Collisional quenching of electronically excited tin atoms, Sn(5p2 3P1

P. D. Foo, J. R. Wiesenfeld, M. J. Yuen, and D. Husain. J. Phys. Chem. , 1976, 80 (2), ... Paul J. Dagdigian and Mark L. Campbell. Chemical Reviews 19...
1 downloads 0 Views 865KB Size
T H E

J O U R N A L

OF

PHYSICAL CHEMISTRY Registered in U.S. Patent Office 0 Copyright, 1976, by the American Chemical Society

VOLUME 80, NUMBER 2 JANUARY 15,1976

Collisional Quenching of Electronically Excited Tin Atoms, Sn(5p2 "1) by Time-Resolved Attenuation of Atomic Resonance Radiation

and Sn(5p2 3P2),

P. D. Foo, J. R. Wlesenfeid,' M. J. Yuen, Department of Chemistry, Cornel1 University, Ithaca, New York 14853

and D. Husain The Department of Physical Chemistry, The Universtty of Cambridge, Cambridge CSZ IE?, EnQland (Received September 15, 1975) Publication costs assisted by the Petroleum Research Fund

An investigation of the collisional behavior of the gaseous tin atom in the and 5p2(3Pz)spin-orbit states, 0.210 and 0.425 eV above the ~ P ~ ( ~electronic P O ) ground state, respectively, has been carried out. The transient, optically metastable atoms were generated by the pulsed irradiation of SnMe4 and monitored photoelectrically in absorption by the time-resolved attenuation of atomic resonance radiation derived from a microwave-powered sealed discharge. Modification of the lifetime of the two spin-orbit states by added quenching gases has been studied in detail and has led to absolute rate constants ( k ~at) 300 K for collisional quenching of both the 3P1and 3P2states by the gases Ar, Kr, Xe, H2, Dz, HD, Nz, 02,CO, NO, COz, CHI, CF3H, CF4, C2H4, CzD4, CzH2, CzD2, and SnMe4. The results are compared with the analogous data for the low-lying spin-orbit states of the lead atom, P b ( 6 ~ ~ ( ~ P 1 , 2Discussion )). of quenching by the noble gases is principally in terms of the Hund's coupling case (c) components arising from the interaction on collision; detailed consideration is given to quenching by the hydrogen isotopes especially in terms of the effects of long-range quadrupole-quadrupole coupling and selection rules for rotation for near-resonance transfer processes on collision; diatomic molecules in general are discussed in terms of correlation diagrams based on ( J ,n) coupling and the effects of E + V,R transfer; quenching by polyatomic molecules is briefly considered, principally in terms of the interaction on collision and the observation of an approximate trend of k~ with the ionization potential for the larger molecules.

Introduction

vast body of data, for example, on quenching of electronically excited mercury atoms.s Indeed, the now large numWhile there has been some recent development in direct ber of absolute rate measurements on various highly enermeasurements of the collisional behavior of heavy atoms in gized spin-orbit states arising out of the electronic ground electronically excited states, many of the results have been state configulation, such as I(5p52P1/z) (0.943 eV),399-11 dealt with on an individual basis, with little reference being Tl(6p 2P3/~) (0.966 eV),12-14 Te(5p43 P ~ , l(0.584 ) and 0.589 made to a unified treatment for considering the relationeV, respectively),15 and Br(4p52P1/2) (0.457 eV)3J6J7 also ship between electronic structure and atomic rea~tivity.l-~ fell outside this general structure, which, by definition, The fundamental reason for this is that the most widely must omit the effects of J splittings. used general structure for discussing atomic reactivity in Very recently, Brown and Husainls have shown that ( J , excited states has been based on correlations between iniSi) coupling, presented in general form for heavy atomtial and final states derived from the weak spin-orbit coumolecule collisions by Husain,lg provides the best framepling a p p r o ~ i m a t i o n With . ~ ~ the main exception of recent work within which to consider the chemistry of the np2 lS0 work by Callear and M ~ G u r k who , ~ have considered in a and 'Dz states of atomic tinl8Pz0and lead.20~22 Similarly, we detailed manner the nature of the molecular orbitals inhave shown in a preliminary investigationz3 that the lowvolved in the collision complex, this hitherto prevailing atlying spin-orbit states of the tin atom, Sn(5p23P1) and titude toward heavy atom chemistry applied even to the Sn(5p23Pz), respectively, 0.210 and 0.425 eV above the

92

Wiesenfeld et al.

5p2 3 P ground ~ state,24 and also those of Pb(6p23P1,2),25 can be best discussed within the same context. This present paper describes detailed and extensive absolute rate measurements for the collisional quenching of the 53P1 and s3P2 states of the tin atom. The resulting data for deactivation by small molecules are compared with those for Pb(63P1,2) and are discussed within the (J,Q) framework, the nature of the interaction on collision with polyatomic molecules also being taken into account. The overall objective of this present work should be viewed within a broader program of seeking a general fundamental structure for considering the reactions of both light and heavy atoms. One may also stress, in the present context, current interest in electronic transition atomic lasers and especially visible chemical lasers arising from heavy atom-molecule collisions.26-29

where the new t would be of the standard dimensions, (c0-l. Secondly, the plots used to determine y, i.e., the first-order intercepts, In (In (ZolZt,))t=o vs. In ( ~ s ~ Min ~ ~ ) this instance, simply constitute a device for determining relative values of atomic concentrations, using the weak light absorption approximation (by the parent molecule), for decays which are found experimentally to be kinetically first order. Thirdly, Bemand and C l ~ n have e ~ ~used the better procedure of employing a power series for resonance absorption by O ( 2 3 P ~where ) the absolute concentrations can be determined by titration techniques. Unfortunately, we have no chemical titration technique for S I I ( ~ ~and P~) Sx1(5~P2),analogous to that for the oxygen atom in a discharge flow system, and eq i or ii is, a t present, the only convenient route to relative measurements of the concentrations of the excited atoms. Materials. Ar (Research Grade, Matheson Co.) was used Experimental Section directly. Kr, Xe, H2, D2, N2, 0 2 , CO, NO, COP, CHI, CFBH, CF4, C2H4, CzH2, and SnMe4 were prepared essentially as The general nature of the experimental arrangement has been given hitherto for the kinetic study of T1(62P3/2)13J4 described in previous publications (ref 18 and references and briefly referred to in our preliminary c o m m ~ n i c a t i o n . ~ ~contained therein). HD (Merck and Co., 98% mole purity) was used directly, as was C2D2 and C2D4 (Merck and Co.). We will here limit consideration to the salient features of the system and a few aspects not discussed previously. The and Discussion basis of the method is to generate Sn(53P1) and S I I ( ~ ~ P ~Results ) Figure 1 shows typical oscilloscopic traces indicating by the pulsed irradiation ( E = 400 J) of low pressures of both time-resolved resonance absorption by S I I ( ~ ~ as P ~an) , SnMe4 (1.5-3.5 X Torr) in the presence of excess example of the type of raw data obtained in these experiargon buffer gas ( ~ A ~ : ~= sca.~ 1OOOOO:l) M ~ ~ in order to ments, and also modification of the lifetime of the tranprevent any significant rise above ambient temperature on sient species by the addition of carbon monoxide. Figure 2 photolysis. Einstein coefficients, both magnetic dipole and shows first-order kinetic plots derived from the data of Figelectric quadrupole for spontaneous emission from these ure 1. Similar sets of data to those presented in Figures 1 spin-orbit states, have been calculated by Garstang30 and and 2 were obtained for the decay of Sn(53P2)and Sr1(5~Pl) indicate quantitatively the high optical metastability of in the presence of the various quenching gases. When these species. Thus they can be readily monitored photoeSnMer is flash photolyzed in the Spectrosil region (A 2 165 lectrically in absorption by time-resolved attenuation of nm), all of the states, Sn(5lD2), Sr1(5~Pz),Sn(53P1), and the following two resonance transitions: Sr1(5~Po),are generated in significantly decreasing 10D8gA, y i e l d ~ ; l * ,Sn(5lSo) ~~ is not produced in any significant Transition A , nm sec-’ 3 1 YZ3 yield.ls The present method, however, was not sufficiently sensitive to permit detection of population following relax10.0 0.77 + 0.04 ~ s ( ~ P ; )~- P+ ’ ( ~ P , ) 270.65 21.0 0.67 k 0.09 ~ S ( ~ P ; )~- P ~ ( ~ P , )284.00 ation from higher states and the decays for both Sn(53Pz) and Sn(53P1) are taken to be first order. The sensible linThese were derived from a microwave-powered (incident earity of plots of the type given in Figure 2 for both spinpower = ca. 10 W) sealed source (E.M.I. electrodeless disorbit states in the presence of all the added quenching charge tube) and optically .separated by means of a 0.5-m gases indicates the reasonable nature of the approximation. grating monochromator. The resonance absorption signals The slopes of the first-order plots in individual experiwere detected by means of a photomultiplier tube (RCA ments (e.g., Figure 2) are given by -yk’ where k’,the overCorp., 1P28) mounted a t the exit slit, monitored on an osall first-order decay coefficient derived by means of the apcilloscope and photographed for subsequent kinetic analypropriate value of y (see earlier), is taken to satisfy the sis. form The appropriate plots leading to the above y values for k’ = k ~ [ $ ] K (iii) the conditions employed using the now standard modified Beer-Lambert law32 and k~ is the absolute second-order rate constant for collisional removal of the individual spin-orbit state by the I+,= IO exp(-t(cl)y) (9 added quenching gas, Q . K is taken to be a constant in a (where the symbols have their usual ~ i g n i f i c a n c e ~have ~) given series of kinetic runs in which [Q]is varied. I t combeen given in the preliminary c o m m ~ n i c a t i o n Various .~~ prises contributions to the first-order decay from weak aspects of this modified law have been discussed in a numspontaneous emission,3° diffusion, and quenching by the ber of papers concerned with attenuation of resonance rabuffer gas, impurities, products of photolysis, and the undiation (see references in reviews, ref 1and 2). Three points dissociated parent molecule. In fact, it is the latter that merit mention. First, t in eq i is not the standard extinction principally governs the magnitude of K in these expericoefficient but an arbitrary constant, of dimensions ( 4 - 7 , ments. Figures 3 and 4 show examples of the variation of in a given set of experiments, depending on the value of y. yk’ for the two spin-orbit states with added quenching We have by custom adopted eq i here but could equally gases, including carbon monoxide. The slopes of the plots employ the equation given in Figures 3 and 4, in conjunction with the appropriate values of y, yield the absolute values of k~ for collisionIt, = 10 exp(-(tcl)r) (ii) ~~

+

The Journal of Physical Chemistry, Vol. 80, No. 2, 1976

Collisional Quenching of Electronically Excited Tin Atoms

93

(bl

Figure 1. Typical oscilloscopic traces for the decay of Sn(53P2)in the presence of carbon monoxide obtained by attenuation of atomic resonance radiation at A 284.00 nm. ~ s = ~3.3 X M lov4 ~ Torr, ~ ptotal with Ar = 30 Torr; E = 400 J. (a, b) 100 psec/division: (c, d) 50 psec/divislon lo3 pco (Torr):(a)0.0; (b) 3.2;(c)7.4; (d) 9.8.

05 r -

-

IW 2m

3m LOO 500 TI& ips

600

Pseudo-first-order plots for the decay of Sr1(5~Pp) in the presence of different pressures of carbon monoxide. 1O3pc0 (Torr): (0)0.0; (0)3.2;(A)7.4; (A)9.8. Flgure 2.

103pc0,10'pc

Pxe

Psuedo-first-order rate coefficients (yk') for the decay of Sn(Lj3P1)in the presence of different quenching gases. the analogous data for Pb(63P1,2) obtained by Husain and Littler35 and subsequently by Ewing et a1.36337 Noble Gases. The general behavior for the collisional quenching of Sn(tj3P1) and Sr1(5~P2)by the noble gases, Ar, Kr, and Xe, essentially follows that observed hitherto for the analogous states of the lead atom (Table I). First, quenching is relatively inefficient. Secondly, the more energized 3Pz state is deactivated more efficiently than the 3P1 state for a given noble gas deactivator. Thirdly, the deactivation efficiency increases with increasing atomic weight of the noble gas partner. The standard, general principle in such processes must clearly operate, namely, that the transfer of large quantities of electronic energy to translational energy can only occur with any significant probability if the potential curves describing the initial and final states either cross or approach sufficiently closely for quantum mechanical tunneling to become significant. Further, such crossing in a low order to approximation must be accompanied by a mixing following the inclusion of higher order terms in the Hamiltonian in order for a so-called "nonadiabatic transition" (NAT) to take place. If we designate states in Hund's coupling case (c) in the standard manner,38 then the molecular fl values clearly indicate the mechanism for relaxation from the 3P2 to the 3P1 level. Thus, for Sn(3P2) noble gas SO), the states arising from 2, 0, are 2, 1, and 0+, while the Sn(3P1) noble gas interaction, 1, 0,, yields 1and 0- and A 0 = 0 for fl = 1 1. By contrast, no common case (c) components arise from the 3P1and 3P0states as the 3P0 IS0 interaction corresponding to 0, + 0, yields only O+. Alternatively, designating the states in Hund's cases (a) or (b),38we see that 3P, 'S, 38- + 311. On this basis, Ewing et have constructed arbitrary repulsive curves for the P b noble gas atom interactions, designating 3P1and 3P0 noble gas in case (c) a t large interatomic distances, merging into a common 3 2 state at closer internuclear separation. Critical to such a discussion are the regions of crossing, which, as Ewing et al.36 point out, could be probed by temperature dependent studies. Similar considerations will, of course, apply to the lighter tin atom where the case (a) and (b) designations at close internuclear separations will constitute somewhat better descriptions of the interactions. Further, a partial minimum, estimated by an approximate Lennard-Jones int e r a ~ t i o nshould ~ ~ assist such a mechanism by an acceleration effect. The foregoing discussion may be contrasted with the similar but far more detailed argument for the

+

-b x

+

-

+

+

L-

1OAPCHL

Torr

-

Pseudo-first-order rate coefficients (yk') for the decay of Sn(53P2)in the presence of different quenching gases.

Figure 3.

/torr

Figure 4.

:KO 1

2 L

a1 removal of S I I ( ~ ~ Pand ~ ) S I I ( ~ ~ PTable ~ ) . I includes all the quenching data for Sn(53P1,2) obtained in this investigation and the preliminary c o r n m ~ n i c a t i o n .It ~ ~constitutes, to the best of our knowledge, the only body of rate data for these two spin-orbit states. Included in Table I are

+

+

-

+

The Journal of Physical Chemistry, Vol. 80, No. 2, 1976

Wiesenfeld et al.

94

TABLE I: Rate Constants ( h ~cm3 , molecule-' sec-', 300 K ) for the Collisional.Removal of the Spin-Orbit States of Atomic Tin (n = 5 ) and Lead (n = 6)by Various Gases (M)' -_I__._

M

Sn(53Pl)(0.210 eV)

Ar

< 5 x 10-16b

Kr

1.1 t 0.3 x 10-15 3.2 i. 0 . 3 x < 2 x 10-12b

Xe HZ

1.5 e 0.4 x 10-I' < 2 . 5 x 10-13

HD N2

8 . 2 f 0.5 x 10-"b

0'

co

-

I

1.62 e 0.5 x 10-l1b 5.4 k 0.8 x lo-'' < 2 . 9 x 10-13 i

___I______.

I

1.6 i 0.2 x 10-15 1 . 2 e 0 . 3 x 10-14 1.15 e 0.3 x 10-lzb

4.91

P b ( 6 T Z )(1.320 eV)

Pb(63P1)(0.969 eV)

< 1 x 10-'6b

1.09 e 0.3x 10-1*b

UZ

Sn(53P,) (0.425 eV)

0.3 x lO-'*b

1.7 e 0.2 x lo-" 6.7 i: 1 . 0 x lo-" 3.2 i: 0 . 2 x 10-13 1.5 i 0.2 x lo-" 7 . 6 e 0.5 x loq1' 2.6 i 0.2 X lo-'' 5.7 i. 0.4 x lo-'' 3.6 i: 0.7 x lo-" 1 . 4 i: 0.1 x lo-'' 8.1 i. 0.1 x lo-'' 2.00 i 0.05 x lO-'Ob

0 i 1.0 x 1O-l6c < 2 . 3 X 10-l6d