i~iergyfor 0-0 Iscind clenvage based on simple peroxide anodel:?. i n Imic/ elk m d tetrasubbtitutcd ozonides sfcr is inhibited or iniposeible plai~?~icn.a cliscretc hiradical
"IE',
J
~
Acknowledgment. The authors appreciate R useful conversation with Dr, R. l\Xnrray. This n o & was supported by the Environmental Pt-otectron Agency Ihrougli the Office of Air Programs under Grants S o . hJ'00018 and AP01044, for which WE are grateful
Mercury Atoms by Isotopic Aromatic Molecules
~ J. Mains ~ and b Mendel ~ Trachtman" ~ t
DepaTtmsnts of Chemistry, Philadelphia College of Textiles and Science, Philadelphia, Pennsylvania Oklahoma State University Stillwater, Oklahoma 74074 (Receiaed October 22, 1971)
29144, a n d
Publkalion costs assisted by Philadelphia College of Textiles a n d Science
Quenching data are reported for benzene, nine deuterium-substituted benzenes, and three deuteriumsubstituted toluenes. Attempts to correlate the observed square of collision diameter for the quenching process, crz, in. terms of simple C-H and C-D bonds and in terms of CH-CH, CH-CD, and CD-CD pairs were unsuccessful. The data were shown to be consistent with a previously proposed mechanism invoiviiig exciplex formation and intersystem crossing. An interpretation involving 3 P Hg ~ atoms could not be ruled out but was considered unlikely.
Inl ermoleculer transfer of electronic energy is a powerful tool in photochemistry, spectroscopy, and chemical kinetics. M o r e specifically, the technique of mercury photosensitization has proven an invaluable means of generating and studying free radicals near room temperature. Indeed, much of our early knowledge of rate constants for free-radical reactions was obtained this nay, arid the late E. W. Steacie's book, "Atomic and Free TZadical Reactions," was a tribute to the power of the method. While much effort has been directed toward the accumulation of data in the subsequent two decades, little progress has been made toward a theoretical understanding of the energytrnrisfer process. Because of the size of the mercury atom and, hence, the number of assumptions which must be mad., therc seems little hope of an ab initio approach through g uantum mechanics even with the high speed computation Facilities currently available. Yang' attempted w rationalization of the differences een the energy-transfer (quenching) rates of hydrogen and paraffin8 in terms of symmetry correlations, treating the (R--H--X€g*)complex as a nonlinear triatomic system. Yikis and r\'doser2 have approached the tlicoretical problem from Ihe viewpoint of unimolecular rate theory, treating the (R-H-Hg") complex as tong lived, and assumed applicability of the RRKM thkory. Straiusz3 considers the quenching reactions of
a.
excited mercury with paraffins in ternis of a simple hydrogen-transfer reaction and has computed potential energies of activation by a modified bond energy-bond order method. Whereas there i s an abundance of quenching data for the excited mercury-paraffin systems, such is far Crom the circumstance in excited mercury-aromatic hydrocarbon systems. Except for one measurement for benzene, the only broad study of quenching of the 3P1Hg" atoms by aromatic molecules was an earlier report from this l a b ~ r a t o r y . ~Bn ihis earlier n ork the reported squares of the collision diameters, 0 2 , for the quenching process were shown to be more than 50% great,er than those estimated from van der Waals radii. The mechanism which was tentatively advanced to explain these observations, especially the unexpectedly larger quenching rate of C6D6as compared with C6H6, involved the formation of a long-lived excited mercuryaromatic exciplex, Hg"-Ar. It was notcd then that the role of 3P0Hg atoms in the system was unknown and, until more data were obtained, there \%asno point in speculating further. The purpose of the present paper is to provide additional Hg" quenching data for (1) K. Yang, 3.A m e r . Chem. Soc., 89, 5344 (1967). (2) A . C. Vikis and H. C. Moser, J . Chem. Phys., 53, 2333 (1970). (3) 0. P. Strausz, private communication, 1972. (4) G. J. Nlains and M. Trachtman, 3 Phot. C h e m , 14, 1647 (1870). T h e Jozkrnal of Physical Chemistry, Vol. 76, N o . 19, 1.972
GILBERTJ. MAINSAND MENDELTRACHTMAN
2666 partially deuterated benzenes which, admittedly, will not resolvc the question of the role of Hg atoms in thcsc systems, but will contribute information about the cffects of deuteration on the quenching process and lend support to the proposed mechanism. (An apparatus to determint. the role of 3 P 0 Hg atoms in these EIg*-aromatit hydrocarbon systems is under construclion and results from that study xi11 be reported in the futurr.) Tho prescnt research is intended as a further contribution to ihc data on quenching of 31’1 Hg* atoms by aromatic hydrocarbon molecules.
Experimental Section The apparatus was dcscribcd p r e v i ~ u s l y . ~Fluores-
those r e p o r t 4 p r e v i ~ u s l y . ~We nom report u2 values of 49 f 3 A2 (cornpared with 39 arid 74 2 (compared with 65 A?). The discrepancy is attributed to a lack of statistical precision in the earlier measurements. To avoid such fluctuations in the present work, the data reported in Table I rrpresent no fen er than ten independent experiments for carh compound, and the uncertainty limits reflect a 95Yo confidence level based upon the least-squares fit of thc data to cy 1.
Az)
*
Table I: Quenching Data for (3P1)EIg a t 23”
c ~ n c emcasumirimts, z.e., the determination of t h r fluoresccnt photocurrent, Q, were made randomly at high and lon pressures of quenching gas, AI, to avoid systematic wrors. The data so obtained were found to be consistent $1 ith the modified Stern-Volmer forniiiia developcd by Yang.6 The data were plotted arid fittcd by t h r ~nwthud of Icast squares to the eq 1, whew Q and QOar(i tlie fluorescent photocurrents in the prwii - Q / Q ~ I - -=~ a - p[ir]-i
(1)
cncc and absc-nee. of the quenching gas; [MI is expressed i n molcs/litrr; mid a and 0 arc> constants related to thc queric3hing ratc constant, c/, and the mean lifetime of -VI Hg* atoms i n thch (+ell,7 , by the equation p = arkq. T h c ratio of thc. slope to the intercept, alp, is a direct measure of the product, dc,. Since the mean lifctime of t h c excited mercury atoms can bc calculated from the formula derived by Yang5 for a cc~llof similar goometry, z.?., 7 = 7 0 / 1 0.25 X 104pr],uhcrt. 7-0 is thc mean radiative lifetimt of an isolated mercury atom. 0.108 pscv, p is t h e prcmurc of mercury vapor, 1.85 X Torr, and I* is t h r distance b e t u c m the irradiated slab of I-Tg vapor and the u indow of emerging fluorescence, 3.X) cm in this research, li, may be calctulatcd from thcl obscrvations. Values for the qwnrhing diameter squartd, u z . can be readily calculated from lc, from c~lcnimtary collision thcory6 and are simply relatc d i o the qucnrhing cross section, ~ u * / 4 . Thc isotopic purity of the benzrnes was determined by mass spcvtromctry using a Consolidated Electrodynamics Corp. :Llodtll 21-490 mass spectrometer; isomeric- purity ivah mcmurcd by gas chromatography. ‘I’hcl isotopic bcnzrnc.s v crc found to be 99.5% isotopically pur(’ as wrtiticd. KO impurities w r e detwtcd by gtis cliromatography, arid the compounds w crv thcrt.forc uwd as rcwived.
+
Results and Discussion Thci rcsults. tabulated as u2 at 23”, arc presented in ‘I’abie I for benzcm. and nine deuterated benzenes. Also included arc data for three isotopically substituted toluenes. It necd bc riotcd that the collision diameters for C6HG and C6D6 rcportrd in Tablc I arc largcr than T h e Journal of f’ilysical C h e m i s t r y , Vol. 76, No, 19,1972
Certain observations need bc made regarding the data of Table I. First, while thew seems to be a gcnera1 increase in C? as dcuterium is added to benzene, the variation is not a linear function of tho composition. Thus, any attempt to correlate collision diametws with the numbers of C-H and C-D bonds, successfully applied to aliphatic hydrocarbons7 is doomed to failure ab initzo. The differencc in g 2 for the ortho, meta, and para isomers of C6H4D?arc quitr outside expcJrimental uncertainties and a clear indication of structural effects. Similarly, the differences between the quenching diameter squarcld for t h r t n o structural isomers of CcHZDj reinforces the oxistcncc~of structural eff ects. The remarkable enhancement of quenching efficiencies whenevcr t\?o deuterium atoms were substituted in adjacent positions on thc benzene ring, c y . , 1,2C6H4D2, 1,2,4,5-C6H2D4. l,3,4,5-C0HzD4, and higlicr deutcratcd bcnzt:ncs, suggested that the cnergy-transfer process might be linked to p a i w of dcuterium atoms located on thc aromatic ring. Accordingly, an attempt was made to corrrlate the observed c2 data in tcrms of contributions from the ring structurv plus contributions from adjacent CH--CH, CH-CD, and CD-CD sites 011 the quenching molecule. For c~xample,the square of the quenching diamc1tc.r for o-C6HJ12 would be n rittcn as thc sum ~ U I I D ~ 3 ~ 1 3 ~C‘ U D D ~ . I n this system on(’ would expect identical quenching cross sections for vi- and p-CsH4D, and, indeed, this is observed within the statistical prwision of t h t b data. T T n fortunately, one would also predict identical valucs of
+
+
(5) K. Ymg, J. A m e r . Chem. Soc., 87, 5294 (1065). (6) C . K. Yang, ibid., 88, 4575 (1966). (7) 11. E. Gunning and 0. 1’. Strxuaa, Adcun. I’hotochm , 1, 253 (1963).
THEQUENCHISG or
MERCURY ATOMS
2667
u2 for the two isoiners of C6H2lj4,and this is far from the observation Wit1 pars isomer (for H's) has a sigriiiirantly 1argI.r rju~nchicgtliamcter squarcd than thv zwtit isomer, 72 u:>. 00 ,12! T I ~extcrision P of t i l e corrrlation to l s r p ( ~striictural features, for wample to C'H 4211- C1-f typc. sites, -was distasteful and hardly w t i r r m t d b y th:. d:tta currcritly available. Wliilc riot simply correlatable, t h r data arc uriderstunduLlc izi trrrns of t hc mechanism advanced earlicr, tZ,:
Hg
+ h v --+ HR*(~I'~)
Hg* + Hg 4- h u Hg*
+ Ar. --*
=
I,
rate = kr[Hg*]
(Kc:" Ar)
(€Xg*-Ar) =- (Hg
rate
ratc
=
Icc[€Tg*][Ar1
AT")
rate
=
Ic,[Hg*-Ar]
(€Ig*-hrj
=
Hg"
-t Rr
ratr
=
kd[Hg*-Ar]
(Hg-Rr*)
=
€ 1i ~ Ar"
ratc
=
kt[Hg-Ar*]
hr.*= Ar
rate
=
k,[Ar*]
where it WE suggcstc>dthat variation in k Q , Le., in could be rationalized in tcrms of the fraction of cxciplex complexes, Hg*- h r , which undcrgo a typt of intersystem csrossing t o thc cxciplcx, Hg-Ar*, bcfore drcomposing back to Hg" and unexcited aromatic molecule, Ar. Since t he enrrgy Ivvrls of aromatic molecules, cspechially bcnzcnr, arc' ncll understood, i t is possible to idcntify Ar" as thtl 31"lu statc which, likc tlic IBz,,, can only bc coupled to tlii' ground statc with an ci2p vibration. If v e assum(' the vibration is thc v18 vibration, as suggested in Bcrzbrrg's book8 and diagrammed carlicr by I-Ierzbrrg,Y on(' might rxpect IC, to be c.xtremcly scnritive to t.ubstitution in tlic 1, 2 , 4,and 5 positions. lndcrd, t h r data of Table I ronfirm this cxpectation. With the. cxccpt ion of pcrdeutcriotoluenc., n hioh may be rnlrd out 011 othcr bases, all molcculcs which have a t least thf. 1, 4 a r d 5 sites dcutcratcd rxhibit u2 values ncar '70 ,A2. I'urtlicrmorcb, all molccules which h a v ( ~at least thc I , 2 , 4, and 5 positions protiatcd cxhibit o2 ncar 50 12. Intermediate cases with fcirer than four deuterium atoms at the 1, 2 , 4, and 5 sites arc' intrrmrdiate in 2 and often overlap, so that the ordrr of incrvasirig d is uncertain. Thus, the data of Table I arr at least consistcnt with the mwhariism pr0posc.d earlier
2,
It would bc satisfying if the u2 data correlated with thr vaiucs of k , deduced from pliosphorcscencc lifctimc m w u r c m r n t s and which formed tiic basii fox prcrtir.for partially dixiitcratcd tions by Jlartin and I