0 FIG. 4.-Potential curves for the X'X+, B3110+ and P O + of IBr. The arrows show how a two photon experiment might break the Franck-Condon selection rule against photodissociation to produce I* atoms from the ground state. iodine atoms to negligibly small Franck-Condon overlap with the zero point wave- function of the IBr ground state.Some years ago I made a detailed analysis of the predissociation from the IBr B'(O+) state,' which suggests a possible two-photon experiment to detect these Franck-Condon forbidden states. Two features of the analysis' are important. The first is that the absorption* and magnetic rotation3 spectra show occasional sharp levels, so sharp in fact that Weinstock and P r e s t ~ n ~ . ~ have been able to observe laser induced fluorescence. The second feature is that the quantitative analysis ' required the adoption of a coupling scheme intermediate between the diabatic and adiabatic limits. This implies substantial wave function amplitude at all three of the turning points shown in fig. 4. M. S.Child, Mol. Phys., 1976, 32, 1495. L. E. Selin, Arkiv. Fys., 1962, 21, 479. W. H. Eberhardt, Wu-Chich Cheng and H. J. Renner, J . Mol. Spectr., 1959,3, 664. E. M. Weinstock, J . Mol. Spectr., 1976, 61, 395. E. M. Weinstock and A. Preston, J . Mol. Spectr., 1978,70, 188.GENERAL DISCUSSION 353 The suggested experiment involves one photon tuned to B' f- X (20- 1) absorption band at 2 17 500 cm-l, used by Weinstock and Preston1i2 to observe the fluores- cence and a second photon to probe the excited state. Dr. C. Fotakis, Dr. M. Martin and Dr. R. J. Donovan (Edinburgh) said: We would like to comment on the work of Baughcum et al., concerning the photolysis of alkyl iodides with a rare gas halide excimer laser. Using the unfocused output of a KrF laser (2 = 248 nm, E = 38 mJ) we have observed chemiluminescence from the reaction, 1(52P+) + I(52P;) + M + 12(B3no+J + M following the photolysis of CF31 and CH31.This luminescence was first observed by Abrahamson et al.3 using a high energy (broad band) flash photolysis system, and more recently by Stephan et al.4 using a quadrupled Nd-YAG laser. In our experiments the 12(B-X) emission, following photolysis of CFJ(0.6 kN mV2) was observed to increase with time over the first few milliseconds after the laser pulse and then decline. This is readily understood in terms of the mechanism: CF31 + hv(A = 248 nm) -+ CF, +- I(5,P,) -+ CF, + I(5,P;) 1(52P+) + Q -+ I(5,P;) + Q I(52P+) + I(52P;) + M -+ 12(B31To+u) + M. The I,@-X) emission is observed immediately after the laser pulse as some ground state atoms are produced in the primary step; however, the maximum in the chemi- luminescence is not observed until 2 2 ms (depending on conditions), when the con- centrations of excited and ground state atoms are equal, as the relaxation of I(52P,) by CF31 is very inefficient.If a sample of CF31 is subjected to a second laser pulse the peak in the chemiluminescence is observed at shorter times, due to the more rapid quenching of 1(5,P,) by I,. For CH31 the 12(B-X) emission is observed to peak shortly after the laser pulse due to the efficient relaxation of I(52P,) by CH31. We would therefore emphasise that the observation of emission from molecular fragments, from molecules such as CH21Z, does not necessarily arise from a primary photochemical step.Indeed, we would expect chemiluminescence from secondary reactions, such as those outlined above, to be a general feature of excimer laser photochemical studies due to the high radical concentrations that can be produced. The excimer laser should therefore be a powerful tool for the study of new chemiluminescent reactions. Dr. A. Ding (Berlin) said: It has been mentioned that certain types of experiments, particularly scattering and spectroscopic polarization experiments, give insight into vector properties of the collision dynamics, which allow predictions about the direction of the angular momenta of the collision encounter. It may be noted that emission spectra of polyatomic, especially triatomic, species contain similar information. The correlation between the two rotational quantum numbers J and K are such a vector property.They describe the mode of rotation of the product and can be used E. M. Weinstock, J . Mot. Spectr., 1976, 61, 395. E. M. Weinstock and A. Preston, J . Mol. Spectr., 1978, 70, 188. E. W. Abrahamson, D. Husain and J. R. Wiesenfeld, Trans. Faraday SOC., 1968, 64, 833. K. H. Stephan and F. J. Comes, in Laser Induced Processes in Molecules, ed. K. L. Kompa and S. D. Smith (Springer-Verlag, Berlin, 1979), p. 301.354 GENERAL DISCUSSION to give information on the collision dynamics. In the case of triatomic molecules, in particular, one is able to distinguish between in-plane and out-of-plane collision geometries. An example for such a system is the ion-molecule reaction Hg + H 2 + H 3 + H for which experiments have been performed by measuring the infrared chemilumines- cence of the H i product.The reaction took place in a large vessel filled with low pressure H2 (w 5 x lo-' Torr), where Hg-reagents were produced by electron impact, and subsequently reacted with the unionized H2-gas. Infrared emission was measured in the 2.5-5.5 pm region with the use of a high throughput double monochromator' specially built for this experiment. So far the spectroscopy of the H i system is not yet completely known, as this is the first time such ion spectra have been recorded. However, with the help of ab initio calculations2 one can reach at least qualitative conclusions, which show a non-statistical behaviour of the appropriate distributions. Particularly one can conclude that the 3rd vibrational level of the asymmetric stretch mode is considerably populated, and the rotational distribution of the H'; shows maxima for levels with J z K , indicating that the rotation is mainly about the axis perpendicular to the molecular plane.This is in agreement with trajectory studies on the same s y ~ t e m , ~ and would lead to the conclusion that the reaction predominantly proceeds via an in-plane collision encounter. Dr. M. Martin, Mr. M. Trainer and Dr. R. J. Donovan (Edinburgh) (communicated) : We would like to mention some recent results which support the findings of Baughcum et al. concerning the low yield of HBr from the reaction of Br(2P+) with H2S. In a wide range of studies involving I(52P+) interacting with hydrides (e.g., CD3CN, C6HsCH3, CH,CHO, C6HSCHO) for which exothermic chemical reaction to produce HI is thermodynamically favoured, we find that the dominant removal process is physical quenching.The fact that the electronic excitation energy in these systems cannot be used efficiently for reaction is not in fact surprising as a non-adiabatic transition from the excited entrance channel hypersurface to the ground state exit channel surface is required before the hydrogen halide product can be formed. Thus adiabatic correla- tion rules provide a good guide to the branching into product channels for these, and many other We would emphasise that where adiabatic reaction on an excited hypersurface is possible, reaction proceeds efficiently. Examples of this behaviour are given by Baughcum et al.for reaction of Br(42P+) and I(52P+) with halogens and interhalogens. Further examples are provided by the reactions of F(22P+) and C1(32P+) with hydrogen halides where the thermal population of the 'P+ state, at 300 K, is sufficient to give rise to substantial yields of the excited halogen atom product via the adiabatic ~ h a n n e l , ~ X("+) + HY -+ HX + Y(2P+). Dr. J. Wanner (Munich) said: In this Discussion Clyne and McDermid reported on an improved method for the determination of bond dissociation energies of the A. Ding and A. Redpath, Proc. I.C.P.E.A.C., 1977,10, 759. G. D. Carney and R. N. Porter, J. Chem. Phys., 1976,65,3547. R. J. Donovan and D. Husain, Chem. Rev., 1970,70,489. C. Fotakis and R. J. Donovan, J.C.S. Faraday 11, 1979, 75, 1553.' J. Muckerman, personal communication.GENERAL DISCUSSION 355 ground state of interhalogen molecules from the observation of the onset of predisso- ciation in the B state under collision free conditions. This method may be compared with independent experiments of laser-induced fluorescence product state analysis. So far we are only able to comment on the bond dissociation energy D: (IF). We have performed a crossed molecular beam study of the reactions F + CH31 --f I F + CH, and F + CF,I -+ I F + CF, using thermal reagent beams at 300 K.' The total available energy for internal I F product excitation E,,, = 45.3 and 54.9 kJ mol- ', respectively, can be calculated using reagent bond dissociation energies Di (CH,-I) = 229.4 kJ mo1-' and Dg (CF,-I) = 219.9 kJ mol-' given by Okafo and Whittle2 and the recently improved value Dg (IF) = 268.0 kJ mol-' by Clyne and M~Dermid.~ Vibrational product excitation in the electronic ground state of I F should thus be possible up to u = 6 and u = 7 for the reactions with CHJ and CF31 as reagents, respectively.This has been found in consistency with our experimental observations. It should be added that the earlier value for 08 (IF) = 272.4 kJ mol-' determined by Clyne and McDermid at a higher pressure4was partially in disagreement with our results since population up to u = 8 should have been observed in the reaction with F + CF31. Dr. M. A. A. Clyne (Queen Mary College, London) said: The elegant crossed- beam experiments of Wanner5 provide confirmation of the value of (22 333 + 2) cm-' for Do0 (IF) reported by Clyne and McDermid., In our work,, the dissociation energy of ground-state I F X'C+ was determined from the energy of onset of predisso- ciation in the rotational structure of I F B3n(O+) - X'C+, observed in laser-induced fluorescence (LIF).Collision-free conditions were used,3 which resulted in observa- tion of a sharp fall in excited-state lifetime between J' = 6 and J' = 7 of the level u' = 9. we observed LIF of I F at higher pressures ( z 1 Torr as compared with \< 1 mTorr in our later, definitive work3). Collisions modified the initial rotation- al ~ t a t e , ~ so that onset of predissociation was not sharp. Rotational (or vibrational) relaxation results in a considerable overestimate of the energy of onset of predissocia- tion, because of the smearing out of the initial energy distributions.This point can be illustrated by the historical decrease in estimated DO, (IF) values, as a function of reducing pressure. Durie's6 value for DO, (IF) was <23 341 cm-', based on data from atmospheric-pressure flames; the first Clyne and McDermid value4 was <(22 700 15) cm-' from LIF data near 1 Torr. The definitive Clyne and McDermid value3 of (22 333 & 2) cm-' was obtained from LIF data below 1 mTorr. Earlier data on other dissociation energies obtained from the onset of predissocia- tion may well prove to give overestimated values. The overestimate may be serious when the excited state is long-lived (7, > 1 p s ) , or when measurements are made at higher pressures. A critical re-examination is desirable in such cases.Thus, very close limits for Do0 (IF) could be obtained. In earlier Prof. P. E. Siska and Prof. M. F. Golde (Pittsburgh) said: The electronic struc- The outer s ture of the metastable rare gases Ne* to Xe* is . . , (n - l)p5 ns. L. Stein, J. Wanner, H. Figger and H. Walther in Laser-induced Processes in Moleciiles, ed. K. L. Kompa and S. D. Smith, (Springer Berlin, 1979), vol. 6, p. 232. E. N. Okafo and E. Whittle, Int. J. Chem. Kinetics, 1978, 7, 273. M. A. A. Clyne and I. S. McDermid, J.C.S. Faraday 11, 1978,74, 1644. M. A. A. Clyne and I. S. McDermid, J.C.S. Faraday II, 1976, 72,2252. J. Wanner, previous comment at this Discussion. R. A. Durie, Canad. J. Phys., 1966, 44, 337.3 56 GENERAL DISCUSSION electron endows these highly energetic species with chemical properties remarkably similar to those of the alkali meta1s.l The imaginative and elegant experiments of Rettner and Simons as well as the copious and informative rate measurements re- ported by Setser thus herald the beginning of a " neo-alkali age ".From such studies we are likely to learn much about the alkali metal reactions that has long lain hidden, as well as a rich new chemistry engendered by the electronic excitation. Though the s electron is the major influence on metastable rare gas properties, the rare gas ion cores are isoelectronic with neutral halogen atoms, and thus are ex- pected to resemble these atoms more than the corresponding alkali metal ions in their chemistry. This halogen-like property may influence van der Waals forces at inter- mediate range involving the metastable atoms, and also may make chemi-ionization channels such as Ar* + Cl, -+ [Ar+Cl-] -+ ArC1+ + C1- energetically possible owing to the likely large bond energy of ArCl+ relative to Ar+ + C1.Evidence for the existence and stability of rare gas halide positive ions comes both from mass spectral plasma studies and ab initio calculation^,^ which suggest they are at least as stable as the isoelectronic halogen or interhalogen molecule. The well-known stability of ArH+ makes the Ar+ + HC1+ ArH+ + C1- reaction exoergic by 1.2 eV. With polyatomic halogen-bearing molecules, more bizarre ion chemistry may occur. We plan to look for production of such ions under single- collision conditions in a crossed-beams apparatus.Prof. D. W. Setser (Kansas) said: Siska and Golde suggest that the isoelectronic nature of the rare-gas ion core and halogen atoms may lead to chemical properties that resemble halogen atoms, such as formation of ArC1+ in some reactions. A re- lated suggestion to this was made by Thrush at the 1972 Faraday Discu~sion.~ We agree that the electron deficient core can be important. For example in some cases incipient ion-molecule chemistry may be important in the quenching mechanisms for the excited states of the rare gases.5 Another example is the results of ab initio calculations6 of excited states of KrF* which predict a high lying bound state that corresponds to the first Rydberg state of the series that terminates in KrF+. We have begun a search for ionic exit channels from quenching of Ar(3P,) using a flowing afterglow monitored by a mass spectrometer (see tables 2 and 3 of the Introductory Lecture to section 3 of this Discussion).The only ion found from the AI-(~P,) + Cl, reaction was trace amounts of C1+. Thus, the chemi- ionization reaction yielding ArCl+ and C1- suggested by Siska and Golde does not occur at thermal energies. The small C l i yield, which arises from Penning ionization is a measure of the colli- sions which stay on the V(Ar*, Clz) potential rather than branch to the V(Ar+, Cl;) potential because Penning ionization occurs primarily at the repulsive wall of V(Ar*, Cl,). The observation of a very low yield of C1+, thus, confirms the adiabatic pathway for the reactive quenching of AI-(~P,) + C12 at thermal energy.By analogy to the For a review, see M. F. Golde in Gas Kinetics and Energy Transfer, (Spec. Period Rep., The Chemical Society, London, 1976), vol. 2, p. 123. I. Kuen and F. Howorka, J . Chem. Phys., 1979,70, 595. For a review, see C. Thomson in Theoretical Chemistry, (Spec. Period. Rep., The Chemical Society, London, 1974), vol. 2, p. 83. B. A. Thrush, Faraday Disc. Chem. Soc., 1972, 53, 121. J. E. Velazco, J. H. Kolts and D. W. Setzer, J . Chem. Phys., 1978, 69, 4367. P. J. Hay and T. H. Dunning, Jr., J . Chem. Phys., 1977, 66, 1306.GENERAL DISCUSSION 357 reactions of alkaline earth metal atoms.' There are likely to be other examples for which chemi-ionization is an allowed exit channel. Prof. M. H. Alexander (Maryland) and Prof.P. J. Dagdigian (Johns Hopkins) said : In connection with the discussion of Simons and coworkers and the subsequent in- formal comment by Herschbach, we wish to point out that the orientation dependence of nonreactive molecular collisions could well provide a sensitive probe of the aniso- tropy in intermolecular potentials, especially if the torque exerted by the collision part- ner is substantially enhanced for particular approach geometries. Several recent theoretical s t ~ d i e s , ~ - ~ of varying degrees of sophistication, indicate that both rota- tionally inelastic and elastic collisions can lead to significant polarization and align- ment of the final molecular rotational angular momentum, even for unpolarized reactants. The nature and magnitude of these effects appear to be quite sensitive to the potential surface.3* The orientation dependence of rotationally inelastic collisions can be studied using for state selection and detection either electric quadrupole fields6 or lasers. As we and others have in the latter case the polarization of the radiation provides a natural means to detect nonuniformities in molecular mj-state distributions. An excellent example of this type of experiment is provided by the beautiful recent work of McCaffery and coworkers9 on collisional depolarization in excited electronic states of diatomic molecules. Prof. J. C. Polanyi (Toronto) said: Rettner and Simons'O note that the " repulsive " component of the energy-release in the exothermic reaction Xe*(3P2) + Br, --f XeBr* + Br could contribute to the XeBr* rotational excitation, J ' .As they surmise, this can be a significant factor under circumstances where the initial orbital angular momentum, L, is small; i.e., when the collision-energy imparted by their paddle- wheel is modest. A model study (3D Monte Carlo trajectories) for an equal-mass exchange reaction A + BC -+ AB + C showed markedly enhanced J' at low L on a more-repulsive surface as compared with a more-attractive p.e. surface." These model surfaces were both of the LEPS variety, and consequently favoured collinear reaction. In the case of Xe* + Br, the intermediate could be bent," thus channelling product repulsion still more efficiently into rotation.'* The same model study was used to examine the factors governing the polarization of J',11913 which is one of Rettner and Simons's variables." Particular attention was paid to the angle x between the product molecule angular-momentum vector J' and G.J. Diebold, F. Engelke, H. U. Lee, J. C. Whitehead and R. N. Zare, Chem. Phys., 1977, 20, 265. M. H. Alexander, J. Chem. Phys., 1977,67,2703; Chem. Phys., 1978,27,229. L. Monchick, J . Chem. Phys., 1977, 67, 4626. S. R. Kinnersley, Mol. Phys., in press. U. Borkenhagen, H. Malthan, and J. P. Toennies, Chem. Phys. Letters, 1976, 41, 222; J. Chem. Phys., in press. ' M. H. Alexander, P. J. Dagdigian, and A. E. DePristo, J . Chem. Phys., 1977,66, 59. D. A. Case, G. M. McClelland, and D. R. Herschbach, Mol. Phys., 1978,35, 541. S. R. Jeyes, A. J. McCaffery, M. D. Rowe, and H. Kato, Chem.Phys. Letters, 1977, 48, 91; M. D. Rowe and A. J. McCaffery, Chem. Phys., 1978, 34, 81 ; Chem. Phys., in press. M. H. Hijazi and J. C . Polanyi, J . Chem. Phys., 1975, 63, 2249. ' M. H. Alexander and P. J. Dagdigian, J . Chem. Phys., 1977,66,4126. l o C. T. Rettner and J. P. Simons, Faraday Disc. Chem. SOC., 1978,67, 329. l2 D. S. Perry and J. C. Polanyi, Chem. Phys., 1976, 12, 37. l3 M. H. Hijazi and J. C. Polanyi, Chem. Phys., 1975, 11, 1.358 GENERAL DISCUSSION the initial relative velocity vector, urel, since the measurement of this angle had been pioneered by Herschbach's The optical method of determining polariza- tion, exemplified in Rettner and Simons' work, is likely to be capable of substantially greater sensitivity and precision than the electric-field deflection method that preceded it.It may, therefore, be timely to point out that, according to the model calculation alluded to above, the best indicator of repulsive energy release as the source of product rotation was a tendency for x = 90" (highly polarised product4) to be observed in the sharply backward-scattered component of the reaction product but not in the less backward-scattered product. What are the prospects for measuring the polarization at more than one scattering angle? It is a pleasure to be in the position of asking for still more-refined data, rather than being asked for it. Dr. J. P. Simons, Dr. Y. Ono and Mr. R. J. Hennessy (Birmingham) said: Since we have extended the work to completing the paper included in this include a study of the reaction Xe(3P2,0) + IC1 --f XeI* + C1 ( 1 4 XeCl* + I.(1b) The chemiluminescence spectrum recorded from the cross-beam interaction at thermal energies shows fluorescence from both XeCl(B, C) and XeI(B, C) but with reaction ( l b ) the dominant branching channel. [In contrast, in the reactions with ClCN and BrCN, no trace of XeC1* or XeBr* has been detected, though there is strong emission from CN(A, B), presumably formed through predissociation of XeCN*, cf. SetserSb]. Preliminary measurements of the polarisation of the XeCl(B+ X ) emission as a function of the collision energy show it to lie close to the curve obtained for the reaction of Xe(3P0,,) with CCl,. In the case of IC1 however, there is no possibility of angular momentum disposal in the atomic product, and the low polarisation can be attributed solely to the " blocking " effect of the bulky I atom.Finally, the close similarity between the behaviour of Xe(3P3,0) and CC1, and Cs and CC1, reported by Riley, Siska and Herschbach,6 is gratifying, particularly on a Jubilee oc~asion.~ C. Maltz, N. D. Weinstein and D. R. Herschbach, Mol. Phys., 1972, 24, 133. D. S. Y . Hsu and D. R. Herschbach, Faruduy Disc. Chem. SOC., 1973, 55, 116. D. S. Y. Hsu, G . M. McClelland and D. R. Herschbach, J. Chem. Phys., 1974, 61,4927. D. S. Y. Hsu, N. D. Weinstein and D. R. Herschbach, Mol. Phys., 1975, 29, 257. (a) C. T. Rettner and J. P. Simons, Furaduy Disc. Chetn. SOC., 1979, 67, 329. (6) D. W. Setser, T. D. Dreiling, H . C. Brashears, Jr, and J. H. Kolts, Foraduy Disc.Chem. SOC., 1979, 67,255. S . J. Riley, P. E. Siska and D. R. Herschbach, Faruduy Disc. Chem. SOC., 1979, 67, 27. ' T. H. Bull and P. B. Moon, Disc. Faraday Soc., 1954, 17, 54.GENERAL DISCUSSION 3 59 ADDITIONAL REMARKS Mr. S. J. Buelow, Mr. D. R. Worsnop and Prof. D. R. Herschbach (Harvard) said: As emphasized by Prof. Siska, metastable argon atoms show a schizophrenic chemical personality: alkali-like at long range, halogen-like at short range. This prompts us to mention recent work which perhaps exemplifies the halogen-like character of argon ions and also may serve as a reminder that physical chemists still occasionally use the word “ state ” to refer to the macroscopic phase of matter. Our study concerns the interaction of argon ions with clusters of perchloroethylene molecules, (C2C14),,.The motivation stems from an exotic solar neutrino experiment.’ This employs a large tank of cleaning fluid (400 000 litres of C,Cl,!) in which a few neutrinos are captured (with cross section M cm’!) by chlorine-37 nuclei to form argon-37 ions via u + 37Cl + 37Ar+ + e-. The expected capture rate is about three neutrinos per week. The experiment requires circulating helium through the tank to flush out the 37Ar atoms (that presumably result from neutralization of the ions), which are detected by radioactive decay. The experi- ment has been pursued and refined for over ten years, with results which perplex physicists and astronomers. The data indicate a neutrino flux only about one third the predicted amount. The possibility that 37Ar+ ions produced by neutrino capture might be trapped in a molecular cage or compound has been suggested’ but discounted in view of a gas phase mass spectrometric experiment3 which found no evidence for argon molecule ions of the form ArC,,Cl,+.In this experiment, which employed a “ high-pressure ” ion source operated at M 1 Torr, the only observed process induced by Arf ions was charge transfer to produce C2Cl,‘ ions followed by drastic fragmentation to form CC1+, C2C1+, CCl,’-, C2Cl,’- and C,Cl;. However, this result may arise merely from the very large difference in ionization potentials of Ar and C2C14, x6.4 eV. The dis- posal of such a large exoergicity is a dominant factor under gas phase conditions, but need not be in the liquid medium which is M 106-times denser.In our experiment, we employed a supersonic nozzle (diameter 0.1 mm, operated at 110 “C with Ar at 20 atm and C2C14 at 10 Torr) in order to generate van der Waals clusters, Arn(C2C14)m. Ionization of such clusters by electron bombardment (at ~ 2 5 V) permits argon-containing molecule ions to be formed without ionizing argon in the clusters and hence avoids entirely the problem of disposing of a large reaction exoergicity. Indeed, we observed large yields of cluster ions of the form Ar,(C,CI,)+, with n = 1-29 and nz = 1-4. Still larger ions were surely present but beyond the range of our mass spectrometer. The role of ions such as Ar(C,Cl,),+ in the solar neutrino detector of course remains an open question. In these ions, the argon atom must be essentially neutral (by virtue of its high ionization potential); the binding nergy results from polarization by the nearby organic cation.On the long time-scale of the solar neutrino experiment, such polarization seems very unlikely to prevent flushing out the 37Ar by the carrier gas. If the 37Ar is nonetheless tied up by an ionic complex, it should be possible to release the argon by electrolytically discharging the ions. For recent reviews, see J. N. Bahcall and R. Davis, Science, 1976, 191, 264; J. N. Bahcall, Rev. Mod. Phys., 1978, 50, 881. K’. C. Jacobs, Nature, 1975, 256, 560. J. J. Leventhal and L. Friedman, Phys. Rev. D., 1972, 6 , 3338.360 GENERAL DISCUSSION Dr. G. M. McClelland (Stanford) and Prof. D. R. Herschbach (Haruard) said: The elegant experiment of Rettner and Simons has provided the first information about the energy dependence of angular momentum alignment in reactive collisions.Such directional or vector properties of reactions offer dynamical information comple- mentary to scalar properties such as rate constants and product energy distributions. Indeed, a potential energy surface in effect acts as a polarizing lens which induces anisotropies and correlations among the directions of relative velocity vectors and angular momentum vectors of the reactant and product molecules. The theory of angular correlations has proved very fruitful in analysis of nuclear reaction processes' and should have a comparable role in treating angular momentum properties of chemical reactiom2 Laser methods now offer the prospect of deter- mining angular correlations among the directions of the initial and final relative veloci- ties k and k' and the rotational angular momenta j and j ' of reagent and product mole- cules.In anticipation, we have calculated angular correlations for an A + BC + AB + C atom exchange reaction for both statistical3 and impulsive models4 Here we give some results calculated for comparison with the data of Rettner and Simons. The rotational alignment coefficient obtained from their data is the second Legendre moment of the angular correlation of j ' and k , defined by a2/ao = 5(P,(j'-k)). This coefficient is large and negative ( j ' tends to be perpendicular to k ) at high collision energies, but the alignment decreases rapidly and becomes very small as the collision energy descends into the thermal regime.Our calculations of the alignment employ a variant 2-4 of the impulsive DIPR model (formulated originally by Kuntz and Polanyi); this amounts to the spectator stripping model with addition of repulsion between the products. The repulsion is released when atom A approaches BC within the covalent-to-ionic crossing distance, R,. The mean repulsive energy released was taken as 35 kJ mol-' for the Br, reaction (by analogy with photodissociation) and 41 kJ mol-' for the CCl, reaction (chosen to simulate product translational energy data reported at this Discussion for the analogous Cs + CCI, reaction). In the latter case, the product radical CCI, was treated as an atom. Except at the limit of complete alignment (a2/a0 -+ -5/2), the DIPR model strongly over- estimates the alignment coefficient.According to this impulsive model, the align- ment persists even when the collision energy becomes negligibly small compared with the product repulsion energy. This result, which we find is essentially unchanged on introducing various forms for the orientation dependence of the reaction, arises from a Jacobian factor. The " dart board " distribution of initial impact parameters must be projected onto a sphere of radius R, centred on the BC molecule. A Jacobian weight- ing proportional to k*R, is thereby introduced and this anisotropic factor produces the residual rotational alignment obtained at low collision energies. We conclude that the experimental observation of near-zero alignment at low energies cannot be accounted for by simply invoking product repulsion.It will be interesting to see if other models or classical trajectory calculations can find a way to obtain nearly isotropic product rotation without going over to a statistical rather than L. C. Biedenharn, in Nuclear Spectroscopy, Part B, ed. F. Azzenberg-Selove (Academic Press, New York, 1960), D. 732. D. A. Case and D. R. Herschbach, J . Chem. Phys., 1978, 69, 150 and references cited therein. G. M. McClelland and D. R. Herschbach, J . Phys. Chem., 1979, 83, 1445 and references cited therein. D. S. Y . Hsu, G. M. McClelland and D. R. Herschbach, J . Chem. Phys., 1974, 61, 4927; G . M. McClelland, Ph.D. Thesis (Harvard University, 1979). Fig. 5 compares our results with the data of Rettner and Simons.GENERAL DISCUSSION + 361 -2.0 - 0 0 - 0 .5 ~ 1 0 20 40 60 80 100 120 collision energy, Et / k J mol-' FIG. 5.-Alignment coefficient for rotational angular momentum of XeBr* and XeC1* from reactions of Xe* with Br2 and CCl,. Points show experimental data of Rettner and Simons. Curves are calculated from DIPR model. direct reaction mechanism. For the pertinent trajectory calculations available at present,' product repulsion has at least qualitatively the same role as in the DIPR model. To the extent that the analogy between reactions of metastable rare gas atoms and alkali atoms is valid, the product angular distributions (correlation of k and k') should be quite anisotropic for the Br, and CCI, systems. Thus it would be parti- cularly interesting to see what kind of trajectories can give nearly isotropic product rotation accompanying strongly anisotropic reactive scattering. Rettner and Simons remark that the mass combination in the M + CH3X reaction forces the product angular momentum to show large polarization and thereby reveals little about the dynamics. Out of loyalty, we are compelled to note that in fact this proves not to be the case; the kinematic constraint resulting from the light mass of the methyl group is not very severe. An electric deflection study' of the Cs + CHJ --f CsI + CH, reaction found that the j ' , k correlation was only modest. However, it proved feasible also to determine a coefficient related to the triple vector correlation of j r , k r , k . This was stronger and corresponds to preferred orientation of the product rotational angular momentum perpendicular to the plane containing the initial and final relative velocity vectors. Model calculations and trajectory results' indicate such alignment is characteristic when strong repulsion occurs between the products. The preliminary results mentioned by Simons for the ICI reaction indicate a particularly striking comparison with the corresponding alkali reaction^.^ The dif- ference in electrophilic character of I and C1 has an interesting consequence. The attacking M atom transfers its valence electron mainly to the I atom, but the charge usually shifts to the C1 atom as the intermediate (IC1)- ion dissociates in the field of the M+ ion. Even when the electron jump occurs, however, the migration of charge to M. H. Hijazi and J. C. Polanyi, Chem. Phys., 1975, 11, 1 ; J. C. Polanyi, remarks at this Discus- sion. D. S . Y . Hsu, G. M. McClelland and D. R. Herschbach, J. Chem. Phys., 1974, 61, 4927; G. M. McClelland, Ph.D. Thesis (Harvard University, 1979). G. H. Kwei and D. R. Herschbach, J . Chem. Phys., 1969, 51, 1742.362 GENERAL DISCUSSION the C1 atom and dissociation of (IC1)- may be inhibited or precluded for a certain class of collisions. This is expected when the closest approach in the reaction tra- jectory occurs for a configuration in which the I (or I - ) atom blocks the excess of M (or M+) to the C1 atom. There is experimental evidence for this “ blocking” mechanism.’ Thus it is pleasing to see that the contrast between the rotational alignment observed for ICl and Br, offers further evidence for such a process. G. H. Kwei and D. R. Herschbach, J . Chem. Phys., 1969,51, 1742.

ISSN:0301-7249    

DOI:10.1039/DC9796700343

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

CLOSING REMARKS BY STUART A. RICE Department of Chemistry and James Franck Institute, University of Chicago, Chicago, Illinois 60637, U.S.A. Received 14th May, 1979 After I accepted the invitation to say a few words tying together and summarizing the various aspects of this Discussion, I searched the printed records of previous Discussions for hints as to how to do it. I found that each of my predecessors had a different view of the requirements, and that the only common element in the many approaches is the lament that it is impossible to do justice to the range of ideas exposed, and the many facets of the discussion that swirled around them, in the short time allotted for these closing remarks. The structure of this Discussion was a little different from that of earlier ones; the change has made my task easier by virtue of the existence of the Summary papers presented by Prof.Grice, Dr. Smith and Prof. Setser, who have done much of what would have been my work. In any event, I will pick out only a few points for comment. As I listened to the papers and the remarks that followed them, it seemed to me that this meeting had two broad components. One set of contributions concerned the results obtained from mature techniques, such as molecular beam studies of reaction dynamics, the use of trajectory studies to model reactions, the nature of reactions involving atomic species in well defined electronic states and the like. These tech- niques are still yielding information of enormous value and the potential for further important contributions remains great, especially as new technology, such as the use of laser induced fluorescence for detection of product states, is added to the armoury of methods applicable to particular problems.The phenomenological theory under- lying the mature techniques is very well developed, so disagreements arise only when there are discrepancies between the results of different investigators, or when the robust deductions from experiment are used to test microscopic models or to suggest microscopic models. In this Discussion one manifestation of the maturity of this component of the field was an obvious lack of contention. The remarks made following a paper were, almost exclusively, of the nature of short reports of new work akin to that in the printed paper.The second component of the work reported at this Discussion is contentious. I refer specifically to the papers and comments dealing with various aspects of multi- photon dissociation of molecules and of photoselective chemistry. In these cases there were frequent disagreements about how the system is prepared, what happens to it and how the results are to be interpreted. Clearly, there is need for the intro- duction of new concepts which will organize our understanding of these matters. At this Discussion we have heard about many aspects of reactive scattering. The several papers dealt with revival of an old but underutilized technique, the roto- accelerator, about the effects of vibrational and translational energy on reaction rates, about the nature of energy disposal in reaction products, about the comparative reactivities of alkalis and electronically excited rare gas atoms and more.The use of potential energy surfaces, both for qualitative interpretation and for quantitative364 GENERAL DISCUSSION computer simulation of trajectories, provided a central organizing theme for all these studies. I found the discussion of the influence of trapped trajectories on the reaction cross-section, by Child and Whaley, particularly interesting. This work, which illustrates the analytical considerations of Pechukas, is a welcome example of the triumph of intellect over brute force. Leaving aside the various uncertainties that still bedevil the study of multiphoton dissociation, we heard of the possibility that both fluence and intensity are important in determining the rate of reaction, which is a departure from the predictions of the available theoretical models of the process.We also heard that there are now a very few hints that carefully prepared initial states of a molecule can lead to photoselective chemistry. There has been con- siderable scepticism, based on the predictions of statistical models of unimolecular reactions, that photoselective chemistry is possible. It is important that the limits of applicability of various models of reaction dynamics be determined, and several reports at this Discussion have dealt with studies of the domain of validity of the hypothesis that energy randomization is faster than competing reaction processes, which is the basis of the RRKM model.As in other studies of this type, neither protagonists nor antagonists of the hypothesis clearly prevailed. Despite the richness and variety of the offerings at this Discussion, I found there were enormous gaps in the coverage of the subject Kinetics of State Selected Species. I believe we should have heard much more about the following: (1) Insufficient attention was devoted to the nature of the prepared state, how it evolves and how this evolution influences what we observe. Although we have learned from the theory of radiationless processes of the intimate relationships between the character of the excitation source and the state of the system prepared, and how the system behaviour sometimes cannot be well represented as a sequence in which preparation and evolution are disjoint, our consciousness concerning this has not been sufficiently raised.Much of the contention in the interpretation of multiphoton dissociation can be traced to a lack of understanding of the nature of the prepared state. In this sense, George’s analysis of laser assisted collision processes is an ex- ample of properly placing the emphasis on the joint properties of the molecule and the electromagnetic field. We must learn how to interpret, and manipulate, the dressed states representing the interaction between a molecule and some exciting field, since an understanding of these will enable us to direct the chemical processes if and when that is possible. (2) Continuing in the same vein, there was insufficient attention focused on how to generate wave packet initial states, which must be used if we hope to excite bonds, or localized regions of large molecules.Even when considering the reactions of tri- atomic molecules, excitation of a nearly stationary state that involves contributions from nuclear motion delocalized over the entire molecule is less likely to lead to photoselective chemistry than is excitation of a particular bond. (3) Throughout this Discussion very little has been said about the properties of large molecules. I believe that the prospects for selective photochemistry are most promising in the category of bimolecular reactions of excited polyatomic molecules, rather than in the variants of multiphoton induced unimolecular decomposition. The key to developing photoselective reactions will be to establish conditions under which there is efficient competition between a given reaction and energy transfer, and this should be easiest when the excitation is in the lower part of the manifold of states where the dynamics of the coupled oscillators of the molecule can be described as quasiperiodic.GENERAL DISCUSSION 365 (4) We have heard nothing at this Discussion concerning collision-induced intra- molecular energy exchange, particularly when one of the collision partners is electron- ically excited. The very limited data available indicate that the cross-sections for internal energy transfer in that case are very large, yet strong propensity rules govern the pathways of energy transfer.Indeed, it seems possible that, with clever manipula- tion of the properties of the system, we can generate selective collisional excitation of levels that cannot be photoexcited, and use the molecules so prepared as reactants.( 5 ) We have also heard almost nothing of the influence of initial rotational state on the rate of unimolecular reaction, disposition of energy amongst the products, and so on. The very limited data available have thus far not yielded up their secrets. Al- though conservation of angular momentum must influence the reaction dynamics, in the cases for which data are available no systematic pattern of dependence of the rate of the process on initial rotational state has yet been discerned. Unravelling this puzzle will surely yield rich rewards in the understanding of reaction dynamics. (6) Finally, I was surprised that we did not hear much more about the dynamics of nonlinear systems, and the transition between quasiperiodic and stochastic be- haviour in such systems.Understanding the dynamics of coupled nonlinear oscilla- tors goes to the heart of the reaction dynamics of polyatomic molecules. We need further development of the analytical and topological theories of nonlinear dynamical systems, a better treatment of the relationship between the quantum mechanical and classical treatments of these systems, more studies of models that include rotation and the coupling of vibration and rotation, analytical and numerical studies of large amplitude motion in systems of nonlinear oscillators, a better understanding of the nature of and the conditions that determine the transition from quasi-periodic to stochastic behaviour, a simplified but accurate way to describe the dynamics of polyatomic molecules, e.g., by the use of an effective Hamiltonian, and more.I have now spoken about a number of subjects not adequately addressed in this Discussion. My purpose in doing so is not to criticize the contributions of the past few days, but rather to remind all of us that a full understanding of the subtleties of reaction dynamics requires a very broad view, and that contributions to that under- standing can be made by investigations covering the entire range from simple experi- ments to arcane mathematical analyses. I will close my remarks with a story which is intended to tickle the consciences of all participants in the ongoing debate over the nature of intramolecular dynamics and the influence of that dynamics on the rate of reaction.This story concerns a renowned art historian, acclaimed all over the world for the perceptiveness and subtlety of his interpretations of iconography. Some of these interpretations could only be called devilishly clever, and all formed part of a philosophical overview of the place of iconography in the social and pyschological structure of the local population where that art form flourished. Indeed, it seemed that no new find, however strange or seemingly inconsistent with the philosophical overview he had developed, could not somehow be fitted into his world view. At a meeting of art historians, a meeting analagous to this one, he was approached by a young man who professed his admira- tion for the work of the master. How, the young man asked, did he manage to always find the clue that permitted fitting every example into one all encompassing theory? Did the search for such clues require an unusually broad background in psychology, classical studies and the like? Was there never a case that did not fit no matter what considerations were brought to bear? The art historian smiled broadly and said: You make too much of it. It’s really straightforward - in all of my interpretations I simply bend the nail until I hit it squarely on the head!

ISSN:0301-7249    

DOI:10.1039/DC9796700363

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

INDEX OF NAMES* Addison, M. C., 286, 346. Alexander, M. H., 141, 357. Allison, J., 124. Aoiz Moleres, F., 138. Ashfold, M. N. R., 204, 247. Atkinson, G., 240. Band, Y. B., 297. Baughcum, S. L., 306. Berry, M. J., 188. Bowen, K. H., 145. Birkinshaw, K., 1 15. Black, G. W., 118. Bly, S. H. P., 221. Brashears, H. C. Jr, 255. Brophy, J. H., 114. Buebio, S. J., 359. Buss, R. J., 162. Butler, J. E., 346. Child, M. S., 57, 119, 343, 352. Clough, P. N., 114, 223. Clyne, M. A. A., 316, 355. Coggiola, M. J., 162. Connor, J. N. L., 120, 123. Dagdigian, P. J., 141, 357. DeVries, P. L., 90. Dickson, L. W., 221. Ding, A., 353. Dixon, R. N., 349. Donovan, R. J., 286, 346, 353, 354. Dreiling, T. D., 255. Durkin, A., 248. Fluendy, M. A. D., 41, 116, 1 18, 343. Fotakis, C., 353. Freed, K.F., 133, 230, 231, 297, 350. FuB, W., 180. Garraway, J., 286. George, T. F., 90, Golde, M. F., 355. Gonzalez Urefia, A., 138. Gorry, P. A., 115, 347. Grice, R., 16, 115, 248, 347. Hancock, G., 204, 236> 247, 345. Hase, W. L., 226. Hennessy, R. J., 358. Herrero, V. J., 138. Herschbach, D. R., 27, 1 14,145,250,251,359,360. Hippler, H., 173. Hirst, D. M., 115. Hofmann, H., 306. Holmlid, L., 228. Husain, D., 273. Hynes, A. J., 114. Jackson, W. M., 235, 344, 351. Jaffer, D. J., 212. Jakubetz, W., 120, 123. Jarrold, M. F., 1 1 5. Johnson, M. A., 124. Johnston, J., 114. Ketley, G., 204. Kneba, M., 223. Kompa, K. L., 180, 253. Lagana, A., 120, 123. Lawley, K. P., 41, 116, 118. Lee. H. U., 127. Lee, Y . T., 162. Leone, S. R., 306. Levy, M. R., 243. Liesegang, G. W., 145.Lubman, D. M., 238. Luther, K., 173, 229, 230, 235, 238. Manz, J., 120, 123. MacDonald, R. G., 128. Mangir, M., 243. Martin, M., 353, 354. McCafferey, A. J., 112. McCall, J., 41. McClelland, G. M., 251, 360. McCormack, J., 112. McDermid, I. S., 316. Menzinger, M., 97, 136, 142. Miller, C. M., 245. Moore, C. B., 146. Morse, M. D., 297. Naaman, R., 236, 238, 242. Nesbitt, D. J., 306. Nomura, Y., 221. Norris, P. E., 273. Nowikow, C. V., 115, 347. Ono, Y., 358. Polanyi, J. C., 66, 110, 122, 129, 227, 249, 357 Quack, M., 229. Reddy, K. V., 188. Reisler, H., 243. Rettner, C. T., 329. Rice, S. A., 363. Riley, S. J., 27. Roberts, A. J., 247. Rynefors, K., 228. Schatz, G. C., 140. Schreiber, J. L., 66. Setser, D. W., 126, 241, 255, 356. Sholeen, C., 41. Simons, J. P., 329, 348, 358. Siska, P. E., 27, 144, 225, 355. Skrlac, W. J., 66. Sloan, J. J., 128, 226. Smith, D. J., 248. Smith, I. W. M., 146, 212, 223. Sutton, D., 41, 116, 343. Tablas, F. M. G., 180, 246. Tanin, A., 136. Trainer, M., 354. Troe, J., 173, 238. Veltman, I . , 248. Walsh, R., 237, 238. Wanner, J., 354. Wellhausen, U., 223. Whaley, K. B., 57. Whitehead, J. C., 120, 123. Wittig, C., 243. Wolf, R. J., 226. Wolfrum, J., 223. Wong, J. C., 136. Worsnop, D. R., 359. Wren, D. J., 97, 142. Yu, M. H., 243. Yuan, J.-M., 90. Zare, R. N., 7, 124, 236, 238, 242, 245. * The page numbers in heavy type indicate papers submitted for discussion.

ISSN:0301-7249    

DOI:10.1039/DC9796700366

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

GENERAL DISCUSSIONS OF THE FARADAY SOCIETY Date 1907 1907 1910 191 1 1912 1913 1913 1913 1914 1914 1915 1916 1916 1917 1917 1917 1918 1918 1918 1918 1919 1919 1920 1920 1920 1920 1921 1921 1921 1921 1922 1922 1923 1923 1923 1923 1923 1924 1924 1924 1924 1924 1925 1925 1926 1926 1927 1927 1927 1928 1929 Subject Osmotic Pressure Hydrates in Solution The Constitution of Water High Temperature Work Magnetic Properties of Alloys Colloids and their Viscosity The Corrosion of Iron and Steel The Passivity of Metals Optical Rotatory Power The Hardening of Metals The Transformation of Pure Iron Methods and Appliances for the Attainment of High Temperatures in a Laboratory Refractory Materials Training and Work of the Chemical Engineer Osmotic Pressure Pyrometers and Pyrometry The Setting of Cements and Plasters Electrical Furnaces Co-ordination of Scientific Publication The Occlusion of Gases by Metals The Present Position of the Theory of Ionization The Examination of Materials by X-Rays The Microscope : Its Design, Construction and Applications Basic Slags : Their Production and Utilization in Agriculture Physics and Chemistry of Colloids Electrodeposition and Electroplating Ca pi 1 lari t y The Failure of Metals under Internal and Prolonged Stress Physico-Chemical Problems Relating to the Soil Catalysis with special reference to Newer Theories of Chemical Action Some Properties of Powders with special reference to Grading by Elutria- The Generation and Utilization of Cold Alloys Resistant to Corrosion The Physical Chemistry of the Photographic Process The Electronic Theory of Valency Electrode Reactions and Equilibria Atmospheric Corrosion.First Report Investigation on Oppau Ammonium Sulphate-Nitrate Fluxes and Slags in Metal Melting and Working Physical and Physico-Chemical Problems relating to Textile Fibres The Physical Chemistry of Igneous Rock Formation Base Exchange in Soils The Physical Chemistry of Steel-Making Processes Photochemical Reactions in Liquids and Gases Explosive Reactions in Gaseous Media Physical Phenomena at Interfaces, with special reference to Molecular Atmospheric Corrosion. Second Report The Theory of Strong Electrolytes Cohesion and Related Problems Homogeneous Catalysis tion Orientation ' Crystar Structure and Chemical Constitution 1929 Atmospheric Corrosion of Metals.Third Report 1929 Molecular Spectra and Molecular Structure 1930 Colloid Science Applied to Biology 193 1 Photochemical Processes 1932 The Adsorption of Gases by Solids 1932 The Colloid Aspect of Textile Materials 1933 Liquid Crystals and Anisotropic Melts Volume Trans. 3 3 6 7 8 9 9 9 10 10 11 12 12 13 13 13 14 14 14 14 15 15 16 16 16 16 17 17 17 17 18 18 19 19 19 19 19 20 20 20 20 20 21 21 22 22 23 23 24 24 25 25 26 26 27 28 29 29368 Date 1933 1934 1934 1935 1935 1936 1936 1937 1937 1938 1938 1939 1939 1940 1941 1941 1942 1943 1944 1945 1945 I946 1946 1947 1947 1947 1947 1948 1948 1949 1949 1949 1950 1950 1950 1950 1951 1951 1952 1952 1952 1953 1953 1954 1954 1955 1955 1956 1956 1957 1958 1957 1958 1959 1959 1960 1960 1961 1961 1962 1962 1963 GENERAL DISCUSSIONS OF THE FARADAY SOCIETY Subject Free Radicals Dipole Moments Colloidal Electrolytes The Structure of Metallic Coatings, Films and Surfaces The Phenomena of Polymerization and Condensation Disperse Systems in Gases: Dust, Smoke and Fog Structure and Molecular Forces in ( a ) Pure Liquids, and (b) Solutions The Properties and Functions of Membranes, Natural and Artificial Reaction Kinetics Chemical Reactions lnvolving Solids Luminescence Hydrocarbon Chemistry The Electrical Double Layer (owing to the outbreak of war the meeting The Hydrogen Bond The Oil-Water lnterface The Mechanism and Chemical Kinetics of Organic Reactions in Liquid The Structure and Reactions of Rubber Modes of Drug Action Molecular Weight and Molecular Weight Distribution in High Polymers.(Joint Meeting with the Plastics Group, Society of Chemical Industry) The Application of lnfra-red Spectra to Chemical Problems Oxidation Dielectrics Swelling and Shrinking Electrode Processes The Labile Molecule Surface Chemistry. (Jointly with the Societe de Chimie Physique at Colloidal Electrolytes and Solutions The Interaction of Water and Porous Materials The Physical Chemistry of Process Metallurgy Crystal Growth Lipo-Proteins Chromatographic Analysis Heterogeneous Catalysis Physico-chemical Properties and Behaviour of Nuclear Acids Spectroscopy and Molecular Structure and Optical Methods of Investi- gating Cell Structure Electrical Double Layer Hydrocarbons The size and shape Factor in Colloidal Systems Radiation Chemistry The Physical Chemistry of Proteins The Reactivity of Free Radicals The Equilibrium Properties of Solutions on Non-Electrolytes.The Physical Chemistry of Dyeing and Tanning The Study of Fast Reactions Coagulation and Flocculation Microwave and Radio-Frequency Spectroscopy Physical Chemistry of Enzymes Membrane Phenomena Physical Chemistry of Processes at High Pressures Molecular Mechanism of Rate Processes in Solids Interactions in Ionic Solutions Configurations and Interactions of Macromolecules and Liquid Crystals Ions of the Transition Elements Energy Transfer with special reference to Biological Systems Crystal Imperfections and the Chemical Reactivity of Solids Oxidation-Reduction Reactions in Ionizing Solvents The Physical Chemistry of Aerosols Radiation Effects in Inorganic Solids The Structure and Properties of Ionic Melts Inelastic Collisions of Atoms and Simple Molecules High Resolution Nuclear Magnetic Resonance The Structure of Electronicallv-Excited SDecies in the Gas-Phase was abandoned, but the papers were printed in the Transactions) Systems Bordeaux.) Published by Butterworths Scientific Publications, Ltd.1963 Fundamental Processes in Rahiation Chemistry Volume 30 30 31 31 32 32 33 33 34 34 35 35 35 36 37 37 38 39 40 41 42 42 A 42 B Disc. 1 2 Trans. 43 Disc. 3 4 5 6 7 8 Trans. 46 Disc. 9 Trans. 47 Disc. 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36GENERAL DISCUSSIONS OF THE FARADAY SOCIETY 369 Date 1964 1964 1965 1965 1966 1966 1967 1967 1968 1968 1969 1969 1970 1970 1971 1971 1972 1972 1973 1973 1974 1974 1975 1975 1976 1977 1977 1978 1978 1979 1979 Subject Chemical Reactions in the Atmosphere Dislocations in Solids The Kinetics of Proton Transfer Processes Intermolecular Forces The Role of the Adsorbed State in Heterogeneous Catalysis Colloid Stability in Aqueous and Non-Aqueous Media The Structure and Properties of Liquids Molecular Dynamics of the Chemical Reactions of Gases Electrode Reactions of Organic Compounds Homogeneous Catalysis with Special Reference to Hydrogenation and Bonding in Metallo-Organic Compounds Motions in Molecular Crystals Polymer Solutions The Vitreous State Electrical Conduction in Organic Solids Surface Chemistry of Oxides Reactions of Small Molecules in Excited States The Photoelectron Spectroscopy of Molecules Molecular Beam Scattering Intermediates in Electrochemical Reactions Gels and Gelling Processes Photo-effects in Adsorbed Species Physical Adsorption in Condensed Phases Electron Spectroscopy of Solids and Surfaces Precipitation Potential Energy Surfaces Radiation Effects in Liquids and Solids Ion-Ion and Ion-Solvent Interactions Colloid Stability Structure and Motion in Molecular Liquids Kinetics of State Selected Species Oxidation For current availability of Discussion volumes, see back cover.Volume 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67

ISSN:0301-7249    

DOI:10.1039/DC9796700367

出版商:RSC    

年代:1979

数据来源: RSC