RD-i148 814 COM1PARA~TIVE STUDIES ON THE ELECTRONIC STRUCTURES OF i/i

AD-1i~l B4 2(02CH)4 AND &2(02CH..U) INDIANA UNIV AT BLOOMINGTONKLAS F IED DEPT OF CHEMISTRY M D BRAYDICH ET AL. 29 NOV 84

UNCLSSIIED NDUDC/T-84 -C N 888479C044 F/6l 712 L

mhhEmhhmhEmhhEEEEEEEEmmohE

- 4.u! ,

"-o

% o.

-:.. 1 I

44 V32

.444 *36 -

"- fl. I I &10

1 11111

MICROCOPY RESOLUTION TEST CHARTNATIONAL BUREAU OF STANDARDS 1963 A

%4

.4.

'

...

d

OFFICE OF NAVAL RESEARCH

4.. Contract No. NOOO14-79-C-0044

*Task No. NR 056-703

TECHNICAL REPORT NO. INDU/DC/TR-84/3-MC

- COMPARATIVE STUDIES ON THE ELECTRONIC STRUCTURES OF W2 (O2CH) 4

AND W2(O2CH) 4 (CH3 )2 BY THE RELATIVISTIC Xa-SW METHOD:

A d3-d3 METAL DIMER WITH A QUADRUPLE METAL-METAL BOND?

by

0

M.D. Braydich, B.E. Bursten, M.H. Chisholm and D.L. Clark.

prepared for publication

41

in

Journal of the American Chemical Society

4.S _Department of Chemistry';-" Indiana University D

Bloominpton, IN 47405 ECTE

November 29, 1984

*- Reproduction in whole or in part is permitted forany purpose of the United States Government.

.4C. This document has been approved for public release

and sale; its distribution is unlimited.

0

84 12 17 007

% . % % % f %~~: ~.~ ;~

SECURITY CLASSIFICATION OF THIS PAGE (W*en Da Entered)

RP TO M TI PGREAD INSTRUCTIONSREPORT DOCUMENTATION PAGE BEFORE COMPLETING FORM

I. REPORT NUMBER Z. GOVT ACCESSION NO. 3. RECIPIENT'S CATALOG NUMBER

% NDU/DC/TR-84/3-MC Ab ig I3 S lq

4. TITLE (and Subtitle) Comparative Studies on the S. TYPE OF REPORT & PERIOD COVEREOElectronic Structures of W (0 CH)_ and W 2(O C11) 4

CH ) bv the Relativistic Xa- W Method: A Technical Report 490d'3 G". PERFORMING ORG. REPORT NUMBER

d -d ' etal Dimer with a Quadruple Metal-Metal Bo d?.. ..__________________________I NDU/DC/TR-84- 3-MC7. AUTHOR(s) 8. CONTRACT OR GRANT NUMIER(*)

M.D. Bravdich, B.E. Bursten, M.H. Chisholm<" . N00014-79-C-0044

and D.L. Clark

9. PERFORMING ORGANIZATION NAME AND ADDRESS 10. PROGRAM ELEMENT. PROJECT. TASKAREA & WORK UNIT NUMBERS

Department of Chemistry

Indiana University

Bloomington, IN 47405II. CONTROLLING OFFICE NAME ANO ADORESS 12. REPORT DATE

Office of Naval Research November 29, 1984

Department of the Navy 13. NUMBER OF PAGES

Arlington, VA 22217 3214. MONITORING AGENCY NAME & ADDRESS(I dilferlnt from Controlling Office) IS. SECURITY CLASS. (of this repon)

ISa. OECL ASSI FICATION/000WNGRAOINGSCH4EoULE

" * 16. DISTRIBUTION STATEMENT (of this Report)

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

distribution is unlimited.

* Z 17. DISTRIBUTION STATEMENT (of the abetract entered l Block 20, it different from Report)

10. SUPPLEMENTARY NOTES

19. KEY WORDS (Continue on reverse side if neceooary and Identify by block number)

tungsten-tungsten triple bonds, quadruple bonds, carboxvlates, Xn-SW calculation

20. ABSTRACT (Continue on revese oide It necessary and IdentIfy by block number)

0 >The bonding in W (OCR) and in the recently characterized W1(O'CR)jR' mole- % cules are compared via XK- W calculations with quasi-relativistic corrections o !... model system (I) and W(0' CH) (CH' (II Several questions con-

cerning the electronic structure of I Eve been addressed; in particular, the ~IapparentLy strong W-W bond in the presence of strong W-C bonds was of interest.It has been found that II is best considered as a W '(OIH) fragment interactinwith two Ci!? radicals, a description consistent with the'photochemical decom-

%" orI position of W'V,,(oCEt), (cr,Ph)',. The resulting W-W bond still retains the IV . , "3 3 "

DD , AN73 1473 EDITION OF I NOV 65 IS OBSOLETE1W 6/N JAN-1- 0 7

SIN 0102-014-6601 SECURITY CLASSIFICATION OF THIS PAGE (Whom Date ENa~

I-, L.1.1J4jTY CLASSIFICATION OF THIS PAGE(Whun baa. Iavr

ZANH ~'0 -essential components of the quadruple bond in I. The W-W bonding remainsstrong in spite of strong axial ligation because of involvement of a higher-

lying s-s & bonding orbital, an orbital whose contribution is more importantin II than in I. It is the presence of this orbital, which is lower-lyingfor third-row metals than for first, which is believed to account for thestructural difference between Crl2(O27R>';Lrand the W;(O'CR) R' systems2 22 (2 R4R2 syts:

Ao'esslon For

XTIS QRDTIC TAB

. -"a emnnouced 0Justifloatlon

"'"" By _

Distribution/

V"Avail and/orDit Special

V'2.

%P 4rIo

• -o'

N'"

,% / $1[SCUNITY C L.ASSlIICATION OF ?141S P A6C(Wl~ m D~ Eee em )

0

COMPARATIVE STUDIES ON THE ELECTRONIC STRUCTURES OF W(?U) AND

W?(OCH)4(CH3 )? BY THE RELATIVISTIC Xn-SW METHOD: A d 3 -d 1 METAL DIMER

WITH A QUADRUPLE kIETAL-METAL BOND ?

• ' MD.Brydc l a .E Brtn l a * l

M.D. Braydich , B.E. Birsten M.H. Chisholmlb and D.L. ClarklC

Con'ribiition from the Department of Chemistrv, Ohio State Univers itv,

Col imbtis, Ohio '43210, and the Department of ('hemist rv and Molecu lar

Strtuctitre Center, Indiana University, Bloomington, Indiana 74 05

ABSTRACT

The bonding in W)(O 2 CR) 4 and in the recentlv characterized

W-' (OCR)4R',? molecuiles are compared via Xa-SW c-alctlations with q,,asi-

relativistic corrections on the model system W.,)( )CH), (I) and

W9 (O 9 CH)4(CH3 ) (II). Several questions concerning the electronIc

s tructure of 11 have been addressed; in particitlar, the apparent ly

strong W-W bond in the presence of strong W-C bond- was of interest. It

ha- been found that II is best considered as a W9 (O9 CH) 4 fragment

interacting with two CH3 radicals, a de-rription ronsistent with (he

ihotochemical decomposition of W2(O2CEt),(CHPh)?. The reslting W-W

bond st iI I retains the essent ial components of the qpadr,, Ie bond in 1.

The W-W bonding remain; strong in spi te of strong axial lIgation berause

, invoIvement of a higher-lying s-s o bonding orbital, an orbital whose

ront ribiit ion is more important in II than in I. It i; the ipresence of

h hi i orbital, which is lower-lying for third-row meta I; than for fir,;t,

* whic h is believed to account for the strictra I di fFerence bet ween

p Cr?(OCR)4L? and the W-?(OCR)4 R systems.

INTRODUCTION

The rapid growth in the ;ynthetic and stricttral -hemisLrv of

din'iclear transition metal complexes containing -trong metal-meta1l bonds

has provided a wealth of information about the natuire of metal-ligand

and metal-metal interactions. 2 The elegant story of the Cr-Cr qiiadr iple

bond as told by rotton and Walton 3 beautiful lv emphasizes this point.

A The general conclusion gained through the structutral tudties of these Cr

n011_npoinds is that it is not the electronic properties of the rhelating

ligands that determines the Cr-Cr bond lengths, buit rather the lresence

or absence of axial ligands. Indeed, the iinfai ling occurrence of axial

ligation in the dichromiitm tetracarboxylates and the enormous range of

Cr-Cr distances has posed exceptional chal lengo, to the theory of the

electronic structure of these compounds. Numerous electronic str,,ctural

• . calcilations have appeared in the literattire31 over the past derade

and both SCF-HF-CI and SCF-Xa-SW calctulations have arrived at

_ satisfactory descriptions of the Cr-Cr quadrulle bond. For axial

ligands, SCF-Xo-SW calculations i n present the description that weak M-M

- bonding in the dichromiim system resultI in a very low ling !1-:-M r,

r. orbital. Axial ligand donation into the * orbital, as well a,

destabi I izat ion of the ,1-M r orbital weakens and lengthens the 'r-Cr

bond. It thus appears to be a generaI result that axial ligat ion

weaken% and lengthens the Cr-Cr quadriiple bond. llowever, it is not safe

to assume that axial ligatton weakens and lengthen, metal-netal bond,; in

general. The dirhodtiim tetracarboxylates have strong metal-metal bonds

and axial ligation. We are in need of new gtructuiral data In order to

gain better insight into the problem.

0V 4

%

0

-J The recent addition of the dit'ingqten (11) tetracarboxvlatts.1 -1

" -is not only of historical Interest, b,t contributes structural data and

new tha I lengts with which to test and expand ouir theories about ax a Ia.

ligation. One of the most fascinating new -tructttral deve lopments in

this area is the recent report by Chisholm and coworkers; of a new class

Of d 3-d3 dimer where strong axial ligation is observed to a ditiingsten

tetracarboxylate center, and resutlts in extremely short W-W distances. I

These new coml,ouinds, bis(aIkyl) tetra(carboxvlato) dittingsten (IIl),

"aN .. R(O?R) , (R = CHPh or CHBut, and R' = Et) are wel I haracterized

in the solid state as having the centrosymmetric structtre depicted ini

Fixire I.

The striking structkiral featutre of these compoinds is that the W-W

distance of 2.19 A and the parameters of the central W(OCEt)4 core are

essentially identical to those seen in W9 (O9 CEt)4 (WW). The obviots

q,estion raised by this data is how (or why) does this molecile exist

with strong metal-metal bonding, and strong metal-ligand bondiny (W-C

19*9 A) in the axial position

In the initial report, 16 extended-Huckel calcilations predicted a

6 electronic configutration for these compounds which did not provide

a satisfactory explanation for the shortness of the W-W bond. It was

suggested that s, pz and dZ? mixing could lead to a TT !C configutration

more in accord with the shortness of the observed bond length.17 It is

apparent, however, that for any molecule containing atoms of suich high

atomic number as t,ngsten, valence corrections for relativistic eftects,

which can amount to an electron volt, are likely to be Important towards

ii-ining a satisfactorv model for the electronic structure of such

.-. 3

%We e t oi0t titte o ic

ba a V'*., -o.Pp/'fV *P

In reson- e to the sittiatton jiiit described we have carried mt,

and report here, the resit Its of comlarat i ve, re lat i vi sti r SCF-Xii-SW

-a I -it lat ions on W)(O2CH), and W?(OCt)l,(CHI),. The detai Is 4)t

ierforming a calrilation on a species of low symmetrv (C,)) wi I ! bepreqented, and the effect of relativistic corrections will be discussed

and demonstrated to play an important role in the description of the

v- erlectronic st r ucttre of these compotnds. It wi I I further be shown that

the resuilts are consistent with strong W-W bonding in the presence of

the ;tronglv 7-donating CH3 groutis in the axial positions. Final Iy, we

wil I comment on the relationship of oitr calcitlations to those l'reviotsly

reported for other M,'9 (02CH), and :12(0?CH),,L? systems and t rv to

S. formtlate some general conclusions aboitt axial ligation.

04

-'C'

C'1

.N

.

C.'.

-:!

".2

4"-4

COMPOTAT WEAL PROCEDUES

Initial Parameters

The W-W and W-C bond lengths itsed in the ca Icoilat ion duf the

electronic strutctiire of W?(O 9 CH) 4 (CH 3 )9 were taken from the crystal

~t ricttire of W?)(OCrE ) 4 (CH?)Bt)1. The W?)( O?(Hb4 fragment was

idealized to D 4h point symmetry while the enti re molecic was idealIized

t o C.)h pIIoint symmet ry with the two CH3 grotips oriented in an axi alI lv

staggered geometry. The bond lengthq and angles of the W?-(0 2C)04

fragment are within the ranges of stritct'iral parameters general lv folind

in dittingsten tetracarboxylato compotnd. 1 19 The bond length,; and

angles lised in the calcutlation are summarized in Table 1.

* An initial molecular charge density and potential were construlcted

from a superposition of Herman-Skillman 1 neuitral charge densities for

*W, 0, C and H. The a exchange parameters were taken from Srhwarz with

the tutngsten a va I lie extrapolated to O.h93 19. A valIence-e I ect ron

weighted average of atomic a values was uised for the inter- and outer-

sphere regions. Over]lapping atomic sphere radii were taken to be 89Z~ of

the atomic number radii in accordance with the nonem 1irical proceditre ot

Norman. ?0 The ouiter qphere was made tangential to the outaermost atomic

spheres. The sphere radii and a parameters utsed are sulmmarized in Table

The symmetry adapted linear combinations of atomic orbitals For allI

- alcitlations (D ' h and C~h) inc bitded %, pit d and f type sphericalI

0 harmonics on the tuingsten atoms, s and p) on C and 0 atoms, s on H atoms)

and spherical harmonics throiigh 1 9 on the outter sphere. Core energy

* - levels; wert. never frozen; in each iteration they were calcruliuted

5

% %%

S.%

Sexplicitly using only the surro,,nding atomi c-sphere potent ia for th

atom in question.-. 4

The iteration to self-consistency on W.)(OCH)4 was started

nonrelativistical iv ising a 5% mixing of the new potent ial into the old

to generate the -tarting potential for the next iteration. This mixing

*iii wag gradual ly increased to a maximum valute of 15% as the calcilat ion

neared convergence, which was asstmed when the maximium shift in the

jotential from one iteration to the next was less than 0.i)f) Ry. The

' virial ratio (-?T/V) at convergence was 1.00011.

The converged nonrelativistic potential of W?(O ?Cli) 4 was used as a

starting potential for the relativistic calctulation. The formalism of

Wood and Boring 21 was used to incorporate the relativistic effects.

The core levels of all atoms and the valence levels of tungsten

explicitly included these effects which were slowly mixed into the

* potential over 10 iterations. The virial ratio at ronvergence increased

to 1.0417 as a consequence of the relativistic formalism. This

-" onverged relativistic potential was used as a starting potential for

the W2(O2CH) 4 Fragment in W2(O 2 CH) 4(CH 3 )?. Likewise, an Xa calettlation

- was performed on "elongated" ethane using the same outer sphere radii,

spherical harmonics and outer sphere a value as was used in W?(O (Cl1)/.

This amounted to having two ClHI fragments in a staggered D"d geometry

with the same C-C distance as in the W9(O9 CH)4 (C.-3 ) 2 molecuile. This

calculation was converged nonrelativistical ly and the converged

potential was ,used as a starting potential for the ((113)2 fragment in

W?(O 2 CH)4 (CH 3 )? •

Executing the calculation in this manner is not only convenlen!,

bit very important in terms of establishing a one-to-one correspondence

)f the energy level s of W2 (O2 CH) 4 from higher to lower symmetry. This

.# .. ..

-"% ' .' "- %-* - ' '' : ; ;' - / . v . . ,, .,,-

minimizes the possibility of "missing" an energy level in tihL e ne rgv

se arch, a pervasi ve lrob Iem in Xa-SW cali I at i on' on lw-vmmvt rv

vs tems.

*1"?.

m'p

p

I'.

-.

oV or .0P

.4.

*1.;."

A A1_.-4g

RESULTS AND DISCUSSION

-. ,,.'. W2(O2CHI)2 (c) 4

The result,; of our nonrelati viqt ir and re lat i vist ir 'al ii lt ion'

ire compared in Fi.gure ? and Table I I for the ortlipi ed va lenre and

• -. o!)we-;t virtiial orbitals of W.)(O.,CH) ,. The bondin r haracteri,4t I.- o f

the molecu.lar orbitals of W ,(OCH) , are essential Iv the same a4 fur

.Io )(OCH), and have been discussed at length bv Norman, et al . We

wi I I focus our discussion on those orbitals irimari Iv responsible for

metal-metal bonding. Xc-SW calcullation,; of M.)(O(CH), sv-;tems in

-eneral yield two components of the metal-metal 7 bond, the al ind

,a, molecoilar o)rbltals. One can envision the formation of these

orbital s a% being derived from Interaction of the nearly pitre dj. nz

orbital of a W,) fragment with the lone pairs of the four formate

I igands. This retilts in the formation of the -aI and 5 a molecular- . gIg

orbitals of W)(O0,CH), which are W-O bonding and antibonding relpectively

-- (r.f. Figiires 3 and 4 and Figure 3 of ref. 22). It shot ild be noted that

to a first approximation, since both the 'al and 5a 1 moIrc lar

orbitaIs are occupi ed, this interaction wi l I have no effect on tie

_ metal-metal - bond. However, there is a second process occurrinp in

these alg interactlons which can contribitte to a net stabi lization of

tie W-W and W-L bonding, namely the involvement of virtuial W 6'4.- '

orbitals. In the relativistic calcullations the 5al, orbital has 69% W

rharacter of which 38% is W s. This mixing of virtial W s c-haracter

into the 5al, orbital has several important ramifications. First, it

mit iiates the W-0 antibondtng rharacter of the orbital; and second it

idds to the net W-W -T bonding, an effect which increases in importance

w'en re lativisti coi rections are applIi ed, a- will be disrussed below.

- %% %

A7O

',e believe that this additionalI s character, in1 ,1 conin rt i)n w i tl hbt.t t L-r

over l ap dite to the great er orb italI ext ensio)n of e dthn t t, r

t he ' d or 3d orbi talI s, Ileads to thle conr I its io(n (hat t lhe .4-1. hondi wi

h e W comI Iex shoii I d be stronger and more co va l ent t han e it her t he 1o or

C r sv st ems. The W-0 bond l engths of 2.()8) are shortecr than t het

oberved M1o-0 distances of ?.I I Aconsistent with this view. The v

molecuilar orbital is 51% W in character of which 937 is W d. We IFeel

hat it is th is ' a , component that makes the ma jor cont r ibit ion t() the

'.;-W bond (s;ee Figure 3, and Figutre ? of ref. ))). Qui aIi t at iv e Ilv we%.

ha ve the ex;iected meta 1 -met a I qiiadr'iq' I e bond of e I ect roni c con f i gli ra t ion

alIthouigh it mutst be emphasized that di sciss i n thle confi girat ion1

in this manner is somewhat of an oversimpl ificat ion.

Relativistic Corrections on W 2( 2 CH)4

The re lati vist ic correct ion,; cause larve energy shi fts in the

*tungsten core levels and similJar shifts; in valence orbitals containing

-;ignificant tuingsten character, whi le, as expected, the I-rimari Iv 0-,

C-, and H-based molecuilar orbitals are scarcely effected. The observed

differential shiftq are simi lar in magnituide to those )b,;krvt~d in

3) 5previolit molIecul1ar ralIcu11a tifons; iing t h i reIa t iv is t ic f ormalIsm.

The changes i ndiiced i n t he bond ing4 pi ct itre uipon thle i nc I its i on of

relIat iv ist ic( effe- ts- are consi stent wi th t le ex~ect ed i nf )luence oif mass-

ve loci ty correct ions4 on the W atomic orbitalIs. The inner s-orb tt 1-4,

hiiving the hii ght.st c lassicral ve locit ies, are the most profouind ly

effected orbi titIs. The relIat iv i sticr mas s i ncrea se result i n a

-ontraction of allI of the s-orbitals w:tli a concomitant decrease in

lit i r ,rbi t at I eihkrgik.es. Th is t-f t e.crt i s m im ic-ked byv t he v -o rhbi ta I s

ogh bot h t he cont ract ion and st abi I i '/at i on of t hese a re levs

9

%0% %

.% j- j '* ~ ~ . ** .~ . ,.

ronouinced t han for the the q-orbi ta I s. The met a I d- and f e I ec t rons,

which have a mutch smalIler probabiIitY of attaining a rclassicalI ve locit y

lose to c, are primari Ivy inf I uenced by the cont ract ion of the s- and

or bitalIs. The contraction of these tatter orbital s resul ts in an

expansion and rise in energy of the d and f orbita Is, i.e. the reverse

of the effect seen for the s-orbitals.

In Figuire ?, it is seen that the relativis;tic7 shifts in the orbital

energy are in the expected direction. The valence levels containing

significant W 5d character rise in energy with the exception of the 5alg

orbital ; without relati vistic corrections this orbital ha,; 1 7% W and

38% W d character, and the substantial s ch1)ar acrt er in t h is or b italI

rauses it to drop, in energy upon the inc lusion of re lativist ic effects.

A similIar resul It has been observed in a recent Di rac scat tered-wave

(DSW) treatment of W2 C18 4 in which the relativistic effects are treated

more properl i using four-component spinors. 17We find it encoutraging

* that the quasi-relativistic corrections employed here mimic the effects

- -(in the orbital energies which are found iinder a more romp let e treatment.

Upon stabilization, the 5a,, orbital acquiires significantly more W

character 06h%) and somewhat 1e-s W d character (36Z) The ; orbital

contribution to this orbital represents a mixing of the f-is-bs a bonding

i rbital, normal ly unoccuipied, with the 5d-5d n bonding orbital. Th is

s-4 7 bonding orbital is important in explaining the extremely short

.'))rd distances, fouind for naked metalI diatomics such as 1o9 ',? and we

* believe it serves a major role in the shortness of the W-W bond in the

t41)(0 9 rR) 4 )R 2 systems.

% Z-V. r.

* % 0. S*'~* %dB.~- '.? .AY~ * .

.. W 2 (0 2CH)4 (CH 3 )2

The correlation of the molecuilar orbital; of W(OCH),d(CH 3 .-1 with

those of its component fragments Wg,(O 9 CH)!. and (CH3), . are shown in

Figitre 5. The energy levels of W,(OCH)4 and (CH 3 ), do not represent

the ;elf consistent levels of these two neutral fragment;. Rather, we

have obtained orbital energies appropriate for direct comparison to

those in the W2 (O.CH) 4 (C 3 )2 complex in the fol lowing manner: Fol lowing

the convergence of the relativistic potential of W?(OCH)(CH 3 )), the

W"?(OCH) and (CH3 )9 portions of the potential were searched geparately

for energy level, inder C.h symmetry. The resilting energy levels,

s hown in Figutre 5, represent the levels of the W-,(OCH)j, and (('13)?

fragments in the same potential as the entire molecutle and is th,, a" method for construicting a molecular orbital correlation diagram ,,sing

the Xa method.

The first feature evident in Figire 5 is that the M-M n, 6, 6* and

levels, as well as the M4-L and formate level- in W.,(O 9Ct) 4 (CH 3 ) -, are

.essential Iv inpertturbed by interaction of W9 (O,(CH), with axial ligands.

Secondly, the six C-H bonding levels in (CH 3 )9 (2a + a,, + b + ?b,,

were found to be entirely noninteracting with the rest of the molecule.

Therefore, we shal l concentrate our discussion on the W-C1- nl

interaction and its effect on M-M a bonding. The W-CH 3 interact ion

manifests itself in two part,, an a and bI interaction.

The (CH3 )9 a orbital Interacts most strongly with the 5alg orbital

of the W)(OCH)A molecule, this interaction being both energetical Iv and2 _/a MO I er I er

;pat iall Iy fa vored over that wi th the 4ag orbi talI. This Interaction

restults In formation of the W-C o bonding 13a and ('ant ibonding 16a

IM) lecit ar orbi t I s of W ((CIl).(('H 3 ) , which are ocrcupied and tinocc ilel,

re,;pectively. That the 1 3 a and 16a molecuilar orbitals; are derived

. * *% .

/ from interaction with the 5aig orbital of W2(O2)CH)4 Is readi Iv seen by

-*the comarison shown in Fi gtitre 4. The filIled Il3 a orbit a I has 35%. W

character (Table Ill) of which h3% is W s. As was the rase for its 5a,,

*;reciirsor, the W4 q character adds to N-M cy bonding a It holigh the NM

overlap is muich less than war, observed in the 5a,, orbital of W,?(OCH) 4.

Perhaps the single most important observation in the overall a 9inter-

action is that the 4 al, molectilar orbital Of W-,(O?)CH) 4 i s actuially

;tabi Ilized by the interaction and picks uip 10% more W character (6h7% W,

91)%.d) to give the 10ag molecul Iar orbital of W,(OICH) 4 (CH3 ) 2. Other

t han t hat , the !&a I . orb it al IIs es sent ialI I y uinpe rtuitrbed by thle axialI ag

interaction and can be described as both >-M T and M-L 3i bonding.

Compari-son of these orbitals in Figuire 3 iilltistrates this point rather

nicely. I t is important to recall that the 4 a Ig molIecul Iar orbitalI i s

the major contributtor to M-N a bonding in W2)(O2)CH),. The fact that this

orbital is actual ly stabilized by the ag interaction mean-; that the

1 )(O H)4 (rHI)?) system actuial ly has a filled molecuilar orbital involved

in strong M-M u bonding.

The (CH3) 2) bi orbital interacts with the 5a,,, H-M (3 orbital of

W-,(O 9-CH), generating the W-C a bonding and antibonding orbital-, which

are occutpied and utnoccutpied respectively (Figuire 6 and Table Ill). The

f fillIed 15b,, orbital I s 49% C and 23% W of which, this small amouint of W

character is a] located between s, p, and d angutlar contribuitions,. This

orbital is primari ly H-L 7 bonding butt has some N-N (Y character as

* well.

For W(OH)( 3Y we are presented with the rather peculiar

rcs'lIt of st rong? M- bonding in consor wi th strong M- odIri nIh

* ~;ixi al p'osi tion. This resitIt is perhaps coiinterintilt iye in light of the

chemistry of ('r-Cr qitadrtiple bonds, wherein the bonding of axia' IJigands

both weakens and lengthenq the M-M bond. The strong W-W and W-C bonding

in W,(O2C)H)4 (CH3 )? is the resti t of several factors. I) The fi rst

important M-M bond weakening ,lion axial ligation restiIts from ligand

- . donation into the '.-M c orbital. For the present calcuilation, the

. magnitude of this donation can be gataged by the amount of M-4 (1

character fouand in occupied orbitals of b,, symmetry. Figure h and Table

..l sygest that the amount of N-M a * character in the 15b orbital is

rather smal l, and thus the M-M bond is not significantly weakened by the

interaction. 2) The 4 alg orbital of W?(OCH),, which comprises the

.J major component of the 5d-5d a bond, is scarcely affected by the axial

ligation and thus remains a strong, occiupied component of W-W 7 bonding

in W?(OCH)4 (CH 3 )2 . 3) The major interaction of the a orbital of

( H3)2, is with the 5al, orbital of W2 (O?,CH),, res ,lting in a significant

contribution of the W 6s orbitals In both the W-W and W-C a bonding.

- The importance of these last two points needs some amplification.

If the 9-s a bonding orbital is inimportant, as it most certainly is in

Cr-Cr quadruiply bonded complexes, the interactions of both the symmetric

and antisymmetric (L)2 orbitals with a M-M qtiadrulJe bond mist

necessarily weaken the 1-i bond. The antiqsmmetric combination donates

into the N-M c orbital, an interaction which obviotsly will weaken the

%-,4 a bond. The qymmetric combination wil I part icipate in a "fi I led-

filled" interaction with the M-M a bond, reilting in M-L bonding and

ant ibonding orbita I s which are occupied and unocctipied, respectively.

Thus a portion of the M-11 a bond Is found In itnoccupied orbitalq, again

weakening the N-M interaction. The importance of the s-s a bonding

o rbital In tungsten systems, which is duie in large part to the

relativistic stabilization of the W fis orbitals, is that it provides

0 %

- % %4 a N

Jb

e.*. another mechanism by which the symmetric (L), orbital can interact with

the dimetal core. As i- apparent in the character of the 13a orbital

,)f W(OCH) , (CH 3 )?, the s-; cy bond is the principal 1-M contrib'ition in

the interaction with the symmetric (at) (CH 3 ), orbital, and that the

-,articipation of the s-s a bonding orbital interaction is greater in

3)2 than in W(0 2 CH) . Thus, the M-M1 weakening "fil led-

filled" interaction described above is effectively replaced by a

"fi l led-empty" interaction which actual ly increaqes the amount of

bonding between the metal centers.

In view of the above factors, it becomes inreasingly diffictilt to

describe the bonding in terms of a simple electronic configisration. It

* waq qhown, for example, that in M?(O 2 CH)! system, the M-M (I bonding

character is really allocated among two orbitals, and for W, appreciable

%%%- s character becomes mixed in. This makes description of the electronic

configuration in terms such as cz 2

,n42 an oversimplification. The

description of the electronic configuration of W?(O2CH)4 (CH3 ), in these

terms is real ly not possible. Formally the molecule might be

onsidered as having a W) 6 + core, thus a d 3 -d 3 dimer. Yet this

description is not adequate since the molec,,le clearly has orcupied

molecular orbital- of 1-M. r, 7 and 6 symmetry; effectively, the CH,

* eroups are not behaving as anionic ligands but rather appear to be axial

one-electron donors. Thus, we offer the fol lowing alternative

description. The bonding is consistent with a neuttral W9 (02 CR)4 moiety

[ interacting with two alkvl radicals. The major W-C interaction occurs

via CH 3 donation into the M-M o' and empty s-s (7 orbitals of W2(O2CH)4 ,

s'. " id th ,, t,) a first * lilroximation, the *-'M bond order Is %t I I Ior.

* This des;rription is consistent with the observed insensitivity of the W-

•,%"

p%.

%, %%. , . . . -" 5 .%, .% % %, . --.

W bond length and also with the experimental observationh that

W (CH. Pl)?(OCEt , upon photo I vsI g in hydrocarbon so I vents, vie I ds

W)(O.)CEt) 4 and dibenzyl by homolytic cleavage of a W-C bond.

An intrigiiny possibility suggested by our Iceilations relates to

t he bonding in W.(O 9CR) 4 L? systems where L is a nelitraI two-electron

donor Ii gand sich as PPh 3 or an oxygen donor I igand sich as THF. If the

ligand-metal interactions are similar to those of CH3 with W, a

is. sibi lity which ;eems likely for a strong donor ligand such as a

94 2 *?phosthine, the resulting electronic configuration would be a 71- S

i.e. a W-W triple bond bearing a striking electronic similarity to d 5-d

,+ 3triple bonds such as those based on the Re9)+ cure.

It Is important to -mphasize that the types of interactions seen

here between axial ligands and an M2(O 2 CR)4 framework have been observed

. .- 10, 30- 3?previos ly for >'I1(0 2 CR) 4 L2 systems. The magnitude of the

interactions are apparently quite variable, however, as exemplified by

,revious Xx-SW ca l cu lat ions on Rh 2 (O 2 CH),, 32 Rh?(O(2CH) 4 (?O)? , 3 and

Rh.(09 CH)4(PH 30 In these systems, the dominant Rh-L interaction is

ligand donation into the Rh-Rh a orbital. We find this interaction to

be less important in the tungsten system, a result which is doubtless

dependent on the accessability of the q-s bonding orbital. It is the

iC contribution of this orbital which we believe wi l I be largely

responsible for the electronic qtructural differences between first-row

and third-row mltiply metal-metal bonded systems.

'.a

Acknowledgement. '1.D.B. gratefuil lv acknowledges the sulpport of the

UInited States Air Force. This research was supported by the Office of

';.tv I Res.c.rlrh it Indiana Uni ver It v. D.L.C. Is the 1984/85 Indlana

University SOHIO Fellow.

. IV

' REFERENICES

I a) Ohio State University; b) Indiana Universtty; e) 1983/84

Indiana University SOHIO fellow.

?. Cotton, F.A. Ace. Chem. Res., 1978, II, 225.

3. Cotton, F.A.; Walton. R.A. "Mitltiple Bonds Between Metal Atoms",

Chapter 4, ohn Wi ley and Sons, New York, 198?.

4. Benard, M. .1. Am. Chem. Soc., 1978, 100, 2354.

5. Benard M.; Veillard, A. Nouv. .1. Chim., 1977, 1, 97.

h. Garner, C.D.; Hi I let, I.H.; Guest, M.F.; Green, j.C.; Co I eman,

A.W. Chem. Phys. Lett., 1976, 41, 91.

..;-7. Cotton, F.A.; Stanley, G.G. Inorj•_ fChem., 1977, lb,- ?bh8.

f 8. Guest, .1.F.; HI Ilier, I.H.; Garner, •D. Chem. Phys. Let t., 1977,

48, 587.

9. Benard, M. J. Chem. Phyj., 1979, 71, 2546.

10. Bursten, B.E.; Cotton, F.A.; Stanley, G.G., res'lts to be piblished.

II. Sattelberger, A.P.; McLaughlin, K.W.; Hiiffman, I.C. .1. Am. Chem.

Soc., 1981, 103, 2880.

)2. Sant~ire, D.J.; McLatighl in, K.W.; Hliffman, .J.C.; Satte I berger, A.P.

Inorg. Chem., 1983, 22, 1877.

13. Santure, D..1•; Hitffman, .1•C.; Sattelberger, A.P. Inorg. ('hem.,

1984, 23, 938.

14. a) Cotton, F.A.; Wang, W. lnory. 'hem., 1982, 21, 3859.

" b) Ibid, 1984, 23, 1604.

15. Chisho Im, M.H.; Chiv, H.T.; Huffman, .I.C. Polhedron, 1984, 3(61,

,,.759.

%,

9-.

9. . . . . . . . .

lh. Chisholm, M.H.; Hoffman, D.M.; Hoffman, .1C.; Van Der Sliv;, W.(;.;

Riisso, S. I. Am. Chem. Soc•, 1984, 10h, 5186.

17. It is well established that the 5 component add- little to the

.trength of a '1-M qadrttple bond. See for example, Cotton, F.A.

(hem. Soc. RLv., 1983, 12, 35.

18. Herman, F.; Ski I Iman, S. "Atomic Stritcturu Calriilation,;,"Prent ire--

Hal]:Englewood Cliffs, N.J., 19h3.

19. Schwarz, K. Phys. Rev. B, 1972, 5, 2466.

?0. Norman, I. .,Ir.; ('hem. Phys., 1974, h, 13'.

21. Wood, J.H.; Boring, M.A. Phy2.. Rev. B, 1978, 18, 271.

2?. Norman, .1.6., Jr; Kolart, H•.J. .1. Am. Chem. Soc., 1978, 100, 791.

?3. Birsten, B.E.; Cotton, F.A.; Fanwick, P..; Stanley, G•G. I. Am.

Chem. Soc., 1983, 1%)5, 308%.

-.-. Bursten, B.E.; Cotton, F.A.; Green, I.C.; Seddon, E.A.; Stan ley, G.G.

.I Am. ('hem. Soc., 1980, 102, 955.

?5. Cotton, F.A.; H,,bbard, .I.L.; Lichtenberger D.L.; Shim, I. .1. Am.

Chem. Soc., 1982, 104, 679.

r. 'lany excel lent reviews exist, see for example a) Pitzer, K.S. Acc.

Chem. Res., 1979, l?, ?71; b) Pyykko, P.; Desc laix, 1. Acc. Chem.

Req., 1979, I?, 27h; c) Sntjders, I.G.; Pyykko, P. ('hem. Phys.- . - . . .. , . . . . .

Lett., 1980, 75, 5; d) Ziegler, T.; Snijders, I.G.;Baerends, E.I. Chem. Ph s. Lett., 1980, 75, I.

27. Arratla-Perez, R.; Case, D.A. Jnorj. Chem., 1984, 23, 3271.

8. Birsten, B.E.; Cotton, F.A. Svml.. Faraday Soc., 1980, 14, 180.

9. sB,trten, B.F.; Cotton, F.A.; Hal I, M.B. .1. Am. Che . Soc., 1980,

% %"%.x .:

,'." " I9 , / ' 17

"o'" ,. ,r tn , . " ot o , F A. I r . em , 1 8 , 2 , l ?

0:"

,..-

* ,

31. Norman, .J.G., Jr.; Renzont, G.E.; Case, D.A. "1. Am. ('hem. Soc.,

1979, i0, 525h.

3?. Norman, fI.G., Jr.; Ko I ari, H.I. J. Am. Chem. Soc., 1978, Io)(),

791.

jl'.°~

'.,.-

11+

18

* %

-...-.

t0?':• ,0'

V

" Table 1: Bond Lengths and Angles, Atomic Sphere Radii, qnd

Statistical Exchange Parameters for W 2 (O2 CH) 4 (CH3 ),)

Atoms Length, . Atoms Angle, degree

W-W 2.186 W-W-0 90.5

W-C 2.190 W-0-C 11H.6

W-0 2.085 O-C-H 119.2

'-.." O-C 1.270 W-W-C 180.0

C-Ha 1.090 W-C-H 115.2

.-H 1.090 H-C-H 109.0

. .

Atom sphere radius, bohr a

Outer Sphere 8.4988 0.74519

W 2.4710 0.69319

0 1.6876 0.74447

Ca 1.5787 0.75923

Ha 1.2963 0.77725

* Cb 1.8143 0.75923

H" 1.2895 0. 77725

,,formate

b alky1

I0.

.%' .•

a.

. . Table 2: Upper Valence Molecular Orbitals of W2 (02CH) 4

% Contributionsb W Angular Contributionso

Energylevela IV W2 (02CH)4 INT OUT p d

A. Nonrelativistic Calculation

5a2, -1.5526 41 5 37 16 5 92 S

4a2 u -3.46'5 24 3 64 1 5 18 14

-4.3453 81 4 15 0 1 9;

2blu -5.1995 73 12 9 0 1 )(

2b2g -6.6775 73 10 16 0 100

" -8.6564 74 15 10 0 98 2% ° U

5ag -9.7335 59 35 6 0 29 6 63 2

la -10.1293 0 84 15 0

4e -10.2099 4 81 14 0

5e -10.2436 15 72 12 1 24 12 4

3eg -10.5364 1 81 18 0

; a2 -10.7882 7 77 15 0

-b -11.1002 12 76 12 0 99

Ib u -11.4074 10 68 22 0 100

. 4bIg -11.7065 21 67 11 1 99 1

I-F. -11.9946 54 42 3 0 3 94 5

lie

.%pa..

-'I-

.4.

:0

Table , continued

B. Relativistic Calculation

5R2u -1.3865 40 4 38 17 6 91

a2u -3.6441 22 3 67 8 15 23 61 5

-e 3 3.89C) 79 4 17 0 1 97 2

2b.u -4.7354 79 11 10 0 100

2 b2g -6.2695 72 10 17 0 100

be -e.2742 78 11 11 0 100

1ain -10.1927 0 84 15 0

4e -10.3101 3 81 14 0

5eu -10.3792 12 69 12 1 34 60 6

3e -10.6111 1 81 18 0g

5aig -10.9982 69 27 4 0 38 8 52 2

3a2 u -10.9984 8 77 15 0

3b2u -11 .I0,35 11 76 12 0 99

Iblu -11 .41-74 69 22 0

4 big -11.6596 19 68 11 1 ' 9

4a, g -11.8552 51 45 3 0 , 54

aHOM', is the 2 b2g orbital

T - intersphere and OUT = outersphere charge contributions

c Listed only for levels which have 10% or more W contribution

.%.

SL%

& , .' ;. . , ,, ~ ., ,V ... , N i .; .v. ', " , ,'''.'Jr ,, f .b , .-A.k,£

Table 3: Upper Valence Molecular Orbitals of W,(O 2 CH) 4 (CH-) (Relativistic)

% Contributionsb W Angular Contributions(

nergyLevel eV W2 (2C)4 (CH) 2 IT )JT s p ( r

I "a 15 1?U 1 r47

goS3 3 12 01 97 2

, a'.. 9z , 1 51 4 1 4,4 3,9 11

- u -2.4 78 12 0 10 0 100

S -6. O 71 11 0 1 (0 100

15b -6.854 23 1 54 19 14 25 55

14b u -7.886 7( 10 5 10 U 97 5

' 9au -7.888 76 10 5 10 0 97

"au -9.969 0 84 0 15 0

15a -10.027 3 79 4 13 0

'" I -10.04i 3 79 4 13 0

."hb . 10 57 2 12 1 2'

7, -10.0Q3 10 57 2 12 1 yU

g , ,-0.3)6 1 80 0 18 0

-I.a - . o9 1 80 0 18 0ag

- I ?b u -).774 8 77 0 15 0

, ; -YJ.8b4 11 76 0 13 0 9 1

" , -. ' Q , ) 45 7 11 1 63 6 30 1

S1'a -11.165 2 4 88 4 2

b -11.171 2 4 88 4 2

., -- 11't2,)t

T-,n nueil

u -I .9 16 47 26 17 1

a .9 57 1 11 1 99

.a 74 2' 9 go ~ 9(

I-"-" Fb ,rtbital

-' T-:9r.zphere and "JT outersphere charge contrihution.s

List :d only for levels which have 10% or more W contribution

'% % 4:V

0

'P.,

V°.

,0°

U...S,.(

r. :

CAPTIONS TO FIGURES

Fimgure I: An ORTEP view of the cent rosymmet ri c W?(CH?Ph )?(Og('Et)

molecule from reference 16.

ig'itre 2: Nonrelativistic and relativistic converged Xa-SW elgen

valuies for W 2 (O2 CH)4 - Primari Iv W-based level,; are in

bold face along with their percentage of W o.ontribit ton.

, -ire 3: Comparison of contotr plots of the 'a molecu lar (orbital

of W?(O2CH) 4 and the la moleci Iar orbita I of

SW 9 (09C-H) 4(CH3) 2 . These plots and al I stibseqiient [lots

are in the horizontal mirror plane containing the W

atomq, two of the formate ligandg, the axial C atomq, and

two of the C-Il bonds. Contour valuie. for this and

s" ubsequent plots are +I, +?, +3, +. +0.0?, +-f.0),

+().08, +0.16 e/A ', respectively.

Figitre 4: Contoor plots of the 5alg orbital of W,(OCH).! and the

13a and lUag orbials of W2(OClI)4(CH 3)?

IiJ

a'

0

k-a

4%*%2

0A

-.- -igtre 5: Re l tz of the re ]at i vist ic S('F-Xct-SW ca I -liJt ion- on

- (O.- H)(c 3 )9 . This diagram shows the corre lat ion of-.the orbitals of W? (OCH)4(CH-)? to those of W-(OCH) 4 and

. . (Cti 3 )?. Only those levels involved in W-W bonding or

ant ibonding are shown. The 8b orbital is the highest

oceupied orbital of W(OC)l ('l'3),;o

-ig.re 6: Contour p lot of the 15 b m. Iecu lar orbi ti I of

N"' 2 (07CH) 4 (CH 3 ) 2 •

lid

% "i%

Ja 1.i1S%%z-

N.

0l I i d l d 1 d d l ~ I | -k~ l I d i • . t -' . . - --

0

0

(*

/

IN

~ -I

0

0

0

S W2(02CH) 4

-1Re I at i v ist I c

2-22 a2 9 223b 29 2___ 9~7 e u 7 .u

.6i 6aig

.4-e -e

8 - 2blu-5~u

-69. -2b 29 72% W

Ix 73%W 2b 2 -

Z-8.

if -6eu 78%W74%W 69u

-9

59%W ag-10- lalu.

'F ~ 4 0 9

3a2u - - a1 0 69%W- 3 82u3b2u-

3 b2ulu I l

4big 4 - big-121 54%W 4& 1 9 -- a a, 51 %W

------ - - -- ------ 1 !-'.'5

I ,),

S.-'-" '-. - -.

*"A *Q |

>D-

'aI .

I I o - - 5-'

I . , . -- ,

, , -. ;,S.- , , .. , -T

I,* ( I 5' -

'v,- "* "*"

• : . .. . .

I"S *

(-a .

.S

P ''i ' /- * :,. . &. :.," - . - -.. .-.. '. .. p.5,j-s%:Js.*- :

'CD

-- --.

:,- ,. ,- ,

-D

I CD

- ", .- - - - --- - -

2 -N V".

" " "-" ' ... "-"

.... . -- - - - I ,o.

[.5.-,,

.-'2-. •, *

-i ... .. .. ., ,-.. .. i _

I (.,

it "S.

,°u

oPo

[o * --- *

coo)

* CN)

,rw Np

beO*

* w2(O2CH) 4 w2(O2CH) 4(CH3) 2 (A 3)0

-2

-3-1 7a9

- 9bg59 -

-4- 16ag

* 2bl. -

-5

cc2b 2 , - 8bg - agwi 72%W 71%W

-15b,-7 23W

14b,-8 78%W76%W 9au

-9

-10-

* -11 5a,9g 69%W la35%W3a

a 51Y*W - 089-1216%

- -- -- --- -

N N

.6 ME j&N

FILMED

1-85

D1 ICW.I 1 Iw