OFFICE OF NAVAL RESEARCHCD

GRANT N00014-88-K-049300o R & T Code 412mOO8N

Technical Report No. 5

MULTIPHOTON IONIZATION OFGROUP VI B HEXACARBONYL VAN DER WAALS CLUSTERS:

TRENDS IN INTRACLUSTER PHOTOCHEMISTRY

by

William R. Peifer and James F Garvey*

DTICS E1ECTE

S JUN 06 1989 D

Prepared for Publicationin

The Journal of American Chemical Society

Acheson HallDepartment of Chemistry

University at BuffaloThe State University of New York at Buffalo

Buffalo, NY14214

Reproduction in whole or in part is permitted for any purpose of the UnitedStates Government

This document has been approved for public release and sale; its distributionIs unlimited

b9 6 05 157

U3CRITY LA IICAT15N OF THIS PAr

REPORT DOCUMENTATION PAGEIs REPORT SECURITY CLASSIFICATION 1b. RESTRICTIVE MARKCINGS

-Unclassified_________________ _____

2a. SECURITY CLASSIFICATION AUTH4ORITY 3. DISTRIBUTION /AVAILABILITY OF REPORT

2b. ECLS$IICAIO~D~wNRA~NG CHEULEApproved for public release;ib.DECASSFICTIODONGRDIN SC4EDLEdistribution unl imited

4. PERFORMING ORGANIZATION REPORT NUMBER(S) S, MONITORING ORGANIZATION REPORT NUMBER(S)

Technical Report #5

6a NAME OF PERFORMING ORGANIZATION [6b OFFICE SYMBOL ?a NAME OF MONITORING ORGANIZATION

SUNY/Buffalo 1 (if ppiScble) Office of Naval Research

k. ADDRESS (City, Stott. and ZIP Code) 7b ADDRESS (City, Stote nd ZIP Code)

Dept. of Chemistry, Acheson Hall, Chemistry ProgramSUNY/Buffalo, Buffalo,'NY 14214 800 N. Quincy St., Arlington, VA

____________ 22717Be. NAME OF FUNDING /SPONSORING O b OFFICE SYMBOL 9. PROCUREMENT INSTRUMENT IDENTIFICATION NUMBER

ORGANIZATION I (if applicable)Office of Naval Research #N00014-88-K-0483

8c. ADDRESS (City, State, and ZIP Code) 10 SOURCE OF FUNDING NUMBERSPROGRAM PROJECT TASK WOKUNiT

Chemistry Program, 800 N. Quincy St., ELEMENT NO NO NO ACCESSION NOArlington, VA 222317

CE

1I TITLE (include Security Classification)ultiphoton Ionization of Group VIE Hexacarbonyl van der Waals Clusters:rends in INtracluter Photchemistry

12 PERSONAL AUTHOR(S)Willian R. Peifer & Jamnes F.-Garvey

13a TYPE OF REPORT 13b TIME COVERED 4, DATE OF REPORT (Year. Month, Day) IS PArc COUNTTechnical IFROM _____ To 26ISUPPLEMENTARY NOTATION

Journal of the American Chemical;oit

7.COSATi CODES 16 SUBI TERMS (fontinue rreverse if necessary end id*pVIfyby bi k number)Fl LD GROUP SUS.GROUP Ci,1e~cJ' ccr S~cN -pCl e-,4-

ABSTRACT (Continue on reverse Nf necessary and Identify by blok numnber)an der Waals clusters of M(C0)n (M - Cr, Mo. W.) generated in the free-jetxpansion of a pulsed beam of seeded helium are subjected to multiphotononization (HPI) and the product ions analyzed by quadrupol~ mass spectrometry.eprevioulsy observed efficient production of HoO. and MoO!Dfollowing MPIf Mo(CO)r? van der Waals clusters. We proposed that these ions arise throughovel reactions between a neutral photoproduced metal atom and the ligands,f an adjacent metal carbonyl "solvent" molecule within the cluster. In order t(test some of the predicitons of thismnodel, ye have examined the MPI of vander Waals clusters of the other Group VIB hexacarbonyls. We find the same novelbehavior (viz:, efficient production of metal oxide ions) in the Wr(CO)g systemso that previously observed for Mo(CO)tl. However, we find no evidence of suchbehavior In the CT(CO)rl system. These abservations suggest that the reactivity.6f first-row transition metal atoms may be fundamentally different from that

a~eod- or third-row In tala. _____1__

20 DISTRIBUION I AVAE1ILI1Y OF AIISTRACT 121. A"STRACT SECURITY CLASSIFICAIO1URCLASSIFIEDIUNWLED 0 SAME AS 07P? DTIC USERS Unclassified

22a NAME OF RESPOnSBLE PNDiviDuAL 122b. TELE PHONE PMcl& AWe COe)21c. OFFICE SYMBOLDr. David L. Nelson 1 (202) 696-4410]1

Submitted to J.A.C.S. , 5/13/89

Multiphoton Ionization of Group VIB Hexacarbonyl vander Waals Clusters: Trends in Intracluster

Photochemistry

William R. Peifer and James F. Garvey*D)<par L ;..on t of 1hem i_:.

St't&ir 11niv rsizy of Ae'w York at huff!,bu'GF.o. New York 142124

Abstract.

..... .. i~ c M .2.) t,(Lr. t,:,, N) g.erer.'t .-J in the I : _--

ex:,an -. ,,f a r.rei bem. of s.-edeJ .-iu are subjected to HLitiphtr.

i.,niFatic. ! ,.,1 ) anj the pro .uct iorin analyzed by qundrup._,e mts

s,,tro~rtry. We previou.]y rep;ortedi the o. ervation of

Ct M.,_ :. , fL1,,win- -'i ,-f Xo(O'k) van der Wag2s clusters, ard

prrs :i that these ions arise through nove' reactions between a neutra'1

phctopro-uced metaE atcn anj the liga:.ds o±7 an adjacent metal carbonyl

stlvent 'r,:;Wule within th? cluster. In order to test somc of thie

,'edicr.i,,_ - this model, w- have now exa:.,ineJ thr rI of van d--r Wacalj

clusters of th, other Group VIE, hexacarbunyls. We find the same nCv..w

behavior (viz., effic.ient prcduction of metal oxide ions) in the V:.C.)e, Fye".e-.

a7 that previously observed for Nc()e. Howev_-r, we find no evidence of su:1h

behavior in the Cr(OO)e system. BaFed on these observatirns, we suggest. tht

the rea,.tivity of first-row transition metal atoms may be fundamentally E

different from that of second- or third-row metals. These differences ale

discursed in terms of the occupancy and relative size of the metal d orbitals.

Avalability Codes

CPYDist A.Idd~114 SPecC 1-

6

Introduction

Transition metal carbornyl complexes have been the-- subject of intense _r-

search by both theorists1 and exTperimentaliStS.2 .3 The variety exhibiteJ in.

the type:- of b:)rding between metalS and carbronyl lig7 nds c-_halleriges Cour Che.: --_

cal in tU4 tion. 4 Many c,:,- dinat Jvely unsat urateJ meta' carL:!ny.5

thought toD Play impcrtant rroes in cataly-i3y- o:.........

spe,:- Le Z can be ef f i -ien tly ge-nerated in tChe g as p hase:- v ia rulsL iae.r::4

ly :i of Lh- 0c-rrr__-Pcn1Hin s3turated speci- s; in fa%-i * thr- r:'.~c

phtcf are~o irreact iois, provide a convenient mEan for tes in -T 1 -f: I

ing~ :,tat 4St i:a I re-i -tion rate theories. b

'I'iie-resc~lveL Lifrared, spectroa&.,_py (TI)has: been uses' t(- st-Lv -

dvriayri.-s of excir-rlaser phtlysis uf transit-icn no;ta carbenviS. .e

taR I c f ene rgy parti'ion ing a lor; d the r-a-Atioi icoordinate , and the reonb-ina-

t:*on kinetics :of pririsry photocproDucts with various ligandr.7-1 The temporal

de _Pen~jene Of certain, features in the infrared spectra suggest that cocrdima-

t~vyu-siz--urated metal carbonyl species undergci clusterin.g- reaActins with

te> atLurated r'rec'--ursojrs at approximately gas kinetic- rates.SlO or xam

Pic, the chr_-terir.9 reaction

CrrCO0'e + Cr(-CO))4 -- Cr2(C)lo (1

proceeds ir, the gas phase with a rate conistant of 1.8 x 107 Torr- , se:-. at.

3C-i K.9 GrouIp th,:aretiual analysislu of Ul spectroscopJiC dat-, (i.e. , the u

ber an:.' rel-ative intensitis.:s oif observed absorption bands) is especially ase

ful in assigning struCtures to the mononuclear carbonyl fragments. However,

the products of clustering reactions such as the one above are less amenable

to such an analysis because of the lower symmetry of these binuclear species-

and overlap with the spectral features of other species. It is necessary tc;

study these metal carbonyl clustering reactions by some complementary tech-

3

nique if one hopes to derive detailed information on structure and bonding.

One technique which holds promise as a structural probe of these clusterin

reactions is multiphoton ionization mass spectroscopy (MPI/MS). The multiph--

ton disso,:iatior and ionization dynamics of mononuclear 2 0 - 2 2 and coval.-::'-

bound miitinuclear 2 -21 transition metal carbonyis is well understoic: i:-t'

multiphoton dissoci-ation (MPL" of the metal carLcnyl results ini c'ort

and st-ipping. leavin- behind a nak:ed metal ato. which is subse;2e;,tiy _!,

i=-! Ccnse]uwntly, tho MYI mu.ss spe:trum is dominat- a--- I :,-.i- :i"e y

thm [,,etal ion swial. On the c, th,:r hwand, the mltiphctcn phctop.y :s of \,&;

der Wa-,ls cmilexes o" transition _aeta] carboryls is not so thorouihly ch:'-

a' z.eried. Su.?h van der Wa ls coml.-lexes are inherently interestin', Sinct

nultiphoton processes within these complexeF may lead to the production of

ordiratively unsaturated transient species clustered within a she! of satu-

rated "solvent' molecules. These photoproduced complexes could serve a- s- el

systems in the study of phenomenri suc:h as heterogeneous catalysis or

chemiscrptionk on metal surfaces. Clustering reactions (such as the-c irfrre

frcn, TRIS experiments) within these complexes should be easily stimulated. arid

would perhaps give rise to anomalous frament ion yields in the mazssp,_ . ..

Indeec, n:,vt-l intracluster chemistry has beer. observej to accor-any the t

photon dissociation cf mixed van der Waals complexes composed of Fe(CO)5 an0

small oxygen-containing molecules. 2 7

in order to explore the interplay between unimolecular photodissociation

processes and bimolecular reactivity in metal carbonyls, we have examined the

MPI/MS of van der Waals complexes of Group VI hexa.carbonyls. We recently re-

ported the observation of novel behavior in the multiphoton dissociation and

ionization of Mo(CO)s van der Waals complexes.28 Efficient production of MO-

and M002- ions following multiphoton excitation of the metal carbonyl van der

4

Waa Lr comclexes led us to propose the intermediacy of a structure (or struc-

tures) containing doubly-bridging carbonyls acting as four-electron ligndE,

as illustrated in Figure 1, where MI is a nascent metal atom which interaotF

strongly via d-r* back-bonding with the oxygen ends of adjacent carbonyl lig-

arids within the neutral van der Waals complex. One of the preictionr o t 1 i.

model is that for smaller metal ators , the metal d orbitals will be more cn -

traz:ted and the bridging carbonyls will thus be forced into closei Lr-x:ity.

Crowding of the bonding ny orbitals of the bridging carbonyl liga:-ds (witi..-.

the -G-Vi-O2 plane) will tenl t;o be destabilizing, and if this rer.ulsive ir-

tera -tion is not overcome by the strength of the Mw-O bonds, this tye cf

doubly-biidged structure will nc.t form. Consequently, yields for MU nd MOI!

should be diminishingly small, if not totally absent, when M is a metal whose

d shell is of smll diameter. in order to test this structural model, we have

exairineJ the MFI/M.S of van der Waals complexes of metal carbonyl-s other than

that of mcl.bdenum. We describe herein our results for Cr(CO)e and W(CO);

complexes, and discuss implications for intracluster reactivity and bondir..

Experimental Section

The metal carbonyls, Cr(CO)e, Mo(CO)s, and W(CO)e, were obt.tinW frcc.

Aldrich in purities of 99%, 98%, and 99%, respectively. Prior to use, ear-h

compound was further purified by several freeze-pump-thaw cycles at 77 K. Cur

cluster beam photoionization mass spectrometer is shown in Figure 2 and hac

been described elsewhere in detail.28 Briefly, helium seeded with a metal

carbonyl compound at its room temperature vapor pressure (typically a few hun-

dred mTorr) is admitted into the low-volume stagnation region of a Newport BV-

100 pulsed molecular beam valve fitted with an end plate having a 0.5 mm dia-

meter, 30' conical aperture. The stagnation pressure of the seeded helium is

1.3 atm. Metal carbonyl van der Waals complexes are formed ini the free-j.et

expansion of the pulsed beam of seedLd helium. The background pressure inside

the vacuum chamber with the pulsed bela valve off can be maintained at lesc

than 11- 7 Torr, aiid operation of the valve at 1 Hz leads to max.imum chuiibr

pressures of about. ; x 10-6 Torr. The cluster beam pulse is direcze! :ia_!.

into the ion source of a Dycor M20CO, quadrupole mass spectrometer, wnere -

interscted by the focused output from a Lambda Physik DIG 150 excimn-r ias. ,

operatei on the Kr'* transition at a pulse errgy of ca. 1,, u;J. Sy-t.hrr_-

ticn of the laser and the m.:,lecular beam valve is accomplished through th-n-Se

.c- a. ete:'nal timii circuit with ajn adjustable delay. The timira ci t

set to fire th- 20 nsec laser pulse on the leading edige of the ca. 15r 4sec:

duration molecular beam pulse. Ions produced in the ion soar,.e are se] ct.

by the quadrupole mass filter rd detected by a low gain (ca. IO0.X) elect:.;

muliiier. Output from the multiplier is converted to a voltage, ani;3ifi-

by a fast 1OX az.Flifier, and averaged with background subtra:.tion by a bD--:cur

avera.er (EG&G Frinceton Applied Research, Model 4420). Typically, the mass.

filter is scanne-d af a rate of 0.04 amu/sec over a l00-amu range so that spe:-

trt r.iy be collected and averaged for 2500 laser shots.

Electron impact (El) mass spectra of the van der Waals corrqlexes

also be collected by leaving the excimer laser off and energizing a thcriated

iridiu. filament within the ion source. Typically, we operated this source at

an electron energy of 70 eV and an emission current of I mA. In additicn.

mass spectra of the unclustered metal carbonyls (i.e., "monomers") may be co.-

lected by admitting the neat sample vapor directly through an effusive inlet.

In order to properly interpret data from cluster beam ionization exeriments.

one must verify that the observed ion signals result from ionization of neu-

tral species within the molecular beam, rather than ionization of residual

molecules within the vacuum chamber. Since we are able to adjust the delay

time between the triggering of the pulsed bear. valve and the firing of th-

laser, we can determine the temporal behavior of the ion signal and there e:

ensure that the ions we are creating actu--.Iy arise from neutrals in thi-

molecular beam.

Res-lts ar i Discussion

Relevant portiins of the MF] mass spec[.tra of the Van der Waas o

of Mc(CDO)s. W(cO)s, and Cr(O9)s are shown in Figsr-.- 3. 4, L r,

ively. The M 'I mass spec-tra cf the correspondin7g. unciust red ;:e .r.....

are shnow l alongsid_ for conparison. We have previously di usseJ our resukt

fo the Hc(.C.)s system. 2 0 We observed the efficient poduc.ion cf K:4)- a..d

*? ions following multiphoton ionization of van der Waals complexes c.f thi

hexacrbony. We noted that such ions are not observed in the MFI mass spec-

trum of the unclustered, monomeric hexacarbonyl 20 nor are they observed ir

e ,ectrin impact (El) mass spectra of either monomers cor clusters. From the

temporal behavior of these oxo- and dioxomolybdenum ions, we reasont-i that

these ions were originating from novel intracuster photoche.nical reactions,

re;..tions which most likely take place between two neutral participants with:

the van der Waals complex. Based upon orbital symmetry arguments 3rJ aralo-

gies fron surface s.2ience 2 G and synthetic organometallic chemistry, 3 0 . 3 i we

propose-] the intermediacy of a structure su:-h as 1 (vide suJipror, tc. acccunt i,.::

the observed photochemistry. Similar types of Etructures have beer. propose.!

to account for observed trends in ion-molecule reactivity among coordinatively

unsaturated transition metal carbonyls.32

As discussed above, one of the implications of our structural model is

that bridging by two carbonyl ligands will be stable only if the dx%- orbital

on M* is sufficiently large to preverit crowcing-_ of the occjpied-- ny ortit-c1s c-i

the bridging carbonyls. We would expect, based on our results with Mc, thr

an ator. suoh as W might tie able tu intera-t with the oxygen ends of tht tw-

bridgiug carbonyls via ba-:-k-dornation from its vair.ence si.-, orbital. We :---

fore expect v.-- den W- ic cornplex-s of W%(C D F to exhibit the Fair.;- type&

nov-l phcoto-:-Ihermal behavior 1foK lowing nil- Itip hc'Lol excitati-;w-: a..: that -.. r

pr-v ous y i, the- HcCC,6 system. In fa-t, we d , obs-rve effic.-icnt p~

aF shc'wn irn Fig'ire 4. Wce ermphasize that- Wo- iF on!;-' observed i:1 -

M~mas.F s~eotruino WCUi)s van der Waals Co'M-z1eXe-S. It is: not oLserveJ

t K Mb maess spec-,:trum of the uncolustered hexacart.bc.,nyl norY is it cb.er.ved- in

any o'f the- El m-a-s sp ctr;,j. We belif-vtz that prod-ct -Iior. of is~ C'. ourriig

thrcu .7h the samte typ of nvcel intracluster phot-,ohe-.mical m--ohanism as th-t.

pr-evilusly propcosed tor thi e H(W)f3 syster.: that is, reaction. betwceen a photo--

prc-duo,-? W a+.om and the oxygen ends of carbonyl liga--iIs or, ari adja.cnt W,-:)r

solventMol.e-ule- Within the vai. der aiF- cotripiex. Our masss'etomt-r7

Conly aet pass ions C-1 -~ 'I'/Z f arnd we are therec-r- u-nable t'17- e

~2- p r c-it i -. (m/ z 2 14 , 21., 2 1e, ar id lK

W.: no-w conrsidvr terermalnirig merter ot theho lous c

Fisgu,"- b sh. rw: a pz'rtion; of the' MFI mass sce:truar. of CrCD)s van der al

complexes. One immEiately notic-es that the photophysical behavior of the-

C.r(,CO)E; system differs from that of the other Group VIB hexacarbonyls. N:7

metal oxide ions are observed in the mass spectrum; instead, van der Waal.,

com~plexes appear to behave just like the unclusterel monomer, yielding a a

spectrum dominated by the metal ion signal. Operation of the pulsed molecular

beam valve at ca. three-fold higher intensity (maximum chamber pressure of

about 10-5 Torr) results in th&' pro~duction of a series of ions, CrOHx+

(x=0,1,2), which presumably form upon MPD/KPI of heterocilusters of Cr(QJ)e and

H20 (prtsent as art iii~.rlty in the heliur, gas coy,'nd#er). Tht~-% ijour are

se-:ri in th- highi-pressure El sp'eotrum. Hc-_wever, MPI 7spectra recor-de2 undwerI

condi:;tions- cf Lower b~~e. nlolrLean,~ inteir_-ity ch.-w no ion sig-.L2.s c~r

spond ing *I-- CrO4 or Cr2 r .SL SJ*E- t Cha we 4:esr l otce

Cr(AJE;va;, 1,_r W~ . -C m c :neXte.; in, our S-md- exi ' -:-rwh-:. L:

is 4 erat -c 2 at iowt-i int --ri i4-1y. H.-w-vei, we C.~ c~~ic !:J-r- t4

r,7--:: t. iust-r fo rir.-zci u in~ .; ~ ~ i. s ri id J- arcw

2 ,C 4r~:r fD ir fr.,_ t' h o jur uI! r g~an] ci lr n Iw'Ii

rew- C)bLs-rve~ rjrys*t:! 1 in,- dr~rit of Cr CC J iziJ-te t:

2.F ~ ofl the r~~~ivalvr- %frf~ s;-ver< i ix-wry ci or.-

havt e Enn- Ir.t e I i th 121r.- ur PIa t Z- :se r L -_/r >4- 7,z .i . 0u L

thb te fK 11: g~c~ t-W eniloy allowdZ f,-Ir facile de~~'t ar-cf

ir ic~orci C4 i-. fr.-!n sr E!td rar~e g,::.- txpansioi;.:. T!.ir. w-

Pry-Va Du- : sh:V tllat 011IS!erilng 7111-t :eO takirlr ece'21 i:- 4-~ J xasc;

(7, 'C i-s~~ ~hLj::%. ba.-i Lhtr tem.porai betiaviu tor the r"-~tolo:e

ph.u~~ fii~-th-r vapicr pre! urer (_T Mo)( I'J)a and CrCDuaec~v~Ic

an- ~~ the c-oidition, c stagriatio.-i pre~'zure andi pulsed bea.T. intensity

du~l"for the CrCQ'C')E expwanslon art, id-entieal to those use-d prtviuury !--

th ~eetaroWe Would 11,Kewar 1e eXP-ect clusterirrg t:'Y take_- pia::_

seeded1 Cr((AJ)s expansion. Finally, since we are 4,pareratly at'le tzo rerezatcf-_

hEterocluzteis of H20) and Cr(QJ)s, even though the number den,_ity cf H1

moilecules in the seeded gas is estimated to be an order of" i.arnitudc, les t

that of the metal. carbonyl, we should have nc difficulty generating o-_ustCzr_

of Cr(OO)8.

Why is it that CrQIXJ)s van der Waals complexes should display photophy~-

ical behavior unique among Group VIB hexacarbonyls? One possibility is that

intracluster bimolecular reactions simply are not taking place withi:, t

l 'eCl) v,-r; der Waais comple-xes, alth::lli-. this is Inci-ris.st. -r-L wit:, t :, --

suits of kine-ti:- FtudieE whi:-h inr-r tn.iat gas-ph --E clusteringretc>. -

twtc Cn ' a:..1' its c rdnivyunisat ur;ated phrc.t,pro-_ducts take pla 54

I': .~..fl7w. ~v~ ~~:t~ y':~~ Iiti.~meritcnsdrt

Cr, CO )k; 's- eIt, ri:.i 1 r-Es o ur .1-:1 V n-,'. r;. -.,: -.1 r 4-:Jf*- :

ir~r::~t~Cr(CJ )u svy.--7 art-i~& i:u7. . -~

15 i~w&'. Ln~r~inercton. edis,2urs ne&- '-w- Itt ritve-: L,

a. ad'*acent x~rc~. rMc>w-thiri the van e zsc-n itx ~

~rucur~a~o~2:~tc "hex oj:.e rr ~Jearlier- fo-r tr~ ti:C

c. _ .Ao ax pMc- ATu a:-. we W~ i s th. constra.rit thait tLm

two er,~~ bri-!;ing c-:.bcr.yls: wh se interatnrjc: ax - wE sho'll tak- to

r~- ~~:: r~r~~-~ are sep-aratedJ 1-- tth.- vai, der Waals tike~c

:r~i~±,a~'xii~it2y .4 . ~ For a pix~struct,_re w Ith C-OD-M* L'c,!d s

~iz f12Veach of tht- two M-Aj bols wod:d haverc a len~th of aLout) 2 A.7f

the separation of the bridging carhonyl.r is maintained arid the 0-M" -0 Lornd-

gle is de-creased from 120' tc. 1LIC" (a, value typical of sonme metal acetylacet-!c

nates3 -4 ), the M-U bond lenigth increases to about 2.2 A. For which (if any)

of the Group VIB metals is this range of M*O bond lengths reasonable- Esti

mated 11-O bond lengths based on atoroic nonpolar covalent radii 3 l are 1.91

0

(M-*Cr), 2.03 A (M*=Mo), and 2.03i A (N*:-W). Eiwerinmental bond lengths fcr 3

variety of dioxygen-metal complexes have been tabulated:36 typical metal-

1 C:

uxyg!en bond lengths for complexes of Cr range from 1.81 tco 1.92 A; ald for

those of M&, from 1.91 to 2.24 A.3 7 Based ori our estimated metal-oxygen bcnd

lengths ald available experime:,tal data-, we believe that ground-state atomt-- cf

Mo or mig'ht rer:sonably be expected to undergn clustering reactions with the

correspoJing metal hexacarbonyls to form structures such as 1, while g:our-

state Cr atoms would be more lik,.!y to for~m some alternative s-_ructute a a

result of clustering reactions with Cr(Cj)E.

O.ne such a]lternative might be a structure in which the Cr atn i:.er.:

via unsyn-ametrical bridging with one (or more) of the CO ligwlds cf the nei-Lj-

bcrir.g htxa.:-arbcnyl, as sugi:ested in Fipire 6. In this bundin scb'-"- t- CZ

iignd is terminally bound t, one merai center, but side bu:ni t,.

This typoe of bondin, has bee-n invoked to explain the rela.vely cOw ga -ha;c:

r itivity of Mr,.CU)s toward recombination with C W. e e cann:t c~ un ,

Fosaihaity that th- two Cr atoms might be bride-_d exulusiv-ly thr.u n t.e

-, 'of a brigin- carbonyl, in a monuhapt.-,- fashion. tsuch

w._-'- involve a fairly substantib.l g.,eometric rearrangement of the

Cr(CO )I3 reactant..

it is interesting to speculate on how the bimolecular reactivity of Cr

would change in the present case if it were to possess occupied 4d ortUs

which could participate in bonding interactions with the ligands of an ad:a-

cent Cr(CO)s molecule. This would require that the nascent Cr photoprodu,:t bc

produued in an excited state. The lowest excitcil state with non-zero occupa-

tion of the 4d orbitals is the e 7D state; its J=1 component is 422b.42 cn -

(ca. 5.24 eV) above the Cr a 73 ground state.3S It is known from emissic;n

studies that KPD of Cr(CO)e at 248 nm produces atomic Cr in a statistical dis-

tribution of states;40 thus, one would expect the ground state of Cr to be the

predominant species produced in our experiment. While the excited e 7D state

'a

is not one-photon accessible from the a 7S ground state, it certainly is ae-

cessible via a two-photon transition; one would therefore expect product ion( of

a significant population of e 71) Cr atoms upon MPD of Cr(CO)s using focused

laser pulses at about 473.3 rnm. These excited Cr atoms, by virtue of their 4.-1

orbital electron density, might display the same kind of bonding intera:tion:

with metal carboryls as we have inferred above for ground-state Mo and V

atoms.

Conclusiois

W- have examined the 248 nm multiphoton dissuciation and icniratJon -

namics of van der Waals clusters of the Group VIE, hexacarbonyls. Mass sp~e-

tral evidence indicates that W atoms undergo the same novul photoinduced bi-

molecular intracluster reactions previously inferred for Mo, while Cr atom= do

not. We suggest that this trend in reactivity is due primarily to changeS in

the spatial extent of the valence d orbitals. Furthermore, such an interpre-

tation may imply enhanced reactivity of higher excited states of third-row

transition metals produced within the cluster environment. We are currently

extending our studies to other third-row transition metal carbonyls ii. order

to refine our understanding of the photochemistry of van der Waals clusters.

Acknowledgment. We gratefully acknowledge the Office of Naval F:esearcl.

and the donors of the Petroleum Research Fund, administtered by the Americ..ul

Chemical Society. for partial support of this research. We alsc' wish to thi

Newport Ccrporation for an equipment grant.

RJFERhVENS

(1) Veillard, A., Ed. NAT[)ASI 1986, Series C.

(2) Moskovits. M., Ed. Metal CluI ers; ,-.rlui Wiley & Sons: New Y: r-.,

1986.

., a-vid 1:. , Ed. Car, i ihs& InorgaTic Chemistry"; 'lenum,. i r'

' a4) d, K. In Transition Metal Clusters, Jo-hiscr., B.F.G., El-.:,.

Wiley & %ons: New York, 198LU; Chapter 1'.

(5 Whetten, R.G.; Fu, K.J.; Grant, E.F. J. A. Chem. So. 1982, >.

(6) Tums, W.; Gitlan, B.; Rosan, A.M.; Yardley, .T. J. Am,-. Che,. Soc.

1982, 104. 515.

(7) Weitz, E. J. Phvs. Chem. 1987. 91, 345, and references therein.

(8) Hollad, J.'.; Rosenfeld, R.N. J. Chem. Phys. 1988, 69, 7217.

(J) Fik:tcher. T.R.; Roserifeld, R.N. .. Am. Chem. Soc. 1985, 101,

(10) Breekenridge, W.H.; Stewart, G.M. J. Ai. Chem. oc. 1986, 17, 3:'4.

(1P'1 Seder, T.A. ; Church, S.'. ; Weitz, E. J. Am. Chem. Soc. 1986, l,.

47

(1">) Ganske, T.A.; Rosenfeld, P.N. 7. Fh.vs. Chem. 1989, 93, 1959.

(13) Ishikawa, Y.; Hackett, P.A.; Rayner, D.M. J. F1hy2js. Chem. 1988, 9-2,

38b3.

(14) Seder, T.A.; Ouderkirk, A.J. ; Weitz, E. J. Chem. Phys. 1986, 65.

1977.

(15) Bogdan, P.L.; Weitz, E.; .. Am. Chem. Soc. 1989, 111, 3163.

(16) Ishikawa, Y.; Hackett, P.A.; Rayner, D.M. .. Am. Chem. Soc. 1987

109. 6644.

(17) Rayner, D.M. ; Nazran, A.S.; Drouii, M.; Hack !tt, Pi.A. .J. [irv.

Chem. 1986, 90, 2882.

(18) It appears to be generally true of neutral metal carbonyls th:t

clustering and li-and recoiination reac:tions occur at or near gas kineti.

rates so lon.f as sp i n ! conservej. See De3rden, 1,.V. ; Hay,,shibharo, K.;

Beaucham , J.L.; Kirhi'er , 1i.J, ; van Koppen, P.A.A'..; ,,wer - Y. J. Ax:. &7tri.

Soc. 1988. 111. 2L•

(1it) Cot ton, F.A. Chemical Applioations of Grou. Th,.;., 2n- ed . ; ..-

Wiley & Sons : New Yirk. 1,<}'/.

'2C:) Duncar , M.A.; Dietz, T.G. ; Smaley, RE. Chet. fhy-. 1979, -14.

(2) rrrity, ).F.; tlberg, L.J.; Va3da, V. Ch.'m. i hys. Lett. 1980.

.4. 1.

(22) Enelking, P.C. Chem. Phys. Lett. 1980, ;4, 2C17.

(22) Lichtin, D.A.. bernstein. F.B.; Vaida, V. J. At. Chem. Soo. 1982,

104, 183;'.

(24) LeD,'2, D.G.; Vaia, V. J. Am. Chem. Soc. 1984, l,, .372. •

(25,) Leutwyler, S.; Eveni, U. Chem. F'hysz7. Lett. 1981, 64, 18.

(21,, Holling7sworth, W.E.; Vaida, V. J. Phys. Chem. 1986, R1, 125.

(27) Whe! -ler, R.G.; Duncan, M.A. J. Phy.s. C'hen. 1986, 90', 2,k.

(28. E-ifer, W.R.; Garvey, J.F. J. Phys. Chem. 1989., 14' (in pre-s).

(2[) Zaera, F.; Kollin, E.; Gland, J.L. Chem. Phys. Lett. 1985, I21,

464.

(2) Ulme-r, S.W.; Skarstad, P.X.; Burlitch, J.M.; Hughes, R.E. J. Am.

Chem. Soc. 1973, 95, 4469.

(31) Blackmore, T.; Burlitch, J.M. J. Chem. Soc., Chem. Coawun. 1973.

405.

(32) Fredeen, D.A.; Russell, D.H. J. Am. Chem. Soo. 1985. 107, 3762.

(3)Pauling, L. The Naiture of the Chemical b~ond; Cornell University

E'resF: Ithaca, 19601.

(34) Fackler, JP. Frovf'. .lnorg. Chem. 1966. 7, KC1.

('3b5 Sanderson. FUP. Inoxzga'iic Chemistry; Rc-inhold: New York, 19 *7.

(37) The Com!-iexe-- of milyLxenurn are well-studie- , dl., ini n,: s-mail parlt

t r~~'nrun,' -i~hnia igniir a~e Unor -nly, there ex-'rtsi 3

y: on~ ir tung! ter- CCO!q~jLeX- F

evr, T. A. Chu rLh, S. F.. Weltz, E. J. A j. Chm Io.18,,5

134zM:.r C.E. Atomic Fnergy Levels (NSRDz/NBS, vcj. 3L); U.S. Gov't.

Fitin~f t'fi-.-t Washiris-,7r2, !: .C. 1971.

(40% Tynd,--ll, G.W. ;Jackson, R.L. d. Ain. Chem. SOu-. 198T7 109, 59

Figure I. Proposed structure for the prnduct of a

clusteriNg reaction between a photoproduced metal atom, ,

and a neighboring metal carbonyl "solvent" molecule within a

van der Waals cluster (see reference 2,). The value of x

may range from 0 to 4. Back-donation from the occupied

metal dxy orbitals to the carbonyl Ry* anti-bonding orbitals

is illustrated.

(CO) xM

Cy

x

Fijurt. 2. Cluster beam photoionization/m&:ss spectrometry

experiment, shown schematically.

CL

ww-J LI)

Yo

Irxw cr 0ix cr0

w <! w cr4c 0

4c E w wU)> 0 z

L14 w0

F1g Ue - . (A) MFI mass spectrum of Mo(CO)e; van der ',laa.!

clusters, 80-100 amu. Signal due to ioris of m/z :IL0 am

is amplified by a factor cf 10. (P) MP! mass spectrum of

the unclustered Mo(CO)s, 80-16 amu, shown for comparistr,.

As above, ion signal is amplified by a factor of 10 for ions

c >f m/ > I01. Note the absence of oxo- and dioxomolybdenum

ions in the mass spectrum of the unclustered Mo(CO)6.

x 10

M+ + a) clustersmMoo

Af a& - IN

-

MO~ b) monomer

0 1

80 120 160

m/ z

Figu:. 4. (A) MFi mass spectrum of W(CO)s van der W -i

uwt-rs. 17,"-- 2 1,_' anu. Signal due to ions of i/z L Y U

is arr.Iified b y a factor of- 2t5o,. WOc ions are clearly

visible, although ]imitations of our quadrupole mass filter

prevent us fro.m detectinr any W02- ions which might be

rrodu;-aed. See discussion, in text. (B) ME"I mass spectrum of

the unclusnred Mo(CO)6, shown for comparison. As above,

ion signal is amplified by a factor of 25 for ions of in/c >

.UJ. Note the absence of WO ions in the mass spectrum .f

the unclustered W(CO)s.

W+ a) clustersr I

4-Wa

:1 x 25

0

0

17 0 210

m/z

Fgure 5. (A) MI mass spectrum of Cr(CC)a van der Wa~2z

clusters, WU-150 amu. Signal due tc ions of m/z ! e" amu iL

amplified by a factor of 2.5. (b) MFl mass spectrum of the

unclustered Cr(CO)s, 10-150 amu, shown for compsrison. As

above, ion signal is amplified by a factor uf 2.5 for ions

of m/a ? 60. Note the exclusive domination of both spectra

by the metal ion signal, and the complete absence of metal

oxide ions in, the mass spectrum of the clusters.

Cr + a) clusters

ow2n5

,0

+j+

-~ x2.5

0F

50- 10015

%=m/e

Figure C. One possible structure for the product of a

tlustering rea-tion between a ground-state Cr atom and a

neighboring Cr(CC)s molecule following multiphotcn

excitation of a Cr(CO)s van der Waals cluster. Clustering

between the nascent. atom and a coordinatively unsaturated

chromium carbonyl would presumably result in a structure

with some degree of metal-metal bond character.

0

III0 ~C% C0

-e

0 11C , /