SYNTHESIS AND REACTIONS OF HYDRIDO

TRIPH=LPHOSPHINE RUTHENIUM(II) COMPLEXES

A thesis submitted

by

ROBERT JOHN YOUNG B.Sc.

for the

DEGREE of DOCTOR of PHILOSOPHY

of LONDON UNIVERSITY

Department of Chemistry,

Imperial College of Science and Technology,

London SW7. January 1976

2

TO MUM AND DAD

AND CLOGGS

3

ACKNOWLEDGEMENT

I would like to thank Professor Geoffrey Wilkinson F.R.S. for

his interest and advice throughout the course of this work.

Many

and David

Thornback

I am also

financial

thanks are also due to Drs. John Bradley, Dick Anderson

Cole-Hamilton for their help and advice, and to Mr. John

and the rest of may colleagues for their friendly encouragement.

indebted to the States of Guernsey Education Council for

support.

Finally I am deeply grateful to my wife for her patience in typing

this thesis.

4

CONTENTS

PAGE ABSTRACT 6 INTRODUCTION 7 CHAPTER I HYDRIDO- AND CARBOXYLATO- TRIPHENYLPHOSPHINE COMPLEXES 21

OF RUTHEIIUM(II), THE ISOLATION AND CHARACTERIZATION OF CATIONIC COMPLEXES DERIVED FROM THEM.

1.1. Introduction. 21 1.2. Catalytic hydrogenation by protonated methanolic 22

solutions of RuH(CO2Me)(PPh3)3 in the presence of

triphenylphosphine.

1.3. Attempts to initiate the activation of saturated 30 hydrocarbons.

1.4. Protonation of ruthenium(II) complexes to yield 31 [RuH(76-PhPPh2)(PPh3)2113F4 and

[RuH(H20)2(Me0H)(Pa3)2>F44, 1.5. Examination of the catalytically active species. 34 1.6. 31P n.m.r. spectroscopic studies on triphenylphosphine 41

ruthenium(II) complexes.

1.7. Conductance measurements on ionic complexes. 44 1.8. Reactions of hydridochlorotris-(triphenylphosphine) 45

ruthenium(II).

1.9. Discussion. 46 CHAPTER II COMPLEXES OF RUTHENIUM(II1 CONTAINING BOTH HYDRIDO- 57

ANDV-ARENE LIGANDS.

II.1 Introduction. 57 11.2 Protonation of RuH2(PPh3)4 in benzene and toluene to 61

yield the [RuHY-arene)(PPh3)2pF4 complexes.

11.3 Protonation of Ru(II) complexes in the presence of 63 potential 77-arene Uganda.

11.4 Summary. 68 EXPERIMENTAL 70 REFERENCES 88, APPENDICES 94

ERRATA et ADDENDA

Page 13, 3rd equation - Right to left arrow in equilibrium omitted.

Page 17, 1st equation - Superscript omitted - should read 2Pr'0H.

Page 18, line 14 - 'carboxylic acid and potassium hydroxide' to be

added after 'trichloridel.

Page 18, line 16 - RuH2(PPh3)3 should read RuH2(PPh3)4.

Page 21, line 5 Ru(002Me)2(ITH3)2 should read Ru(002Me)2(PPh3)2.

Page 53, SCEME 12 - [111111076-PhloPh2)(P)31+ should read

[RuilH(7 6 -PhPPh2)(1)21 4.•

Page 66, diagram (90) - Then-bonded ring should be a pyridine as

opposed to pyrrole as shown.

Page 67, diagram - The triphenylphosphine ligands should read PPh3

as opposed to PPh.

5

ABBREVIATIONS

g.l.c. gas-liquid chromatography

i.r. infrared

n.m.r. nuclear magnetic resonanoe

N.T.P. normal temperature and pressure

sym symmetrical

asym asymmetrical

Me:- methyl; Et - ethyl;

Prl- iso-propyl; Bun- normal-butyl;

But- tertiary- butyl; Ph - phenyl;

Cp - cyclopentadiene

6

ABSTRACT

The action of strong non-complexing acids on hydrido triphenylphosphine

ruthenium complexes such as RuH(CO2Me)(PPh3)3 , RuH2(PPh3)4 and RuH4(PPh3)3

have been studied. The rates of homogeneous hydrogenation for various

alkenes by a catalyst derivedfromthe protonation of RuH(CO2Me)(PPh3)3

with 2-toluenesulphonic acid monohydrate in methanol have been measured.

The catalytic efficiency of the solutions was found to decrease with

time.

The nature of the solutions initially produced on protonation

and of the species isolated from the 'aged' solutions has been studied

in several solvents and under different reaction conditions. The

initial red species in methanol and acetone is shown to be [RuH(PPh3)3(S)1+

where (S) = solvent, and possible catalytic reaction mechanisms are

discussed. The methanol solution ages in the absence of triphenylphosphine

to yellow [RuH(H20)2(Me0H)(PPh3)2]+ but in the presence of an excess of

this ligand the n-bonded [RuH(76 -PhPPh2)(PPh3)2r is formed. In acetone

the end product is [Ru(CO2Me)(H20)(PPh3)3]+.

The species are characterized by i.r., 1H and 31P n.m.r. spectra

and by conversion into other known or easily characterizable complexes

by action of carbon monoxide and acetonitrile.

The action of fluoroboric acid on RuH2(PPh3)4 in benzene or toluene

has been shown to produce Tr-arene complexes of the type

[RuHW-arene)(PPh3)2]BF4. A preliminary investigation of protonation

reactions of Ru(II) complexes in the presence several aromatic species

has been carried out, with the aim of producing novel 7T-arene complexes.

7

INTRODUCTION

8

INTRODUCTION

The introductory chapter to this thesis contains a brief survey

of the known chemistry of hydrido triphenylphosphine ruthenium(II)

complexes. Some species which are not members of this class of compound,

but which may act as precursors to or be derived from such complexes,

are also discussed.

In 1961 Vaska and Diluzio(1)reported the identification of the

first hydrido triphenylphosphine complex of ruthenium(II),

RuHC1(CO)(PPh3)3, prepared by the reaction of triphenylphosphine and

1RuC333H20' in 2-methoxyethanol, a reaction which proceeds by reductive decarbonylation of the solvent. The authors having previously formulated

\( the complex as RuCl(PPh3)32) reconsidered their data in the light of

simultaneous work by Chatt and Shaw(3), which showed that complexes of

transition metal halides with tertiary phosphines or arsines give

hydrido and/or carbonyl complexes when they are heated with potassium

hydroxide (or sodium carbonate) in alcohols. Chatt and Shaw had isolated

RuHC1(C0)(PEt2Ph)3 according to the equations

[Ru2C13(PEt2Ph)6] Cl + 2Et0H + 2KOH

2[RuHC1(00)(PEt2Ph)3] + 2CH4 + 2KC1 + 2H20

and subsequently(4)proposed a mechanism involving base promoted

hydridocarbonylation (SCRFME 1), in which the initial step is the

formation of an ethoxide-ruthenium complex (I), followed by hydride-

transfer to ruthenium to give the acetaldehyde complex (II). (II) then

breaks down to give methane and the carbonyl complex (IV), possibly by

way of an isocarbonyl complex (III):

H \ /

CH3-C-0-Ru- I /

/

CH34.0-Ru-

(I)

/ OEC-Ru-

H

(IV)

SCHEME 1

+CH4

9

RuHC1(C0)2(PPh3)2 was first postulated as an intermediate in the

production of RuC12(C0)2(PPh3)2 by the oxidative addition of hydrogen

chloride to Ru(C0)3(PPh3)25). James et al(6)found that the reaction of

carbon monoxide gas with NN'-dimethylacetamide (dma) solutions of the

complex RuHC1(PPh3)3 (prepared in situ) yields a mixture of the

complexes RuHC1(C0)2(PPh3)2 and RuC12(C0)2(PPh3)2, and were able to

separate the two cis-dicarbonyl products by recrystallization from

methylene chloride/methanol.

RuHC1(C0)2(PPh3)2 is related to the well known rhodium(I) N(7) hydroformylation catalyst RhH(C0)2(PPh312, which is active under mild

conditions. The ruthenium complex however was found to be inactive

as a hydroformylation catalyst under similar conditions(6b)

but

slowly hydrogenated methyl vinyl ketone.

Treatment of RuC12(C0)2(PPh3)2 with lithium aluminium hydride

yields the dihydrido complex, RuH2(C0)2(PPh3)2(8/a), which has also

been produced by the reaction of Ru112(C0)4 with triphenylphosphine(8/a).

The low solubility of RuH2(C0)2(PPh3)2 precluded the observation of the

H n.m.r. high field signal and characterization was on the basis of

i.r. and analytical data. At about the same time the high field I H n.m.r.

triplet was observed and i.r. bands were assigned to both v(C-0) and

v(Ru-H) modes for RuR2(CO)2(PPh3)2 prepared by the reaction at high

temperature and pressure of molecular hydrogen with a tetrahydrofuran

solution of Ru(C0)2(PPh3)3(8b ). The dependence of conversion and

aldehyde ratios on catalyst concentration, temperature, partial and

total pressures, nature of substrate and addition of excess

triphenylphosphine together with other ligands, in hydroformylation

with Ru(C0)3(PPh3)2 has recently been studied. A mechanism involving

RuH2(C0)2(PPh3)2 as the principal active catalytic species has been

proposed(9).

The reaction of Ru(C0)3(PPh3)2 with nitcosonium hexafluorophosphate

has been shown(10)to produce the cationic hydrido species,

[RuH(C0)3(PPh3)2r. The reaction is assumed to involve the formation

of hexafluorophosphoric acid according to the equation:

NOPF6 + Me0H

HPF6 + MeONO

RuH(N0)(PPh3)3 was first prepared by the reaction of ethanolio

potassium hydroxide with trichloronitrosylbis-(triphenylphosphine)ruthenium,

nitric oxide and nitrogen peroxide on aqueous ruthenium trichloride(14)

is treated with an excess of triphenylphosphine in alcoholic potassium

hydroxide:

in the presence of excess phosphine(12)

. The complex may be prepared

in one operation from ruthenium trichloride trihydrate(13);

trichloronitrosylruthenium, prepared by the action of a mixture of

10

The protonated species were also obtained as white crystalline salts

by the treatment of the tricarbonyl complex with strong acids HX

(X PF6, 0104 or BF4 ) in ether.

The reaction of RuHC1(C0)(PPh3)3 with silver perchlorate in

acetonitrile(11)affords the cationic complex, Pla(C0)(MeCN)2(PPh3)2]C104,

which is a particularly versatile intermediate because of the difference

in labilities of the coordinated acetonitrile ligands.

RuHC1(C0)(PPh3)3 + AgC104 + 2MeCN -*•

[RuH(C0)(MeCN)2(PPh3)21C104 + AgC1•PPh3

Treatment of this cation, first with carbon monoxide in a fast reaction,

and secondly with triphenylphosphine in a much slower reaction, yields

the complex cation, [RuH(C0)2(PPh3)3r, in which the carbonyls are trans.

1 RuC13.3H201 NO/NO2 aq. soln.

Ru(N0)C13(aq.)

Ru(N0)C13 + 3PPh3 + 3KOH + 2Et0H

RuH(F0)(PPh3)3 + 301 + 2MeCHO + 3H20

The solid state structure has been confirmed by X-ray crystallography(15)

as a slightly distorted trigonal bipyramid. The complex is an effective

catalyst for the isomerization of olefins and the hydrogenation of styrene.

The dihydrido complex, RuH2(PPh3)4, has been obtained: (a) by the

reaction of 'RuC13.3Hi01 or Ru(acac)3 , triphenylphosphine, and •

triethylaluminium in tetrahydrofuran(16)

RuX3 + PPh3 + AlEt3

RuH2(PPh3)4

I

11

(b) by the rapid successive addition of hot ethanolic solutions of

'RuC13.3H20' and sodium borohydride to a well stirred, boiling ethanolio

solution of triphenylphosphine(17) ; (c) by the interaction of the:.

dichloride, RuC12(PPh3)3, with sodium borohydride in the presence of

excess triphenylphosphine in hydrogen saturated benzene/methanol solution(18) ;

and (d) by the addition of triphenylphosphine to RuH4(PPh3)3 or

RuH2(N2)(PPh3)3(19). Yamamoto et al(16)found that RuH2(PPh3)4 dissociates

in benzene solution into free triphenylphosphine and RuH2(PPh3)3, which

undergoes reversible combination with both molecular hydrogen and

nitrogen to yield the complexes, RuH4(PPh3)3 and RuH2(N2)(PPh3)3.

RUH2(PPh3)4 N2

RuH2(N2)(PPh3)3 + PPh3

RuH2(PPh3)4 H2 RUH4(PPh3)3 PPh3

They had previously reported(20)

these reactions and had tentatively

suggested the ruthenium complex as RuH2(PPh3)4. Subsequently Knoth(19,a)

prepared and isolated the dinitrogen complex, RUH2(N2)(PPh3)3, the

assigned composition being supported by reaction with hydrogen chloride:

RuH2(N2)(PPh3)3 + 2HC1 RUC12(PPh3)3 + N2 2H2

(81%) (94%)

He demonstrated that the nitrogen molecule is reversibly displaced by

ammonia and by hydrogen, forming RuH2(NH3)(PPh3)3 and RuH4(PPh3)3

respectively. Characterization of the latter product as a tetrahydride

was supported by sequential reactions with triphenylphosphine and

hydrogen chloride:

RuHa(Ph3)3 PPh3 ► H2 + RuH2(PPh3)4

(84%)

HC1 RuC12(PPh3)3 2H2

(83%)

The reactions of RuH2(PPh3)4 in benzene with iodine in a ratio of 2:1

and with excess ethyl bromide yield the complexes RuHI(PPh3)3 and

RuHBr(PPh3)3 respectively. When the ruthenium complex, RuH2(PPh3)4,

in benzene solution was allowed to equilibrate with a large excess

of deuterium(16)it was found that the extent of exchange corresponded

to approximately 24-25 hydrogen atoms 22.r mole of complex. On the basis

12

of i.r. and n.m.r.(21)correlations it was assumed that a rapid equilibrium

involving insertion of the ruthenium complex into ortho C-H bonds was

taking places

RUH2(1311h3)3 +PPh3_, - PPh3 Rull2(1)113)4

Ru + D2

Ru

PPh2 D / I

PPh2

ll

(PPh3)2 /

Ru

H/D

PPh2

H ■ (PPh3 )2 H fn„ D ‘rru3/2

I •• /

SCHEME 2

RuH4(PPh3)3 has been prepared directly by treating RuC12(PPh3)3 ( with sodium borohydride in a benzene/methanol solution la)

. An earlier

report(22)

had proposed that the complex was of formula RuH2(PPh3)3,

but confirmation that the species was a tetrahydride was obtained by

Knoth(19). The tetrahydride readily reacts with several small molecules(le)

Molecular nitrogen, nitric oxide and carbon monoxide add with

displacement of hydrogen to yield the complexes RuH2(N2)(PPh3)3,

Ru(NO)2(PPh3)2 and RuH2(C0)(PPh3)3. The reactions with sulphur dioxide

and nitrosylchloride are less well authenticated but appear to yield the

complexes Ru(S02)2(PPh3)2 and RuC13(N0)(PPh3)2. Carbon disulphide

reacts instantly with both RuH4(PPh3)3 and RuH2(PPh3)4, by insertion

into the Ru-H bonds to yield in both cases the bis-dithioformato complex,

Ru(HCS2)2(PPh3)2.

RuH4(PPh3)3 + 2CS2

Ru(HCS2)2(PPh3)2 + + PPh3

RuH2(PPh3)4 + 2CS2 Ru(HCS2)2(PPh3)2 +2PPh3

Nitriles also react directly with the tetrahydride; in benzene solution

acetonitrile and benzonitrile react rapidly to yield RUH 2(UeCN)(PPh3)3

and RuH2(Ph0)(PPh3)3.

13

During studies on carbon dioxide activation by means of transition

metal compounds Vol'pin et al(23)

have reported the preparation of the

complex RuH2(CO2)(PPh3)3:

RuH2(N2)(PPh3)3 + CO2

RuH2(PPh3)4 + CO2

RuH2(CO2)(PPh3)3 + N2

RuH2(CO2)(PPh3)3 + PPh3

and have explained its chemical properties in terms of the equilibrium:

0 (PPh3)3 Ru(H)-00 H (PPh3) 3 RuH2(CO2)

Knoth(19,a)has also replaced the nitrogen molecule in RuH2(N2)(PPh3)3

with benzonitrile to yield RuH2(PhCN)(PPh3)3. The same author has

treated RuH2(N2)(PPh3)3 with inner diazonium salts of 13/0H: , such as

B10H8(N2)2 and N2B10H8S(CH3)2 to yield novel complexes in which dinitrogen

bridges ruthenium and boron. Hydrido species produced in this way were

of the types

RuH2(N2)(PPh3)3 + BI0H8(N2)2 ---1- [(Ph3P)3RuH2N2]2Bioli8

NABH4

OR

112/Pt

[(Ph3P)3RuHC1N2]211418

The action of acids on RuH2(PPh3)4 in methanol(24)

yields solutions

which are active catalysts for the hydrogenation of alkenes.

RuH2(PPh3)4 in benzene has been shown to react with ethylene and styrene

and said to yield the complexes Ru(C2H4)(PPh3)3 and Ru(PhCH=CH2)(PPh3)3(25)

together with equimolar amounts of ethane and ethyl benzene according to

the following equations:

RuH2(PPh3)4 + 2CH2.CH2 Ru(C2H4)(PPh3)3 + PPh3 + C2116

RuH2(PPh3)4 + 2PhCH.CH2 Ru(PhCH.CH2)(PPh3)3 + PPh3 + EtPh

RuBCl(PPh3)3 + Bm11 8(N2 )2

14

More recently(26)RuH2(PPh3)4 has been shown to initiate polymerization

ofm-substituted olefins of high e-value(27)

such as acrylonitrile and to

form unstable complexes with olefins of low e-values. A detailed kinetic

study of the polymerization of acrylonitrile in dimethylformamide (dmf)

solution was reported. On the basis of kinetic results and the

independence of molecular weight of the polymer of time, a reaction

mechanism involving a slow initiation step followed by rapid propagation

and termination steps is proposed:

RUH2(PPh3)4 + dmf RUH2(dmf)(PPh3)3 + PPh3

RuH2(dmf)(PPh3)3 + AN

RuH2(AN)(PPh3)3

RuH2(AN)(PPh3)3 + dmf

INSERTION

n(AN)

PROPAGATION

TERMINATION

(CATALYST INACTIVATION)

SCHF151 3

AN - MONOMER MOLECULE

Such a mechanism is supported by the 1H and 31P nmr spectra of

RuH2(PPh3)4 in dmf solution which indicate the presence of a

cis-dihydridotris-phasphine species:

PPh3

dmf Ph3P

/ Ru / + PPh3

dmf

PPh3

RUH2(1)Ph3)4

When treated with vinyl acetate(20

RuH2(PPh3)4 liberates one mole of

ethylene per mole of complex and yields RuH(CO2Me)(PPh3)3, the suggested

reaction scheme being:

15

P P

P I , P vinyl acetate II* I .'

,.P .. Ru 0....--- Ru•

EL e' I H 0 . C H M I ‘''''' H i

P Me P

,O. I ,P

Me - C , ‘Ru

\Oee l .NH+ C2114

SCHEME 4

Me

C 0 s .13 01 1 Ru CH2„. 02.0/ 1 H

P

(P m PPh3)

The reaction of RuH2(PPh3)4 with hydrosilanes has been reported(29)

to give 'apparently' seven coordinate silyl ruthenium(IV) complexes of

the type RuH3(SiR3)(PPh3)3. Treatment of these compounds with excess

carbon disulphide gave the known complex Ru(S2CH)2(PPh3)2 according to

the reactions

-PhMe2SiH

RuH3(SiMe2Ph)(PPh3)3 RuH2(PPh3)3

2CS21 -PPh3

Ru(S2CH)2(PPh3)2

A recent study(30)has demonstrated that the double bond migration

in n-pentenes is inhibited by N2 in the presence of ruthenium catalysts

whose precursor is RuH4(PPh3)3. Spectroscopic and kinetic data are said

to indicate that the active ruthenium species are RuH2(PPh3)3 and

Ru(PPh3)3; the inhibition by N2 is attributed to the ability of

molecular nitrogen to compete with the olefin for coordination to the

ruthenium.

It has been reported(17)that a reaction using half of the quantity

of triphenylphosphine normally required for the preparation of

RuH2(PPh3)4 resulted in the isolation of a brown powder, which was

formulated as RuH2(PPh3)3 on the basis of analytical data and its

reactions with formaldehyde, triphenylphosphine,

N-methyl-N-nitrosotoluene-17sulphonamide to yield RuH2(C0)(PPh3)3,

16

RuH2(PPh3)4 and Ru(NO)2(PPh3)3. It is interesting that the authors found

no evidence of the production of the dinitrogen complex,

Rua2(N2)(PPh3)3, under these reaction conditions.

The interaction of RuH2(PPh3)4 and tritylhexafluorophosphate leads

to [RuH(PPh3)41PF8 by abstraction of a hydride ligand as H-. This salt

reacts with nitriles to give six-coordinate cationic hydrido species of

formula, [RuH(RCN)2(PPh3)311,F8. On standing in dichloromethane the red

tetrakisphosphine species is converted to a yellow complex of formula

[RuH(PPh3)3]PF8; the tetrafluoroborate salt was also isolated(31) On

the basis of 1H n.m.r. data Sanders(31)suggested that one, of the

phenyl rings of one triphenylphosphine group was bound to the metal as

an arene. Other methods of synthesis of the salt [RuH(T6-Ph-PPh2)(PPh3)2>F4

together with confirmation of its structure, both in the solid state by

X-ray crystallographic studies, and in solution by 31P n.m.r. studies,

have been briefly reported(32)and are discussed more fully in Chapter I

of this thesis.

Wilkinson and co-workers(33)have shown the complex RuC12(PPh3)3in

benzene/ethanol to be an effective homgeneous hydrogenation catalyst.

They showed that the catalytically active species was RuHC1(PPh3)3

produced in situ by base promoted hydrogenolysis. In the mixed solvent

system ethanol acted as base but the catalyst was more conveniently

studied in pure benzene using bases such as triethylamine and sodium

phenoxides

RuC12(PPh3)3 H2 + base -0. RuHC1(PPh3)3 + base • HC1

The intermediate hydrido-species was detected in solution by IH n.m.r.

but not isolated. Subsequently(34)the pure complex was isolated as the

benzene solvate and its chemical and catalytic properties studied.

RuHBr(PPh3)3 was prepared analogously from RuBr2(PPh3)3. Benzene

solutions of RuHC1(PPh3)3 were shown to reac'. with 2,2'-bipyridyl and

norbornadiene to produce RuHC1(bipyr.)(PPh3)2 and the hydrido-alkene

complex RuHC1(C7H8)(PPh3)3 respectively. The reaction of RuC12(PPh3)3

with sodium borohydride in methanol or ethanol was found to yield

RuH2(C0)(PPh3)3. The catalytic behaviour of RuHC1(PPh3)3 was summarized

as: (a) extremely effective and specific for the hydrogenation of

linear alk-1-enes; (b) rapid hydride exchange with both alk-1-enes

and alk-2-enes; and (c) slow isomerization of alk-1-enes and

non-isomerization of other alkenes. The high selectivity for alk-1-enes

17

was attributed to steric interaction(35) of the bulky triphenylphosphine

groups in an alkyl formation step. The X-ray analysis(36) has shown that

RuHC1(PPh3)3 has a considerably distorted trigonal bipyramidal structure.

The hydridochloride reacts with aldehydes losing hydrogen to form

complexes which contain a it-bonded acyl group(37) Products of

formula RuCl(CO)(rr-COR)(PPh3)2 have been obtained virtually quantitively

from reactions of either RuHC1(PPh3)3 or RuHC1(C0)(PPh3)3 with

proprionaldehyde and acetalehyde according to the equations s

RuHC1(PPh3)3 + 4EtCHO RuCl(C0)(17-COEt)(PPh3)2

+ C2H4 + 2Pr OH + PPh3

RuHC1(PPh3)3 + 3MeCHO RuCl(C0)(7/-COMe)(PPh3)2

+ CH4 + Et0H + PPh3

RuHC1(C0)(PPh3)3 + 2RCHO---0. RuC1C0)(rt-COR)(PPh3)2

+ RCH2OH +,PPh3

Similar complexes are produced by the reaction of RuHC1(C0)(PPh3)3

with anhydrides s

RuHC1(C0)(PPh3)3 + (RCO)20 RuCl(C0)(?Y-COR)(PP/13)2

+RCO2H + PPh3

Sulphur dioxide reacts with RuHC1(PPh3)3 in n-hexane to yield one

of the few known sulphur dioxide hydrido complexes, RuHC1(S02)(PPh3)3 (38)

The action of carbon monoxide on RuHC1(PPh3)3 in n-hexane gives

cis-RuHC1(C0)(PPh3)2. Cenini et al(38)have carried out a series of

reactions between RuC12(PPh3)3 and RONa (R a Et, Pri) and have produced

hydrido complexes according to the following equations s

RuC12(PPh3)3 + RONa + PPh3 -42-• RuH2(PPh3)4

RuC12(PPh3)3 + RONa Rull N PPh 2(-2)(---3.) 3

RuC12(PPh3)3 + RONa Ar RUH2(C0)(PPh3)3

18

The diphenyltriazenido anion, Ph2N3 , has been reacted with

RuHC1(PPh3)3 39)to yield the hydrido complex, RuH(N3Ph2)(PPh3)3 by

displacement of the chloride ion s

RuHC1(PPh3)3 + Ph2N3 • RuH(N3Ph2)(PPh3)3 + Cl-

this complex has been further reacted(40)by a process of 'reverse osmosis'

to yield the complex RuH(N3Ph2)(N2)(PPh3)2 s

RuH(N3Ph2)(PPh3)3 N2,-----

THF• RuH(N3Ph2)(N2)(PM3)2

-PPh 3

The acetato and other carboxylato complexes RuH(RCO2)(PPh3)3 have

been obtained by interaction of the dichloride, RuC12(PPh3)3 with the

sodium salt of the carboxylic acid RCO2H, in methanol solution under an

atmosphere of hydrogen(41). There is no necessity to use molecular

hydrogen and the solvent, methanol, can act as the hydride source - an

improvedswihesis following this procedure is described in this thesis.

These complexes have also been prepared by (a) the reaction of RuH2(PPh3)4

with the appropriate carboxylic acid in boiling 2-methoxyethanol(42)

and (b) the rapid successive addition of ruthenium trichloride to a

solution of triphenylphosphine in vigorously boiling ethanol(42). As

already noted RuH2(PPh3)3 reacts with vinyl acetate and vinyl propionate

to yield RuH(CO2Me)(PPh3)3 and RuH(CO2Et)(PPh3)3 respectively(28). The

reaction of the dihydrido complex, RuH2(C0)(PPh3)3 with weak carboxylic

acids in 2-methoxyethanol to form hydridocarboxylato derivatives,

RuH(CO2R)(C0)(PPh3)2 has also been reported(42'b). Wilkinson et al(41)

showed that the complexes, RuH(CO2R)(PPh3)3, are effective catalysts for

the selective homogeneous hydrogenation of alk-1-enes, and studied the

rate of hydrogenation of hex-1-ene with the complex RuH(CO2CF3 )(PPh3)3.

The positions of the symmetric (7) OCO) and asymmetric (7,'as

ymOCO) sym

stretching vibrations of the carbagOate group and the value of

t1(7)as s 000-7;

y OCO) were found to be consistent with a symmetrical

bidentate chelate arrangement of the carboxylate group. The X-ray

crystal structure(43)shows that the complex is monomeric and the metal

atom has a highly distorted octahedral coordination. The three

triphenylphosphine ligands are meridional and the hydride hydrogen atom

is cis to all three. The acetate group is bidentate and weakly held, with

two rather long Ru-0 distances (2.198 and 2.210 10. Methanolic

suspensions of RIIH(CO2Me)(PPh3) were shown(41)to react with 2,2'-bipyridyl

19

or 1,10-phenanthroline to yield complexes of stoichiometry

RuH(CO2Me)(PPh3)2L. The molecular weight of the 1,10-phenanthroline

complex indicated a dimer and an acetate bridged structure was proposeds

Me 1

/ 1 N PPh3 / I

1

N 1 0---C-- 0, -... o,- . ,, . , Ru Ru

H/ I \ 0--- C ' n / 1 NN

PPh3 I

N--""

Me

In 1973 Wilkinson et al(44)reported the reduction of Ru30(CO2Me )6(PPh3)3 by both chemical and electrochemical means and the

use of the reduced species in acidified methanol solutions as hydrogenation

catalysts for alkenes. At the same time they discussed the formation of

similar catalytic species by the action of acids on the complexes

RuH(CO2Me)(PPh3)3, RuH2(PPh3)4 and Ru(002Me)2(PPh3)2.

The purpose of the work described in this thesis has been to

investigate the mechanisms of :,hese protonation reactions and the nature

of the catalytically active species.

20

CHAPTER I

21

CHAPTER I

HYDRIDO- AND CARBOXYLATO- TRIPHENYLPHOSPHINE COMPLEXES OF RUTHENIUM(II),

THE ISOLATION AND CHARACTERIZATION OF CATIONIC COMPLEXES DERIVED FROM

THEM.

1.1. INTRODUCTION

Previous work(44645)has shown that solutions active in the

catalytic hydrogenation of alkenes can be made in methanol by the action

of non-complexing strong acids, such as aqueous fluoroboric,

trifluoromethylsulphonic or 2-toluenesulphonic on the complexes,

RuH(CO2Me)(PPh3)3, Ru(CO2Me)2(PPH3)2 or RuH2(PPh3)4. The proper

characterization of these catalytically active species proved difficult

due to their low solubility and transient nature but it seemed that

they were cationic. It was also noted that on treatment with molecular

hydrogen or with carbon monoxide in the absence of alkene, or on

standing the 'orange/red' catalytically active solution was converted to

an inactive yellow species.

The work described in this chapter deals primarily with studies

into the nature of the catalytically active intermediates together with

the isolation and characterization of several derived species.

One of the most compelling reasons for these additional studies

on the 'acid system' was the observation(46)during 1972 that the

addition to the yellow/orange acidic methanol solutions of

RuH(CO2Me)(PPh3)3 under hydrogen (but not nitrogen or argon) of

saturated hydrocarbons caused a relatively fast (5-10 minutes) change

in colour to purple. However, after cleaning of the reactor vessel

with chromic-nitric cleaning mixture, we were unable to reproduce the

phenomenon despite numerous attempts using different samples of the

components of the system, additions of traces of other elements, free

radical initiators and inhibitors, action of light etc.. Several attempts

to initiate this reaction are described in this chapter and are

generally concerned with mechanisms for the generation of free radicals.

Previously unpublished results of the rates of hydrogenation and

extent of isomerization of alkenes by the 'acid system' are also

reported.

22

1.2. CATALYTIC HYDROGENATION BY PROTONATED METHANOLIC SOLUTIONS OF

Ruii1202,LjiriePPlial3 IN THE PRESENCE OF TRIPHENYLPHOSPHINE.

Several studies of homogeneous catalytic hydrogenation using

carboxylatotriphenylphosphine complexes of ruthenium and rhodium have

been reported, mainly by Wilkinson and co-workers. Catalyst precursors

used have included: RuH(CO2R)(PPh3)41)where (R=Lie,Et,Pr1,Ph,o-HO.C6H4.); \ RuH(CO2Me)(PPh3)3 and Ru30(CO2Me)6(PPh3)3(44) ; Ru(CO2Me)2(PPh3)2 and

Rh(CO214e)(PPh3)3(47). The hydrogenation rates have in general been

determined with the pure catalyst precursor in benzene solution or with

a protonated species in a polar solvent, such as methanol. Homogeneous

catalytic hydrogenation is typically a multistep process as exemplified

by the widely accepted mechanism of the RhCl(PPh3)3 catalysed hydrogenation

of alkenes(48) Such processes are characterized by the nature of the

catalytically active complexes and intermediates, which are typically

labile species coexisting in several forms related through dissociative

equilibria. Such lability is an essential feature of catalytic activity,

prime requirements of which are facile incorporation of reactants into

the coordination sphere, and elimination of products from the coordination

sphere, through substitutional or dissociative processes. Coordinative

unsaturation is also an important feature of catalytically active

species. The coordinatively unsaturated species present in the system

under study are cationic and are apparently produced by the initial

removal of the acetate ligand via protonation followed by phosphine

dissociation. The large variations in relative hydrogenation rates

between various types of substrate for different protonated catalyst

systems (Table 1.1.) is doubtless an illustration of the subtle and

complex interrelation between electronic and steric factors, probably in

alkene and alkyl intermediate formation. Table 1.1. shows relative

hydrogenation rates for a range of unsaturated substrates under standard

hydrogenation conditions, compared with rates for similar types of

substrate obtained in three other studies.

23

Table I.

Hydrogenation rates expressed as gas uptake in as. mid: measured

at 40cm.Hg. pressure at 4000.

ALKENE HYDROGENATION

A(41)

PREVIOUS

RATES

STUDIES

B(44) c(47)

(1) (2)

pent-l-ene 37.0 30.7

hex-1-ene 87.9 21.70 48.0 36.0 7.6

kept-l-ene 57.1 16.38

oct-1-ene 76.1 12.0

dec-1-ene 11.3

undec-1-ene 20.90

cyclohexene 0.9 1.26 (0.1

cycloheptene 2.0

cyclooctene 2.5 0.54 31.9 28.0 4.7

cyclododecene ( 0.5

hex-2-enes ( 0.5 (0.10 ( 0.1 e 0.1 e 0.1

hex-3-enes e 0.5 (0.10

cycloocta-1,3-diene 4.2 ( 0.1 e 0.1 1.8

cycloocta-1,5-diene 160.6 58.8 51.5 40.7

hexa-1,5-diene 128.5 4.45 28.9

hexa-1,4-diene 22.9 1.26

hex-1-yne (0.5 e0.10 (0.1

2-methylpent-1-ene (0.5 (0.1

2-ethylhex-1-ene e 0.5 e 0.1

..i

24

The standard hydrogenation conditions used were 103 M methanolio

solutions of RuH(CO2Me)(PPh3)3 in the presence of Ru:H* and Ru:PPh3

(total) ratios of 1:4 and 1:12 respectively. The protonating acid was

11-toluenesulphonic monohydrate and hydrogenation rates were measured as

gas uptake (mis. min.1 ) at 40am.Hg. hydrogen pressure and 40°C.

Conditions quoted for the other studies were :

A(41). Relative rates of hydrogen consumption (ml. mina) at 50cm.Hg.

hydrogen partial-pressure for substrates under standard conditions;

catalyst RuH(CF3CO2)(PPh3)3 0.625 mmole in benzene, substrate concentration

1.25M. No temperature was quoted.

-1 B(44)(1). Ruthenium concentration 7.5 X 10 4 mole 1; hexene, Imole I 1 ;

rates measured at 40am.Hg. hydrogen pressure; methanol at 40°C using

aqueous fluoroboric acid, Ru:H* of 1:4. (2).Catalyst, 7.5 X 10-4 mole 171 Ru from electrolytic reduction of Ru30(CO2Me)6(PPh3)3 Ru:PPh 3 a 1:2 ;

hexene Imole 11 ; 40°C.; H2 pressure 40cm.Hg.; methanol plus 4e, aqueous

HBF4.

C(47). Ru(CO2Me)2(PPh3)2, 10 3 M; alkene, 1M; solvent (methanol), 50m1.;

hydrogen pressure, 400 Torr, 2-toluenesulphonic acid (H':Ru) 10:1 at 40°C.

As with all the systems quoted in Table 1.1. the hydrogenation rates

for linear terminal alkenes are high but no direct relationship is

obvious between rate and chain length. Under standard hydrogenation

conditions very little competitive isomerization was detected.

Table 1.2. shows the initial composition of alkene substrates

together with product compositions after 10 minutes hydrogenation under

standard conditions. The results are also expressed as percentage

hydrogenation and percentage of isomerization of the initial terminal

alkene. Expressed in this way it can be seen that the relative percentages

of alkene undergoing hydrogenation and isomerization are of the same order

for the three alkenes. The ratio of hydrogenation to isomerization and the

total amounts of substrate reacted after 10 minutes (Table 1.2.), however,

do not exhibit simple relationships with the initial hydrogenation rates,

measured as gas uptake (Table 1.1.).

25

Table 1.2.

Percentage of linear terminal alkenes hydrogenated and isomerized

after 10 minutes under standard hydrogenation conditions.

ALKENE SUBSTRATE

COMPOSITION

PRODUCT

COMPOSITION

HYDR

OGEN

ATIO

N

z o 1-1

IN

cn tl hi '6 m E4 li

r A t-44

hi N E-4

E-1 x eq rd A

5 I4

Hex-1-ene

Hept-1-ene

Oct-1-ene

98.88

98.77

99.66

40.01

1.12

0.21

1.11

0,11

0.13

78.02

77.92

70.66

72.20

61.78

61.50

1.31

1.34

3.41

3.36

3.26

2.95

20.67

20.72

25.92

24.44

34.95

35.55

19.78

19.83

26.13

24.63

34.94

35.54

1.31

1.34

2.32

2.27

3.04

2.75

Previous results(44'45)reported for hydrogenation using a catalyst

generated by the protonation of RuH(CO2Me)(PPh3)3 were obtained in the

absence of excess triphenylphosphine. At the same time it was reported

that in the presence of liquid alkene there was no change in the rate of

hydrogenation with time over a period of an hour or so. Under the quoted

hydrogenation conditions and in the absence of alkene a yellow inactive

species was obtained and formulated as [(Ph3P)2Ru(OH)2Ru(PPh3)21(EP4)2.

Repetition of this reaction has shown the isolated species to be a

mixture of [RuH(76-PhPPh2)(PPh3)21BF4 and [RuH(H20)2(Me0H)(PPh3)21BF4.

26

The catalytic system described here undergoes poisoning by

irreversible production of the catalytically inactive species,

[RuH(1/6-PhPPh2)(PPh3)21BP4. The poisoning effect can be demonstrated in

two ways: (a) by the relationship between 'conditioning time' and initial

hydrogenation rate; and (b) in the presence of alkene, by repressuring

the hydrogenation vessel and measuring the rate over several

hydrogenation runs.

'Conditioning time' was the time allowed for the solution to become

homogeneous between the additions to the acidified methanol of the

catalyst precursor, and the alkene substrate. During this time the

catalyst precursor was undergoing reaction to the inactive species in the

absence of alkene.

Table 1.3.

Relationship between 'conditioning time' and initial hydrogenation

rate for oct-1-ene under standard hydrogenation conditions.

Hydrogenation rate as gas uptake.

CONDITIONING TIME -

[mins.]

INITIAL HYDROGENATION RATE

[mls. milli.]

5 109.2

10 76.1

15 69.5

20 65.7

The effect of poisoning in the presence of alkene was demonstrated

by repressuring the hydrogenation vessel with hydrogen and repeating the

runt when the rates cbserved were lower than those found originally. Such

an experiment was carried out using cycloocta-1,5-diene. The standard

'conditioning time' of 10 minutes was allowed before the first addition

of the alkene, hydrogen uptake was then plotted for a period of 5 minutes. After a further minute the system was repressured with hydrogen and the

gas uptake again measured. This procedure was repeated to give a series

27

of 4 hydrogenation runs at effectively 6 minute intervals. The

hydrogenation rate for each individual run was calculated at a hydrogen

pressure of 40am.Hg.

Table 1.4.

Relationship between hydrogenation rate and time for

cycloocta-1 15-diene under standard hydrogenation conditions.

Hydrogenation rate as gas uptake.

HYDROGENATION

RUN

RELATIVE TIME

[mins.]

HYDROGENATION RATE

[mis. min:]

RUN 1 1 194.7

RUN 2 7 177.4

RUN 3 13 160.9

RUN 4 19 66.6

The production of [RuH(/16-PhPPh2)(PPh3)21HF4 is favoured by an

increase in triphenylphosphine concentration. The poisoning effect is

apparently opposed by the presence of alkene as noted previously(44)

A significant contribution to the reduction in hydrogenation rate for

successive runs for the cycloocta-1,5-diene experiment must result

directly from the decrease in alkene concentration. The decrease in

alkene concentration must also favour production of inactive species as

a result of a reduction in the probability of competitive reactions

yielding alkene and alkene intermediate complexes. Thus although the

mechanisms involved are complex two qualitative observations are apparent:

(a) that the catalyst system undergoes rapia poisoning in the absence of

alkene and the presence of excess phosphine; and (b) that in the presence

of alkene this poisoning is slower but is accelerated under hydrogenation

conditions as the alkene concentration decreases. Both of those

observations are consistent with the poisoning effect being the result of

the competitive and irreversible production of the catalytically

inactive species [RuH(16-PaPh2)(PPh3)2p3F4.

28

From the above data the behaviour of the high phosphine system as a

catalyst may be summarized as follows: (a) Rapid catalytic hydrogenation

of terminal alk-1-enes RCH.CH2 with very slow associated isomerization of

alk-1-enes under hydrogenation conditions; (b) very slow hydrogenation of

internal, cyclic or alk-1-enes of the type R1 R2C=CH2; and (c) rapid

hydrogenation of cycloocta-1,5-diene and hexa-115-diene.

Two specific points worthy of note are apparent from the measurements

of the initial hydrogenation rates: (i) that the rates for linear

terminal alkenes were exceeded only by those for hexa-1,5-diene and

cycloocta-1,5-diene, whereas the rates for hexa-1,4-diene and

cycloocta-1,3-diene were relatively low; and (ii) that the rate of

hydrogenation of cyclooctene was low and of the same order as that for

cyclohexene and cycloheptene whereas in other studies on protonated

catalyst systems the rate for cyclooctene has been reported as much

higher than that for cyclohexene and of the same order as that for

hex-1-ene.

The high rate of hydrogenation for hexa-1,5-diene is probably

largely a concentration effect as a result of the substrate containing

two linear terminal alkene bond:, i.e. the alkene concentration is

effectively doubled. Also if, following hydrogenation of one double bond,

the uptake of hydrogen by the complex was sufficiently fast, the second

double bond of the same molecule would still be in close proximity to the

active catalyst and available for coordination and reaction. It might be

possible to estimate the relative importance of such processes by

determining rates of production of hex-1-ene and hexane. Unfortunately,

no measurements of this type were made. The position may, however, be

further complicated by the possibility of chelate formation by

hexa-1,5-diene. The possibility of chelate formation has previously been

suggested(45)as an explanation for the rapid hydrogenation of

cycloocta-1,5-diene by similar catalyst systems. As the rates observed

for this substrate are vastly in excess of those for both internal

monoalkenes and conjugated dienes (cyclooctene and cycloocta-1,3-diene),

it would appear probable that the 'chelate effect' in terms of the

favourability of complex formation and the stereochemistry of the

resulting alkene complex are important factors. Results obtained with

other systems(44'45)indicate that the cycloocta-1,5-diene undergoes rapid

hydrogenation at both aouble bonds to yield cyclooctane. Such systems

29

were also shown to rapidly hydrogenate cyclooctene but not cyclohexene.

As the system under consideration here only hydrogenates cyclooctene

very slowly it would prove difficult to suggest a mechanism for

hydrogenation of both double bonds in cycloocta-1,5-diene and it appears

probable that only one double bond is hydrogenated.

A rigorous kinetic study of the RhCl(PPh3)3 catalysed hydrogenation

of alkenes has recently been presented(49). No such studies have been

attempted with the cationic catalytic systems produced on protonation of

the carboxylato complexes of ruthenium. Indeed results in this thesis

demonstrating the interrelation, of dissociative equilibria, and the

irreversible production of catalytically inactive cationic species

leading to progressive poisoning of the catalyst system indicate that

such a study would be impossible. It is the poisoning process which in

fact makes the system of relatively little interest as a catalyst, and

thus attention was focused on the mechanisms by which the inactive

species were produced, the nature of the intermediates involved, and

consideration of methods of initiating the saturated hydrocarbon reaction.

Ethylene and isobutylene were hydrogenated by acidified solutions

of RuH(CO2Me)(PPh3)3. Of necesrity, the reaction conditions varied

considerably from those used for the experiments with liquid alkenes.

A similar hydrogenation apparatus was used with the addition of a

300 ml. gas reservoir bulb. Methanolic solutions (50 ml.) containing

0.1 mmole RuH(CO2Me)(PPh3)3, 1 mmole triphenylphosphine and 0.5 mmole

ml....toluenesulphonic acid were prepared and stirred rapidly at 40°C for

ca. 10 minutes under an atmosphere of hydrogen (900 ml. at N.T.P.).

The gaseous alkene (300 ml. at N.T.P.) was introduced into the system

and the gas uptake measured. Initial gas uptake rates appeared

relatively slow (the hydrogenation products ethane and isobutane are

also gases). The initial rate for ethylene (measured at a hydrogen

pressure of 30 cm Hg) was approximately twice that for isobutylene.

Total gas uptake volumes (measured after ca. 18 hours) indicated

almost complete hydrogenation of the alkenes, suggesting that in the

presence of such gaseous alkenes catalytic activity is -retained

for a considerable time.

30

1.3. ATTEMPTS TO INITIATE THE ACTIVATION OF SATURATED HYDROCARBONS.

Attempts to initiate the reaction with saturated hydrocarbons,

typically hexane, to yield the 'purple' species have mostly been

concerned with the generation of free radicals. It has been reported(5o)

that polyethylene, polypropylene and polytetrafluoroethylene are

catalytically active in the oxidation of tetralin, and it has been

suggested(51)that the process involves a sequence of autoxidation

reaction steps initiated by polymer surface free radicals.

The normal hydrogenation technique involves a magnetic follower

(which may have been either polypropylene or polytetrafluoroethylene

covered at various times). After many experiments using aqueous

fluoroboric acid the reaction vessel was severely etched and it was

considered possible that the grinding of the stirrer against the wall of

the vessel and the consequent cleavage of C-C bonds might generate allyl

and alkyl radial groups at the polymer surface(52)

A series of experiments was conducted in which freshly ground

samples of various polymers were added to catalytic solutions prepared

under standard hydrogenation conditions. No definite or consistently

reproducible results were obtained although on several occasions a slight

darkening in colour to red was observed- no solid products were isolated

from these reactions. It has been reported(51)that tetralin oxidation was

significantly affected by the time during which ground polyethylene was

aged in air at room temperature prior to reaction. In view of this it was

decided to attempt production of the free radicals in situ in the reaction

flask. In order to do this reaction vessels were made with ground glass

adhered to the inner surface and set up such that samples of polymer

attached to a high speed mechanical stirrer could be brought into contact

with the abrasive ground glass during the reaction. Experiments were

carried out over ranges of temperature, reactant concentrations and

stirring rates using several types of polymer, but it proved impossible

to generate a purple solution. Trace amounts of additives used in the

fabrication of polymers were also added to reactions. Such additives

include amines and amides, titanium dioxide, and 2,6-ditertbutyl-4-n-

butylphenol. No purple solutions were generated in the presence of such

additives.

31

Unrelated to the possible effects of polymers other unsuccessful

attempts to initiate the saturated hydrocarbon reactions included the

addition to standard hydrogenation runs of (a) platinum black and

(b) tungsten hexamethyl. Reactions were also carried out in hydrogenation

vessels which had been 'conditioned' by refluxing platinum black and

tungsten hexamethyl in acidified methanol, but were not successful in

producing purple solutions.

1.4. PROTONATION OF RUTHENIUM(II) COUPLEXES TO YIELD

1RuH(1/6-PhPPh2)(Pa3)21BP4 AND [RuH(H20)2(Me0H)(PPh3)21BF4.

I.4(i). Hydrido(T6-ohenyl-diohenylohosohine)bis-(triohenylohosohine)-

ruthenium(II)tetrafluoroborate.

The reaction of RuH(CO2Me)(PPh3)3 with a large excess of aqueous

fluoroboric acid in methanol containing excess triphenylphosphine yields

pale yellow needles of the salt [RuH(PPh3)3]13F4, (A). X-ray crystal

structure (see APPENDIX I) and 31Pn.m.r. studies on this species(32)have

confirmed a previous(31)suggestion that one of the phenyl rings of one

triphenylphosphine group is bound to the metal as an arene.

The hydridoacetate is almost insoluble in cold methanol, but on

addition of aqueous (.14) fluoroboric acid and heating to reflux an air

sensitive red solution is produced. Under an atmosphere of argon, nitrogea

or hydrogen, this red solution is short-lived, decaying rapidly to a pale

yellow solution from which (A) is obtained as yellow needles on cooling.

The same product is obtained from the reaction of fluoroboric acid with

(a) RuH2(PPh3)4 or (b) Ru(CO2Me)2(PPh3)2 in the presence of

triphenylphosphine. The use of CF3S03H or 2-MeC6H4S03H as protonating

acid has resulted in the isolation of yellow solids, which have 1 H n.m.r.

spectra similar to that of (A), indicating that these salts also contain

the [RuH(76-PhPPh2)(PPh3)21+ cation. The 1 H n.m.r. spectrum of (A) in

4-acetone shows: (a) a broad absorption band T- 2.5-3 due to the aromatic

phosphine protons; (b) two triplets at 4.60 and 5.68 which are assigned

to the ortho- and meta- protons on the arene ring 76-bonded to the metal;

and (c) a hydride resonance centred at 7 18.67 and split into a triplet by coupling with the two equivalent phosphorus atoms of the two normally

bonded phosphines 235 Hz) and further split by the unique phosphorus

of the 76-bonded phosphine ^'6 Hz).

32

(A) is also precipitated, on standing over several days, from

solutions containing catalytic concentrations of RuH(CO2Me)(PPh3)3

(ca. 163M) and aqueous fluoroboric acid (ca. 4 x 10ft). At catalytic

concentrations the red/orange solutions produced on protonation of

the ruthenium complexes undergo reversible colour changes to

yellow/orange with the passage of hydrogen. These systems remain

reversible and catalytically active over a period of several hours,

after which time the 'aged' yellow solution is produced and (A) is

slowly precipitated under either nitrogen or hydrogen.

RuH(CO2Me)(P) 3

HBF4/(P)

Ru(CO2)403)2(P)2 RUH2(1)4 I or

HBF4/(P) RuH4(P)3

HBF4/(P)

RED/ORANGE

[RuH(16-PhPPh2)(P)T. BF4 1,\\ N2 I 1 H2

Active catalyst for

=LOW/ORANGE the hydrogenation

of alkenes.

(P) Q (PPh3)

SCHEME 5 Methods for the production of the catalytically active

protonated species and subsequently (A) in methanol

solution.

The yellow solution from which (A) is obtained will not act as

a hydrogenation catalyst and does not react with carbon monoxide.

G.l.c. analysis of the solutions produced on protonation of

RuH(CO2Me)(PPh3)3 have shown them to contain methylacetate at ca. 850 of

the levels expected for quantitative conversion of the acetate in the

starting material. The reaction of (A) with excess sodium acetate in

methanol alone or in the presence of triphenylphosphine regenerates

RuH(CO2Ae)(PPh3)3.

(A) is moderately soluble in acetone giving a yellow solution, which

is not an active hydrogenation catalyst. The 1 H n.m.r. spectrum of this

acetone solution is unaltered by the passage of hydrogen, carbon monoxide

33

or ethylene. (A) is soluble in acetonitrile giving a lime green solution,

the high field 1H n.m.r. spectrum of which shows moderately resolved

multiplets in the regions 18-197"and 23-247'indicating the presence of

both (RuH(1-PhPPh2)(PPh3)21+ and [RuH(MeCN)2(PPh3)3r cations in solution.

Attempts to isolate {RuH(MeCN)2(PPh3)33BF4 from reactions of acetonitrile

or MeCE/HBF4 with (A) have, however, proved unsuccessful.

Hydridodiaouomethanolbis-(triphenylphosphine)ruthenium(II)-

tetrafluoroborate.

The reaction of a methanol suspension of RuH(CO2Me)(PPh3)3 or

Ru(CO211e)2(PPh3)2 with high concentration of aqueous fluoroboric acid in

the absence of excess triphenylphosphine leads to the precipitation of a

bright yellow crystalline solid (B). The i.r. spectrum of (B) shows two

pairs of bands at 3640, 3560 and 2060, 2020 cm;', which are assigned

respectively to 0-H and Ru-H stretching modes. Normal bands associated

with triphenylphosphine ligands and the tetrafluoroborate anion are also

present. The 'H n.m.r. spectrum of (B) shows a broad pattern assigned to

the aromatic protons on triphenylphosphine together with resonances

indicating the presence of small amounts of water and methanol, and a

high field signal centred at 18.87'consisting of a doublet of doublets

with very similar hydride-phosphorus coupling constants

(J ,̂32Hz, J "-'38Hz). The proton noise decoupled 31P n.m.r. spectrum of -EP1 HP2 (B) shows a pattern typical of an AB system in which A J and

P1 P2 'PI P2 centred at +55.77ppm. relative to free triphenylphosphine in

deuterochloroform. Selective decoupling of the phenyl protons only

produces a splitting of the pattern into an approximately symmetrical

quintet, with a mean phosphorus-hydride coupling for the central triplet

of 35.3 Hz. The available evidence suggests that (B) is of structure :

H

Me0H I PPh3 If . .- BF4 - Ru

. 1

H2O /I \PPh3

L

(L Q H2O or MeOH)

34

The sixth coordination site appears to be occupied by either a water or

methanol ligand. There are indications that (B) may be a mixture of

several species of general formula [RuH(H20)x(Me0H) (PP113)0F4, where

x + y - 3. Under normal reaction conditions the product appears to contain >9W.[RuH(H20)2(Me0H)(PPh3)2]BF4 , but clearly the water and

methanol ratios can vary. (B) was originally formulated as

bis-(p-hydroxybis-triphenylphosphine)rutheniumtetrafluoroborate(44)and

has also been isolated as a minor component of a mixture produced by the

reaction of AgBF4 and RuC12(PPh3)3 in methanol.

I.S. EXAMINATION OF THE CATALYTICALLY ACTIVE SPECIES.

Studies into the nature of the initial red species produced on

protonation have been hindered by the low solubility of RuH(CO2Me)(PPh3)3

in methanol, and by the transient nature of the intermediate. The rate of

production of (A) and hence the life-time of the red intermediate appears

to be dependent on acid concentration (more probably on the concentration

of BF4 ions), but relatively high acid concentrations are necessary in

order to rapidly dissolve significant amounts of catalyst.

Attempts to isolate the red species initially produced on protonation

both at the preparative and catalytic scales have been unsuccessful.

As it was originally thought that the 'red' species might be a

dication with at least two bulky triphenylphosphine ligands, attempts

were made to stabilize and precipitate this species using large

counter-ions(53) Typically, concentrated solutions of the red species

were prepared by protonation of RuH2(PPh3)4 or RuH4(PPh3)3 in methanol

or acetone. RuH(CO2Me)(PPh3)3 was not used in order to avoid interference

from acetic acid. The counter-ion was then added in large excess and the

solution cooled and reduced in volume in vacuo with rapid stirring.

Reactions with SiF62- and Fe(CN)62-both slowly precipitated yellow solids,

which were obviously not produced directly by precipitation of the active

intermediate.

On addition of PtC162- or PtC142- as the acid or potassium salt

respectively the red solutions were immediately converted to an intense

purple colour, and a microcrystalline lilac solid was precipitated

(in only ^'10% yield) - the solution then aging to brown even under argon.

35

The lilac solid was almost totally insoluble in most organic solvents

and only sparingly soluble in dichloromethane, in which the 31P n.m.r.

spectrum indicated that scrambling had occurred and some phosphine

ligands were coordinated to platinum. The number and complexity of the

far i.r. bands in the 11-C1 stretching region also suggested that chlorine

was bonded to both platinum and ruthenium and probably in some instances

bridging two metal atoms.

In further attempts to precipitate the red species halogenoboranes

derived from the closoborane anion B12 H122- were used. Typically solutions

of B12C112H2 or B 12Briz(NH4)2 were added to methanol or acetone solutions

of the protonated red species, or B12C112112 was employed as the protonating

acid. Results were not consistently reproducible, but, on occasions,

immediate precipitation of red crystalline solids was achieved. The

solids once precipitated were virtually insoluble in all common solvents.

The i.r. data were inconclusive showing a weak broad hydride band and

variable solvent content. Elemental analyses indicated that the products

contained triphenylphosphine and the halogenoborane anion in the ratio of

approximately 5:1 ; and it would seem probable that the products isolated

were mixtures of various combinations of two bis- and/or tris-phosphine

hydrido ruthenium unipositive 0,ttions with one halogenoborane

counter-ion,tRuH(PPh3) n (S)

m 1'

2 [B12X1212-. There was i.r. evidence to

suggest that even the products derived from immediate precipitation

contained some DlaH(,76-PhPPh2)(PPh3)21+ cation.

Attempts to isolate the red species using large unipositive

counter-ions such as BP4-1 PF6 1 and BPh4 have also proved unsuccessful.

Three approaches have been adopted in attempts to surmount the

problems involved with the study of the red species in situ and the

isolation and characterization of derived species: (a) the use of long

chain carboxylates in order to increase solubility of catalyst precursor

eg.RuH(CO2R)(PPh3)3 where (R = -C"H35, -CH•02H5C4H9); (b) the use of

RuH2(PPh3)4 and RuH4(PPh3)3 as the catalyst precursor; and (c) the use of

alternative solvents; predominantlyacetone.

36

I.5(i). Protonation of complexes containing long chain carboxylates.

RuH(CO211)(PPh3)3 with (R s -C17H35, -CH.C2H5.C4H9) were made and

found to be sufficiently soluble in methanol or acetone at room

temperature to enable ready observation of the hydride 1H n.m.r.

resonances. The observed high field signals are symmetrical quartets

centred at's,28T'and having phosphorus-hydride coupling constants of--,26Hz.

These data are in good agreement with those obtained previously(41'42) for RuH(CO2R)(PPh3)3 complexes. Protonation of RuH(CO2R)(PPh3)3 with

aqueous fluoroboric acid in methanol or acetone resulted in an immediate

intense red coloration and the disappearance of the high field signal.

No change in the n.m.r. spectrum was apparent after prolonged passage of

hydrogen although a slight lightening in colour of the solution was noted.

These data might have suggested that protonation involved the removal of

the hydrogen bound to the ruthenium in addition to the removal of the

carboxylate, but it appears probable that the non-observance of the high

field line is due to non-rigidity and exchange processes. (A) was isolated

from the 'aged' solution of the protonated hydridostearate complex.

I.5(ii). Protonation of hydrido species in acetone.

The red species produced on protonation of RuH(CO2Me)(PPh3)3 is longer

lived in acetone than in methanol, but it too decays to give a

yellow/orange solution. Again it has not proved possible to isolate the

initial red species. In acetone, however, the protonated species does not

act as a hydrogenation catalyst for alkenes, and only a very slight

colour change occurs on the passage of hydrogen. (A) cannot be isolated

from or detected in the 'aged' yellow/orange solution. An orange

crystalline solid (C) can be isolated from the 'aged' solution. I.r. and

1H n.m.r. spectra indicate that (C) contains a bidentate acetate group A N (V

'asymOCO 1500CM.

-1 y symOCO 1467cm. 1

,L033cm."*1 ) and is not a hydride

complex. In addition to normal bands associated with triphenylphosphine

and fluoroborate ion the i.r. spectrum also shows several bands

(3585, 3305, and 3195cm."1 ) which may be assigned to hydrogen bonded and

free 0-H stretching modes. Conductivity measurements in both acetone and

nitromethane indicate that (C) is a 1g1 electrolyte. This evidence

suggests that (C) is of formula [Ru(CO2Me)(H20)(PPh3)31 BF4 and that the

cation has structure 1

37

PPh3

,O I ,PPh3

Me — C Ru

N/ I OH2

PPh3

H2O

,0 , I PPh 3 .

Me — C‘. /Ru

0 PPh 3

PPh3

or

Structure (V) appears the more probable and would result from the

direct substitution of a water ligand for a hydride ion in the

hydridoacetate and as in that complex the cation would be expected

to display highly distorted octahedral coordination(43). This

formulation is supported by the fact that the water ligand in (C)

may be readily replaced by acetonitrile or carbon monoxide to yield

the orange crystalline salts [Ru(CO2Me)(MeCN)(PPh3)31BF4 and

[Ru(CO2Me)(C0)(PPh3)3p3F4. The action of sodium borohydride on (C)

in either acetone or methanol solution regenerates RuH(CO2Me)(PPh3)3.

(C) is soluble in acetonitrile giving an orange solution which

changes rapidly to yellow/green on addition of aqueous fluoroboric

acid, on standing a pale yellow crystalline solid (D) is precipitated

from this solution. The 1H n.m.r. and i.r. spectra of (D) indicate

the presence of acetonitrile, water and phosphine ligands together

with fluoroborate ion. There is no evidence of an acetate group.

Conductivity measurements are in the appropriate range for a 2;1

electrolyte of assumed formula [Ru(H20)(MeCN)2(PPh3)3](BF4)2. The

action of sodium borohydride on a methanol solution of (D) leads to

the replacement of the water ligand by hydride ion and yields the

complex [RuH(MeCN)2(PPh3)3]BF4(31)as white microcrystals. The 'H n.m.r.

high field signal of the latter species is a broadened symmetrical

quartet and it would therefore seem probable that the meridional

coordination of the three phosphine ligands is retained throughout

the reaction scheme RuH(CO2Me)(PPh3)3---10.[RuH(MeCN)2(PPh3)31BF4.

The 31P n.m.r. spectra of the complexes in the reaction scheme

(Table 1.5.) are consistent with meridional orientation of the

triphenylphosphine ligands in the tris-triphenylphosphine species.

The 31P n.m.r. spect:um of [Ru(CO2Me)(H20)(PPh3)31BF4 in deuterochloroform

is a single sharp line, 6 +54.01 ppm from triphenylphosphine in CD013, suggesting that dissociation of water leads to a non-rigid 5-coordinate

species in which the phosphorus nuclei are equivalent on the n.m.r.

time scale.

38

(C) is soluble in methanol giving an orange solution, which when

treated with aqueous fluoroboric acid under hydrogen gives immediately

yellow crystals of [RuH(H20)2(Me0H)(PPh3)3133F4.

[RuH(MeCN)2(PPh3)3]BF4 has been made previously by the action of

acetonitrile on [RuH(PPh3)4]73F4(31). When yellow suspensions of RuH(CO2Me)(PPh3) in either acetone or methanol are treated with

aqueous fluoroboric acid in the presence of acetonitrile they give

immediately yellow solutions, from which [RuH(MeCN)2(PPh3)3]IiF4 is

rapidly precipitated as white microcrystals.

The reactions of [Ru(CO2Me)(H20)(PPh3)31BF4 discussed above are

summarized in SCHEME 6. [Ru(CO2Me)(H20)(PPh3)3113F4 has also been reacted with a stoichiometric amount of lithium chloride, in refluxing

methanol under nitrogen. An orange solid was rapidly precipitated.

The i.r. spectrum of this species showed, in addition to normal

triphenylphosphine bands, a shift in the acetate stretching

frequencies (vasymOCO 1513cm.-1, sym000 1464cm.-1), and a band at -1 , which might be assigned to a gRu-C1)mode. There was

no evidence of the tetrafluoroborate anion. The i.r. data together

with elemental analysis (Table E.2.) suggested that the compound was

RuCl(CO2Me)(PPh3)3. This reaction has proved difficult to repeat.

I.r. data and the dark red colour of the products of several

experiments have indicated the presence of RuC12(PPh3)3 as a component

of a mixture.

When carbon monoxide is passed into the red solution produced on

protonation of RuH(CO2Me)(PPh3)3 in acetone (under either hydrogen or

nitrogen) there is an immediate colour change to pale yellow.

Concentration of this pale coloured solution in vacuo and addition

of degassed diethyl ether leads to the isolation of a white crystalline

solid. Infrared spectra of this species show bands characteristic of

the tetrafluoroborate anion and of triphenylphosphine and carbonyl

ligands. The carbonyl bands are at 2070 and 1990 am." the latter

being the stronger. There is also i.r. evidence for the presence of

acetone and water. The V(0-H) band is relatively sharp and centred at

3420 om:1. The crude product has bands at 1715 and 1655 cm:i which are assigned to free and coordinated acetone. On recrystallization from

39

CH2C12/Me011 the band at 1715 cm:1 disappears leaving a broad absorption between 1660 and 1610 cm:1 which is assumed to contain both the v(C-0) vibration of the coordinated acetone and the 6(H0H) mode of the water. Both the 'H and 31P n.m.r. data indicate the presence of two distinct species. The 1 H high field. signal in both d6-acetone and deuterochloroform consists of two separate triplets. In d6-acetone they are at 14.687 (JPH JPH 17Hz) and 14.15r( 19Hz) and. in deuterochloroform the signals are observed at 14.58r(J 18Hz) and 14.48r(JpH 20E1z). The relative intensity of the two signals varied. considerably for several samples made under similar conditions. The proton noise decoupled 31P n.m.r. spectrum in d6-acetone gives two lines (6 +49.52 and +48.51 ppm to high frequency of internal triphenylphosphine). Selective decoupling of all but the hydridic proton produced splittings in these lines of 17.6 and 18.8 Hz respectively. Conductivity measurements on the white species indicate that it is a 111 electrolyte, and on treatment with lithium chloride in acetone it yields white crystals of apparently the same isomer or isomeric mixture (i.r. data identical) of RuHC1(C0) 2 (PPH 3)2 as that reported by James et al(6b). it seems probable that the product of the passage of carbon monoxide through the red protonated species is a mixture of [RuH(S)(C0)2 (PPh 3 )2p3F4 (where S is water or acetone, withwater predominating) and a small amount of [Rull(C0)3(PPh3)21BF4(10).

40

RUH(CO2Me)(PPh3)3

P

O I ,P

Rua.. 0 f I NCO

MeCN via RED

HBF4(aq.) SOLUTION)

CO/MeCH CO

+

MeCN

P O, I ,P

Ru

0 I NeMe

MeCN MeCN

11 4(aq.) HBF4(aq.)

P

MeGN I ,P

Ru

MeGN It/ I NOH2

NaBH4/MeOH

iMeCN

RUH(PPh3),J+

Ph3C+BP47CH2C12

RUH2(Pn3)4

P = PPh3, 0----0 (CO2Me) .

SCHEME 6 Derivatives of [Ru(CO2Me)(H20 )(PPh3)3P3F4, complexes

isolated as tetrafluoroborate salts.

41

1.6. 31P N.M.R. SPECTROSCOPIC STUDIES ON TRIPHENYLPHOSPHINE

RUTHENIUM(II) COMPLEXES%

31P n.m.r. spectroscopy has proved to be probably the most powerful

single technique available for the study of the triphenylphosphine

ruthenium complexes discussed in this thesis. All spectra were proton

noise decoupled in respect of aromatic protons and in the case of

hydrido complexes, selectively decoupled in the region of the Ru-H high

field resonance.

The proton noise decoupled 31P n.m.r. spectrum of

[RuH(76-PhPPh2)(PPh3)2iBF4 in (CD3)2C0 shows a doublet (relative intensity

2, JPP 0.8Hz) at 48.91ppm. to low field and a triplet (relative intensity

1, Jpp 0.8Hz) at 5.77ppm. to high field of external H3PO4 reference.

Selective irradiation at the 1H absorption frequency of the aromatic

protons (ca.7.6ppm. to low field of Me4Si) led to the observation of a

major doublet splitting on each of these lines due to coupling with the

hydridic proton (low field line, JpH 33.3n, high field line, JpH 5.8Hz).

The appearance of the partially decoupled 31P spectrum is critically

dependent upon the centre frequency of the secondary irradiating field in

the broad, unresolved aromatic region of the IH spectrum. The doublet

line positions do not change significantly if this frequency is altered,

but the relative intensity of the components varies greatly. This

phenomenon has been discussed previously(32)and has been attributed to an

Overhauser effect involving a redistribution of the populations of

nuclear magnetic energy levels of phosphorus which are common to both

aromatic and hydridic proton spins. Subsequently, similar effects have

been observed in other hydridotriphenylphosphine ruthenium species, such

as [RuH(S)(CO)2(PPh3)2]BF4.

The 31P n.m.r. spectrum of [RuH(16-PhPPh2)(Pn3)21BP4 is consistent

with the solid state structure(32)which is discussed in APPENDIX I. The

chemical shift of the phosphorus in the arene-bound ligand is very close

(0.57ppm. to low field) to that in free triphenylphosphine added to the

solution of the compl3x in (CD3)2C0 which confirms that the bonding is not

via phosphorus since the two conventionally bonded ligands resonate at

55.25ppm. to low field of free triphenylphosphine. The observation of

spin-spin coupling between the nonequivalent phosphorus nuclei, and

between the unique phosphorus and the hydridic proton appears to indicate

that there is substantial electron delocalization from phosphorus into

the aromaticir-system and onto Ru. The unique phosphorus to hydride

42

coupling is 5.8Hz as opposed to 0.8Hz for the coupling between the

nonequivalent phosphorus nuclei. In the solid state the Ru-H bond is in

almost exactly the same plane as the bond from the unique phosphorus to

the arene-bonded ring. If the stereochemistry is unaltered in solution

this might account for the relatively large coupling between the unbonded

phosphorus and the hydridic proton as compared with the much smaller

phosphorus-phosphorus coupling.

The proton noise decoupled 31P n.m.r. spectrum of RuH(CO2Me)(PPh3)3

in CDC13 (Table 1.5.) is a typical AB 2 pattern (L144? 1p4 p214p4p3)°

Selective decoupling of all but the hydridic proton produces a further

splitting of the pattern and demonstrates that the hydridic proton to

phosphorus coupling constants to the unique phosphorus nucleus

(J10 27.8Hz) and to the two equivalent phosphorus nuclei

(112,413, 25.3Hz) vary by 2.5Hz. These data are consistent with the

1H n.m.r. spectrum of RuH(CO2Me)(PPh3)3 measured at 60 or 100MHz, in

which the high field resonance appears to be a well resolved symmetrical

quartet, produced by the close similarity of the hydride coupling

constants to the three cis-phosphorus nuclei. A 31P n.m.r. study of the

solution behaviour of RuC1TPh3)3, RuHC1(PPh3)3 and RuH(CO2Me)(PPh3)3

has recently been published 57). The proton noise decoupled 31P n.m.r.

spectra of RuH(CO2Me)(PPh3)3 in CH2C12 (in which it is catalytically

inactive) and in TEF (in which it is active) are both in close agreement

with that reported here in CDC13. There was no spectroscopic evidence for

free triphenylphosphine or any other complexes in any of the spectra.

Selective decoupling of the aromatic protons enabled determination of the

coupling of each phosphorus environment to the hydride ligand

(J 1 0 28.4Hz; =J._=25.8Hz). A 220 MHz spectrum of the 1H n.m.r. high j12 is field region (CH2C12, 20°C) of RuH(CO2Me)(PPh3)3 showed that the

amplitudes of the quartet peaks are not 1:3:3:1, the central peaks have

shoulders, identifying the pattern as a doublet of triplets with

deceptively similar JpH values. RuHC1(PPh3)3 and RuH(CO2Me)(PPh3)3 are

both active hydrogenation catalysts in THE but show no detectable

dissociation of phosphine. This would seem to illustrate that 'kinetically

significant concentrations' of catalytically active species (i.e. the

catalyst, as opposed to the catalyst precursor) may be undetectable even

using sophisticated n.m.r. techniques.

43

The 31P n.m.r. spectrum of [Ru(CO211e)(C0)(PPh3)31BP4 in CD3OD

consists of a triplet and a doublet (relative intensities 1:2, Jpp 20.5Hz)

with chemical shifts of 45.75 and 35.08ppm. to low field of PP1131 together with two less intense singlet lines (,,,Z abundance) at 49.33

and 39.55ppm., possibly due to five-coordinate species produced by ligand

dissociation. The 31P spectrum of [Ru(H20)(MeCN)2(PPh3)31(BP4)2 contains

as major components a triplet and a doublet (relative intensities

1:2, Jpp 25.6Hz) at 40.85 and 27.32ppm. to low field of free PPh3. The

spectrum also shows sharp singlets at 55.01 and 45.40ppm. to low field of

free PPh3 together with broadened resonances in the regions of both these

singlets and in the area of free PPh3. The proton noise decoupled

spectrum of [RuH(MeCN)2(PPh3)3]BF4 in CDC13 consists of a triplet and

doublet (relative intensities 1:21 Jiap 31.3Hz) at 63.39 and 52.58ppm. to

low field of free PPh3. Selective proton decoupling gave phosphorus to

hydride ligand coupling constants to each component of

—nr J„i 16.7Hz J 22.4Hz. There was no evidence of phosphine dissociation

or other species present.

Table 1.5.

"P n.m.r. spectra in CDC13. Shifts to high frequency from internal

PPh3 in CDC13; spectra proton noise decoupled from aromatic protons

and selectively decoupled in region of Ru-H high field proton

resonance; temperature 25°C.

COMPLEX 62= 63 64 124=434 112=313 114

RuH(CO2Me)(PPh3)3 +49.53 +85.74 27.4 25.3 27.8

[Ru(C0)(CO2Me)(PPh3)3]BF4(a) +35.08 +45.75 20.5

[RuH(MeCN)2(PPh3)3]BF4 +52,58 +63.39 31.8 22.4 16.7

Plu(420)(Me01)2(PPW3RBF4)2 +27.32 +40.85 25.6

(a)in CD3OD referenced to internal PPh3.

P2 X.... I ...P4

Ru' Y %Z1

X,Y = CO2Me, MeCN

Z = H, CO, H2O P3

44

1.7. CONDUCTANCE MEASUREMENTS ON IONIC COMPLEXES.

The problems involved in the determination of electrolyte type of air-sensitive ionized complexes by conductance measurements in solution in non-aqueous solvents have been discussed elsewhere(58'59)

The electrolyte type of the complex [RUE(76 -PhPPh2 )(PPh3)21BY4 was determined by a method(58)which is unaffected by molecular complexity in solution. The procedure involved the determination of equivalent conductance (Ae) over a range of dilute concentrations. Subsequent electrolyte type determinations for other complexes were made by taking conductance measurements at one concentration (10-3M) and calculating molar conductance (A) using an assumed molecular weight. These values were then compared withAra values for electrolytes of known type(59)

principally with tetrafluoroborate salts of similar ruthenium triphenylphosphine cations and in particular the values were referred toJkla of [RuH(76-PhPPh 2 )(PPh3)2>F4. Typically, determinations were made in redistilled, dry, analar acetone or nitromethane and the solutions were prepared and measurements made under nitrogen.

Table 1.6.

Molar conductances and electrolyte types of cationic ruthenium complexes.

, COMPLEX (km )MOLAR

CONDUCTANCE pmElam mole]

SOLVENT .

TEMP.

rC]

ELEC. TYPE

[Ehaq-PhPPh 2 )(PPh 3)21BF4•MeOH 76-83 MaNO2 23 1:1 108-115 Me2C0 26 1:1

Din(CO2Me)(H20)(PPh3)3]BF4 75-84 1461102 23 1:1 114-122 Me2C0 26 1:1

[liu(CO2Me)(COXPPh3)3iBF4 88-96 MeNO2 23 1:1 [Ru(O02Me)(MeCN)(PPh3)31BF4 86-93 MeNO2 23 1:1 [RUH(E120) 2 (Me011) (PPh 3 )2] BP4 115-127 Me2C0 25 1:1 [Eu(H20)(MeCN)2(PPh3)31 (13F4 )2 138-146 MeNO2 23 1:2 [PhiE(MeCN)2(PPh3)3iBF4 92-98 MeNO 2 23 1:1 kull(MeCN)2 (PPh3)31C1 96-102 MeNO2 21 1:1

1 [RuE(S)(C0) 2 (PPh3)21BF: 115-123 Me2C0 25 1:1

45

1.8. REACTIONS OF HYDRIDOCHLOROTRIS-(TRIPHENYLPHOSPHINE)RUTHENIUM(II).

When RuHCl(PPh3)3 is dissolved in acetonitrile it rapidly produces

a pale yellow/green solution from which the white microcrystalline solid

[Ra(kleCN)2(PPh3)3]Cl isreedily isolated. This complex has a 'H n.m.r.

spectrum similar to that of the analogous fluoroborate salt, but it

displays interesting solvent and temperature dependence. In methanol or

acetonitrile the complex shows a well resolved high field quartet

1'23.6 (JP-H22 Hz). Passage of hydrogen has no effect on the spectrum,

but addition of acetone gives an immediate red coloration and the

appearance of an additional somewhat exchange broadened high field

multiplet in the region 27-282, probably due to the formation of

RuHC1(PPh3)3. [RuH(MeCN)2(PPh3)3]Cl when refluxed in acetone gives a deep

red solution from which RuHC1(PPh3)3 is isolated.

RuHC1(PPh3)3 MeCN acetone [RuH(MeCN)2(PPh3)3]+ + Cl

In dichloromethane [RuH(MeCN)2(PPh3)31C1 is a red solution at room

temperature and above. At +40°C broad high field signals are detected in

the regions 23-24r and 27-28r. On cooling to -10°C, the solution becomes

yellow/green, a well resolved quartet is obtained centred at 23.5rand

the signal to higher field disappears.

[RuH(MeCN)2(PPh3)31C1 CH2 C12,, RuHC1(PPh3)3 + 2MeCN -10°C

The relationships between RuHCl(PPh3)3 and the cationic ruthenium(II)

species discussed in this chapter are clearly shown in SCHEME 7.

RuC12(PPh3)3 (i) ► RuH(CO2Me)(PPh3)3--...(ii)[Ru(CO2Me)(H20)(PPh3) 3 1HF4

j(iii) I(iv)

Racl(pph3)3 4S252__ [RuH(MeCN)2(PPh3)3JBF4

(vi)1 1(vii) (viii )

N ix) [RuH(MeCN)2(PPh3)310 [Ru(H20)(MeCN)2(1Th3)3](BF4)2 SCHEME 7

(i) MeCO2Na PPh3 Me0H; (ii) HBF4(aq.) + Me2C0; (iii) H2 + Et3N + C6H51.11e; (iv) HBF4(aq.) + MeCN + (Me2C0 or Me0H); (v) HBF4(aq.) + MeCN;

(vi) MeCN Me0H; (vii) Me2C0(reflux); (viii) KBF4(xs.) + Me0H;

(ix) NaBH4 + Me0H; (x) LiC1 + Me2C0.

46

1.9. DISCUSSION

In the studies of homogeneous hydrogenation of alkenes, using

RuH(CO2Me)(PPh3)3 as catalyst precursor, the catalytically active species

has previously(44)been assumed to involve bis-(triphenylphosphine)-

ruthenium(II) cations but the nature of the 'red' solutions was not

properly established. The main evidence in support of this suggestion

was the assumption that the same red solutions were obtained directly

from the protonation of RuH(CO2Me)(PPh3)3, RuH2(PPh3)4, or

Ru(CO2Me)2(PPh3)2 and the fact that no high field signal could be

detected in the 1H n.m.r. of the red solutions. The present work has

shown that the [111SOPPh3)21+ moiety is common to all products isolated

from the methanolic reaction mixtures and indicates that the hydride

ligand is probably retained throughout the reaction scheme.

When RuH(CO2Me)(PPh3)3 is protonated in methanol in the presence of

acetonitrile, or acetonitrile is added to the red solution produced on

protonation in the presence of excess triphenylphosphine, the product is

[RuH(MeCN)2(PPh3)31BF4. These reactions indicate that the first stage in

protonation of RuH(CO2Me)(PPh3)3 involves elimination of acetic acid to

yield a red cationic hydrido tris-phosphine species. The acetic acid is

evidently removed as methyl acetate, and g.l.c. study shows that methyl

acetate is produced essentially quantitatively.

r RuH(CO2Me)(PPh3)3 + LRuH(PPh3)3(5)1+ + MeCO2H

MeCO2H + Me0H

MeCO2Me H2O

The tris-species can dissociate so that in the absence of excess

triphenylphosphine the yellow bis-phosphine species is rapidly formeds

[RuH(PPh3)3(S)1+ [RuH(PP113)2(S)31+ + PPh3

where S may be methanol or water as in the isolated yellow salt )

[RuH(H20)2Me0H(PPh3)21BF4.

Sanders(31)suggested that the red colour of some similar ruthenium(II)

species is characteristic of 16-electron systems - [RuH(PPh3)3(S)1+ is

such a species. As a 5-coordinate species it would be expected to be

47

non-rigid and involved in dissociative equilibria with the yellow,

18-electron, octahedral species [RuH(S)3(PPh3)214. and [RuH(76 -PhPPh2)(PPh3)2r. Protonation of RuH(CO2Me)(PPh3 )3 in CD3OD has also shown that the hydridic proton undergoes hydrogen exchange with both

the solvent and the ortho-hydrogens of the phenyl rings on

triphenylphosphine. The formation of [RuH(76-PhPPh2)(PPh3 )2>F4 appears to be favoured by the presence of excess triphenylphosphine - under

similar reaction conditions in the absence of triphenylphosphine the

product in methanol of the protonation of both Ru(CO2Me)2(PPh3)2 and

RuH(CO2Me)(PPh3)3 is [RuH(H20)2(Me0H)(PPh3)21BP4. The reaction mechanism

for the protonation of RuH(CO2Me)(PPh3 )3 may be considered in terms of several competing equilibria:

RuH(CO2Me)(PPh3 )3 H+ [RuH(PPh3)3(S)r + MeCO2H

[RuH(H20)x(Me0H)y(PPh3)2])3F4 -- [RuH(PPh3)2(S)31+ + PPh3 (ii)

Me0H(reflux) (iv) 1 kin) PPh3(xs.)

SCHEME 8 Competing equilibria in methanol on protonation of

RuH(CO2Me)(PPh3)3.

(i) Slow dissociation of phosphine is probably the rate determining step

and is suppressed by excess PPh3.

(ii) is favoured by incr. in [H2O] and deer. [PPh3].

(iii)is favoured by incr. in [PPh3].

(iv) is a slow reaction.

In the presence of triphenylphosphine the red colour is both more intense

and longer-lived, and the overall reaction is slower. These data would be

consistent with the suppression of the phosphine dissociation step (i)

necessary for the formation of either product.

Possible mechanisms for the protonation of Ru(CO2Me)2(PPh3)2 and RuH(CO2Me)(PPh3)3 in methanol in the presence of excess

triphenylphosphine are s

[ RuH(76 -PhPPh 2) (PPh3)2] BF4

RuH(CO2Me)(PPh3)3

111+ [RuH(PPh3)3(S)r + MeCO2H

1 . I -PPh3

[RuH(PPh3)2(S)31+

Ru(CO2Me)2(PPh3)2

MeCO2H + [Ru(CO2Me)(PPh3)2(Or

iMe0H

MeCO211 + [Ru(0Me)(PM3)2(021+

48

+PPh3

[BuH(76-MPPh2)(Pa3)21+

-H2C0

+PPh3 [RuH(PPh3)2(S)3r

SCHEME 9

It is clear that if such a mechanism is operative for Ru(CO2Me)2(PPh3)2

in the.absence of excess triphenylphosphine, then the red species produced

on protonation is not the same as that obtained from the hydrido

complexes. Assuming that the first stage in protonation is again the

elimination of acetic acid, a possible 16-electron, red, species would

be [Ru(CO211e)(PPh3)2(S)r. Hydrida formation via an alkoxide intermediate

is a well-established pathway(4,54). The preparations under nitrogen of

RuH(CO2Me)(PPh3)3 and Ru(CO2Me)2(PPh3)2 in methanol and t-butanol

respectively, require the same stoichiometry of the reactants,

RuM2(PPh3)3 and NaCO2Me-3H20; and Ru(CO2Me)2(PPh3)2 is converted to

RuH(CO2Me)(PPh3)3 in refluxing methanol containing triphenylphosphine.

These reactions illustrate the facility with which such ruthenium species

abstract a hydride ion from methanol. The behaviour of RuC12(PPh3)3 in

several solvents has been reported(55); and its probable behaviour in a

polar solvent such as methanol would be consistent with the following

reaction scheme :

Ru(CO2Me)C1(PPh3)2(S)

-CO2Memethanol

-HC1

49

RUC12(PPh3)3

-PPh31 I

RuC12(PPh3)2

t-butanol

-C11 I-COglie-

Ru(CO2Me)2(1Th2)2

Me0H]

Ru(OMe)(0214e)(1Th3)2 -H2C0

11 14. Ile002H MeCO2Me + H2O

H

H—0 0

H-) I Ru(CO2Me)(PPh3) 2

•H2C0

HRU(CO2Me)(Pa3)2

443Ph3

HRu(CO2Me)(PPh )3

1+PPh3

HRu(CO2Me)(PPh3)2

SCHEME 10

Hydride formation is not possible in t-butanol as the alkoxide complex

is not capable oft -elimination to yield a hydride and an aldehyde.

The dissociative mechanism for the production of

[RuH(76-PhPPh2)(PPh3)2113F4 is supported by protonations of RuH2(PPh3)4

in aromatic solvents, which themselves may act asirr-bonding arene ligands.

In benzene or toluene the protonation results in an immediate red

coloration which is rapidly discharged to a pale yellow, from which

[RuH(116-arene)(PPh3)2iBF4 complexes are readily isolated (see CHAPTER II).

The involvement of the ortho-hydrogen and solvent exchange mechanisms in

the system is demonstrated by the preparation of [RuH(76-PhPPh2)(PPh3)2]BF4

in d4-methanol. The resulting complex has several i.r. bands not

present in the undeuterated species. There are bands at 632, 782, 851 and

874cm.-1 all of which have been observed previously in the products from

deuterium gas exchange studies(16,18,34,56 ) with RuHC1(PPh3)3,

RuH2(PPh3)4, RuH4(PPh3)3 and RhH(PPh3)3, and ascribed to out of plane

deformation vibrations of di- and/or trisubstituted benzene rings. A C-D

stretch is also observed at 2260cm."I (34), but the expected Ru-D stretch

(at 1465-1470cm.-11 assumingPRu1010 = 1.39) is obscured by a strong

50

band at 1454cm.-1 which also appears in the spectra of deuterated

RuH2(PPh3)4 and RhH(PPh3)4(16 ) . It is most likely an aromatic C-C

stretching frequency shifted by deuteration at one carbon atom.

• In acetone, as in methanol, protonatioll of RuH(CO2Me)(PPh3)3 in the

presence of acetonitrile or addition of acetonitrile to the initially

formed red solution yields the complex [RuH(MeCN)2(PPh3)31BF4, which

suggests that the first stage in reaction again produces the cation,

[RuH(PPh3)3(S)r.

It has been noted here and elsewhere(44)that the red protonated

solutions react immediately with carbon monoxide to produce a pale yellow

solution. In acetone a white crystalline solid, [RuH(S)(C0)2(PPh3)21BF4,

is readily obtained from this solution. In the presence of carbon

monoxide the phosphine dissociation is accelerated, and the 6-coordinate

dicarbonyl cation is stable even in the presence of excess phosphine. It

has proved more difficult to isolate a solid from the treatment of the

methanolic solution with carbon monoxide, but a white solid has been

obtained which has a similar i.r. spectrum to that of the acetone

product.

However, the 'aged' yellow/orange solution in acetone does not

contain a hydrido species but gives only [Ru(CO2Me)(H20)(PPh3)3]BF4.

Hence it appears that the slow overall reaction is s

[RuH(PPh3)3(S)1+ + MeCO2H [Ru(CO2Me)(PPh3)3(S)1+ + H2

In acetone there is no mechanism comparable with methyl acetate formation

leading to removal of acetic acid from the reaction sphere. A mechanism

involving re-attack by acetic acid is supported by the fact that

[RuH(Tf-PhPPh2)(PPh 3)21BF4 has been isolated from the protonation of

RuH2(PPh3)4 in acetone. The orthometallation(56)reaction is known to

occur in this type of ruthenium system and it seems likely that the

reaction of the coordinatively unsaturated red species might proceed via

such a mechanism s

I

-H2 .._ (ph3p)2Ru +H2 PPh 2

51

[RuH(PPh3)3(S)r

TICO2Me-

[Ru(CO2Me)(PPh3)3(S)r Ph3P)2(CO2Me)Ru.....,

PPh2

SCHEME 11

Reactions at catalytic concentrations and under conditions found to

give the optimum rates for catalytic hydrogenation of terminal alkenes

must be considered separately from those described for the preparation of

complexes (A), (B) and (C). Catalytic hydrogenations were typically

carried out on 1M solutions of alkene in methanol containing approximately

10311 concentrations of catalyst and Ruse, Ru:PPh3 ratios of 1:4 and 1:12

respectively. Under such conditions the reversible colour change

(red/orange --yellow/orange) associated with the introduction and

removal (by reaction, pumping, or passage of nitrogen) of hydrogen is

most pronounced.

These data suggest that the red protonated species, [RuH(PPh3)3(01+,

is in equilibrium with a lighter coloured species s

[RUH(PPhi)3(S)r

RED/ORANGE

H2 YELLOW/ORANGE

and that the equilibrium is forced to the right by molecular hydrogen. It

is this lighter coloured species produced in the presence of hydrogen

which is apparently active in alkene hydrogenation. Two alternative

equilibria, yielding different catalytic intermediates, have been

considered s

(a) [RuH(PPh 3 ) 3 (5)] + H2

(b) [RuH(PPh3)3(01 4. + H2

RuH2(PPh3)3(S) + H+

[RU.113(PPh3) 31+

(S)+ t [RuH3(PPh3)2(S)1+ + PPh3

52

Rigorous kinetic interpretation of the catalytic system would be

rendered extremely complex by the fact that a 'true catalyst' is not

operative in that the overall reaction,

RuH(CO2Ms)(PPh3)3 + BBF4 [11476-PhPPh2)(PPh3)21BF4 + MeCO2H

is irreversible under the reaction conditions. It has been demonstrated

previously(45)that the hydrogenation rate is acid concentration dependent,

the optimum rate being obtained at a Rue ratio of 1s4 and the rate

thereafter decreasing with increasing acid concentration. These data

might be considered to favour alternative (a) but it appears more probable

that the acid concentration (or in fact the BF4- concentration) is more

directly related to the precipitation of the insoluble and catalytically

inactive [RuOy6-PhPPh2)(PPh3)21BF4 species. The dependence of

hydrogenation rate on phosphine concentration is also probably more

directly related to its effect on the rate of production of the

arene-bonded species, than on any other process.

Mechanism (a) would yield a neutral dihydrido catalytic intermediate

RuH2(PPh3)3(S), which would be similar to the species identified as the

active catalyst in the polymerization of vinyl compounds in

dimethylformamide solution(26), RuH 2(PPh3)3(dmf). The low solubility of

the dihydrido species in methanol and the isolation, almost exclusively,

of cationic species from the methanolic reaction mixtures in the presence

of high acid concentrations would seem to suggest that mechanism (b)

yielding the cationic trihydrido species might be more reasonably expected

to be operative in the acidic medium. The catalytic reaction vessel is

designed such that under normal hydrogenation conditions at a pressure of

40cm.Hg. approximately 0.01 moles of hydrogen are available for reaction.

If the catalytic reaction mixture is stirred in a sealed vessel under a

total pressure of approximately three atmospheres of hydrogen the

solution becomes almost colourless.

A white solid was isolated on addition of hexane (20m1s.), with

extremely vigorous stirring, to the reaction of RuH(CO2Me)(PPh3)3 with

aqueous fluoroboric acid in methanol (50m1s.) in the presence of excess

triphenylphosphine. This product was shown by i.r. spectroscopy to be a

tetrafluoroborate salt containing phosphine ligands and having a strong

broad Ru-H stretching mode centred at 2000cm.-1 and no obvious solvent

bands. The complex was readily soluble in dichloromethane, acetone, and

methanol giving immediately deep red solutions. The red methanol

[Ru"H(P)3 (S)1+

[RUHH(V-PUTh2)(13)3r (iii) -H2+ (P)

(i) [RuwH3(P)3r • + H2

(ii) l- (P)

[ReH3(02(5)]+

/

(vi) + alkene

53

solutions Slowly turned yellow and [RuH(76-PhPPh2)(PPh3)2>F4was

precipitated. No hydride resonance could be detected in the 'H n.m.r.

spectrum of the red solution in 4-acetone. It appears probable that

vigorous stirring effectively increases the available hydrogen

concentration and forces the equilibrium to the rights

[ma(PPh3)3(S)1+ (s

H2 [RUE13(PPh3)3r )

The high concentration of hexane combined with the stirring effect probably aids precipitation of the ionic trihydrido complex,

[Fitild3(PPh3)3]BP4. When this complex is dissolved in a polar solvent

such as methanol, even under a static atmosphere of hydrogen, the red

[Ph1B(PPh3)3(S)]+ species is generated with elimination of hydrogen. The

cation then undergoes slow conversion to [RuH(176-PhPPh2)(PPh3)21BP4.

If mechanism (b) is operative a catalytic cycle can be envisaged

based on phosphine dissociation from the trihydrido species:

[ROH(P)2(0314- (vii)

[Re113(P)2(alkene)]+ - alkane

(viii)\1. alkene (ix) + H2

[ROH(P)2(02(alkene)]+

(P) (PPh3), (S) . solvent (water or methanol)

SCHEAE 12

54

The type of catalytic hydrogenation cycle described in SCHEME 12

is an extremely complex system, and the irreversibility of processes

(iii) and (iv) would preclude its description in normal kinetic terms.

The catalytic efficiency of the system and its active lifetime would

be intimately related to the relative rates of two sets of reactions.

The relative rates of: (a) reactions (iii) and (vi), and (b) reaction

(iv) as opposed to the combined effect of reactions (v) and (viii),

would determine both the rate of production of the inactive species,

[RuH(76-PhPPh2)(PPh3)2]BF4, and the efficiency of the catalytic

hydrogenation process. Under conditions of high hydrogen and alkene

concentrations, the reactions (v), (vi) and (viii) would be favoured,

leading to continuence of the catalytic cycle and preventing the

removal of the catalytic intermediate from the reaction sphere. The

complex roles played by acid and excess phosphine concentrations are

not easily reduced to simple terms but appear to be involved in the

rate of precipitation of the inactive salt, [RuH(76-PhPPh2)(PPh3)2P3F4.

This might indicate that whilst in solution, in the presence of acid

at the concentrations employed in the catalytic system (‘'4 x 15311),

the [RuH(,6 -PhPPh2)(PPh3)21+ cation is, to some extent, capable of

phosphine dissociation and further involvement in the catalytic cycle,

and that it is the increase in BF4 ion concentrationtleading to

precipitation of this complex cation as the sparingly soluble

tetrafluoroborate salt, which produces the decrease in hydrogenation

rate associated with increasing acid concentration.

The reactions of some ruthenium(II) species with aqueous fluoroboric

acid in methanol are summarized in SCHEME 13. The scheme incorporates

reactions at both the catalytic and preparative scales and demonstrates

that under most conditions, in the presence of excess triphenylphosphine,

[RuH(76 -PhPPh2)(PPh3)2]BF4 is the preferred product.

55

(

BBF4%reflux„,..

Ay:F4(xs) ( (P)s H ,(Me0H) 1 Ru

tnNi N k

fu n\ (H20) 12v/

MeCN (P) (p)

Ru I

MeCN (p) H

RuC12(P)3 H20/Me0k

(reflux)

+BP4-

t-BuOH

\NaCOgsle

NaCO2Me BBF4 F4 (reflux) MeCN

BBF4 MeCN

Ru(CO211e)2(P)2

RuH(CO2Me)(P)3

RuH2(P)4

or

RuH4(P)3

(P)(xs) HBP4/(P)(xs HBF4/(P) HBF4 /(P)

RED/ORANGE [RuH(P)3(S)r

+ -PPh2 BF4

Rus

(P)I \(P)

YELLOW/ORANGE [RuH3(P)3r

(P) a (PPh3), reactions in methanol except where stated.

SCHEME 13

N2 H2

56

CHAPTER II

57

CHAPTER II

COMPLEXES OF RUTHELNIUM(II) CONTAINING BOTH HYDRIDO- AND r1-ARENE LIGANDS.

II .1 . INTRODUCTION

Haines et al(62)first reported a neutral complex of Rh(I)

formulated as Rh[P(OMe)3]2BPh4 in which one of the phenyl rings of the

BPh4 group is bonded as an arene to the rhodium atom. About the same

time Schrock and Osborn(63), during investigations into the catalytic

properties of solvated cationic complexes of the type, [Ed12(PPh3)2(S)21 +,

isolated several tetraphenylborate complexes of Rh(I) such as

Rh(BPh4)(PPh3)2 p in which the BPh4 anion is similarly bound to the

metal. They also prepared analogous complexes of Ir(I). The n.m.r. data

for these species were in accord with previous studies, which had shown

that the proton resonances of an arene ring bonded to a metal atom occur

ati; 1 to 2 to high field of those in free arene species(64). It was also

noted(65)that the patterns of the aromatic resonances observed for the

bonded phenyl ring in complexes such as Rh[P(OR)3}2 BPh4 are sL4tilar to

those for arene derivatives of the type M(C6H5R)(CO)3 = Cr, Mo, W;

R s alkyl)(66). For Cr(arene)(C0)3 complexes these shifts have been

correlated(67)with three effects s withdrawal of 7r-electron density from

the ring; the magnetic anisotropy of the Cr(C0)3 moiety; and the quenching

of the ring current.

The catalytically active cationic complexes of Rh(I) and Ir(I) were

prepared directly from the cyclooctadiene complexes, [(diene)MC1}2. The

insoluble nature of the diene-ruthenium complex [(C8H12)RuC12]x (x>2)

precluded a similar route to cationic ruthenium(II) salts. It has been

shown(60, however, that [(diene)RuCld x dissolves in warm methanol in the

presence of hydrazine and NN-dimethylhydrazine to give solutions, from

which cationic ruthenium(II) complexes of the type [(diene)Ru(N2H4)4J(BPh4)2

are_ precipitated on addition of the BPh4 anion. Such complexes react

with PPh3 under reflux to give the neutral hydride RuH(BPh4)(PPh3)2 in

high yield. This species was formulated as an 'arene-bonded zwitterionic

ruthenium(II) complex' s

H

58

I PPh3 / Ru+

PPh3

BPh3

The 'zwitterionic' sandwich species, (/-05H5)Ru(BPh4), has been

prepared(69)by refluxing a methanolic solution of (7r-05H5)Ru(PPh3)2C1(70)

in the presence of sodium tetraphenylborate, and its crystal structure

determined(71). Attempts to produce species similar to

[(rt-05H5)Ru(ii-arene)]Cl(72)by the reaction of benzene and alkyl- or

fluoro- substituted benzenes with methanol solutions of

(rt-005)Ru(PPh3)2C1 in the presence of PF6- ion were unsuccessful.

Bennett et al(73)and Zelonka and Baird(72'74)have reported the

preparation of a range of halogen-bridged dimeric Ru(II) arene complexes

of the types

Cl

These complexes were also shown to undergo reaction with pyridine,

tertiary phosphines, phosphites, and tertiary arsines (L) to yield

air-stable complexes of the type [RuC12(arene) L]. X-ray studies(73,a)on

[RuC12(006)(PMePh2)] and [RuC12(p-MeC04.CHMe2)(PMePh2)] have shown

that they have a half-sandwich structure similar to arenetricarbonyl-

chromium complexes except that in the ruthenium species the arene ring

is slightly but significantly non-planar. I.r. and n.m.r. spectra,

however, gave no indication of non-planarity in the arene ring systems.

Bennett et al(73'b) noted that arenes containing electron-withdrawing

substituents (e.g. Cl, F, CF3 or CO2Et) did not displace benzene or

toluene from[RuC12(arene) L]complexes. Arenes with electron-withdrawing

substituents such as CO2H, CO2Me, or CO2Ph have been shown to replace

co-ordinated benzene or toluene in Cr(C0)3(arene) complexes. This

difference in reactivity between the two types of complex was assumed to

be a reflection of the greater importance of 6- relative ton-bonding

interactions in the rutenium(II)-arene bond as compared with that in the

chromium(0)-arene bond. Benzene was, however, displaced only to a small

extent from the complex RuC12(T-C6H6)(PBe3) by more strongly electron-

59

donating areaes such as toluene and hexamethylbenzene. This overall

effect was assumed to be the result of the increased favourability of

the Tr-bonding interaction being counteracted by steric effects opposing

the co-ordination of arenes containing bulky substituents.

The reactions of tritolylphosphines with the metal hexacarbonyls,

M(C0)6 (M . Cr, Igo, Vi) have been investigated(75), and a novel series of

derivatives with tri-(o-tolyl)phosphine (pot) isolated and formulated as

tr-(pot)M(C0)3. These complexes contain a phosphine ligand bound to the

metal via a 'ff-bonded o-tolyl ring. Subsequent(76)attempts to produce an

analogous triphenylphosphine species resulted in the isolation of the

complex [Cr(C0)2(PPh3)12. Spectroscopic data(77)was consistent with a

molecular arrangement incorporatingv-bonding between chromium and one of

the arene rings of the ligand which is also a-bonded through phosphorus

to the other metal atoms

PPh2

(C0)2Cr CrC0 ) 2

Ph2P

A novel route to a phosphine ligand which is both normallya-bonded via

phosphorus and simultaneously acting as a 6-electron donor to another

metal has recently been published(78):

Li

Cr Ph2PC1

TMEDA PPh2

0///Cr(C0)6

PPh2CO Cu N ; / Cr

"I Piph2c0 co

Only a few chelate complexes, in which the metal is bound both to

a cyclicrr-system and to another functional group which is part of the

same ligand as the ring system, have been reported(80). One such class

Cr n-BuLi

PPh2

60

of compound', are the '8 -alkenylbenzenedicarboulchromium complexes:

X

Cr.... CH CO -*/ CO CH

(X= CH 2, OCH2, (CH2)2-4 )

The complex Mo(We2Ph)3(16-PhPMe2) has recently been reported,

and although it has proved difficult to synthesize the structure has

been confirmed by X-ray crystallography(79). There appears to be no

significant distortion of the co-ordinated phenyl ring from six-fold

symmetry.

The examples cited above are merely a few of the vast range of

complexes, which have appeared in the literature in the past ten years

containing moieties such as Cr(C0)3 bound to six-electron ligands.

Complexes containing the (76-arene)tricarbonylmanganese(I) cation,

which may be considered as isoelectronic with [Rug,6-PhPPh2)(PPh3)2.1 +,

have been known since 1957(81). The ease of nucleophilic substitution reactions has been shown to increase sharply for a series of halogeno-

arene compounds linked to (0C)3Cr < (Cp)Fe+ < (003Mn+; and the high

reactivity of the manganese complexes has been utilised in the

preparation of other functionally substituted arene complexes by

reaction with amines and anionic nucleophiles(82)

On consideration of: (a) the interesting areas of chemistry

available via such arene complexes, (b) the facility with which the

RuH(PPh3)24. moiety appeared able to accept triphenylphosphine as a

six-electron donor ligand under a wide range of reaction conditions

(see CHAPTER I), and (c) the stability of the [RuH(76-PhPPh2)(PPh3)2] BF4

complex, it seemed advisable to examine possible methods for the

production of other [RuH(arene)(PPh3)2]+X- species.

61

11.2. PROTONATION OF RUH2(PPh3)4 IN BENZENE AND TOLUENE TO YIELD THE

ritIO(,sarene)(Pn3)21BF4 CMPLEXES.

RuH2(PPh3)4 is soluble in benzene and toluene giving yellow

solutions under argon. These solutions when treated with aqueous

fluoroboric acid and heated to reflex with rapid stirring, immediately

become deep red. The red coloration is sustained for only a short time

("30 secs.) after which the solution rapidly becomes pale yellow/green.

The reaction mixture separates into a yellow/green hydrocarbon layer and

a colourless aqueous layer. Yellow crystalline solids may be isolated

from the organic layer.

The i.r. spectra of the isolated yellow solids show several

similarities with that of [Ru07/6PhPPh2)(PPh3)2>F4; and conductivity

data indicate that the complexes are 1:1 electrolytes. Certain

absorption bands appear common to the i.r. spectra of several complexes

containingw-arene ligands - no attempt has been made to assign these

bands or to study them rigorously but they have proved useful indicators

of the presence of such ligands. Complexes containing normally bonded

triphenylphosphine have a strong sharp band at",1430cm."1 ; in complexes

containingir-arene ligands, in addition to normally bonded phosphines,

this band is split into two sharp bands of approximately equal intensity

and separated by 3-4cm."'. In the spectra of the 7f-arene complexes the

region 480-550cm."1 is also complicated - usually consisting of three

pairs of sharp bands. Spectral details for complexes of this type are

shown in Appendix II. Similar spectral patterns have also been observed

in rhodium complexes containing arr-bonded phenoxide ligand(83). The

'H n.m.r. spectra of both complexes show high field triplets

characteristic of a hydride coupled to two equivalent phosphorus nuclei

(7'19.0-19.5, Jpii 35-40Hz) and resonances in the region 4.5-5.5t, which are considered to be indicative of arene complexes(66'73). In the benzene

complex the protons of the arene-bonded ring resonate at 4.46I:and the

signal is a slightly broadened singlet. For the toluene analogue the

resonance pattern consists of a doublet atr 5.4, a triplet atr4.7, and

a second triplet atr3.6 of relative intensities 2:2:1 and assigned

respectively to the ortho-, meta- and Para- protons on the ir-bonded

toluene. This shifting of the resonances of aromatic protons to high

62

field by largely differing amounts depending on their position in the

ring has led to the suggestion that complexes of this type might act as

n.m.r. 'shift' reagents for aromatic species. The methyl group of the

arene-bonded toluene gives a sharp singlet resonance at 7.78r. In both

complexes integration of the 1H n.m.r. spectra and microanalytical data

(Table E.3.) are consistent with the formulation of the species as the

cationic arene-bonded complexes

(Hplue) BF4

,Ru.

PPh; ''PPh3

Success in the preparation of the benzene and toluene complexes,

and the mode of reaction, appeared to indicate that the protonation of

RuH2(PPh3)4 in the presence of aromatic species capable of acting as

six-electron ligands might provide a general route for co-ordinating such

ring systems to the RuH(PPh3)2+ moiety. The colour changes observed

during reaction were similar to those which occur in the protonation of,

for example, RuH(CO2Me)(PPh3)3 in the presence of excess

triphenylphosphine to yield [RuH(.76-PhPPh2)(PPh3)2]BF4. It therefore

seems probable that the intense red coloration is again produced by

[RuH(PPh3)35r; the reaction then proceeds, in the presence of a large

excess of aromatic solvent, via phosphine dissociation and the rapid

co-ordination of the aromatic solvent as a six-electron donor. (S.solvent)

Several reactions have been attempted involving various classes of

substituted aromatic substrates. Practical difficulties and lack of

available time have prevented the successful isolation of pure compounds

from these reactions, but tentative suggestions as to the nature of

probable products are given here. Five types of reaction in which

RuH2(PPh3)4 was protonated in the presence of aromatic substrates and gave

interesting results were with :

(i) Condensed polynuclear aromatics;

(ii) Phenol or sodium phenoxide (RuHC1(PPh3)3 also used as substrate);

(iii) 2-Toluenesulphonic acid;

(iv) Allyl phenyl ether;

(v) Heterocyclic aromatics.

63

11.3. PROTONATION OF Ru(II) COMPLEXES IN THE PRESENCE OF POTENTIAL iT-ARENE LIGANDS.

II.3(i). Protonation of RuH2(PPh3)4 in the presence of condensed

polynuclear aromatics.

Large excesses of aqueous fluoroboric acid were added to methanolic

suspensions of RuH2(PPh3)4 in the presence of (a) naphthalene and

(b) indene under argon. The resulting suspensions were heated to reflux

with rapid stirring. The reaction mixtures rapidly gave the deep red

coloration assumed to be characteristic of the protonated ruthenium

species, [RuH(PPh3)3(S)r. The lightening in colour of the solution,

associated in the benzene and toluene cases with the production of the

arene complex cations, [Ru06-arene)(PPh3)21+, was not as rapid or as

complete for these polynuclear substrates. The resulting orange oils,

which were immiscible with the aqueous layer, appeared to contain most of

the ruthenium but also large amounts of unreacted substrate. It was not

possible to isolate solid products from these reactions, and no high

field line could be detected in the 1 H n.m.r. of a sample of Vie orange

oil removed from the reaction mixture of the naphthalene system. The

first bis-arene complex of naphthalene to be prepared was

ku(C10 H8)21(PF6)2(84). There would, therefore, seem to be no reason to

suspect that the complex [RuH(CI0 H8)(PPh3)2P3F4 could not be prepared

under favourable conditions.

II.3(ii). Reactions of phenol or sodium phenoxide with RuH2(PPh3)4 and

RuHC1(PPh3)3.

RuH2(PPh3)4 has been shown to react with excess phenol to yield a

complex of stoichiometry, Ru(PPh3)2.3PhOH. Comparison of the

spectroscopic data for this complex with those for

[RuH(76-PhPPh2)(PPh3)2113F4 led to the suggestion that the former contains

a C6H50- ligand bound as a 76-phenoxide to the RuH(PPh3)24- species. The

complex was therefore formulated as RuH(OPh)(PPh3)2.2PhOH, with the two

phenol molecules hydrogen bonded to the oxygen of the?Y-bonded phenoxide.

This structure was confirmed by X-ray crystallography(85), and the

C6H50 ligand was found to occupy a 'boat' conformation, suggesting that

64

it should more realistically be regarded as a Ts-cyclohexadienylone

ligand. The structure of RuH(71-OPh)(PPh3)2.2PhOH, however, shows one

striking similarity with that of {RuH(4/6 -PhPPh2)(PPh3)2]BP4 in that in

both complexes the bond between the substituted ring carbon atom and the

substituent group, the CO;z and C-P bonds respectively, almost exactly

eclipse the Ru-H bonds.

RuH(OPh)(PPh3)2.2PhOH has also been prepared by the reaction of

RuHC1(PPh3)3 with excess phenol, in argon-purged toluene, in the presence

or absence of base (triethylamine). A complex having a similar i.r.

spectrum to RuH(TI-Ph0)(PPh3)2 has been prepared, in low yield, by the

action of a stoichiometric amount of sodium phenoxide on RuHC1(PPh3)3 in

toluene. This would seem to suggest that if the protonating acid contains

an aromatic ring, and the negative charge of its conjugate base can be

delocalized onto that aromatic ring, then the base will be preferentially

co-ordinated to the RuH(PPh3)2+ moiety even in the presence of a large

excess of a neutral aromatic species such as toluene.

II.3(iii). Reaction of p-toluenesulphonic acid with RuH2(PPh3;4 in

acetone.

The reaction of a 511 excess of 11.-toluenesulphonic acid monohydrate

with RuH2(PPh3)4 in acetone under argon gave a deep red solution. On

reduction in volume of this solution a yellow solid was obtained. The

i.r. spectrum of this species showed a sharp band at 2045cm.-1 , which was

assigned to a relatively strong Ru-H mode, together with bands

associated with triphenylphosphine, 2-toluenesulphonate anion, and water

(relatively sharp at 3440om.-1). There were also bands at 1431 and 1435cm-I

indicating the probable presence of art-arene ligand (as discussed in 11.2.).

The complex was not very soluble in acetone but an orange solution in

4-acetone gave a weak broad triplet high field resonance (T20,4"25Hz).

On the basis of this limited amount of data it is suggested that the

complex contains ail-bonded Me*C6H4.S03 ligand. A possible structure

would be s

Me

Rug

PPh 3 g ' \PPh3 3

SO-3(HOH)x

65

As in the phenol case delocalization of electron charge density might

produce distortion of the aromatic ring from planarity. A possible

complicating factor in this system is the action of acetone on N Ru1I2(PPh3)4(86) , and the reaction has proved difficult to repeat.

II.3(iv). Protonation of RuH2(PPh3)4 in the presence of allyl phenyl

ether.

A large excess of allyl phenyl ether was added to an argon-purged

diethyl ether suspension of RuH2(PPh3)4. On addition of aqueous

fluoroboric acid and heating to reflux a deep red solution was. initially

produced and rapidly lightened in colour to pale yellow. The reaction

mixture then separated into a yellow ether layer and a colourless aqueous

layer. Reduction in volume of the ether layer by pumping resulted in a

sudden exothermic reaction and an associated change in colour to red.

Additional pumping produced a red oil, which did not revert to yellow on

addition of diethyl ether. Allyl phenyl ether has been shown to act as a

six-electron donor(80)1 and it is probable that the first colour change

(red—I.-yellow) involves co-ordination of the phenyl ring to the

' RuH(PPh3)2+ moiety to yield anTr-arene cationic complex of the type:

[ <Z>-0—C—C.0 + Rill.

PM i Ph

3 H ---3

Several possibilities exist for the subsequent reaction steps. The allyl

function has been shown to eliminate a carbon monoxide ligand from the

Cr(C0)3(arene) species to yield a chelated

78-alkenylbenzenedicarbonylchromium species(ao). Elimination of phosphine to yield an analogous system seems unlikely, and a more probable mechanism would be insertion of the allylic double bond into the Ru-H

bond to yield a chelate complex cation s

66

Inter-molecular insertion could lead to polymeric species. The insertion mechanism is supported by the fact that no high field resonance could be detected in the 111 n.m.r. spectrum of the red oil in CDC13.

II.3(v). Complexes in which heterocyclic =wounds act as 5- or 6-electron ligands.

Green(87)has noted that few complexes containingir-bonded heterocyclic ligands are known, although there seems no reason why a wide variety of unsaturated heterocyclics should not form stable,ff-bonded metal complexes. As early as 1962 Pauson et al had preparelir-pyrrole complexes which are isoelectronic with their?'-cyclopentadienyl analogues s

N (88)

Fe

<-:'N (89 ) mia(c0)3

Analogues of benzenechromiumtricarbonyl; vr-C-methylpyridinechromiumtricarbonyl and (C4H4S)Cr(C0)3 have both been prepared s

Me ‹-;\N (90) • s (91)

dr(CO)3 Cr(C0)3

Other complexes which have been prepared include s

4 N ,;: (92) r

06 i 'co PPh3

R<TS (93) cp (94)

Cr(CO) 3 Mn(CO)3

Good reasons for investigating the type of bonding between transition metals and hetero-aromatic systems are provided by their importance in biological processes. The imidazole ring, as a histidine

67

moiety, functions as a ligand toward transition metal ions in a variety

of biologically important molecules including iron-heme systems, vitamin

B12 and its derivatives, and several metalloproteins(95). The chemistry of

ruthenium(II) complexes of pyridine-type bases has been studied(96)and

many of their properties related to strong metal-ligandir-bonding. The

imidazole ring contains a pyridine-type nitrogen, the ability of which

to act as a rt-acceptor, is expected to be attenuated by therr-donor

characteristic of the pyrrole-type nitrogen, which serves as an internal

electron donor substituent. During discussion of nitrogen-bound and

carbon-bound imidazole complexes of ruthenium ammines, the possibility of

the imidazole acting as art-arene ligand was considered but subsequently

dismissed for the species under study(96)

The reactions discussed here of pyrazole, imidazole and pyrrole with

RuH2(PPh3)4 indicate that such compounds may be capable of acting as

arene-bonded ligands in ruthenium(II) complexes.

Reaction of RuH2(PPhn)4 with 3,5 -dimethylpyrazole.

A large excess of 3,5-dimethylpyrazole was reacted with RuH2(PPh3)4

under reflux in argon-purged toluene. The resulting solution was cooled to

ca. — 200C and yellow/orange crystals were rapidly precipitated. The complex

was collected under argon, washed several times with argon-purged diethyl

ether and dried in vacuo.

The isolated yellow species appeared to be moderately stable but

turned green on prolonged exposure to the atmosphere. The i.r. spectrum

of the complex contained a moderately strong, broad band at 1965cm.-1 ,

which was assumed to be an Ru-H stretching mode. The IH n.m.r. spectrum

of the product in CDC13 showed a broadened high field triplet at ,--.27.5t

and J --4Hz, together with broadened lines at--4.8r and 8.0Vin the -PH ratio of approximately 1:6, which might correspond to the resonances

produced by the single aromatic proton and the six methyl protons of

airr-bonded dimethylpyrazolato ring. These data together with elemental

analysis indicated that the product may have been

Me Me

11 4

Ru +

PPh HPPh

68

This reaction has proved difficult to repeat.

The reactionsof large excesses of pyrazole, imidazole or pyrrole with

RuH2(PPh3)4 in argon-purged toluene appeared to follow similar paths. On

heating to ralux the initially bright yellow solutions darkened in colour

to orange/red. On standing and cooling to ca. — 20°C these solutions paled

slightly in colour and small amounts of pale yellow solids were precipitated.

These solid products proved to be very air-sensitive, turning green

immediately on exposure, and have not been isolated. The apparent greater

stability of the dimethylpyrazole product could possibly be the result of

the electron-donating effect of the alkyl substituents, which would reduce

the acidity of the N-H proton yet might stabilize air-bonded complex.

II.4. SUMMARY

It has not been possible to isolate and adequately characterize most

of the products of the reactions discussed in this Chapter.

Thosecamplexes which have been identified indicate that the RuH(PPh3)2+

moiety, in which the ruthenium atom has twelve d electrons in its outer

shell, will readily accept six further electrons from a single donor, thus

attaining, for the metal, the stable eighteen electron configuration and

producing a (rt-arene)Ru(II) complex.

There seems to be no reason why a wide variety of complexes should

not be available via protonation of RuH2(PPh3)4, by strong non-complexing

acids in the presence of potentialiT-arene ligands, or by acids which

themselves contain an aromatic ring system. The problem may well be in

the establishment of suitable reaction conditions, with particular

significance attached to the choice of solvent, on which protonation

reactions of hydridotriphenylphosphineruthenium(II) complexes often appear

to be critically dependent (see CHAPTER I).

69

EXPERIMENTAL

70

EXPERIMENTAL

E.1. GENERAL

E.1(i). Elemental Analysis

Microanalyses were by the mioroanalytical laboratories of Imperial

College; Bernhardt, Wilhelm; and Pascher, Bonn. The results are tabulated

in Table E.2.

E.1(ii). Instruments

IH n.m.r. spectra were recorded on Perkin-Elmer R12A(60MHz) and

Varian HA100(100MHz) spectrometers. 31P n.m.r. spectra were recorded at

40.505MHz using a Varian XL100-12 spectrometer operating in Fourier

transform mode. Perkin-Elmer model 257, 457 and 325 spectrometers were used to record i.r. spectra. G.1.0. measurements were made s (a) by Perkin-Elmer F11 chromatograph with flame ionization detector and a Kent

Chromalog integrator; and (b) by Parkin-Elmer F33 chromatograph and an

Infotronics CRS-208 automatic digital integrator. Conductivities were

measured using a Mullard conductivity bridge type E7566/3 with a matching

conductivity cell.

E.1(iii). Materials and Procedure

All preparations and other operations were normally carried out under oxygen-free, argon nitrogen or hydrogen. Transfers were made by

conventional techniques using thin steel tubing, syringe, serum cap, etc..

Inert gas atmospheres were maintained during some procedures by the use

of Schienk-tube techniques. Reagent grade solvents were used throughout

and were deoxygenated by purging for 5-10 minutes before use. For rigorous

oxygen-free work gases were purified by passage through commercial

deoxygenation catalysts: white spot nitrogen was passed through a Messer

Griesheim Cmb Industriegase OXISORB HD-Absorber, and hydrogen through an

ENGELHARD-DEOXO HYDROGEN PURIFIER. Solvents were degassed by freeze-pump-

thaw techniques.

71

Starting materials and catalyst precursor complexes were in general

prepared by literature techniques, several of which were improved to give

products, in greater yields, or of higher purity, or both. The complex,

RuC12(PPh3)was prepared from Johnson Matthey 'ruthenium trichloride

trihydrate'th)). RuH(CO2Me)(PPh3)3 and Ru(GO2Me)2(PPh3)2 were prepared

from RuC12(PPh) as previouslydescribed(44). RH2PPh and RH4PPh3 3)3

were prepared from RuC12(PPh3)3 following the general procedure described

by Harris et al(18), more crystalline samples of RuH2(PPh3)4 were

obtained, however, if after addition of the NaBH4 the solution was

heated slowly to reflux for ca. 10 minutes and then allowed to cool.

RuHC1(PPh3)3 was prepared as the benzene solvate from RuC12(PPh3)3 as

previously described(34)or as the toluene solvate by the method

previously applied to RuC12(PPh3)4(61)

72

E.2. PREPARATIONS (CATALYST PRECURSORS)

Hydridoacetatotris-(triphenylphosphine)ruthenium(II)

RuC12(PPh3)3 (0.6g., 0.63mmole) and CH3CO2Na-3H20 (0.71g.,5.22mmole) were added to methanol (250cm.3), which had been purged with nitrogen for

ca. 5 mins.. The resulting suspension was refluxed under nitrogen for

45 mins.. The hydrido complex precipitated from the hot solution as

bright yellow crystals. After cooling at room temperature the solid was

collected on a sintered filter under nitrogen, washed successively with 25cm.3 portions of degassed methanol, water, methanol, and diethyl ether,

and dried in vacuo. Typical yields based on ruthenium were 0.47-0.50g.

(80-85g)m.p. 221-223°C.. This preparation has been scaled up by a

factor of ten with no reduction in the purity of product and with a

slight increase in yield (v90%).

Diacetatobis-(triphenylphosphine)ruthenium(II)

Finely ground RuC12(PPh3)3 (2g., 2.1mmole) and CH3CO2Na-3H20 (2.84g., 20.9mmole) were added to t-butanol (100cm.3) at"30°C. which

had been purged with nitrogen. The resulting suspension was heated under

reflux in a nitrogen atmosphere for N1 hour with rapid stirring. The

solution was then allowed to cool to-30°C. and the stirring rate

reduced. The orange crystalline solid precipitated from the warm solution. Warm (,-,30°C.) degassed diethyl ether (100cm.3) was added to the solution

and it was filtered rapidly under nitrogen, washed with degassed 25cm.3 portions of water, methanol (X 2) and diethyl ether and dried in vacuo.

Typical yields were of the order of 1.15g.(75%).

Dihydridotetrakis-(triphenylphosphine)ruthenium(II)

A mixture of benzene (60cm.3) and methanol (100cm.3) containing

triphenylphosphine (3g., 11.4mmole) was purged with hydrogen for ca. 5 mins..

RuC12(PPh3)3 (1.0g., 0.87mmole) was then added and the suspension

stirred at room temperature for ca. 10 mins.. Dry finely ground sodium

borohydride (1.5g., 0.04mole) was added in five portions of approximately

0.3g. over a period of 20 mins. with rapid stirring. The suspension

changed colour from red/brown, through purple and greenish phases to an

off-white. At this stage further triphenylphosphine (3g., 11.4mmole) was

73

added and the suspension heated to reflux for ca. 10 mins. giving a deep

yellow solution, which was allowed to cool. The hydrido complex

precipitated from the hot solution as bright yellow crystals. Degassed

methanol (100cm.3) was added to the solution and the product collected on

a sintered filter under argon, washed with 4rgon purged methanol

(2 X 25cm.3) and dried in vacuo. Typical yields based on ruthenium were

1.02-1.08g. (85-90%).

Tetrahydridotris-(triphenyluhosphine)ruthenium(IV)

A mixture of benzene (60cm.3) and methanol (100cm.3) was purged

with hydrogen for ca. 5 mins.. RuC12(PPh3)3 (1.0g., 1.05mmoles) was then added and the suspension stirred at room temperature for ca. 10 mins..

Dry finely ground sodium borohydride (1.5g., 0.04mole) was added in five

portions of approximately 0.3g. over a period of 20 minutes with rapid

stirring. The red/brown solution slowly became almost colourless and a

white precipitate was produced. Degassed methanol (100cm.3) was added to

the solution and the solid was collected on a sintered filter under argon,

washed with argon purged methanol (3 X 25cm.3) and stored under argon.

Typical yields were of the order of 0.75g.(80%). The complex is extremely

sensitive to air when wet and the repeated washing appeared necessary,

possibly to remove any traces of unreacted borohydride. Drying of the

product proved difficult; pumping directly on the wet solid produced

decomposition and it was found more satisfactory to maintain a slow

stream of argon.

E.3. HYDROGENATION TECHNIQUE

Alkenes for use in catalytic studies?t were purified by a combination

of the methods previously described(44,48) . The substrates were shaken

several times with acidic ferrous ammonium suphate solution, the organic

layer was then separated and passed down an activated alumina column

under nitrogen directly onto sodium. The alkenes were then distilled

from sodium under nitrogen into the hydrogenation apparatus immediately

prior to use. The purity of terminal linear alkenes was determined by

g.l.c. and in general the purity of substrate was controlled by selection

of the appropriate boiling fraction.

74

As discussed in Chapter I the irreversible aging process of the

catalyst required that the timing and conditions for each catalytic run

were rigorously controlled in order that meaningful relative hydrogenation

rate data could be obtained for different substrates. A standard (Table E.1.)

hydrogenation procedure was, therefore, adopzed. The temperature of the

reaction vessel was maintained constant by a thermostatically-controlled

water jacket.

Hydrogenation rates were obtained by determining the gradient of the

pressure versus time curves at a hydrogen partial pressure of 40cm. Hg..

These values were expressed in terms of volume of gas uptake in unit

time corrected to S.T P •_ ••

E.4. PREPARATIONS OF COMPLEXES DERIVED FROM PROTOITATION OF CATALYST

PRECURSORS.

Hydrido(16 -phenyl-dinhenylphosphine)bis-(triphenylphosphine)-

ruthenium(II)tetrafluoroborate

Suspensions of RuH(CO2Me)(PPh3)3, RuH2(PPh3)4, or Ru(CO2Me)2(PFh3)2

in methanol under Ar, N2 or H2 all react with large excesses of HBF4(aq.)

and PPh3 to yield [RuH(76 -PhPPh2)(FPh3)2PF4.

RuH(CO2Me)(PPh3)3 (0.5g, O.53mmole) and PPh3 (0.5g, 1.91mmole) were

added to 42% (aq.) HBF4 (5cm.3) in methanol (50cm.3), which had been

purged with hydrogen for ca. 5 mins.. The suspension was heated to reflux with rapid stirring then allowed to cool under H2. The

precipitated pale yellow crystalline complex was collected under N2,

washed with degassed methanol and ether, and dried in vacuo. The complex

as isolated contains one methanol of solvation (0.35 - 0.40g, 66-75%).

Hydridodiacuomethanolbis-(triphenylnhosphine)ruthenium(II)-

tetrafluoroborate

RuH(CO2Me)(PPh3)3 (0.5g, 0.53mmole) was added to 4e. (aq.) HBF4(5cm.3)

in methanol (15cm.3) degassed with argon. The suspension was heated to

reflux with rapid stirring then allowed to cool under argon. The

precipitated bright yellow microcrystalline complex was collected, washed

with degassed methanol and ether and dried in vacuo (0.31-0.35g, 74-835).

The i.r. spectrum (nujol mull) includes absorptions at 3640, 3560 (v(OH)),

2060 and 2020cm.-1 (P(RuH)). The I H n.m.r. spectrum (at 35°C in CDC13)

75

includes a hydride signal consisting of a doublet of doublets

T18.8 (J a 32Hz, Jam, . 38Hz).

Acetatoaouotris-(triphenylphosohine)ruthenium(II)tetrafluoroborate

RuH(002Me)(PPh3)3 (0.5g, 0.53mmole) was added to 42% (aq.) HBF4

(5cm.3) in acetone (15cm.3) degassed with argon. The suspension was

heated to reflux, with rapid stirring, giving initially a deep red

solution, which rapidly became lighter in colour. The resulting orange

solution was reduced in vacuo to half its volume when the orange

microcrystalline complex was precipitated. This was collected under argon

washed with degassed acetone and ether, and dried in vacuo (0.5g,^190%).

The i.r. spectrum (nujol mull, and KBr disc.) includes absorptions at

3585, 3305, 3195 (v(OH)), and 1498, 1467om.-1 (7) sym s 000) and (7)

y000).

a

Acetatocarbonyltris-(tripherylphosphine)ruthenium(II)tetrafluoroborate

This complex is precipitated as a yellow crystalline solid in high

yield when CO is passed through a saturated methanolic solution of

[Ru(CO2Me)(H20)(PPh3)31BF4 and the solution reduced to half of its

original volume (r."85%). The i.r. spectrum (nujol and hexachlorobutadiene

mulls) includes an intense carbonyl absorption at 1963cm.-1 and bands at

1500 (V'asymOCO) and 1467cm.-1 (Z) asym000).

Acetatoacetonitriletris-(triphenylphosphine)ruthenium(II)-

tetrafluoroborate

The complex is obtained as an orange-yellow crystalline solid when

[Ru(CO2Me)(H20)(PPh3)31BF4 is dissolved in the minimum amount of warm

MeCN and the solution reduced to half of its original volume and cooled

to 0°C. (^8Z). The i.r. spectrum (nujol and hexachlorobutadiene mulls)

includes absorptions at 2285 and 2250 (Tp(CN)), 1500 (pasym

000) and

1468cm.-1 (1) 000). sym

Aquobisacetonitriletris-(triphenylphosphine)ruthenium(II)bis-

(tetrafluoroborate)

[Ru(002Me)(H20)(Fa3)3JB114 (0.25g, 0.24mmole) was dissolved in acetonitrile (15cm.3) and 42% (aq.) HBF4 (5cm.3) added when the orange

solution immediately became pale lime green and the pale yellow

76

crystalline complex was precipitated. This was collected in air, washed

(X 2) with ether and air dried (0.22g, 80%). The i.r. spectrum (nujol mull)

includes bands at 3625 and 3557 (p(OH)), 2320 and 2290 cm.-1(p(CN)).

Hydridobisacetonitriletris-(triphenyll.hosphine)ruthenium(II)-

tetrafluoroborate

(Method A) RuH(CO2Me)(PPh3)3 (0.5g, 0.53mmole) was added to solutions

of acetonitrile (5cm.3) in acetone (or methanol) (25cm.3). The yellow

suspension was vigorously stirred and heated to reflux. Addition of 42%

(aq.) HBF4 (2cm.3) gavea pale yellow solution, from which on reduction

to about half the original volume the white microcrystalline complex was

precipitated. The product was collected, washed with degassed ether (X 2)

and dried in vacuo. (0.30 - 0.34g, 52-59%).

(Method B) [Ru(H20)(MeCN)2(PFh3)31(BF4)2 (0.5g, 0.43mmole) was

dissolved in acetone (15cm.3), which had been purged with Ar. Addition of

NaBH4 (0.1g, 2.63mmole) to the lime-green solution caused an immediate

change to yellow. The solution was reduced to half volume and methanol (10cm.3)

added, to precipitate the white microcrystalline complex which was collected

under argon, washed (X 2) with ether, and dried in vacuo (0.21g, 46%).

The i.r. spectrum (nujol mull) includes a pair of bands (w) assigned to

coordinated acetonitrile at 2295 and 2280am.-1(p(CN)). The 1 H n.m.r.

spectrum in dichloromethane/acetonitrile includes a symmetrical quartet

hydride signal atT23.5 (J 20Hz). (vRu-H 1960cm. 1)

Hydrido(aauo/acetone)dicarbonylbis-(triphenylphosphine)rutheium(II)-

tetrafluoroborate

RuH(CO2Me)(F113)3 (0.5g, 0.53mmole) was added to 42% (aq.) HBF4(5cm.3)

in acetone (15cm.3) under hydrogen. The suspension was heated to reflux,

and carbon monoxide passed through the resulting red solution, which

immediately became pale yellow. The solution was reduced to half volume,

degassed ether (10cm.3) added and the solution stored at -20°C. for 24

hours. The white crystalline complex was collected, washed with ether (X 2)

and dried in vacuo (0.14g, 36%). The i.r. spectrum (nujol mull) includes

a broad hydroxyl stretching band centred at 3420cm. 1 together with

carbonyl bands at 2070 and 1990cm.-1, the lower being the broader and

the stronger, and in some samples a band at 1655cm.-1 assigned to the C.0

stretching vibration in co-ordinated acetone. The 1H n.m.r. as in

77

d6-acetone at 35°C. contains two distinct triplet hydride signals of

approximate relative intensity 1:4 at r 14.15 (Jpi, 19Hz) and r 14.68

PH 18Hz), suggesting that the product is a mixture. The possible -. nature of the components of the mixture is discussed in Chapter I.

Hydridochlorodicarbonylbis-(triphenyl-phos-phine)ruthenium(II)

[RuH(S)(C0)2(PPh3)2]BF4 ("0.2g) was dissolved in methanol (10cm.3),

LiC1 (0.05g) was added and the colourless solution refluxed under N2 for

a few minutes. The white crystalline complex precipitated on cooling, was

collected, washed with degassed methanol and ether and dried in vacuo

(.,65:(1). The i.r. and 'H n.m.r. spectra were in good agreement with those

previously recorded by James et al(6'b)

E.5. PREPARATIONS OF OTHER COMPLEXES DISCUSSED IN CHAPTER I.

Hydridostearatotris-(trichenylohosphine)ruthenium(II)

RuC12(PPh3)3 (2g, 2.09mmole) and sodium stearate (5g, 16.3mmole)

were added to methanol (250cm.3 ) and the mixture heated to reflux. The

orange/yellow crystals of the complex, which rapidly precipitated on

cooling, were collected and dried in vacua (5070). The complex is

moderately soluble in petroleum (60-80O) and a petroleum wash resulted in

some loss of yield. The i.r. spectrum (nujol mull) includes a strong Ru-H

stretching mode at 1965am.-1. The 1H n.m.r. spectrum in acetone includes

a symmetrical quartet hydride signalir28.25 (J 27Hz).

Hydridobisacetonitriletris-(triphenylohosphine)ruthenium(II)chloride

RuHC1(PPh3)3.C7H8 (0.5g, 0.49mmole) was dissolved in acetonitrile

(15cm.3). The lime green solution was reduced to half volume to

precipitate white microcrystals of the complex which was collected,

washed with degassed ether (X 2), and dried in vacuo (0.25g, 52%). The

i.r. spectrum was similar to that of the analogous tetrafluoroborate. The

'H n.m.r. spectrum in methanol-acetonitrile at 35°C includes a

broadened symmetrical quartet hydride signal att 23.6 (JHP

. 22Hz), with

line width r.-5Hz.

78

Hydrido(72 6 -phenyl-dilphenylphosphine)bis-(triphenylphosphine)-

ruthenium( II ) D- t oluenesulphonat e

RuR(CO2Me)(FFh3)3 (0.5g, 0.53mmole) and PPh3 (0.5g, 1.91mmole) were

treated with 11-toluenesulphonic acid monohydrate (1.9g, 10mmole) in

methanol (50cm3), which had been purged with H2 for ca. 5 mins. The

suspension was heated to reflux then allowed to cool under H2. The

solution was reduced to 20cm3 by pumping and cooled to -20°C for 12 hours.

The precipitated yellow complex was collected under N2, washed with

N2-purged methanol and diethyl ether, and dried in vacuo. The complex

appeared to contain methanol of solvation (0.19g -'33g).

ruthenium(II)trifluoromethylsulphonate.was similarly prepared from

CF3-S03H added dropwise with care to the substrate suspension.

E.6. PREPARATIONS OF COMPLEXES DISCUSSED IN CHAPTER II.

Hydrilot6 -benIma&LE:(t221ELDElphosphine)ruthenium(II)-

tetrafluoroborate.

RuH2(PPh3)4 (0.5g, 0.43mmole) was added to benzene (10cm3) which

had been purged with argon for ca. 10 minutes. The suspension was treated

with argon-purged HBF4 ("'4Zaci.) (5 cm3). The resulting suspension was

heated to reflux with rapid stirring, giving initially an intense red

solution. The red colour rapidly changed (ca. 30 secs.) to pale yellow.

On standing the reaction mixture separated into a yellow benzene layer.

The organic layer was separated and reduced to approximately half volume

by pumping. A yellow crystalline solid was precipitated. The precipitated

solid was collected under argon, washed with argon-purged ether and dried

in vacuo. Water was strongly held by the product and only removed slowly

in vacuo at elevated temperature (,..-60°C). (Yield 0.25g 74.

tetrafluoroborate.was similarly prepared from toluene. (Yieldev705).

H.ydrido(' 6-3, 5-dimethylpyrazolatojbis-(triphenylphosphine)-

ruthenium(II).

A large excess of 3, 5-dimethylpyrazole (0.5g, 5.3mmole) was reacted with RuH2(PPh3)4 (0.5g, 0.43mmole) under reflux in argon-purged toluene

79

(20cm3). The resulting solution was cooled to ca. -20°C and

yellow/orange crystals were rapidly precipitated. The complex was

collected under argon, washed with degassed diethyl ether (2 X 20cm )

and dried in vacuo. (Yield 0.1g N32;,).

Hydrido(76-Dhenoxy)bis-(triphenylphosphine)rathenium(II)bis-phenol.

RuHC1(PPh3)3 C7H8(0.5g, 0.49mmole) and excess phenol (1.0g, 10.7mmole)

were suspended in N2-purged toluene (20cm3) and excess triethylamine

(2cm3) added. After refluxing for 2 hours the orange solution was reduced

to one third volume and cooled to ca. -20°C. The yellow crystals of the

complex were collected and washed with N2-purged ether. (YieldnJ30;4).

TIME

ELAPSED

PROCEDURE

47mg. RuH(cO2me)(PPh3)3 weighed into polypropylene

bucket and suspended in reaction vessel.

157mg. PPh3 placed in reaction vessel.

Reaction vessel sealed and evacuated (X 2), hydrogen

atmosphere introduced.

(50-x)cm3 methanol containing 38mg. lirMe.Ph.S03H.H20

placed in solvent burette and purged with nitrogen.

Acidic methanol admitted to reaction vessel.

System evacuated (X 3) and hydrogen atmosphere

introduced.

Catalyst added and solution pumped (X 3) then stirred

rapidly under an atmosphere of hydrogen.

Alkene substrate distilled into burette under nitrogen.

Catalyst solution pumped and pressured to 45(: 2)cm.Hg.

hydrogen.

x cm.3 alkene added (where x cm3 is 0.05mole of

alkene).

Lowest manometer reading noted (after alkene has exerted

vapour pressure).

Manometer readings recorded and plotted against time

(give total gas pressure).

Samples taken for g.l.c. analysis(a)

5 mins.

6 mins.

6 mins.

16 mins.

16.5 mins. 10

17 mins. 11

17 mins. 12

27 mins.

27 mins. 13

80

Table E.1.

Standard hydrogenation procedure at 40°C.

(a) G.l.c. analyses were carried out on a Perkin-Elmer P11 chromatograph

using a silver nitrate-diethyleneglycol column and a flame ionization

detector.

81

Notes to Table E.2.

(a) Examination of the above data and of some of those obtained in

other studies indicate that for ruthenium complexes of this type carbon

analyses tend to be consistently low, probably as the result of

combustion difficulties. (cf. reference 6 (b) for example)

(b) This complex was extremely air and moisture sensitive and may

have decomposed slightly during analysis.

(c) As originally isolated this complex appears to contain a relatively

loosely held methanol molecule in the crystal lattice. Carbon and

hydrogen analyses, performed soon after the isolation of the product,

would contain this methanol of crystallization, but it may have been

lost, to some extent, from the complex in storage before phosphorus

and fluorine analyses were carried out.

(d) This complex should more accurately be regarded as a mixture of

several species of general formula [RuH(S)3(PPh3)2]BF4, in which (S)

may be either methanol or water. The sample as analysed, although

predominantly of the stoichiometry as indicated, probably contained

a mixture of several species, in which the ratio of methanol to

water varied.

(e) (S) may be either acetone or water and the ratio of the two

products appears to vary with reaction conditions. N.m.r. evidence

suggested a water to acetone species ratio of approximately 4:1

in the sample which was analysed.

82

Table E.2.

Elemental analytical data for ruthenium complexes discussed in Chapter I. Found (calculated) ,t.

COMPLEX C(a) H P N X

RuH(CO2Me)(PPh3)3 70.7(70.9) 5.3(5.2) 9.6(9.8)

Ru(CO2Me)2(Pa3)2 64.3(64.6) 5.2(4.9) 8.5(8.3) RuB2(PPh3 )4 75.1(75.0) 5.7(5.4) 10.4

(10.7) RuH4(PPh3)3(13) 71.2(72.7) 5.7(5.5)

F [RuF1(76 -PhPPh2) (PPh3 )2 i BF4 (C) 65.7(66.5) 5.1(4.7) 9.9(9.5) 7.9(7.8)

II •MeOH (65.5) (5.0) (9.2) (7.5)

[RuH(7 6 -Ph.PPh2)(PPh3 )2] E-tolSO 68.1(69.0) 4.8(5.0) 8.3(8.8) F

[RuH(76-PhPPh2)(PPh3)2]CF3•803 63.2(63.6) 4.7(4.5) 8.2(9.0) 6.3(5.5) [RuH(11 20 )2(me0H)(1Th3)213F4 (d) 56.3(56.8) 4.8(5.0) 7.2(7.9)

F [Ru(CO2Me)(H20)(PPh3)31BP4 63.6(63.9) 4.9(4.8) 8.7(8.8) 6.4(7.3) [Ru(CO2Me)(CO)(PPh3)3]BF4 64.5(64.5) 4.7(4.5) 8.9(8.8)

rla(CO2Me)(MeCN)(PPh3)3]BF4 63.9(64.8) 5.1(4.8) 8.6(8.7) 1.6(1.3) F

Piu(H20)(MeCN)2(PPL13)3](BF4)2 58.3(59.8) 4.5(4.6) 7.6(8.0) 2.8(2.4) 13.5 (13.2)

[Thill(MeC1Y)2(pPh3)3]BF4 65.5(654 5.0(4.9) 8.4(8.8) 2.4(2.6)

[Rua(s)(00)2(pph3 )2 113114 (e) 56.8(57.9) 4.3(4.2) 7.5(7.9) ca.

Ruilci(co)2(PPh3)2 62.8(63.6) 4.6(4.3) 8.1(8.6) 5.2(4.9) Cl

RuCl(CO2Me)(Pa3)3 68.1(68.5) 5.2(4.9) 8.9(9.5) 3.9(3.6) RuH(CO2C17H35)(PPh3)3 73.5(73.8) 7.0(6.9) 7.3(7.9) RuH(CO2CH•C 2H5C4 H9 )(PPh3 )3 71.6(72.2) 6.2(5.9) 9.4(9.0)

Cl RuHC1(PPh3)3.C7H8 72.0(72.2) 5.4(5.4) 9.4(9.1) 3.8(3.5)

Cl [RuH(LIeCN)2(PPh3)3]Cl 68.6(69.2) 5.5(5.2) 8.6(9.2) 3.1(2.8) 3.6(3.5)

83

Table E.3.

Elemental analytical data for ruthenium complexes discussed in

Chapter II. Found (calculated) 'co.

COMPLEX C H P N •

[RuH(76-C6H6)(PPh3)2]BF4 62.6(63.1) 4.8(4.6) 7.2(7.9)

[RuH(y 6 -005. Me) (PPh3)2] BF4 64.1(64.2) 4.9(4.7) 7.6(7.6)

[RuH(76-Ph0)(FP113)2 • 2Ph011] 69.5(71.4) 5.7(5.3) 7.1(6.8)

RuH(75-C3N2H(Me)2)(PPh3)2 69.1(68.0) 5.6(5.3) 7.9(8.8) 3.4(3.9)

Table E.4.

Molar conductances (,%1) and electrolyte types of ruthenium complexes

discussed in Chapter II. Measured at 26°C in acetone (163M solutions).

COMPLEX (A-) -1 -1

ohm cm mole

ELECTROLYTE

TYPE

[RuH('/6 -PhPPh2)(1)Ph3)2>F4 108-115 1:1

[RuHW-C6H6 )(PPh3)2]B54 120-125 \

1:1

[RuH(,6 -C6H5•Me)(PPh3)2]BF4 116-124 111

Ruid05-C3N2R(Me)2)(13Ph3)2 8 NON-CONDUCTOR

84

Table E.5

I .r data for ruthenium complexes. Bands expressed in cm-1

COMPLEX /) (Ru-H) ( a)

( b) P s(OCO) ,Pa(OCO) ( c) PC;;C (ring)

OTHER BANDS

[RuH(76-PhPPh2)(PPh3)2] BF4 fi itle011

2035w ( c) 1436, 1432

( g) 3495

[ R1111(7 6 -Pb.PPh2)(PPh3)212-tolS03 2030w ( c) -1438, 1434

[ Rai ("6 -PhPPh 2) (PPh3)2] CP3•SO3 2035w ( c) 1437, 1433

[RUH(H20) 2 (de0H) (PPh3 )2] BF4 (d) (g) 3420

[Ru(CO2Me)(1120)(PPh3)3.1 13P4 (b) 1498, 1465 (g) 3585, 3305 3195

[Ru(002Ye)(C0)(1Th3)3.1BF4 (b) 1500, 1467 ( h) 1963

[Ru(CO2Me) (MeCN) (PPh3)31 BF4 (b) 1500, 1468

[Ru(H20)(mecli)2(Prh3)31(33F4 )2 (g) 3625, 3557

[Raii(mecN)2(PPh3)3pF4 1960s

[ RuH(MeCN) 2(PPh3)31 Cl 1960s

[RuH(S)(C0)2(PPh3)2] B114 (e) (f ) ( 0) 3400

RuHC1 (CO) 2(Pa3)2 (f) (i) 278, 288

RuC1(CO2/1e)(PPh3) 3 (b) 1515, 1464 (1) 317

Ra(c020171135)(1Th3)3 1965s (b) 1515, 1437

Ra(c02cli-c2m3c4119)(PPh3)3 1975s (b) 1517, 1434

[Ruil((/ 6 -C6H6)(PPh3)3]BF4 2000m ( c) 1436, 1430

[RuH(7 6 -C6H5Me ) (PPh3 )3] BF4 2015m (c) 1437, 1432

ROW -C3N2H(Me)2)(PM3)3 1965m

85

Notes to Table E,5.

(a) The strengths of theAP(Ru-H) bands are quoted relative to the strength

of this band in RuH(CO2Me)(PPh3)3 and not relative to the *strengths of

other bands in the individual spectra (w=weak, m=medium and s=strong).

(b) The bands quoted (vs(0C0),Pa(OCO)) are the symmetric and asymmetric

stretching vibrations in the carboxylato ligand. The values observed for

these bands and forAP in these complexes are characteristic of chelating

carboxylates.

(0) The bands quoted are considered to be characteristic of complexes

containig//-bonded phenyl rings.

(d) This complex has two weak bands at 2060 and 2020 cm.-1, the higher

band may be associated with methanol and the lower one is probably due

to theP(Ru-11) vibration.

(e) AB note (e) to Table E.2.

(f) These complexes have several bands between 2040 and 1950 cm.-1 which

are assigned to carbonyl and metal-hydride vibrations, the spectrum for

RuHC1(C0)2(PPh3)2 was similar to that observed previously(6b)

(g)p(OH).

(h)p(C0).

(i) p(Ru-C1).

86

Table E.6.

I H n.m.r. data for hydridoruthenium(II) complexes also containing

arr-bonded ring. Spectra recorded in nitrogen- or argon-purged

CDC13 at 35°C and 60 MHz.

COMPLEX HYDRIDE

(4PH)

T/-BONDED RING(a) (Intensity)

a P. m

[RuH (176 -PhPPh 2) ( PPh3 )2] BF4 18.7(13) (33) 5.68t(2) 2.888(1) 4.60t(2)

PluH(76 -PhPPh2)(PPh3)21P:-toIS03 18.5(13)(35) 5.70t(2) 4.62t(2)

[RuH(T6 -PhPPh2)(1Th3)2] CF3803 18.6(b)(33) 5.66t(2) 4.60t(2)

[ho(T6 -061.16)(pph3)21BE4 19.ot(36) 4.463(6)

[Ruli(1/6 -co:15- me) (PPh3)21BF4 19.4t(36) 5.42d(2)13.63t(1) 4.70t(2)

RuHW-C3N2H(ke)2)(PPh3)2 27.5t(24) 4.808(1)(C)

Notes to Table E.6.

(a) The couplings between protons bonded to adjacent ring carbon atoms

are between 5 and 6 Hz.

(b) The structure of this spectrum is discussed in I.4(i). and illustrated

in Appendix III.

(c) Thew-bonded dimethylpyrazolato ring would contain a single laromarict

proton.

87

REF ERUIC ES

88

REFERENCES

1. L. Vaska and J. W. Diluzio, J. Amer. Chem. Soc., sa (1961), 1262. 2. (a) L. Vaska, Z. Naturforschg., 15b (1960), 56.

(b) L. Vaska and E. M. Sloane, J. Amer. Chem. Soc., 82 (1960), 1263.

3. (a) J. Chatt and B. L. Shaw, Chem. and Ind., (1960), 931. (b) J. Chatt and B. L. Shaw, Chem. and Ind., (1961), 290.

4. J. Chatt, B. L. Shaw, and A. E. Field, J. Chem. Soc., (1964), 3466.

• 5. J. P. Collman and W. R. Roper, J. Amer. Chem. Soc., a (1965), 4008. 6. (a) B. R. James and L. D. Markham, Inorg. Nucl. Chem. Letts., 1 (1971),

373. (b) B. R. James, L. D. Markham, B. C. Hui and G. L. Rempel, J. C. S. Dalton, (1973), 2247.

7. C. K. Brown and G. Wilkinson, J. Chem. Soc. (A)., (1970), 2753.

8. (a) J. D. Cotton, M. I. Bruce and F. G. A. Stone, 32 ghtatSaczLLI.,

(1968), 2162. (b) F. L 1 Eplattenier and P. Calderazzo, ILI2Eajltpja., 1 (1968), 1290.

9. (a) R. A. Sanchez-Delgado, private communication. (b) R. A. Sanchez-Delgado, J. S. Bradley and G. Wilkinson,

J. C. S. Dalton, (in press).

10. B. F. G. Johnson and J. A. Segal, J. C. S. Dalton, (1973), 478.

11. B. E. Cavit, K. R. Grundy and W. R. Roper, Chem. Commun., (1972), 60,

12. S. T. Wilson and J. A. Osborn, J. Amer. Chem. Soc., 21 (1971), 3068. 13. J. S. Bradley and G. Wilkinson, Inorganic s, Vol. XVII

(in press).

14. J. M. Fletcher, I. L. Jenkins, F. M. Lever, F. S. Martin,

A. R. Powell, and R. Todd, JL.InoEg.t Nucl. Chem., 1 (1955), 378.

15. C. G. Pierpont and R. Eisenberg, Inorg. Chem., 11 (1972), 10940 16. T. Ito, S. Kitazumey A. Yamamoto and S. Ikeda, J. Amer. Chem. Soc.,

22. (1970), 3011. 17. J. J. Levison and S. D. Robinson, J. Chem. Sq21C/11.1 (1970),2947.

18. R. O. Harris, N. K. Hota, L Sadavoy and J. M. C. Yuen,

J. Organometal. Chem., 54 (1973), 259. 19. (a) W. H. Knoth, J. Amer. Chem. Soc., a (1972), 104.

(b) W. H. Knoth, Inorganic Syntheses, vol. XV (1974), 31.

20. A. Yamamoto, S. Kitazume and S. Ikeda, J. Amer. Chem. Soc., .92 (1968), 1089.

21. G. W. Parshall, J. Amer. Chem. Soc., 20 (1968), 1669.

89

22. T. I. Eliades, R. O. Harris and M. C. Zia, Chem. Commun., (1970),

1709. 23. (a) I. S. Kolomnikov, T. S. Lobeeva and M. E. Vol'pin,

Izvest. Akad. Nauk. SSSR. Ser Khim., (1972), 211. (b) M. E. Vol'pin and I. S. Kolomnikov, Pure and Awl. Chem., 3

(1973), 567. 24. R. W. Mitchell, A. Spencer and G. Wilkinson, J. C. S. Dalton,

(1973), 846. 25. S. Komiya, A. Yamamoto and S. Ikeda, J. Organometal. Chem., 42.

(1972), C65. 26. S. Komiya, A. Yamamoto and S. Ikeda, Bull. Chem. Soc. Japan, .48,

(1975), 101. 27. J. Brandrup and E. H. Immergut, 'Polymer Handbook', Interscience

Pnblishers, New York, London, Sydney (1966), 11341. 28. S. Komiya and A. Yamamoto, Chem. Commun., (1974), 523. 29. H. Kono and Y. Nagai, Chem. Soc. Ja an. Chem. Letts., (1974), 931.

30. F. Pennella, Coord. Chem. Rev., 16 (1975), 51.

31. J. R. Sanders, J. C. S. Dalton, (1973), 743. 32. C. McConway, A. C. Skapski, L. Phillips, R. J. Young and

G. Wilkinson, Chem. Commun., (1974), 327. 33. (a) D. Evans, J. A. Osborn, F. H. Jardine and G. Wilkinson,

Nature, .208 (1965), 1203. (b) P. S. Hallman, D. Evans, J. A. Osborn and G. Wilkinson, Chcm. Commun., (1967), 305.

34. P. S. Hallman, B. R. McGarvoy and G. Wilkinson, 11 212pat 122.1 1A1., (1968), 3143.

35. C. 0 ' Connor and G. Wilkinson, LClIcat Eocz LAI., (1968), 2665. 36. A. C. Skapski and P. G. H. Troughton, Chem. Commun., (1968), 1230.

37. R. R. Hitch, S. K. Gondal and C. T. Sears, Chem. Commun., (1971), 777.

38. S. Cenini, A. Mantovani, A. Fusi and M. Keubler, Gazzetta Chim. Ital., 125. (1975), 255.

39. W. H. Knoth, Inorg. Chem., 12 (1973), 38. 40. L. W. Gosser, W. H. Knoth and G. W. Parshall, J. Amer. Chem. Soc.,

(1973), 3436.

41. D. Rose, J. D. Gilbert, R. P. Richardson and G. Wilkinson, J. Chem. Soc. (Al., (1969), 2610.

90

42. (a) S. D. Robinson and M. F. Uttley, Chem. Commun., (1972), 1047. (b) S. D. Robinson and M. F. Uttley, J. C. S. Dalton, (1973), 1912.

43. (a) A. C. Skapski and P. G. H. Trougnton, Chem. Commun., (1968), 1230. (b) A. C. Skapski and F. A. Stephens; J. C. S. Dalton, (1974), 390.

44. R. W. Mitchell, A. Spencer and G. Wilkinson, J. C. S. Dalton,

(1973), 845.

45. R. W. Mitchell, Ph. D. Thesis, Imperial College, Univ. of London (1972).

46. by R. W. Mitchell and A. Spencer.

47. A. Spencer, J. Organometal. Chem., 9.2 (1975), 389.

48. J. A. Osborn, F. H. Jardine, J. F. Young and G. Wilkinson, J. Chem. Soc. (A)., (1966), 1711.

49. J. Halpern, Proc. Jap.-Am. Seminar, Honolulu, May (1974), Y. Ishii and M. Tsutui (Eds.), publ. Plennum Prens N. Y. and Lond. (1975),

P.109,

50. (a) W. F. Taylor, J. Catal., 16 (1970), 20. (b) W. F. Taylor, J. Phys. Chem., a (1970), 2250.

51. P. S. Mody, N. D. Orlando, J. M. Schultz and B, C. Gates, J. Catal., a (1973), 372.

52. H. H. Kausch-Blecken von Schmelling, J. Macromol. Sci. Rev. Macromol. Chem. 4 (1970), 243.

53. F. Basolo, Coord. Chem. Rev., a (1968), 213.

54. E. Benedetti, G. Braca, G. Sbrana, F. Salvetti and B. Grassi; J. Orgurometal. Chem., E (1972), 361.

55. B. R. James and L. D. Markham, 222MELM202., la (1974), 97.

56. (a) G. W. Parshall, W. H. Knoth and R. A. Schunn, J, Amer. Chem, Soc., a (1969), 4990. (b) G. W. Parshall, Accounts Chem. Res., a (1970), 139. (c) R. A. Schunn, Inorg. Chem., 2 (1970), 2567. (d) G. W. Parshall, Accounts Chem. Res., 8 (1975), 113.

57. P. R. Hoffman and K. G. Caulton, J. Amer. Chem. Soo., 22 (1975), 4221.

58. R. D. Feltham and R. G. Hayter, J. Chem. Soc. (1964), 4587.

59. W. J. Geary, Coord. Chem. Rev., (1971), 81.

60, P. S. Hallman, T. A. Stephenson and G. Wilkinson, Inamllaj,Artheses, Vol. XII (1970), 238.

61. R. A. Schunn and E. R. Wonchoba, Inorganic Syntheses, Vol.XIII (1972), 131.

91

62. M. J. Nolte, G. Gafner and L. M. Haines, Chem. Commun., (1969), 1406.

63. R. R. Schrock and J. A. Osborn, Inorg. Chem., 2 (1970), 2339. 64. W. McFarlane and S. 0. Grim, J. Organometal. Chem., 5, (1966), 147. 65. L. M. Haines, Inorg. Chem., 10 (1971), 1685. 66. W. R. Jackson, W. B. Jennings, S. C. Rennison and J. Spratt,

J. Chem. Soc. (B)., (1969), 1214. 67. R. V. Emanuel and E. W. Randall, J. Chem. Soc._., (1969), 3002

and references therein.

68. J. J. Hough and E. Singleton, Chem. Commun., (1972), 371. 69. R. J. Haines and A. L. Du Preez, J. Amer. Chem. Soc., 93 (1971),

2820.

70. J. D. Gilbert and G. Wilkinson, J. Chem. Soc. (A)., (1969), 1749. 71. G. J. Kruger, A. L. Du Preez and R. J. Haines, J. C. S. Dalton,

(1974), 1302. 72. R. A. Zelonka and M. C. Baird, J. Organometal. Chem., (1972),

383.

73. (a) M. A. Bennett, G. B. Robertson, and A. K. Smith, J. Organometal. Chem., /I (1972), 041. (b) M. A. Bennett and A. K. Smith, J. C. S. Dalton, (1974), 233.

74. (a) R. A. Zelonka and M. C. Baird, J. Organometal. Chem., 35, (1972), C43. (b) R. A. Zelonka and M. C. Baird, Canad. J. Chem., 90 (1972), 3063.

75. J. A. Bowden and R. Colton, Aust. J. Chem., _al (1971), 2471.

76. J. A. Bowden and R. Colton, Aust. J. Chem., 26 (1973), 43.

77. G. B. Robertson and P. O. Whimp, J. Organometal. Chem., 60 (1973),

011. 78. (a) Ch. Elschenbroich and F. Stohler, J. Organometal. Chem.,

(1974), C51. (b) Ch. Elschenbroich and F. Stohler, /gew. Chem. Int. Ed., (1975), 174.

79. A. N. Nesmeyanov, M. I. Rybinskaya, V. Vkrivykh and

V. S. Kaganovich, J. Organometal. Chem., 23 (1975), C8.

80. W. S. Trahanovsky and R. A. Hall, J. Organometal. Chem., 25. (1975), 71, and references therein.

81. T. H. Coffield, V. Sandel and R. D. Closson, J. Amer. Chem. Soc.,

12 (1957), 5826. 82. P. L. Paulsonand J. A. Segal, J. C. S. Dalton, (1975), 1677 & 1683.

92

83. D. J. Cole-Hamilton, R. J. Young and G. Wilkinson, in preparation.

84. E. O. Fischer, Ch. Elschenbroich and C. G. Kreiter,

J. Organometal. Chem., 1, (1967), 481.

85. J. C. McConway and A. C. Skapski, J. C. S. Dalton, in preparation.

86. D. J. Cole-Hamilton, private communication.

87. G. E. Coates, M. L. H. Green and K. Wade, 'Organometallic

Compounds', Vol.2 (3rd Ed.), Chapman and Hall (London) (1972).

88. K. K. Joshi and P. L. Paulson, Proc. Chem. Soc., (1962), 326.

89. K. K. Joshi, P. L. Paulson, A. R. Quazi and W. A. Stubbs,

J. Organometal. Chem., 1 (1964), 471.

90. E. O. Fischer and K. 6fele, Z. Naturforsch., 14.6 (1959), 736.

91. E. O. Fischer and K. 6fele, Chem. Ber., 21 (1958), 2395.

92. R. B. King and A. Efraty, J. Organometal. Chem., 20 (1969), 264.

93. C. Segard, C. Pommier, B. Rogues and G. Guiochon,

J. Organometal. Chem., E (1974), 49 & 59.

94. F. Mathey, J. Organometal. Chem., 22 (1975), 377.

95. R. J. Sundberg and R. B. Martin, Chem. Rev., 7A (1974), 471. 96. R. J. Sundberg, R. P, Bryan, I. F. Taylor and H. Taube,

J. Amer. Chem. Soc., 26 (1974), 381.

93

APPENDICES

94

APPENDIX I

THE X-RAY STRUCTURE OF THE HYDRIDOTRIS-(TRIPHITYLPHOSPHINE)RUTHENIUM(II)

ION, 1RuH0/6-PhPPh9lIPPhalL:

The X-ray crystal structure(32)of hydridotris-(triphenylphosphine)- ---- ruthenium(II) tetrafluoroborate confirmed the suggestion based on

spectroscopic studies(31)that one of the phenyl groups of one PPh3

ligand is bound as an arene to the metal. As shown in Figure I the

complex cation also has two PPh3 ligands bonded normally via phosphorus

atoms, and a hydride hydrogen atom completes the coordination about the

metal atom.

The mean Ru-C distance is 2.28X and the distance from the metal

atom to the centroid of the TI-phenyl ring is 1.78A. The distortion of

the ii-phenyl ring from planarity is minimal ( 0.021). The two Ru-P

distances are 2.332 and 2.311X, and the angle between them is 98.7°.

The hydride hydrogen is approximately symmetrical with respect to the

two metal-bonded phosphorus atoms, with H-Ru-P angles of ca. 78°. Thus

the presence of a large flat ligand close to the ruthenium atom pushes

the other three ligand atoms together to give less than tetrahedral

angles between them.

Figures II and III are views from above and below thelq-bonded

ring in the [RuH(76-PhPlph2)(PPh3)2]+ and RuH(176-Ph0)(PPh3)2.2PhOH species

respectively. They illustrate the similarity between the two structures

in that the bond between the substituted ring-carbon and the substituent

group eclipses the metal-hydridic proton bond in both species.

C22

4 C62

C81

C71

C92 C91

sk•

C21 21 q k 1-;

...,,,,,- Ci i i,„,....,,o,sAI82 3: C52 '''' .,...,.--;.--'.' C51

C12

4 C61

C72

Elzure I. illull(76 -.PhPPh2)(PPh3 )2 ]+

95

96

[RuH(76 -PhPPh2)(PP113)2r

97

Ei&ure III. RuH(16-Ph0)(PPh3)2.2Ph011

400

( 4)

(3) (1) 3585 (2) 3305

(2) (3) 3195

98

APPENDIX II

SCHEMATIC REPRESENTATIONS OF SEVERAL INFRARED BANDS OF SPECIFIC COMPLEXES

(1) 1436 (2) 1432

(1)

2)

utoo 660

Figure IV. Bands in [RuH0 6 -PhPa2)(PPh3)2] + characteristic of a Tr-arene ring.

- cm 1 3500

Figure V. p(OH) bands in [Ru(CO2/1e) (1120) (PPh3 )3] BF4 .

- CM

1

99

APPENDIX III

SCHEMATIC REPRESENTATIONS OF SOME N.M.R. RESONANCES OF SPECIFIC COMOCES

J\ 33H. re„, 5.8H.

Element High field resonance for [RuH(/6-POTh2)(PPh3)2J BF4

centred at 18.7Vmeasured in CDC13 at 60 Hz and 35°C.

32]Hz--1

EiBure VII. High field resonance for [RuH(H20)2(,1e0H)(PPh3)21BF4 centred at 18.8T measured in CDC13 at 60Mz and 35°C.

Lizare VIII, Resonances of protons of the rr-bonded ring in [Rull(y 6 -PhPPh2)(PPh3)2]+ in CDCl3 at 100MHz and 30°C.

100

Fic,:ure IX. Resonances of protons of their-bonded ring in {Rul3(2 6 -07118)(PPh3 )2 ]+ in CDC13 at 60NIciz and 35°C,