Interplay of direct and indirect coupling contributions in proton hyperfine splitting constants of alkyl radicals

Interplay of direct and indirect coupling contributions in proton hyperfine splitting constants of alkyl radicals

-: ‘. ENTERPLAYOFDIR~~AND,I[NDIRECTCOUPL~GCONTR~~IONS .'~~NPR~TONH~~~INIE SPLIT~INGCONSTAN~S~FALKYLRADI~AL~ : ., :. ‘. '. .: Y. ELIJNGER,A:l$A...

332KB Sizes 0 Downloads 49 Views

-:

‘. ENTERPLAYOFDIR~~AND,I[NDIRECTCOUPL~GCONTR~~IONS .'~~NPR~TONH~~~INIE SPLIT~INGCONSTAN~S~FALKYLRADI~AL~

:

.,

:.

‘.

'.

.:

Y. ELIJNGER,A:l$ASSAT,R.SUBRA .., : .‘kafnxa~oire de Chimfe Urganique [email protected] du Centre d’Etuds Nucfeaires, _<8 - Grenoble, France

.’

,’

'1

.'

..

G.BERTHIER Lnborurofre de Chfmie Quantique,

.

13 rue Pierte et Marie Curie, Wis 5t?me, France

‘.

and’,

P.MILLIE La&war&e de Chimie, E.N.S., 24,tue Lhoywnd, Paris .%me, France

Reck&cd26 Jufy 1971 ,. An ah i&tio

catcufation’onthe CHs-&& radical swests that the spin-pokrization and spin&ocalization con-

_isibutions arc both equd.Iy impottznt nnd shouid be simuIianrortsly

hyperfine

mnridered

in the thcorctial

nrwlysis of proton

coupling constants oialkyfradicals.

The semi-empirical relationship i;(NB) = Bd + B2 c0s2e ,

(1)

suggested by McConnell [I] has frequently been us& in electron sp;k resonrnce spectroscopy to correlate pproton’hyperfiie splittings a([email protected] with the conformation of free radicals (e being the d&edral angle between th! p-orbitaf conning tee unpaired efectron and the C-C,[email protected] plane)., In most of these studies 80 is assumed to be negligible relaiive to B2. However, in some ra+als it has been found that the term B, in eq. (1) is sight f2-51, This point has’bcen ten- ,’ tatively attribvted to a residual motion of +e ppio-, -. tons-in riiti,oalkane radical anions [6] and an inversion niotion,of [email protected] N-0 bond in n&oxides (73’. ... Fr,om a theoretical point $f.‘vjew,~there ha? been,;

calculation of hyperfme splittings of the ethyl radical, using a restricted open-shell formalism. The wavefunction has been computed by means of the IBMOL progrctm [9] _The closed and open-shell SCF equations have be’en solved following Roothaan’s method using a b&is set of gaussian orbitals containing 9 s and, 5 p type functions on carboq atoms ancl4 s functions on ~l~droge~ atoms:The molecular geometry has been adapted from ethane and’only the planar form of the paramagnetic centre has been examined. The htierftie splittings aH are the sum of two kontributions, .adirect ~~~b~tion ad Fomputed from the SCF wavefunct’ion $_and an indirect contribution Oi determined here by a fist-order +rtuibaii.&i treatment ttiking into .a&ou.nt spin-polarizatidn effe& [ 4 11.. This leads ia ‘.

‘cals, thii &$est:radi,cal

&wing

both Q’- and’@ctiu-‘,.

.with



:plings[8];. ” .,_ I-., .“) .;’ ‘_, :: ; :,.‘.,_ ..pd,:~.i~u(‘fI),~.‘.. We report here &+resuI&, of the first noriempu;lcaI .. : -, ;_ _ .. ..

,.

::

:‘. ,,,

;.:y : ,.,. ,:,,I : ;

I.‘, ... .,_ .._ 1.. (3) ,, ‘.; ,’ ..,_. :..: ._ ; ;-. : ._&‘:. ..i :.L .:’ /..’ ;_ _,‘. ., : .. ~+y,, ‘., ,, ,:;:, ..; ‘I ;‘yI,.,,..,. ,_~,~~,.‘,._: .,..,I. :“;, ,_,.:_. (,, ‘,I. _C’ ,.’ _’ -: :. .., ,.: ,. ,’ .. .. ,, ,. .~ -. ‘; I’ . . ._.. _‘;’ .:. ‘. . :. -’ .: _, .-.-: . ‘: 1 :::“. _ <:.:>.‘, i,. ‘1,’,_ ..:’ .;::_. -, ‘, .;.,,:,._ :_.:_, ..‘y : : :_., _. .‘;I .;.,; :,.,,: ,_ ,__.__, :,; .;., .,. ___. _,.; _ _:.: ,= .. .-‘..‘. ,, ,; ,_.. -:,’ ,-I ; _, .Y.. : :‘. ..,;: -.-:: _:,.,,,. ..l,.._,.‘ ‘... . ;I’ _:... ; .. ,.,.., /.,_,, . ;:. ., .’ . ,. ‘. .., .,, ..’ .::. ‘-,I .,, ,I .,., : ..’ ‘: ,. : 1,’ .. I. ,,_,..; .-.- ,. ,. ._. ‘_

1

Volum;

11, number

3

CHEMICAL

PHYSICS

c d

c (QV*~u’~u~d)OdO,)Q,.(rH) v+ EO -Q-v*

,’

(4)

where @dare doubly ocCupied orbitals, &,* virtual or- ._ bitals and 9, the orbital containing the unpaired electron. In order to be able to describe the.origin of hy; pen%e splittings in terms of localized interactions, we have determined equivalent lo&ed orbitals according to the procedure of Foster and Boys [ 121. Consequently, every term of the perturbation expansion may be related to a specific mono-excited state (d+v*). The main resuIts of the computation are f 13] : (i) The direct contribution to a-proton (Ha and Hb) splittings is zero. The direct contribution to &proton splittings (fig. 1) fits eq. (1) with B. = 0 and B2 = 19.20 Oe. (ii) As in the methyl radical [l l] ?, the indirect contribution to the a-couplings is negative. In confort Further details on hyperfiie splittings in alkyl radicals incIuding long range interactions will be published elsewhere 1131.

1971

a(Ha) =&lb) = -33.82 Oe and in conformaOe and a&b) = -34.04 Oe. = -33.60 tion II, Q(h) The indirect contribution to &proton splittings (fig. 1) fits eq. (1) with B, = -0.96 Oe and B, = 16.19 Oe. (iii) The calculated total hyperfiie splittings can be fitted to eq. (1) withBo = -0.96 Oe and B, = 35.39 Oe. This leads to an average value of 16.7 Oe for the freely rotating methyl group. (iv) The observed Q- and /$couplings are respectiveljr -22.38 Oe and 26.87 Oe [14]. Proton hyperfine splittings computed here are correct in sign but are slightly apart from the experimental values as in the ab initio calculations for methyl using SIater orbitals r151* (v) In all cases the dominant contribution to indirect couplings (90%) is the electron excitation from the bonding CH-orbital to the corresponding CH*antibonding orbital. (vi) In the absence of any vibronic effect, a P_ proton in the nodal plane of the 2p orbital containing the unpaired electron (cos8 = 0) would present a negative coupling due to spin poIa.rization. These results demonstrate that detocalization and spin polarization contribute almost equalIy to /?-hyperfiie splittings. This particukr point gives a better mation

pi = -2

15 October

LETTERS

I,

understanding

for the successful

development

of semi-

empirical methods in which orJy one of the two contributions is taken into account: if numerica agreement between observed and calculated hyperfme splittings is achieved by suitable parametrization, the cos2B variation is obtained independently of the actual nature of the coupling.

-lid

460

-60'

b’

Ongk

?H,

Ha Y&tH-320

Hb :

n120

:

I

6.0.

120.

160’

References [ 1 ] C. HeUer and H.M. hi&onnell, J. Chem. Phys. 32 (1960) 1535. [2] A. Hordield, JR. Mortonand D.H.Whiffen, hloL Phys. 5 (1962) 115. [3] D. Pooley and D.H. Whiffen,Tran+ Faraday Sot 5’; (1961) 1445. [S] E.W. Stone and A.H. hfaki, I. Chem. Phys. 37 (1962) 1326. 15] G. Cbpebt-Letourneux, H. Lernatcc, R. L-enk, M.A. hiare’cinl and A. l&sat, Bull. Sot. Chim. France (1968) 3963. [6] E.W. Stone and AH. Maki. 3. Chem. Phys. 37 (1962) 1326.

Fig. 1. 363

[email protected]

‘,

- ,’

ll,‘numb&

:

3

:

.‘.

thier, k&L Phyi (B].A$.k I+,ach&

::



,,

CXEkICALPHY~ICS

.-.:

17(.1969) 2!?. MoLF’hya 1 (1958) 233;

m3;chemut, 4. chm. phyS..~9(1958) 43; “~‘Pi;..Lykos;f.Cfierit;Phys’ 32fl960)625; -‘.“:i _.’ -,’ ‘DH. Levy,Mdt$‘h~s.10 (1966) 233; ,.. ., D.~]tad&tsand M:Karpkis,J~ Chem. Phys. 44 (1966) ,[email protected]’ :,: :,;“, G.k ~etersson and A.D. Mc,Lachtan, J. Chem. Phys. 45 (1966)628. 1 i9J ‘A. Veihl, &~iD&d atid P:MiBi&, lBi&L UX! 36t%



:

:

LETTERS

_‘.

:

” _,

I5 oct&er.1971

[I.1 1 P. NW, 8. L&&d G. krthier, IntqI, J. _. ~‘Chem.,tobe&bLished.. J.N.,Fostei and.S.if;.Boii, Rev: pd. Ptiys. : .., ‘. 1740. 1x31 Y. EUinger, A; &sat; R. S&a, G. ~e&er ..MiilitS, to be published. [!4] R.W. Feswden and R.H. Sch&, J. Chem. ‘(1963) 2149: ., [ISI S.Y. Chang, ER. Davidson and G. Vinck,

Quantum 32 (1970) and P. ‘.,’ Phys. 39 J. k!hkm,

Phys:52 (1,970) 1740.

version, Laboratoire de Chin& de?EaAe Norma& Su., @&; ,‘, .’

‘(

.,

:

:

_’ :.

._

: ;

..

.’

; : .:’

.’

.’

:



.

.’

‘.

:

: ..

..

“.

.’

..

_

.’ .

‘I-

‘.

::.;.

:

,,,

(.,’

.’

., .‘.,

.

.,

;

,_

: .:.,,:

:’ :.’

-’

.’

.._

j_’

)

%.

.,

L:

.~’