Oxidative-addition reactions of organic halides and interhalogens with tetrakis(trip-henylphosphine)platinum(o.).

MISSING IMAGE

Material Information

Title:
Oxidative-addition reactions of organic halides and interhalogens with tetrakis(trip-henylphosphine)platinum(o.).
Physical Description:
xii, 100 leaves. : ill. ; 28 cm.
Language:
English
Creator:
Lee, Tong Wai, 1943-
Publication Date:

Subjects

Subjects / Keywords:
Halides   ( lcsh )
Halogens   ( lcsh )
Platinum   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis--University of Florida.
Bibliography:
Bibliography: leaves 97-100.
Statement of Responsibility:
By Tong Wai Lee.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000580718
notis - ADA8823
oclc - 14076358
System ID:
AA00003945:00001

Full Text













OXIDATIVE-ADDITION REACTIONS OF ORGANIC HALIDES AND INTERHALOGENS
WITH TETRAKIS(TRIPHENYLPHOSPHINE)PLATINUM(O)






BY






TONG WAI LEE


A DISSERTATION PRESENTED TO THE GRADUATE
COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL
FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY




UNIVERSITY OF FLORIDA


1974




























DEDICATION


The author proudly dedicates this dissertation

to his mother and to the memory of his father.





























"I can see farther because I stand on the shoulders

of giants."


Issac Newton












ACKNOWLEDGMENTS


The author wishes to express his appreciation to his research

director, Dr. Carl Stoufer, for his interest, support and guidance

for the duration of this work. Thanks are also due to Dr. George

Ryschkewitsch and Neil Weinstein for stimulating and helpful discussions.

The author wishes to thank Dr. John F. Baxter who taught him

how to teach and in the process imparted much chemical knowledge.

Gratitude is also expressed to the United States Government for

a Fulbright-Hays Award, which enabled the author to study in an

American institution.

Finally, the author expresses a very special appreciation and

gratitude to his wife for not only typing this manuscript, but above

all for her love, patience, understanding, encouragement and support

without which this dissertation could not have been written.















TABLE OF CONTENTS


ACKNOWLEDGMENTS . . .

LIST OF TABLES . . .

LIST OF FIGURES . . .

ABSTRACT . . .

INTRODUCTION . . .

EXPERIMENTAL . . .

Materials . . .

Elemental Analysis . .

Dry Box . . .

Spectrometers . . .

Kinetics . . .

Syntheses and Reactions . .

Synthesis of 4-Pyridyldiphenylmethyl Chloride .

Synthesis of 4-Pyridy1di(p-tolyl)methyl Chloride

Synthesis of 4-Methyl-2-thiazolyldiphenylmethyl

Synthesis of 2-Thiazolyldiphenylmethyl Chloride

Synthesis of Di(2-pyridyl)phenylmethyl Chloride
Di(2-pyridyl)(p-methoxylphenyl)methyl Chloride

Synthesis of "Molecular" Silver .


* .

. .

* .

. .



* .

. .

. .

* .

. .

. .

. .




. *



Ch1




and
* .


Page

iv

.viii

ix

xi

S. 1

12

12

12

12

13

13

14

14

15

oride. 16

. 16



. 17


Synthesis of Tetrakis(triphenylphosphine)platinum(0) ..

Synthesis of Cis-diiodobis(triphenylphosphine)platinum(II)






Page
Synthesis of Dichlorobis[4-pyridylbis(p-methoxyphenyl)-
methyl chloride]Pd(II) . . 20
Reaction of Tetrakis(triphenylphosphine)platinum(0) With
Iodine Monochloride. . 20

Reaction of Trans-diiodobis(triphenylphosphine)-
platinum(II) With Iodine Monochloride ... 22
Reaction of Trans-dibromobis(triphenylphosphine)-
platinum(II) With Iodine Monochloride . .. 22
Reaction of Cis-chloroiodobis(triphenylphosphine)-
platinum(II) With Iodine Monochloride .. 22

Reaction of Cis-diiodobis(triphenylphosphine)platinum(II)
With Iodine Monochloride. . 23

Reaction of Tetrakis(triphenylphosphine)platinum(0) With
Iodine Monobromide . . 23

Reaction of Trans-diiodobis(triphenylphosphine)-
platinum(II) With Iodine Monobromide . .. 24

Reaction of Cis-bromoiodobis(triphenylphosphine)-
platinum(II) With Iodine Monobromide . .. 24

Reaction of Cis-diiodobis(triphenylphosphine)platinum(II)
With Iodine Monobromide . .25

Reaction of Trans-diiodobis(triphenylphosphine)-
platinum(II) With Bromine . .. 25

Reaction of Tetrakis(triphenylphosphine)platinum(0) With
Iodine . .. .. 26

Reaction of Tetrakis(triphenylphosphine)platinum(0) With
Bromine . . . 26

Reaction of Tetrakis(triphenylphosphine)platinum(0) With
Chlorine . . . 27

Reaction of Tetrakis(triphenylphosphine)platinum(0) With
Diphenylmethyl Chloride . .. 27
Reaction of Tetrakis(triphenylphosphine)platinum(0) With
Triphenylmethyl Bromide and Triphenylmethyl Chloride .. 27
Reaction of Tetrakis(triphenylphosphine)platinum(0) With
4-Pyridyldiphenylmethyl Chloride . .. 28







Page


Reaction of Chlorocarbonylbis(triphenylphosphine)-
iridium(I) With a Mixture of Trityl and Diphenyl-
methyl Bromide . . .. 29

Reaction of Chlorocarbonylbis(triphenylphosphine)-
iridium(I) With Trityl Bromide . .... .29

Reaction of Chlorocarbonylbis(triphenylphosphine)-
iridium(I) With Diphenylmethyl Bromide ... 30

RESULTS AND DISCUSSION . .. .31

Synthesis . . 31

Generation of Free-Radicals by Reaction of Pt(O) With
Triarylmethyl Halides . .... 33

Visible Spectra . .. 33

Esr Spectroscopy . .... 43

Kinetics and Mechanism of the Reaction of Organic Halides
With Tetrakis(triphenylphosphine)platinum(O) ... 52

Reactions of Tetrakis(triphenylphosphine)platinum(0) With
Interhalogens . ... .... .. .72

Oxidative Addition of Halogens to Tetrakis(triphenyl-
phosphine)platinum(O) . .... 84

Substitution Reactions of Dihalogenobis(triphenylphosphine)-
platinum(II) Complexes. . . 87

CONCLUSION . . ... ........ .94

REFERENCES . . ... ........ .97











LIST OF TABLES

Table Page

1 Amax in the Visible Spectra of Triarylmethyl Radicals
in Benzene . . .. .. 42

2 Rate Constant Dependence on Initial Phosphine
Concentration . . ... .. 53

3 kobsd as a Function of [Ph2CHBr]/[PPh3] . 55

4 Second-Order Rate Constants for the Reaction Between
Pt(PPh3)4 and Organic Halides in Benzene ... 59

5 Activation Parameters for the Reaction Between
Pt(PPh3)4 and Organic Halides at 250 ... 64


viii











LIST OF FIGURES

Figure Page

1 Infrared Spectrum of 2-thiazolyldiphenylmethyl 34
Chloride . . .

2 Infrared Spectrum of Di(2-pyridyl)(p-tolyl)methyl
Chloride . . .. 35

3 Infrared Spectrum of 4-pyridyldiphenylmethyl
Chloride . . 36

4 Infrared Spectrum of 4-pyridyldi(p-tolyl)methyl
Chloride . . 37

5 Infrared Spectrum of Dichlorobis(4-pyridyldiphenyl-
methyl chloride)palladium(II) . .. 38

6 Infrared Spectrum of Dichlorobis[4-pyridyldi(p-tolyl)-
methyl chloride]palladium(II) . .. 39

7 Infrared Spectrum of Dichlorobis[4-pyridylbis(p-
methoxyphenyl)methyl chloride]palladium(II) .. 40

8 Esr Spectrum of Tri(p-tolyl)methyl Radical at 230 45

9 Esr Spectrum of 4-pyridyldiphenylmethyl Radical at 23 46

10 Esr Spectrum of 4-pyridyldi(p-tolyl)methyl Radical at
10 . . 47

11 Esr Spectrum of Radical Derived From the Reaction of
Pt(PPh3)4 With PdC12L2, L = (C5H4N)Ph2CCI. Temp 230 48

12 Esr Spectrum of Radical Derived From the Reaction of
Pt(PPh3)4 With PdCl2L2, L = (C5H4N)(p-CH3C6H4)2CC1.
Temp 23 . . .. 49
13 Esr Spectrum of Radical Derived From the Reaction of
Pt(PPh3)4 With PdCl2L2, L = (C5H4N)(p-CH30C6H4)2CC1.
Temp 230 . . . 50






Figure Page

14 Plot of kobsd vs [Ph2CHBr]/[PPh3] in Benzene at 25 54

15 Plot of k'(K + [L]) vs [L] in Benzene at 250 .. 57

16 Plot of -log k' vs 1/T for the Reaction of Pt(PPh3)4 With
Ph3CC . . 61

17 Plot of -log k' vs 1/T for the Reaction of Pt(PPh3)4 With
CH31 . . . 62

18 Plot of -log k' vs 1/T for the Reaction of Pt(PPh3)4 With
Ph2CHBr . . . 63

19 Infrared Spectrum of (A) Cis- and (B) Trans-PtBr2(PPh3)2 74

20 Proposed Mechanism for the Oxidative Addition to IC1 to
Pt(PPh3)3 in the Absence of Free Phosphine .. 80

21 Proposed Mechanism for the Reaction of Cis-PtC1I(PPh3)2
With IC1 . . . 91

22 Scheme for the Reactions of Pt Complexes With Halogens
and Interhalogens. . .. .. 93







Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

OXIDATIVE-ADDITION REACTIONS OF ORGANIC HALIDES AND INTERHALOGENS
WITH TETRAKIS(TRIPHENYLPHOSPHINE)PLATINUM(O)

By
Tong Wai Lee

December, 1974
Chairman: R. Carl Stoufer
Major Department: Chemistry
The oxidative-addition reactions of some halogens, interhalogens

and arylmethyl halides with tetrakis(triphenylphosphine)platinum(O)
were investigated. The arylmethyl halides used were triphenylmethyl

chloride and diphenylmethyl bromide. Activation parameters for the

reaction of tetrakis(triphenylphosphine)platinum(0) with methyl iodide,

triphenylmethyl chloride and diphenylmethyl bromide were obtained. The

values of the second-order rate constants at 250(M-lsec-1), activation

energies(kcal/mole) and entropies of activation(eu), respectively,

were determined to be: triphenylmethyl chloride, 2.6 x 10-1, 9.4 and

-32; diphenylmethyl bromide, 2.9 x 10-1, 8.2 and -32; and methyl iodide,

2.6 x 10-3, 5.2 and -54. From these data and other evidence, a free-
radical mechanism for the oxidative addition of aromatic tertiary and

secondary halides is proposed.

Several 4-pyridyldiarylmethyl chlorides and their palladium(II)
complexes were synthesized. It was discovered that the reaction of

these compounds with tetrakis(triphenylphosphine)platinum(0) provided
a rapid method of generating 4-pyridyldiarylmethyl radicals.

The oxidative-addition reaction of iodine monochloride to







tetrakis(triphenylphosphine)platinum(0) was found to yield cis-

chloroiodobis(triphenylphosphine)platinum(II), trans-diiodobis(tri-

phenylphosphine)platinum(II) and trans-dichlorobis(triphenylphosphine)-

platinum(II), respectively, when the mole ratio of the Pt(O) complex

to interhalogen was 1:1, 1:3 and 1:6. The formation of such diverse

products merely by changing the stoichiometric amounts of the two

reactants was successfully rationalized and a mechanism for the

initial oxidative addition of iodine monochloride to tetrakis(tri-

phenylphosphine)platinum(O) is proposed. This mechanism predicts

the oxidative addition of iodine, bromine and chlorine to tetrakis-

(triphenylphosphine)platinum(O) would give trans-dihalogenobis-

(triphenylphosphine)platinum(II), in direct conflict with numerous

reports of the isolation of cis-isomers in the literature. These

predictions were verified under carefully controlled conditions.

Iodine monobromide and iodine monochloride were found to react

analogously towards tetrakis(triphenylphosphine)platinum(0).

Many reactions of halogens and interhalogens with PtX2(PPh3)2
and PtXY(PPh3)2, where X and Y = C1, Br or I, were found to be facile

substitution reactions.











INTRODUCTION


Platinum(O) complexes containing triphenylphosphine and related

ligands were first synthesized by Malatesta and co-workers1'2 nearly

17 years ago by reduction of platinum(II)-phosphine complexes with

hydrazine, alcoholic potassium hydroxide and excess triphenylphosphine.

It was first contended3'4 that these were hydrido-complexes of platinum(II)

with coordination numbers of 5 or 6 because hydrido-complexes had been

obtained under similar reduction conditions and because of the similar

stabilities with respect to oxidation. Between 1961 and 1963 several

groups of workers5'6'7 demonstrated conclusively that these complexes
were bona fide zero-valent compounds (defined as compounds having zero

oxidation states) rather than hydrido-complexes; and since then, much

attention has been devoted to the chemistry of zero-valent metal

complexes.

It is now known that the first stage of the reduction of these

triphenylphosphine-platinum(II) complexes is the formation of a hydrido-

complex of platinum(II); in the second stage, the reducing agents, also

bases, dehydrogenate the species:

KOH/EtOH KOH/(n-2)(PPh3)
cis-PtC12(PPh3)2 trans-PtHC1(PPh3)2 Pt(PPh3)n (1)
-Cl- -HC1


N2H4
cis-PtCl2(PPh3)2--- cis- and trans-PtHCl(PPh3)2

N2H4/(n-2)PPh3
2H4n-2) Pt(PPh3)n n = 3, 4 (2)
-HC1







The detailed mechanism of these reductions is complex but some labile

intermediates in the hydrazine reduction have been isolated8 such as

that represented below:

N2H4 N2H4
cis-PtCl2(PPh3)2-----[PtCl(N2H4)(PPh3)2]C--- >



N NAH4 2NH
(PPh Pt Pt(PPh Cl PPh3)PtPt(PPh3)2 Cl2 (3)


NH


-N2 / N2H4

PtHC1(PPh3)2

As a matter of fact, the isolation of either hydrido-complexes of

platinum(II) or platinum(O) depends upon the temperature, the con-

centration of the reducing agent and upon the reaction time.
Zero-valent compounds dissociate in solution giving rise to co-

cordinately unsaturated species via the equilibrium:
M(PAr3)4,= M(PAr3)3 + PAr3= M(PAr3)2 + 2PAr3 (4)

In solution, these low-valent, coordinately unsaturated species such as
bis(triphenylphosphine)platinum(0) are solvated centers to which "oxi-

dative addition" is attributed. The term oxidativee addition" is used

to designate a broad class of reactions, generally of low-spin, transi-
tion metal complexes,in which oxidation (i.e., an increase in the

oxidation number of the metal) is accompanied by an increase in the

coordination number of the metal.







It has been asserted that the unusual stabilities of the coordinate-

ly unsaturated species with respect to tetrakis(triphenylphosphine)-

platinum(0) is a consequence of an excess of negative charge on the

zero-valent metal.9 Thus, the small tendency of the metal ion in these

zero-valent complexes to increase its coordination sphere arises from

an electrostatic repulsion of the a-electron pair of the free ligands

by the non-bonding electrons on the metal. The basic properties of some

dlO, zero-valent systems have been demonstrated recently in complexes

with phosphines in which stable hydrido-compounds such as Pt(H)(X)(PPh3)2

are formed with acids HX where X = C1-, N03-, CN-, SCN-, C104-. Thus,

ligands which increase the electron density on the metal atom cause the

metal to be more susceptible to oxidative addition reactions which results

in the relieving of the high electron density on the metal atom. Con-

versely, ligands which decrease electron density at the central metal

atom decrease the tendency of the metal to undergo oxidative addition.

Indeed, Nixon and Sexton11 have found no evidence for the formation of

any tri- or diccordinate fluorophosphine platinum(0) species (fluoro-

phosphine = PF3, CF3PF2 or (CF3)2PF) in a solution of the tetra-

coordipate complexes. Moreover, in contrast to the triphenylphosphine

platinum(0) complexes, the fluorophosphine analogs do not react with

alkyl halides and hydrogen chloride. This difference in behavior is

attributed to the differing donor and acceptor abilities of the ligands.

With fluorophosphines which are weaker o-donors but stronger r-acceptors

than triphenylphosphine, the electron density at the metal is lowered,

thus reducing its tendency to dissociate and to undergo oxidative-
addition reactions.
The major effort of research activity on zero-valent complexes

has been concerned with the preparations and reactions







of complexes of diverse types and with the elucidation of structures.
Complexes with many different and often novel types of organic and in-

organic ligands have been conveniently synthesized by utilizing oxidative-

addition reactions. Indeed, many of the products formed are inaccessible
by other synthetic routes. For a d10 configuration, such as that repre-

sented by tetrakis(triphenylphosphine)platinum(0), oxidative-addition
reactions are known in which halogens, metal halides, alkyl halides,

aryl halides, acyl halides, sulfonyl halides, inorganic and organic

acids, and other organic molecules serve as the oxidants. Relatively
few mechanistic studies on these reactions have been reported. Further-

more, the dependence of reactivity upon the electronic and structural
factors of reacting species is, in the main, still not well understood.
The oxidative additionsof methyl iodide to tetrakis- and tris-

(triphenylphosphine)platinum(O) yield the o-organometallic compound

of the type PtICH3(PPh3)21213. This type of reaction has been extended

to methyl, styryl and cyclohexyl bromidesl4; benzyl 15, phenyl and per-

fluoroalkyl halides6. Pearson and Rajaram7 have reported the
oxidative addition of methyl iodide to tetrakis- and tris(triphenyl-

phosphine)platinum(O). They found that there is a preliminary

dissociation of tetrakis(triphenylphosphine)platinum(0) yielding un-
saturated, solvated species which subsequently react with the methyl
iodide:
fast
Pt(PPh3)4 Pt(PPh3)3 + PPh3 (5)

K
Pt(PPh3)3 t Pt(PPh3)2 + PPh3 (6)

Pt(PPh3)3 PtICH3(PPh3) + PPh (7)
Pt(PPh3)3 + CH31 + PtlCH3(PPh3)2 + PPh3 (7)






k2
Pt(PPh3)2 + CH3I PtICH3(PPh3)2 (8)

From the kinetic data, they estimated k, and k2 to be 3.5 x 10-3 M-1
sec-1 and 2.0 x 10-2 M-1 sec-1,respectively.
A similar mechanism is proposedl5 in the oxidative-addition reac-
tions between ethylenebis(triphenylphosphine)platinum(0) and methyl

bromide, benzyl bromide and ethylene diiodide (which gives ethylene

and the diiodo-complex):
K
Pt(PPh3)2C2H4 > Pt(PPh3)2 + C2H4 (9)


Pt(PPh3)2 + CH3I PtICH3(PPh3)2 (10)

Although the overall reaction scheme has been established for

these compounds, nothing is yet known of the nature of the transition
state for the alkyl halide addition to bis(triphenylphosphine)platinum(0).

Cook and Jauhal14 have proposed that there is an intermediate or transi-
tion state of the type (A) by analogy with the iridium(I) addition
18
reactions .

[(PPh3)2Pt .....R..R.....
(A)

Another feasible polar mechanism can involve transition state of type

(B), which is a transition state of low polarity in which two addendum

R
M:' -

SX

(B)
atoms interact with the metal simultaneously. However, Pearson and







Rajaraml7 found that, in the case of methyl iodide and tetrakis(tri-

phenylphosphine)platinum(0), the rate of reaction increases with an

increase in the polarity of the solvent, an observation consistent with
a dipolar transition state (A).

Two-electron oxidations of Pt(PPh3)3 by primary alkyl halides
have been reported based on spin-trapping experiments,19 but the results

have been invalidated.20 However, Halpern and his co-workers21,22 and
others23 have demonstrated that alkyl radicals are involved in the one-

electron oxidations of Co(II) complexes with alkyl halides, RX, to give
Co(III) alkyl compounds. The stoichiometry of the reaction

2[Co(CN)5]3- + RX [Co(CN)5R]3- + [Co(CN)5X]3 (11)

and the rate law
-d[Co(CN)53 ]/dt = 2k[Co(CN)53-][RX] (12)

were shown to be in agreement with the mechanism:
k
[Co(CN)5]3- + RX [Co(CN)5X]3- + R* rate-determining step (13)

[Co(CN)5]3- + R*- [Co(CN)5R]3- (14)
The evidence cited for the proposed mechanism22b,23 is as follows:

(1) The isolation of alkyl-acrylonitrile adducts from the reaction of

n-propyl or isopropyl iodide with pentacyanocobaltate(II) in the presence
of excess acrylonitrile indicate the scavenging of alkyl radical inter-

mediates. (2) Dimeric species presumably from free-radical precursors
were found in reactions where the oxidizing agents were trityl and
tropylium halides. (3) When benzenediazonium chloride was the oxidizer,

nitrogen was evolved and a complex was formed which appeared to contain
a phenyl-cobalt bond. In similar processes, aryl radicals are known to

be involved.24 (4) The observed reactivity pattern of pentacyanocobal-

tate(II) toward alkyl halides was found to exhibit an inverse dependence






on the carbon-halogen bond (i.e., kR-C1 < kR-Br < kR-I), an observation
;-compatible with the proposed mechanism. (5) The observed trends in rate
constants (kX-CH2CH2COO- < kX-CH2CO0-) reflect greater stabilility of
the resulting free-radical;trends analogous to those in (4) and (5) have
been found for the rates of halogen abstraction from organic halides by
sodium atoms2, and by organic free radicals26
The oxidation of pentacyanocobaltate(II) by hydrogen peroxide,
hydroxylamine and cyanogen iodide has been proposed27 to proceed by an
analogous free-radical mechanism. Halpern and co-workers found that the
presence of an added amount of iodide ion did not affect the rate of the
reaction between pentacyanocobaltate(II) and hydrogen peroxide, but
it did alter the stoichiometry of the reaction which is:

2Co(CN)53- + H202 + I- + Co(CN)513- + Co(CN)50H3- + OH- (15)

They contended that the iodide ion acted as a scavenger for the hydroxyl
radicals and that the resulting iodine atom subsequently reacted with
pentacyanocobaltate(II) to yield pentacyanoiodocobaltate(III). The
mechanism of the reaction is thus:

Co(CN)53- + H202 + Co(CN)50H3- + -OH rate-determining step (16)

Co(CN)53- + -OH Co(CN)5(OH)3- (17)

*OH + I- OH- + I. (18)

I. + Co(CN)53- Co(CN)513- (19)

The only other postulated free-radical mechanism found in the
literature 28 involves the two-electron oxidation of the primary halides







II a-c to trans-chlorocarbonylbis(triphenylphosphine)iridium(I) resulting
in III C6H5


R1 R2
1 Y2



FX H


M

II, M = Br

a, R1 = R2 = H

b, R1 = H; R2 = D
c, R1 = D; R2 = H

III, M = IrBrC1(CO)(PMe3)2
A mechanism for the oxidative addition to Ir(I) was proposed from the

Ir + Q- Ir -Q Q = initiator (20)

IrII-Q + RBr + Br-IrIII-Q + R- (21)

Ir + R- Ir I-R (22)

IrII-R + R-Br -* Br-IrII-R + R* (23)

following:

(1) The rate of addition is enhanced by free-radical initiators such as

benzoyl peroxide but is retarded by radical scavengers. (2) There is

racemization at the carbon atom in the products III, an observation

consistent with reaction (21) and (22). (3) Competitive experiments







indicate that the rate is enhanced by electron-withdrawing substituents

in the organic halide. This observation has precedence in related radical

processes involving alkyl halides.29
Although the reaction of tetrakis(triphenylphosphine)platinum(O)
with alkyl halides usually results in the addition of both the halogen

and the alkyl moiety to the platinum atom, it was found in some cases

that cis-dihalogenobis(triphenylphosphine)platinum(II) is formed. This
was observed in the reaction of the platinum(O) complex with carbon

tetrachloride,30,31 cis-1,2-dichloroethylene,14 hexachloroethane,14

chloroform14 and bromotrichloromethane.14 Another peculiarity is that,

whereas, tetrakis(triphenylphosphine)platinum(0) has been reported by

Cook and Jauhal14 to react smoothly with trityl bromide to yield the
usual organobromo-platinum(II) complex, the same reaction with trityl

chloride yields trityl free-radicals and dichlorobis(triphenylphosphine)-

platinum(II).19 The factors which favor one organic halide to react to
yield the usual organohalogeno-platinum(II) complex from the reaction

with tetrakis(triphenylphosphine)platinum(0) and another analogous halide
to yield cis-dihalogenobis(triphenylphosphine)platinum(II) should be
investigated.

In the case of the reaction of tetrakis(triphenylphosphine)-

platinum(O) with carbon tetrachloride, cis-dichlorobis(triphenylphos-
phine)platinum(II) and hexachloroethane are formed. One might propose

that the reaction first involves an oxidative addition to form the

normal platinum(II) complex, PtC1(CC13)(PPh3)2, followed by another
oxidative addition to form a platinum(IV) species which subsequently

undergoes a reductive elimination yielding the final products.
In the light of these and similar observations, it seems to be







of interest and of importance to investigate and attempt to elucidate

the mechanism of the reactions between selected organic halides and

tetrakis(triphenylphosphine)platinum(0) which yield dichlorobis(tri-

phenylphosphine)platinum(II) and other products and also to determine

if this behavior arises as a consequence of a new free-radical pathway.

Such an investigation should include kinetic studies of the reactions

and the evaluation of activation parameters from which data on the

nature of the transition states may be inferred. Further, an attempt

to correlate reactivity and mechanism with structural aspects of the

addendum molecule would shed more light on these oxidative-addition

reactions. This work will be extended to the investigation of d8

systems such as Rh(I) and Ir(I) and a comparison of the reactions and

mechanisms with those of d10, platinum(O) systems.

The work of Lappert and Lednor19 involving the reaction of tetra-

kis(triphenylphosphine)platinum(0) with trityl chloride to form trityl

radicals and cis-dichlorobis(triphenylphosphine)platinum(II) was repeated

and confirmed. Logically, the report of Cook and Jauhal involving the

reaction of tetrakis(triphenylphosphine)platinum(0) with trityl bromide

was also repeated in this laboratory but no organohalogeno-platinum(II)

complex could be isolated as claimed by the authors; instead trityl

radicals and cis-dibromobis(triphenylphosphine)platinum(II) were formed

as is to be expected by analogy with the reaction using trityl chloride

and the platinum(O) complex. Moreover, the rate of these reactions was

found to be relatively fast, the reaction being completed in minutes.

Thus, it would be useful to investigate the feasibility of the general

applicability of this method for the generation of triarylmethyl radicals

as this would be an improvement to the classical method of Gomberg32




11


who used molecular silver and the triarylmethyl chloride, a reaction

which takes up to 4 days for complete reaction.33

Finally, the objective of this work is the synthesis of some 4-

pyridyldiarylmethyl and di(2-pyridyl)arylmethyl chlorides and the

reaction of these with the platinum(O) complex in the attempted syn-

thesis of the corresponding, hitherto unknown, free-radicals (a reaction

analogous to that between platinum(O) complex and trityl chloride) so

that their stabilities and other properties can be evaluated.












EXPERIMENTAL


Materials


Common chemicals were of reagent grade and were used without

further purification unless otherwise specified. Palladium complexes

were synthesized according to the method of Richardson34 by reacting

tetrachloroplatinate(II) with the appropriate ligand in methanol.

Solvents used in the synthesis of tertiary organic halides and in

reactions involving the use of interhalogens were dried by distilla-

tion over a suitable drying agent. Iodine monochloride and iodine

monobromide were obtained commercially and used without further

purification. Chlorocarbonylbis(triphenylphosphine)Ir(I) was also

obtained commercially and was recrystallized before use.


Elemental Analysis


Elemental analyses were performed by Galbraith Laboratories,

Inc., Knoxville, Tennessee. All samples for analyses were dried in

vacuo at the boiling point of acetone, n-heptane or xylene to remove

solvent molecules from the crystals.


Dry Box


All reactions involved in the synthesis of tertiary organic

halides and others requiring a dry or inert atmosphere were carried

out in a Vacuum Atmosphere dry box (model: Dri-Train).







Spectrometers


All infrared spectra were obtained using a Beckman IR-10 spec-

trophotometer. Samples of the compounds were prepared for analysis

as KBr discs.

Visible and ultraviolet spectra were obtained with a Cary 15

recording spectrophotometer. Quartz cells of 1.00 cm path lenghts

were used.
H nmr spectra were obtained at 60 Hz using a Varian A60-A

spectrometer. The chemical shifts are reported in Hz from tetra-

methylsilane which was used as the internal reference.

Electron spin resonance spectra were obtained using a Varian

E-3 model spectrometer.

Kinetics


All the kinetic measurements were made at 250 in benzene solu-

tion unless otherwise stated. The reactions were followed

spectrophotometrically using a Cary 15 recording spectrophotometer

with the cell holder thermostated to within + 0.100. The rates of

the reactions were measured by following as a function of time the

disappearance of the tetrakis(triphenylphosphine)platinum(0) at 415 nm

(c 1.5 x 103 M-1 cm-1) in the case of triphenylmethyl chloride and

methyl iodide and at 425 nm (e 1.3 x 103 M-1 cm-) in the case of

diphenylmethyl bromide. Customarily, the initial concentrations of

the platinum(0) complex employed were between 5.0 x 10-4M and 1.0 x

10-3M, those of triphenylphosphine between 1.0 x 10-3M and 1.0 x 10-2M

and those of the reacting halide were > 50 times that of the platinum(O)

complexes to maintain a constant concentration of reacting halide.







Benzene was dried and deoxygenated by distillation over sodium

benzophenone ketyl and bubbling a stream of pre-purified nitrogen

through the distilling solution. Methyl iodide was distilled under

vacuum and degassed by freeze-thawing. The stock solutions were

prepared in a dry box and 1 ml of the platinum(0) complex and tri-

phenylphosphine solution were pipetted into a 1 cm cuvette which,

together with the flask containing the organic halide stock solution,

was then sealed with a rubber serum cap. The cuvette and flask and

a hypodermic syringe were put into glass tubes immersed in the thermo-

stated water bath for 15 minutes to achieve thermal equilibrium of

reactants. Using the syringe, 1 ml of the organic halide stock was

quickly transferred to the cuvette and the absorbance was plotted as

a function time by the spectrophotometer. A plot of log (A-Am) vs

time was found to be linear for at least 75% reaction.


Syntheses and Reactions


Synthesis of 4-Pyridyldiphenylmethyl Chloride

A suspension of 4-pyridyldiphenylmethanol (10 g, 0.038 mole)

in 100 ml of carbon tetrachloride was brought to reflux and 9.0 ml

of freshly distilled thionyl chloride in 3 ml aliquots was added

cautiously at 5-minute intervals. The solid dissolved and the solu-

tion was refluxed for one hour after which it was concentrated by

distillation to about 20 ml. Addition of benzene and cooling gave

crystals of the hydrochloride salt of 4-pyridyldiphenylmethyl chloride.

Recrystallization from methylene chloride-benzene afforded 7 g (70:;)

of colorless crystals, mp 1870(dec.).

To a methylene chloride solution of the hydrochloride salt (6.3 g)







prepared above was added an excess of 2,6-1utidine (12 ml), whereupon

colorless crystals of 2,6-1utidinium chloride (mp 234-5 dec) deposited.

Petroleum ether (10 ml) was added to effect more complete precipita-

tion of the lutidinium chloride. The precipitated solid was removed

by filtration; and the filtrate was evaporated to dryness under

reduced pressure to yield 4-pyridyldiphenylmethyl chloride. This could

be purified by recrystallization from carbon tetrachloride, petro-

leum ether or by sublimation in vacuo to give 2.5 g (42%) of product

of mp 89.5-90. Anal. Calcd for C18H14NC1: C, 77.28; H, 5.04; N, 5.01;

C1, 12.67. Found: C, 77.86; H, 5.08; N, 4.93; Cl, 12.74.

Synthesis of 4-Pyridyldi(p-tolyl)methyl Chloride

A stream of dry HC1 gas was slowly bubbled through a solution

containing 6.0 g (20 mmole) of 4-pyridyldi(p-tolyl)methanol in a mixture

containing an equal volume of CHC13 and CH2C12. After 2.5 hours, the

solution was evaporated to dryness under reduced pressure to give the

hydrochloride salt of 4-pyridyldi(p-tolyl)methyl chloride. This was

recrystallized by dissolving in methylene chloride and precipitating

with benzene or n-hexane.

To a solution containing 2.5 g of the above recrystallized

hydrochloride in 20 ml of methylene chloride was added 5 ml of 2,6-

lutidine. The agitated mixture was allowed to stand for 10 minutes

during which colorless crystals of 2,6-lutidinium chloride deposited

and was removed by filtration. On concentrating the filtrate under

reduced pressure, more 2,6-1utidinium chloride deposited and it was

again filtered off. The mixture of solvents was then completely

evaporated off under reduced pressure to yield a rather thick and

yellowish viscous liquid. After standing the flask over a cold plate,







slightly yellow-colored crystals slowly formed. The solid was broken

up with a spatula and washed with a small quantity of cold petroleum

ether to give crystals having a melting point of 68-740. Recrystalli-

zation from petroleum either yielded 1.6 g (72%) of colorless crystals,

mp 74-75. Anal. Calcd for C20H18NC1: C, 78.04; H, 5.89; N, 4.55;

Cl, 11.52. Found: C, 77.94; H, 6.11; N, 4.45; C1, 11.41.


Synthesis of 4-Methyl-2-thiazolyldiphenylmethyl Chloride

To a refluxing solution of carbon tetrachloride (15 ml) containing

4-methyl-2-thiazolyldiphenylmethanol (1.1 g) was added 2 ml of thionyl

chloride. After 30 minutes, most of the liquid was distilled off and

the solution was evaporated to dryness under reduced pressure to give

a pale yellow oil. This was dissolved in a little carbon tetra-

chloride and on adding petroleum ether and standing on a cold plate,

colorless crystals deposited, mp 62-64. The crude product was then

purified by sublimation in vacuo to yield 0.7 g (65%) of pure product,

mp 64-65. Anal. Calcd for C17H14ClNS: C, 68.10; H, 4.71; N, 4.47,

Cl, 11.83. Found: C, 67.88; H, 4.92; N, 4.47; C1, 11.66.


Synthesis of 2-Thiazolyldiphenylmethyl Chloride

A suspension of 2-thiazolyldiphenylmethanol (2.0 g) in 20 ml

methylene chloride and 5 ml of carbon tetrachloride was brought to

a reflux and 5 ml of thionyl chloride was then added. The solution

was refluxed for another 4 hours and the solvent was then evaporated

off under reduced pressure to give a purplish and then a greenish gum.

After the gum was dissolved in methylene chloride, 5 ml of 2,6-1utidine

was added and the solution concentrated to give a white solid which





17


was removed by filtration. The filtrate was then evaporated to dryness

to give a very dark brown solid. To this solid was added 10 ml of

carbon tetrachloride whereupon most of the solid dissolved. Activated

charcoal was added and this mixture was boiled for 5 minutes. On

filtration and concentration of the solvent, petroleum ether was added

and the solution was placed on a cold plate. Crystals formed on the

side of the flask were separated by filtration and purified by sublima-

tion in vacuo to give a colorless solid, mp 68.5-69.50. The mass

spectrum of the product gave two peaks of mass 287 and 285 corresponding

to m/e of the 2-thiazolyldiphenylmethyl chloride ion containing 37C1

and 35C1, respectively. Anal. Calcd for C16H12CINS: C, 67.24; H, 4.23;

N, 4.90, C1, 12.41. Found: C, 67.13; H, 4.27; N, 4.75; C1, 12.14.

Synthesis of Di(2-pyridyl)phenylmethyl Chloride and Di(2-pyridyl)-
(p-methoxylphenyl)methyl Chloride

The reaction between 4.0 g of di(2-pyridyl)phenylmethanol and

8 ml of thionyl chloride gave the hydrochloride of di(2-pyridyl)-

phenylmethyl chloride. Treatment of this with excess lutidine and

work-up in a similar manner as for the preparation of 2-thiazolyldi-

phenylmethyl chloride gave colorless crystals of melting point 107.5-

108.5. The mass spectrum of the product gave two peaks of mass 282

and 280 corresponding to m/e of the molecular ion containing 37CI

and 35C1, respectively. Anal. Calcd for C17H13C1N2: C, 72.73; H,

4.67; N, 9.98; Cl, 12.63. Found: C, 72.79; H, 4.69; N, 9.87; Cl,

12.71.

Similarly, using di(2-pyridyl)(p-methoxyphenyl)methanol and the

same procedure yielded a white solid of melting point 91.5-92.5. The

mass spectrum of the product gave two peaks corresponding to m/e




18


296 and 294 of the parent molecular ion containing 37C1 and 35C1,

respectively. Anal. Calcd for C18H15C1N20: C, 73.34; H, 5.13;

N, 9.50; C1, 12.03. Found: C, 73.13; H, 4.90; N, 9.36; Cl, 12.27.

Synthesis of "Molecular" Silver

This element was prepared by internal electrolysis following the

procedure of Gomberg.32 Pure, well-washed silver chloride (100 g) was

placed in a beaker and covered with water and a finely porous procelain

cell which contained water and several zinc bars, was placed upon the

silver chloride. A piece of platinum sheet was put into the silver

chloride and the zinc bars and platinum sheet were connected by a wire.

Several drops of concentrated hydrochloric acid was then added to the

water in the cell allowing the initial rate of the reaction to increase

substantially. The reduction was completed in about 2 days.

The gray, powdery silver was first washed with water, 6 M NH3,

then again with water and finally with ethanol and ether. After

drying in vacuo over sulfuric acid, it was heated to 1500 and finally

forced through a 100-mesh sieve. The product is more reactive than

the less finely-divided, commercially available "silver powder."


Synthesis of Tetrakis(triphenylphosphine)platinum(0)

This compound was prepared according to the procedure35 of

R. Ugo, Cariati and La Monica except that the preparation was carried

out under nitrogen. Triphenylphosphine (15.4 g 0.0588 mole) was

dissolved in 200 ml of absolute ethanol and the solution heated to

650. Potassium hydroxide (1.4 g) in 32 ml ethanol and 8 ml water was

then added. Then 5.24 g (0.0126 mole) of potassium tetrachloroplati-

nate(II) dissolved in 50 ml of water was added dropwise to the







alkaline triphenylphosphine while stirring at 650. Addition was

completed in 20 minutes. After cooling, the yellow compound which

precipitated during the addition was recovered by filtration under

nitrogen, washed with 150 ml of warm ethanol, then with 60 ml of cold

water and again with 50 ml of cold ethanol. The resulting yellow

powder was dried in vacuo at room temperature for 2 hours. The yield

was 12.5 g (80%).

Synthesis of Cis-diiodobis(triphenylphosphine)platinum(II)

A modification of the method of Mastin36 involving the metathesis

reaction between cis-dichlorobis(triphenylphosphine)platinum(II) and

sodium iodide was employed. A mixture of 0.34 g of cis-dichlorobis-

(triphenylphosphine)platinum(II) and 2.6 g of sodium iodide (1:40 mole

ratio) in 40 ml of a solvent mixture containing equal volumes of chloro-

form, acetone, ethanol and water was refluxed for 4 hours. After

separating from the aqueous layer, the bright yellow organic layer was

evaporated to dryness. The bright yellow powder was then washed with

water, benzene and finally with ethanol to give 0.43 g (90%) of product,

mp 307-3090 (lit37 mp 303-3040). The ir spectrum in the 650-350 cm-1

region showed 4 strong absorptions characteristic of the cis-isomer and

compared to only 3 for the trans-isomer. On refluxing a chloroform

solution of this cis-isomer, the orange trans-isomer (identical to

the product from the reaction of tetrakis(triphenylphosphine)-

platinum(O) with iodine) is obtained. The cis-trans isomerization

can also be effected merely by heating the solid cis-isomer at 200.







Synthesis of Dichlorobis[4-pyridylbis(p-methoxyphenyl)methyl chloride]-
Pd(II)

A suspension of PdC12L2 (1.7 g), where L = 4-pyridylbis(p-methoxy-

phenyl)methanol, in 10 ml of carbon tetrachloride was brought to a

reflux and 5 ml of redistilled thionyl chloride was added. After

1/2 hour, during which all the solid went into solution, the solvent

was evaporated off under reduced pressure whereupon a yellow solid

precipitated out. The yellow solid was recrystallized twice in

methylene chloride-hexane to yield 0.6 g (35%) of the product; mp 118-

1210. Anal. Calcd for C38H36C14N204Pd: Cl, 16.55. Found: C1, 16.60.

Similarly, PdC12L2 where L = 4-pyridyldiphenylmethyl chloride or

4-pyridyldi(p-tolyl)methyl chloride were prepared. Anal. Calcd for

C36H30C14N2Pd: C, 58.67; H, 3.83; N, 3.80, C1, 19.24. Found: C, 58.67;
H, 3.80; N, 3.78; C1, 18.96. Anal. Calcd for C38H36C14N2Pd: C, 59.84;

H, 3.96; N, 3.67; C1, 18.59. Found: C, 59.87; H, 4.55; N, 3.88;,

C1, 18.43.

Reaction of Tetrakis(triphenylphosphine)platinum(0) With Iodine
Monochloride

A solution of 0.50 g of tetrakis(triphenylphosphine)platinum(0)

in 15 ml of benzene was mixed under nitrogen with a solution of 0.39 g

of iodine monochloride (1:6 mole ratio) in 10 ml of ether. The mixture

was then shaken for 5 minutes during which the color of the solution

became rather dark and a nearly black solid began to form on the

surface of the flask. Methanol (70 ml) was then added and the solution

was shaken for 5 minutes during which the color of the solution became

reddish brown with the settling down of a fine yellow precipitate.

After standing for 5 minutes, the yellow precipitate was separated by






filtration and washed with methanol. The yield was 0.305 g (95%) of
trans-PtCl2(PPh3)2, mp 3120 dec (lit36 310-314). The ir spectrum of
the product was identical to that of an authentic sample of trans-

PtC12(PPh3)2 and showed an absorption at vmax 340 cm-1 indicative of
the CI-Pt-C1 asymmetric stretch (lit37 342 cm-1). Recrystallization
from benzene gave finally 0.240 g (74%) of lemon-yellow crystals.
Anal. Calcd for C36H30C12Pt: Cl, 8.97. Found: C1, 8.88.

Similarly, by following the above procedure, the reaction of
0.50 g of Pt(PPh3)4 and 0.195 g of IC1 (1:3 mole ratio) yielded a
yellow solution and a yellow-orange precipitate. Recrystallization
from chloroform-ethanol or from benzene resulted in 0.32 g (81%) of
trans-PtI2(PPh3)2, mp 304-3050 (lit36 307-3080). It was identified
by elemental analysis and by its ir spectrum. Anal. Calcd for

C36H3012Pt: I, 26.05. Found: I, 26.47.
The reaction was repeated for a third time using Pt(PPh3)4 (1.0 g)
and ICl (0.18 g) in a 1:1.1 mole ratio. The yield of a yellow solid,
sparingly soluble in benzene, was 0.34 g. A small quantity (50 mg) of
trans-PtCl2(PPh3)2 was also isolated by extracting the precipitate
with 50 ml of hot benzene. The ir spectrum of the yellow solid in-
dicated that Pt(II) complex had a cis-configuration, and the formula
cis-PtC1I(PPh3)2 was assigned to it. The elemental analysis was
consistent with this complex formula, but contaminated with PtI2(PPh3)2.
However, when this yellow solid was reacted with ICl in a 1:2.5 mole
ratio, cis-PtC12(PPh3)2 was isolated in 89% yield. This observation
lends further support to the assigned formula, cis-PtClI(PPh3)2, for
the yellow solid.







Reaction of Trans-diiodobis(triphenylphosphine)platinum(II) With
Iodine Monochloride

To a chloroform solution containing 269 mg of trans-diiodobis-
(triphenylphosphine)platinum(II) was added an ethereal solution

containing 300 mg of iodine monochloride (1:6 mole ratio). The mixture
was shaken for 10 minutes during which the solution turned dark violet

in color. On adding 75 ml of methanol, a yellow solid precipitated

out from the brownish-orange solution. After filtration, the solid

was washed with methanol yielding 185 mg (77%) of trans-dichlorobis-

(triphenylphosphine)platinum(II), identified by its ir spectrum and by

elemental analysis. The yellow product dissolved completely in benzene
and was recrystallized in this solvent, mp 308-3110. Anal. Calcd for

C36H30C12Pt: C1, 8.87, Found: C1, 9.18.

Reaction of Trans-dibromobis(triphenylphosphine)platinum(II) With
Iodine Monochloride

The reaction of 100 mg of trans-dibromobis(triphenylphosphine)-

platinum(II) with 45 mg of iodine monochloride (1:2.2 mole ratio)

yielded, after the usual work-up, 74 mg of trans-PtC12(PPh3)2 (84%),
mp 309-3110. Analysis of the sample obtained after recrystallization
from benzene gave the following results: Calcd for C36H30C12Pt: C,

54.69; H, 3.82; C1, 8.97. Found: C, 54.05; H, 3.94, Cl, 8.93.

Reaction of Cis-chloroiodobis(triphenylphosphine)platinum(II) With
Iodine Monochloride

A mixture (1:2.5 mole ratio) of PtClI(PPh3)2 (43 mg) dissolved

in 10 ml CH2C12 and IC1 (20 mg) dissolved in 2 ml ether was stirred
for 5 minutes. The solution was reduced, under reduced pressure,

to 1/2 of its original volume and,after addition of 30 ml of ether,







white crystals of cis-PtCl2(PPh3)2 precipitated out. The product was

identified by its ir spectrum. Recrystallization from chloroform-ether

yielded the pure product (32 mg, 89%).

Reaction of Cis-diiodobis(triphenylphosphine)platinum(II) With
Iodine Monochloride

A 20 ml solution of CH2C12 containing 70 mg cis-PtI2(PPh3)2 was
mixed with 2 ml of an ethereal solution containing 36 mg IC1 (1:2.5 mole

ratio). After stirring for 5 minutes, the solvent was concentrated

to 2 ml by a stream of nitrogen. On addition of 25 ml of ether and
25 ml of ethanol, a white precipitate settled out and was separated

by filtration, washed with ethanol and then with ether. The yield of

cis-PtCl2(PPh3)2 was 55 mg (96%); and the product was identified by

melting point (3050) and by its ir spectrum.

Reaction of Tetrakis(triphenylphosphine)platinum(0) With Iodine
Monobromide

The same procedure for the reaction of Pt(PPh3)4 with ICl was

followed. The reaction of 0.50 g of the Pt(PPh3)4 with 0.50 g of

iodine bromide (1:6 mole ratio) afforded 0.360 g of yellow trans-
dibromobis(triphenylphosphine)platinum(II) having a melting point of

312-314 (lit36 312-3140). On recrystallization from chloroform-

ethanol, 0.311 g (86%) of the product was obtained. Anal. Calcd for

C36H30Br2Pt: C, 49.16; H, 3.44; Br, 18.17. Found: C, 48.75; H,
3.39; Br, 17.99.

By using a 1:1 mole ratio, the reaction for 5 minutes of
Pt(PPh3)4 (0.93 g) with iodine monobromide (0.163 g) in benzene

solution caused the precipitation of 0.48 g of a yellow precipitate.







After washing with EtOH, the solid was heated with 70 ml of benzene

and the insoluble portion of the solid was then recrystallized from

methylene chloride-benzene to give 0.20 g (30%) of cis-bromoiodobis-

(triphenylphosphine)platinum(II) having a melting point of 308-3090.
The cis-isomer was deduced from the ir spectrum. Also isolated was

70 mg of trans-diiodobis(triphenylphosphine)platinum(II). Anal. Calcd

for C36H30BrlPt: Br, 8.63; I, 13.70. Found: Br, 8.47; I, 13.59.

The stoichiometric amounts of the reactants were further varied

by using 1:2 and 1:3 mole ratios of Pt(PPh3)4 to IBr. In each case

a mixture of cis-PtBrI(PPh3)2 and trans-PtI2(PPh3)2 were obtained.

In the 1:2 reaction, approximately equivalent amounts of each product

were isolated, whereas in the 1:3 reaction, 40% of cis-PtBrI(PPh3)2

and 60% of trans-PtI2(PPh3)2were obtained.

Reaction of Trans-diiodobis(triphenylphosphine)platinum(II) with
Iodine Monobromide

A benzene solution containing 100 mg of trans-diiodobis(tri-

phenylphosphine)platinum(II) and an ethereal solution containing

50 mg of iodine monobromide (1:2 mole ratio) was shaken for 10 minutes.

Addition of 50 ml of methanol caused the precipitation of 71 mg of a

yellow precipitate of trans-dibromobis(triphenylphosphine)platinum(II)

(79%), melting point, 313-315. After recrystallization from benzene,

the sample was sent for analysis. Anal. Calcd for C36H30Br2Pt:

Br, 18.17. Found: Br, 17.86.

Reaction of Cis-bromoiodobis(triphenylphosphine)platinum(II) with
Iodine Monobromide

To 50 mg of cis-bromoiodobis(triphenylphosphine)platinum(II)







dissolved in 10 ml of methylene chloride was added an ethereal solution

containing 25 mg of iodine monobromide (1:2.2 mole ratio). After

shaking the mixture for minutes, the solution was concentrated to

one-half the original volume under reduced pressure. On addition of

35 ml of methanol, 39 mg (84%) of cis-dibromobis(triphenylphosphine)-

platinum(II) precipitated out, and was identified by its ir spectrum.

Reaction of Cis-diiodobis(triphenylphosphine)platinum(II) with Iodine
Monobromide

A methylene chloride solution containing 60 mg of cis-diiodobis-

(triphenylphosphine)platinum(II) was mixed with 2 ml of an ethereal

solution containing 30 mg of iodine bromide. The mixture was stirred

for 5 minutes after which the volume of the solution was reduced to

about 3 ml by a stream of nitrogen. On addition of 40 ml of methanol,

a pale yellow powder (45 mg, 83%) precipitated out, which was subse-

quently recrystallized from chloroform-methanol, melting point, 307-

309 and identified by ir spectroscopy as cis-dibromobis(triphenyl-

phosphine)platinum(II).

Reaction of Trans-diiodobis(triphenylphosphine)platinum(II) with Bromine

Trans-PtI2(PPh3)2 (150 mg) was dissolved in a minimum amount

of benzene and an ethereal solution containing 48 mg of bromine was

then added to it. The mixture was stirred for 10 minutes after which

methanol (50 ml) was added to it to cause the precipitation of trans-

PtBr2(PPh3)2 (0.130 mg, 96%). The product was removed by filtration,

washed with methanol and ether, and subsequently recrystallized from

benzene.







Reaction of Tetrakis(triphenylphosphine)platinum(0) with Iodine

The reaction of tetrakis(triphenylphosphine)platinim(0) (0.50 g,

0.4 mmole) dissolved in 20 ml of benzene, and 0.30 g (1.2 mmole) of

iodine produced a pale orange precipitate. After 10 minutes, 40 ml

of methanol was added and on standing a further 10 minutes, filtering,

and washing with methanol, 0.41 g (86%) of an orange precipitate was

collected. Recrystallization from chloroform-methanol gave a bright

orange precipitate, melting point, 314-315. Heating the solid in

vacuo at 1000 resulted in loss of solvated chloroform. This compound

is also soluble in benzene and its ir spectrum is identical to that

of an authentic sample of trans-diiodobis(triphenylphosphine)platinum(II)

in the 650-350 cm-1 region36

Reaction of a suspension of tetrakis(triphenylphosphine)platinum(0)

in ethanol with iodine according to the method of Tayim and Ak129 also

gave the same product, even though these authors did not specify which

isomer was obtained.

Reaction of Tetrakis(triphenylphosphine)platinum(0) with Bromine

A 1:4 mole ratio of Pt(PPh3)4 to Br2 was used. Pt(PPh3)4 (0.529 g)

was dissolved in 15 ml of benzene and an ethereal solution containing

0.256 g of Br2 was added to this solution. The mixture was stirred

for 3 minutes after which 30 ml of MeOH was added to cause more

complete precipitation of the product, trans-PtBr2(PPh3)2 (0.371 g,

98%). The ir spectrum of the precipitated product, which was identical

to that of an authentic sample, did not indicate the presence of any

cis-isomer. Recrystallization from benzene eventually yielded 0.240 g







(65%) of the pure product. Anal. Calcd for C36H30Br2Pt: C, 49.16;

H, 3.44; Br, 18.17. Found: C, 48.94; H, 3.36; Br, 17.96.

Reaction of Tetrakis(triphenylphosphine)platinum(0) with Chlorine
Chlorine gas was bubbled for 2 minutes through 3 ml of benzene.

The yellow solution was then stirred and 15 ml of a benzene solution

containing 0.50 g of Pt(PPh3)4 was added in a fast dropwise fashion.
After 1.0 minute, 50 ml of MeOH was added to the mixture to precipitate

the product, trans-PtCl2(PPh3)2. No cis-isomer could be detected from
an ir spectrum of the product which was recrystallized from benzene

(yield: 70%). Anal. Calcd for C36H30C12Pt: C, 54.62; H, 3.82;

Cl, 8.97. Found: C, 54.58; H, 3.85; C1, 8.84.

Reaction of Tetrakis(triphenylphosphine)platinum(0) with Diphenyl-
methyl Bromide

Tetrakis(triphenylphosphine)platinum(0) (0.5 g, 0.40 mmole) in

20 ml of benzene was mixed with a benzene solution containing 0.20 g
(0.85 mmole) of diphenylmethyl bromide, the yellow Pt(0) complex
solution was decolorized and on standing for 1/2-3/4 hour, the pale

yellow solid was filtered off to give 0.31 g (88%)of cis-dibromobis-
(triphenylphosphine)platinum(II), melting point, 307-308. No esr

signal could be detected from the reaction mixture and work-up of
the mother liquor yielded 80 mg of sym-tetraphenylethane having a
melting point of 209-2110 (lit38 208-210)

Reaction of Tetrakis(triphenylphosphine)platinum(0) with Triphenyl-
methyl Bromide and Triphenylmethyl Chloride

The reaction between 0.95 g (0.79 mmole) of tetrakis(triphenyl-

phosphine)platinum(0) dissolved in 20 ml of benzene and a 10 ml solution








of benzene containing 0.40 g (1.24 mmole) of triphenylmethyl bromide
afforded, after 1/2 hour, 0.54 g (80%) of crystals of cis-dibromobis-

(triphenylphosphine)platinum(II). The solution gave a very intense

but poorly resolved esr signal; however, the resolution was much

improved when the radical concentration was diluted. Examination

of the esr spectrum showed it to be the triphenylmethyl by comparison

with the literature spectrum39. The radical was also characterized

as ditritylperoxide,identified by its melting point, 178-180 (lit40

178-179) and by comparison with the ir spectrum40 of an authentic

sample of ditritylperoxide.

In a similar procedure, the reaction of the Pt(0) complex with

triphenylmethyl chloride for 5 minutes was shown to yield cis-

dichlorobis(triphenylphosphine)platinum(II) (43%) and triphenylmethyl

radicals.

Reaction of Tetrakis(triphenylphosphine)platinum(0) with 4-Pyridyldi-
phenylmethyl Chloride

Tetrakis(triphenylphosphine)platinum(0) (0.93 g) was dissolved

in 15 ml of benzene and an equimolar amount of 4-pyridyldiphenylmethyl

chloride (0.204 g) was added as a benzene solution. The yellow

solution turned dark brown and a precipitate came down. On standing

for 1/2 hour, the solid was filtered off to yield 2.30 g of cis-

dichlorobis(triphenylphosphine)platinum(II), identified by its melting

point of 3050 and an infrared spectrum. The reaction mixture gave a

strong esr signal, indicative of free-radicals.

When the mother liquor was evaporated to dryness in the dry box,

a yellow gum resulted. This was then redissolved in some benzene,







adsorbed on alumina and eluted with benzene. The eluent from an

intense yellow band was evaporated to dryness to give sticky, orange
crystals; melting point, 200-2300. Mass spectrum analysis of this
solid gave intense peaks at m/e = 488 and 244, consistent with the
assignments C36H28N2+ and C18H14N+ respectively.

Reaction of Chlorocarbonylbis(triphenylphosphine)iridium(I) with
Trityl Bromide
Benzene solutions containing IrCI(CO)(PPh3)2 (0.272 g) and
Ph3CBr (0.250 g) in a 1:2.2 mole ratio were mixed in the dry box and
stirred. The yellow solution turned orange and after 5 minutes a
sample of it was taken out of the dry box and examined by esr
spectroscopy. A strong signal was obtained and the detected radical
was identified as Ph3C*. The solution was concentrated to 10 ml under
reduced pressure. After 3 hours of stirring, filtration of the
reaction mixture yielded 0.270 g (80%) of a yellow precipitate,
melting point, 304-3080. The product was recrystallized from an
equivolume of CHC13 and CH2C12 (mp 310-3120) and was identified as
IrBr2C1(CO)(PPh3)2 by comparison of its ir spectrum with that of an

authentic sample prepared by reaction of IrCI(CO)(PPh3)2 with Br2.41

Reaction of Chlorocarbonylbis(triphenylphosphine)iridium(I) with
Diphenylmethyl Bromide
Chlorocarbonylbis(triphenylphosphine)iridium(I) (0.270 g) was
dissolved in 50 ml of benzene and 0.191 g of Ph2CHBr was added to
the solution. The mixture was stirred and after 16 hours the yellow
precipitate (0.262 g),identified by ir spectroscopy as IrBr2C1(CO)(PPh3)2,
was removed by filtration. The filtrate was then concentrated under

reduced pressure to a small volume (2 ml). Addition of methanol (20 ml)







yielded more yellow solid (55 mg) which was filtered off. The total

isolated product containing Ir was 0.310 g (95%). The mother liquor

was then evaporated to dryness yielding colorless crystals of sym-

tetraphenylethane (mp 208-210).

Reaction of Chlorocarbonylbis(triphenylphosphine)iridium(I) with a
Mixutre of Trityl and Diphenylmethyl Bromide

Equimolar quantities of the three reactants were used in this

experiment. A benzene solution containing trityl and diphenylmethyl

bromide (114 mg and 87 mg, respectively) was added to a benzene

solution of IrC1(CO)(PPh3)2 (275 mg). The yellow solution turned

orange and after several minutes the yellow color returned. After

2 hours, the reaction mixture was examined by esr spectroscopy. No

radicals, however, could be detected. The reaction was stirred over-

night and the precipitated product (0.299 g, 90%) IrBr2C1(CO)(PPh3)2,

was removed by filtration. The solvent was then removed from the

mother liquor to give 169 mg of pale yellow solid which was dissolved

in a minumum amount of cyclohexane. This solution was then filtered,

and a little petroleum ether was then added and the solution allowed

to stand on a cold plate whereupon pentaphenylethane precipitated

out, mp 168-1800 (lit42 168-1840). Mass spectrum analysis of this

compound showed intense peaks corresponding to m/e of 243 and 167,

corresponding to Ph3C+ and Ph2CH respectively. Weak peaks at m/e

410 and 333 assigned to the molecular ion and Ph4C2H+ were also

observed.













RESULTS AND DISCUSSION


Synthesis


The triarylmethyl chlorides, intended precursors in the generation

of the corresponding triarylmethyl radicals, were prepared from the

corresponding methanols by chlorination with either hydrogen chloride

or thionyl chloride. In each case, the nitrogen atom in the hetero-

cyclic ring is protonated and the major problem which had to be solved

in the synthesis involved the deprotonation of the nitrogen atom which

had to be done in non-aqueous medium since the triarylmethyl chlorides

are sensitive to moisture. Initially, attempts were made to achieve

deprotonation by sublimation in vacuo and in the presence of sodium

hydroxide. This was undertaken with the expectation that the hydro-

chloride salt of the triarylmethyl chloride might dissociate into the

triarylmethyl chloride and free hydrogen chloride during sublimation

and that the hydrogen chloride would be absorbed by the sodium hydroxide

before it could recombine with the free triarylmethyl chloride. This

procedure proved to be unsuccessful. Eventually, the problem was

solved elegantly by employing 2,6-lutidine, a stronger base than the

triarylmethyl chlorides. The general procedure involved dissolving

the hydrochloride salt in a minimum of methylene chloride and subsequent

addition of a 3- to 5-fold excess of 2,6-lutidine to the solutions

until precipitation of lutidinium chloride occurred. A non-polar

solvent, such as benzene, was used to effect more complete precipitation







of lutidinium chloride. Filtration and subsequent evaporation of

volatile, filtrate components of all solvents yielded the desired

product. The triarylmethyl chlorides, all soluble in benzene or n-

hexane, either can be recrystallized from these solvents or sublimed

in vacuo to yield a pure product. In some instances, sublimation of

the impure product containing colored impurities led to decomposition

of the products as evidenced by the formation of more colored de-

composition products. The colored impurities also seem to lead to the

formation of oils in the recrystallization process. However, it was

found that treatment with animal charcoal was effective in removal of

most of these impurities; after their removal, a pure crystalline

product could be obtained. The hydrochloride salt of 4-pyridyldi-

(p-tolyl)methyl chloride contained a colored impurity which could not

be removed either by treatment with animal charcoal or by recrystalli-

zation. However, with HC1 gas used as the chlorinating agent,

deprotonation of the resulting salt followed by recrystallization

yielded the pure product without difficulty.

The palladium complexes oftriarylmethyl halides were prepared from

the corresponding completed carbinols by reaction with thionyl chloride.

The complex was usually suspended in a halocarbon such as CC14 and,on

addition of excess thionyl chloride, the complex dissolved. It is

worthwhile to mention two observations. Firstly, the complex of 4-

pyridyldiphenylcarbinol crystallizes out with solvent of recrystalli-

zation (methanol). If this solvent of recrystallization was removed

by heating, the desolvated complex would not dissolve in thionyl

chloride and chlorination could not be accomplished. Secondly, attempts

to prepare 4-pyridylbis(p-methoxyphenyl)methyl chloride from the methanol






by chlorination with HC1 or SOC12 were unsuccessful because a reddish

brown gum resulted each time. The completed methanol, however, could

be easily chlorinated. The ir spectra of some triarylmethyl chlorides

and the palladium(II) complexes of three of these are shown in figures

1-7.

Generation of Free-Radicals by Reaction of Pt(O) With
Triarylmethyl Halides

Visible Spectra

A benzene solution of triarylmethyl halide, Ar3CC1, was syringed

into a 10-3M benzene solution of Pt(PPh3)4 contained in a serum-capped

cuvette and the visible spectrum of the resulting mixture was monitored.

The visible spectra of solutions resulting from the reaction of the

Pt(O) complex with Ph3CCI, (k-C1C6H4)3CC1 and (p-CH3C6H4)3CC1, respec-

tively, showed the same absorptions as the solutions resulting from the

reaction of the same triarylmethyl chlorides with "molecular" silver.

The latter series of reactions typify the classical method of Gomberg

for the generation of triarylmethyl radicals. Thus, the reaction of

Pt(PPh3)4 with triarylmethyl halides must have produced the correspond-

ing triarylmethyl radicals. This is further substantiated by the

observation that the spectra of the reaction mixtures of Pt(PPh3)4

with Ph3CC1 and with Ph3CBr were observed to be identical and that the

lowest energy absorption at 513 nm compares well with the literature

value of 515 nm43 for Ph3C-. The Pt-containing product from these

reactions is cis-dihalogenobis(triphenylphosphine)platinum(II), which

was identified by ir spectroscopy. The reaction of Pt(PPh3)4 with

these triarylmethyl halides forming cis-dihalogenobis(triphenylphosphine)-

platinum(O) and triarylmethyl radicals (in equilibrium with their






























s -o
C3r


Co o






4-)

E
m ai

co



~---o





LC
4-)
40 >)
C3C


0




E


o4-
4-

0
0E

C) 4-)
C

CL


Oo





4-
L





35

















-0










~------ 5
r-












oo
CJ
















CC


0

\ 'i



I*' 0
5-












i I 1 Ii -
/ *






























'I

r



















i --
t

















-C

E
-r









0
--4















co
o u



















ve ca.
En
__ __ o -u

















eq
0 -
-o
0

o a)
o -u

c, 0..

o o


-cr
o .-



o




\o










o
U~I

LJ

j0
o 0Y
0


































'a
L--
.3








IL'I





a 0











10 S-
o



LJ
I 0







V)












I-
-c

Co LL-
U
-oc~

01





U

V)

-% .-



o C




-oY

















































- I-I~ jr..Lu






dimers), contrasts markedly with the classical oxidative addition of

alkyl halides, e.g., CH3I, to Pt(PPh3)4 forming trans-PtICH3(PPh3)2.

The report14 that trityl bromide reacts with Pt(PPh3)4 to yield
PtBr(Ph3C)(PPh3)2 was found to be in error.

When the triarylmethyl halides containing a heterocyclic ring

such as 4-pyridyl were reacted with Ag, no characteristic absorption
indicative of the formation of free-radicals was noted in the visible

spectra. The same observations were made with 4-pyridyldiphenylmethyl,

4-pyridyldi(p-tolyl)methyl and 4-pyridylbis(p-methoxyphenyl)methyl

chlorides coordinated to Pd(II). However, when these triarylmethyl

halides, free or completed with Pd(II), were reacted with Pt(PPh3)4,
peaks in the 500 nm regions characteristic of free-radicals, appeared

immediately. Thus, the formation of free-radicals from triarylmethyl

halides is faster by using Pt(PPh3)4 than with Ag. The visible spectral

data of the various free-radicals and their mode of generation are

tabulated in Table 1. The possibility that these peaks could be

attributed to carbonium ions can be ruled out for the following

reasons: (1) The visible spectrum of Ph3C+ (generated from Ph3COH

and concentrated H2SO4 or from a solution of Ph3CPF6 in CH2C12) shows

two broad absorptions at Xmax 404 and 431 nm compared to Amax 513, 485

and 475 nm for Ph3C.. Similarly, spectra of (p-CH3C6H4)3C+ and

(p-C1C6H4)3C+ show absorption peaks at 452 nm and 465 nm, respec-
tively, whereas those for the corresponding radicals are at 526 nm and
533 nm, respectively. Thus, no carbonium ions are formed in the
reaction of Pt(0) with triarylmethyl halides. (2) The presence of

free-radicals in the solutions is demonstrated by esr spectroscopy

which will be described in the next section.
















a 0 aO
<3- C-l fl2





ilr LO LO L O LO l Lr) LA



*
YCM


E0 0
C0




)o


























a -C=
0 L
4C ) o 4 4
0- C L CL 0 0















































.) "I I I
S -- 0 C
<) I O C
M a- = u -


























oo 0



u k- ko 0i Cd (

























C I I I _
CM



0j
C-)













-o- a- t- Qj-







0 '.0





I I



CM -C o4






l. C) 0) 0) 3: I I I
ro o-) m ) ) v
-~ a- ..c oj 11 11

Q- I > 0. ^-~
0 0 0 Z Z-
a 0 C I d- 3- CM CM CM







Finally, the reaction mixture of the Pt(0) complex with chlorides

containing two pyridyl rings or a thiazolyl ring did not show any ab-

sorption peaks around the 500 nm region characteristic of free-radicals

even though the yellow color of the Pt(0) complex turned orange when

the two solutions were mixed. The reason is probably due to the great

instability of the free-radicals formed.


Esr Spectroscopy

Benzene solutions of the triarylmethyl radicals were usually

generated by adding excess organic halide to 10-3 10-2M Pt(PPh3)4

and the resulting solution scanned at room temperature. For radicals

which are unstable with respect to the corresponding dimers or dis-

proportionation products and whose esr signals decreased rapidly with

time, the spectra were run at 100. No esr signal could be detected

in mixtures of Pt(0) solutions and chloride solutions containing two

pyridyl rings or a thiazoyl ring. It was found also that, for the

case of 4-pyridyldi(p-tolyl)methyl chloride, an esr spectrum could be

obtained by reaction with silver even though the visible spectrum did

not indicate the formation of free-radicals. The esr signal strength

did decrease significantly within a matter of minutes. These observa-

tions seem to reflect the fact that esr spectroscopy is a more sensitive

method than visible spectroscopy and, hence, can detect the small

concentration of free-radicals formed from the slow reaction of the

chloride and silver. All the spectra are characterized by a g-value

of 2.0 and by extensive hyperfine structure. The plethora of lines

sometimes exceeded one hundred. Even so, this represents but a fraction

of the total number of lines theoretically calculated to be 343 and 2025






for triphenylmethyl and 4-pyridyldiphenylmethyl radicals, respectively.

Whenever a solution spectrum did not contain hyperfine structure, the

solution was usually diluted to prevent "exchange narrowing," which

has the effect of masking hyperfine structure by collapsing the esr

signal into a single line. In some instances, better resolution was

achieved by dilution. The esr spectra of Ph3C* and (p-C1C6H4)3C. are

identical to published spectra of the same species44. The others are

shown in figures 8-13. No attempt was made at interpretation of these

complex spectra because it lies outside the scope of this work.

The reaction of triarylmethyl halides is faster with Pt(PPh3)4

than with Ag. Whereas no advantage is gained by using the Pt(O)

rather than Ag for generation of radicals from halides containing no

heterocyclic rings, the use of Pt(O) is critical in the case of those

halides containing heterocyclic rings or in completed halides. In the

former case, the presence of an electronegative nitrogen atom in the

heterocyclic rings appears to lower the stability with respect to

dimers or disproportionation products of the generated radicals. If

the rate of generation of the radicals is slow as is the case when Ag

is used, the rate of disproportionation, for example, may be comparable

to it so that the concentration of free-radical is too low to be

detected by visible spectroscopy. When a halide is completed to a

Pd(II) atom, the results indicate that its reactivity with Ag is too

slow to be of any utility in generating radicals, but by using Pt(O),

radicals,probably completed radicals,are formed rapidly. Hence, the

reaction of Pt(PPh3)4 with a triarylmethyl halide appears to be a

superior method for the generation of free-radicals than Gomberg's

method.






45




















cmm
CC





4Ji
I













V) ~ 2_O CI u
L) 0
U






-



"- ______---.---- .1c 1



__- --- ----=.- E-- -
____ S.-----rr c= 1
~ -------- ----- -Doc-
___~'" "= ;; ;~ __ __0
_____ __ __ ___ U
______~~1 5--- i-- P -
____-- __
______-- ~ _____ ________________ LzJ-T.i -






_______- LL~~oo






46

















0

C\I
to


r-
Cr






4o










_E
a)


















-- 4-
ca






















n-
U,
0,




















i-
i~~- ii ,iiii--~Q






-- _- i--



















































I



.E17


1 1I-











rrji^ lliiilllin '*i^^^^^^g^^ ,/


;P~LIIIIL
---~-
;~P






---







-~p----------
-V -
I~C--~-~---
~- ~IY


-II --

---i -------
~-~-U~------
i,-xr~

I -- I--

-I-r--------
--




;i~
IL:==ILPlli-IIII~-








--
-~D-~-


--------


;I
~"'
~_~IP-
--C-r

--
-~r;L'L
-I
n

-~ ~,~


"-~p~







q







48




















C,)
U) )E











4-)


0





>~ CL)
M =a
cl.















4-) 9l-

_SL

V)
a-















LO
'-C
C)
___ __ __ **~ I
go-

'4 C
0 .





a)E
C,)




L





U-)r






49



















cn


a-
4-


0
0








S.--- U
a
0 CIQ

o

co





-- "..----' "1
E ,._j
~r'-__ -






















*r -
0 -j








5-
U,









S-

.f-






50

















4:n
Ln =3
4-F)





4-
00
cn


)0)
iin








u 4)
r0)?-.
-I--









CY,



C.-)
00
'a C











0-I



cu
IIr


















V) 0'-



LA)
a)
ro




















LL-







The visible spectral data show that the absorption of longest

wavelength is shifted to lower energy with the change from the cations

to the corresponding radicals. This shift is 82 nm, 76 nm and 68 nm

in going from Ph3C+, (p-CH3C6H4)3C+ and (2-ClC6H4)3C+ to the respective

radicals. There is also a parallel shift to lower energy with the change

from the completed carbonium to the completed radicals. The absorptions

of lowest energy for the completed (4-C5H4N)Ph2C+, (4-C5H4N)(p-CH3C6H4N)2C+

and (4-C5H4N)(p-MeOC6H4)2C+ are 457 nm, 492 nm and 537 nm, respectively,34

whereas the values for the corresponding completed radicals are 520 nm,

531 nm and 542 nm, respectively. The shifts are 63 nm, 39 nm and 5 nm,

respectively. There is also a shift to lower energy from free base

radical to coordinated base radical. The mangitude of the shifts are

19 and 21 nm, respectively, when the radicals are 4-pyridyldiphenyl-

methyl and 4-pyridyldi(p-tolyl)methyl. No explanation, at present, can

be advanced for the magnitudes of such shifts. The absorptions of the

uncomplexed cations containing a heterocyclic ring cannot be compared

conveniently with those of the corresponding radicals because the values

are reported for the protonated cations and the effect of this protonation

on the position of the absorption peaks has yet to be evaluated. Thus,

the only consistent trend observed is that the lowest energy visible

absorption peaks of the free-radicals appear at longer wavelength than

those of the corresponding cations.

The absorptions for the coordinated radicals also appear at longer

wavelengths than those of the corresponding uncoordinated radicals.

There is no simple explanation for these observations.

From the visible spectra it is deduced that the 4-pyridyldiaryl-

methyl radicals are less stable with respect to their dimers than are







the triarylmethyl radicals. This is indicated by the rapid decay of

intensity of the absorption spectra to almost zero absorbance in the

case of 4-pyridyldiphenylmethyl and 4-pyridyldi(p-tolyl)methyl radicals

presumably because of the following equilibrium which

2 radicals dimer (24)
lies farther to the right than for the non-heterocyclic triarylmethyl

radicals. This result is expected on account of the electronegative

N atom in the pyridyl ring which destabilizes the radical and hence
favors the higher degree of dimerization.

Kinetics and Mechanism of the Reaction of Organic Halides
With Tetrakis(triphenylphosphine)platinum(O)

The kinetics of the reaction were followed using 1.23 x 10-1M

triphenylmethyl chloride, 1.91 x 10-3M tetrakis(triphenylphosphine)-

platinum(O) and various concentrations of triphenylphosphine. The

reaction followed second-order kinetics, first order in each reactant.

Using a large excess (> 50-fold) of the halide, pseudo-first-order

kinetics was obtained. The slope of the linear plots of log (A-Am) vs

time equals kobsd/2.30 and the rate constants so obtained were found

to be different at different [PPh3] (Table 2).

For the reaction of diphenylmethyl bromide with tetrakis(tri-

phenylphosphine)platinum(O), it was found that a plot of kobsd vs

[Ph2CHBr]/[PPh3] was linear; but different slopes were obtained for

each phosphine concentration. Two characteristic sets of data are

presented in Figurel4. The data for these plots are tabulated in

Table 3.

The observed kinetic behavior can be accommodated by the mechanism:






















u


v,




I-



II














I











u
0



0

0-

















CY)
-a



























X




i-'

4o
a-


















4r.
C


















'-4
I-1,

-c

Q-
1-


ee3
4-'
+J
r-


co -;t om a- q co C\M
a- CM CO '0 00 C9 0































CM C\ CM m















-o
CV








0 0 a- Co U> 0

C 0 C n Co C,- ,-- a

ar- E


o
x











i-







*-





















0







4-,
0
40
c'J













I-0
0







r-




















E
CM




















ia -
rc












4- *
o
a-m



w t
4->r





(0 0
4-> S
tfL
LC: 1
0 t03

U





54

















Ur,
x













OC
I-
u







5-P
U







C)

a~ (0
N S-

C:
o











CL S
o a


-6- )(









L-J
4)










C\J
>1
N L~
0 0















4- 4
C, S-
CE I r (0

a U,



U


0C,

C)
0r

Du -
L.J
'0






4-



0


r- -


N-






0,
U-



















TABLE 3

kobsd as a Function of [Ph2CHBr]/[PPh3]


a[Ph2CHBr]/[PPh3] akobsd x 102


312 9.2
208 6.3
104 3.1
55 1.8
36 1.2
31 1.0
49 8.1
25 3.9
9.5 1.4
6.3 0.95
5.4 0.83

a For first 6 entries, [P(C6H5)3] = 8.0 x 10-4

For last 5 entries, [P(C6H5)3] = 4.7 x 10-3






fast
PtL4 PtL3 + L (25)

K
PtL3 PtL2 + L (26)

kl fast
PtL3 + RX -* intermediates -> .... products (27)

k2 fast
PtL2 + RX intermediates .... products (28)

where L = PPh3 and R = organic group.

This gives rise to the rate law:

Sd[Pt(O)] = (kl[L] + k2K)[RX][Pt(O)] = kobsd[Pt(O)] (29)
dt K + [L]
Normalizing to a constant RX concentration,

kobsd kl[L] + k2K
k (30)
= [RX] K + [L]

As [L] k' k1

As [L] 0, k' k2

Rearranging equation (30) gives

k'(K + [L]) = kl[L] + k2K (31)

From this, if K is known, the values k1 and k2 can be computed.

Using the kobsd values of Table 2, one can plot k'(K + [L]) vs [L]

to obtain a straight line (Figure 15). From the intercept, the slope

of this plot and the value of K = 1.6 x 10-4M obtained by Halpern and

his co-workers45, values for kl and k2 were calculated to be 2.1 x 10-1

M-1 sec-I and 7.3 x 10-1 M-1 sec-1, respectively.

In order to obtain some thermodynamic data from which one might

be able to infer the nature of the transition states involved in the

reactions between the Pt(0) complex with trityl chloride, benzhydryl

bromide and with methyl iodide, the second-order rate constants, k', were




























-I 15
r-
X

L-J
+


- 10








5









2 4 6 8

[L] x 103, M


Figure 15. Plot of k'(K + [L]) vs [L] in Benzene at 250






determined at several temperatures. The results of these determinations

are tabulated in table 4. Since tris(triphenylphosphine)platinum(0) is

in equilibrium with bis(triphenylphosphine)platinum(0) and free phosphine,

an initial concentration of the excess triphenylphosphine was used to

ensure that at least 98% of the platinum(O) complex was in the trico-

ordinated form. This calculation was based upon a constant of 1.6 x 10-4M

reported by Halpern and his coworkers.45 This condition was satisifed

by using an initial concentration of 1.90 x 10-3M Pt(PPh3)4 and 3.76 x

10-2M PPh3 and was shown to be true from the estimated second-order

rate constant which were experimentally the same at a given temperature

when the excess triphenylphosphine concentration was either 3.76 x 10-2M

or 1.16 x 10-1M. Similar concentrations of excess triphenylphosphine

and tetrakis(triphenylphosphine)platinum(0) were used for the methyl

iodide and diphenylmethyl bromide reactions.

In the case of methyl iodide, the kinetic runs were stopped when

a precipitate of [PPh3CH3]I appeared (after about 30% reaction). However,

the salt has very low solubility (concentrations > 5 x 10-5M bring about

precipitation) and Pearson and Rajaraml7 have suggested that the reac-

tion leading to the formation of this phosphonium salt may be a mildly

competing reaction, but will have a negligible effect on the data since

no spectral changes occur in the region of measurement.

Using the Arrhenius equation (32),

lo k' = log A AE 1 (32)
2.30 2.30 R T
where k' = second-order rate constant

A = pre-exponential factor

AEt = activation energy









TABLE 4

Second-Order Rate Constants for the Reaction Between Pt(PPh3)4
and Organic Halides in Benzene


XY Temp, C k', M-1 sec-la


(C6H5)3CC1b








CH3Id






(C6H5)2CHBre


20.1

22.4

25.0

27.6

30.2

22.9

25.0

27.7

30.8

23.0

25.0

26.6

28.6

31.4


1.49

1.75

1.91

2.26

2.55

3.28

3.59

4.16

4.65

2.63

2.93

3.17

3.38

3.89


10-1
10-1

10-1

10-1

10-1

10-3

10-3

10-3

10-3

10-1
10-1

10-1

10-1
10-1


(1.51

(1.68

(1 .91

(2.31

(2.58


x 10-1)c

x 10-1)

x 10- )

x 10-1)

x 10-1)


a Average from two kinetic runs.
b Initial concentrations: 1.23 x 10-1M Ph3CC1, 1.91 x 10-3M PPh3.

c Values in parenthesis refer to the rate constants for 1.16 x 10-1M
PPh3. The average value for the 2 concentrations of PPh3 was
used to determine the best fit plot.
d Initial concentrations: 7.3 x 10-1M CH3I, 9.0 x 10-4M Pt[PPh3]4
and 1.31 x 10-2M P(C6H5)3.
e Initial concentrations: 7.1 x 10-2M Ph2CHBr, 1.07 x 10-3M
Pt(PPh3)4 and 1.37 x 10-2M PPh3.







R = ideal gas constant

T = absolute temperature

a plot of-log k' vs 1/T was found to be a straight line whose slope
equals AE*/2.30 R and whose intercept equals log A/2.30 from which
AE and A can be calculated. Using the relation

AE* = AH* + RT (33)

and the Eyring equation

k kT eAS /R e-AH/RT
k' e (34)

where k = Boltzmann's constant

h = Planck's constant
AS* can then be calculated. AG* was obtained from the relation
AG4 = AH4 TAS4 (35)

The plots of the kinetic data are shown in figures 16-18 while the

activation parameters estimated are listed in Table 5.

The kinetic behavior of the reaction of Pt(PPh3)4 with Ph3CC1 and
Ph2CHBr resemble that previously observed by Pearson and Rajaram for

the reaction of Pt(PPh3)4 with CH37 These reactions follow overall
second-order kinetics, first-order in Pt(PPh3)4 and first-order in the

halide. The kobsd values are found to be dependent linearly on the

value of [RX]/[PPh3], but the slopes of the linear plots are different
for different phosphine concentrations (Table 3). Thus, the kinetic

behavior of these reactions are consistent with the mechanism shown in

equations (36) and (37)

Pt(PPh3)3 + Ph3CC1 PPh3 + other products) (36)

k2
Pt(PPh3)2 + Ph3CC1 products (37)


























































































CC) C
00 0 0


0 0. Q
0" LO

0 0 0


,) 60[-


0.
0



*r)
-C





4-
c'









4-
O
.,




0




4-









0
I-








0
0















































L.0
L O
COY --
co
X



CY







cr-




C0
CO
'ci ,--


,i 6oL-


en~ enc
(%J cJ cuJ c%
























s-



CL
CL.
-C






0





4-C




4-)
O







I

















-tj
o-





4--



0
4-
0






(-3
0



4-






0
Ic






4-'









u
a-

I-
L^

o/
>|



_It

01
0,

<


4->






I-


,5 601-


LO C) LO
LO C))
o 0 -0 I


o 0 6 o
0 O

























0
o
LO






0)
Oa










9--



a--
0










C-
I-







4-
0






O






0

r-


a)







c
n






-E



0-
4-















*-
4-)-















<4)
C(

U)
ct:
0)

4-
s-

t-
Ll
s-

4-'
01

(0






oc


r0

E

r-r





























a)
U









0
E










,-
X

































































U
I-



I-^







.4-




r-"










-IH-
<3














v(
Q,



















2:

U


O': C r
r- r- CM














CM CM 1-4
C') C) LO















*Z --
en 00 LO







The rate constants k1 and k2 for the reactions were found to be 2.1 x
10-" M-1 sec"- and 7.3 x 10-1 M-1 sec-l respectively. The correspond-

ing values of Pearson and Rajaram for CH3I are 3.5 x 10-3 M-1 sec-1
and 2.0 x 10-2 M-1 sec-. Even though the faster rate of reaction
of Pt(PPh3)3 and Pt(PPh3)2 with Ph3CC1 as compared to the reaction of

the Pt(0) complex with CH3I is consistent with the lower carbon-halogen
bond dissociation energy of Ph3CC1 compared to that of CH3I, no correla-
tion between rate and bond dissociation energy can be made. The
reason is because the mechanisms for the reaction of these halides
with Pt(PPh3)4 are different, as will be demonstrated later.

Although the reaction of Pt(PPh3)4 with CH3I, Ph3CC1 and Ph2CHBr
exhibit similar kinetic behavior, the products are not analogous.
Whereas the reaction of Pt(PPh3)4 with CH31 yields the normal oxidative-
addition product PtICH3(PPh3)2, the reaction of Pt(PPh3)4 with Ph3CC1
and Ph2CHBr yields Ph3C* (and the dimer) and Ph2CH-CHPh2, respectively.

The other products are cis-PtCl2(PPh3)2 and cis-PtBr2(PPh3)2, respec-
tively. Thus, the details of the mechanism of these reactions must
be different. The reaction of Pt(PPh3)4 with CH3I has been proposed
to proceed via a polar transition state on the basis of an increase in
rate with an increase in polarity of the solvent usedl7; and as stated
earlier, one objective of this work is to determine if the mechanism
of analogous reactions with tertiary and secondary halides such as
Ph3CC1 and Ph2CHBr proceed via free-radical intermediates.

In the estimation of the activation parameters of the reaction of
Pt(PPh3)4 with the three halides, the concentration of PPh3 was chosen

such that the tricoordinated Pt(0) complex, Pt(PPh3)3,was equal to the
total initial Pt(0) concentration. Under this condition the activation







parameters obtained are for the principal reaction between Pt(PPh3)3

and RX, where RX = Ph3CC1, Ph2CHBr and CH3I. The values of k' at 250
(M-1 sec-1), of AE* (kcal/mole), of AS* (eu) and of AG* (kcal/mole),

respectively, were determined to be: Ph3CC1, 2.6 x 10-2, 9.4, -32

and 19; PhCHBr, 2.93 x 10-1, 8.2, -32 and 18; and CH3I, 2.59 x 10-3

5.1, -54 and 21. The AE* values of between 5.1 to 9.4 kcal/mole are
in the same range as those cited for the reaction of other organic
halides with Co(CN)53-; but, no correlation between the enthalpy of

activation can be deduced from the data. However, the free energy of
activation, AG*, in the order Ph2CHBr < Ph3CC1 is consistent with the

reverse order Ph3CC1 < Ph2CHBr for the rate constants. Furthermore,

the order for the AG* values parallels the order for the bond dis-
sociation energies of Ph3CC1 and Ph2CHBr. Thus, the reactivity of
Ph3CC1 and Ph2CHBr with Pt(PPh3)3 depends, at least in part, on the

carbon-halogen bond strength of the halides. It should be noted that
these correlations cannot be extended to CH3I since a different mecha-

nism is involved.

The entropies of activation (-32 eu for Ph3CC1 and Ph2CHBr and
-54 eu for CH3I) are unexpectedly negative for reactions involving

uncharged reactants and products. It is of interest to note that AS*
for CH3I, in which PtICH3(PPh3)2 is the oxidative product, is some

22 eu more negative than that for Ph3CC1 and Ph2CHBr in which the
dihalobis(triphenylphosphine)platinum(0) is the Pt(II) product. A

large negative entropy of activation has been attributed to either a
marked increase in polarity, or unusually stringent stereochemical

requirements, in going from reactants to the transition state846
For CH31, the increase in rate with increase in polarity of the solvent
has been adduced to a polar transition state of type (I) rather than




67

type (II)17

.CH3
[(PPh3)2Pt+ ....CH3.... I]* (PPh3)2PPt


(I) (II)
The reaction of trans-IrCl(CO)(PPh3)2 with CH3I has been shown to in-
volve an analogous polar transition state, the estimated AE and AS*
values of 6.2 kcal/mole and -51 eu, respectively, being remarkably close
to the corresponding values for the Pt(O) reaction with CH3I. Analogous
transition states cannot be ascribed to reaction of the Pt(O) with Ph3CC1
and Ph2CHBr for steric reasons. Indeed, solvolytic displacement reac-
tions of Ph3CC1 and Ph2CHBr proceed via an SN1 rather than an SN2 mechanism.

As mentioned earlier, there is a relatively large difference of
some 22 eu in ASt when the halide is changed from CH3I to Ph3CC1 or
Ph2CHBr. There is also an increase in rate of reaction with increase
in polarity of the solvent. (The magnitude of this increase has not
been determined because the rate is too fast to be measured by con-
ventional methods.) If it is first assumed that a polar transition
state, analogous to the one in the case of CH3I is formed, the less
negative AS* (compared to CH3I) cannot be rationalized on the basis of
a more constrained transition state since one would expect then that
ASt for the larger Ph3CC1 and Ph2CHBr would be more negative than that
obtained for CH3I. However, one can argue that, since the reactions
are carried out in benzene, and Ph3CC1 and Ph2CHBr are aromatic, the
solvent molecules around the halide molecules are somewhat ordered
but are "squeezed" out in the transition state resulting in a less nega-
tive AS*. One would also have to submit that this solvent effect
predominates over the steric factor in the transition state so that







the net result is a less negative AS* for Ph3CCl and Ph2CHBr. If this
were the situation, then it also follows that AS* for Ph2CHBr should be

more negative than Ph3CC1 since the solvent effect should be less with
Ph2CHBr which has only 2 aromatic rings compared to 3 in Ph3CC1, but AS*

is the same for both. Hence, transition state of type (I) may not be
operative in the reaction of Pt(O) with Ph3CC1 or Ph2CHBr on the basis

of activation parameter data also. The second possible transition state

(type (II)) is relatively non-polar and contradicts the experimental
fact that the rate of reaction is enhanced by using a more polar solvent.

Moreover, such a transition state would dictate a more stringent stereo-

chemical requirement in the transition state for the larger Ph3CC1 and
Ph2CHBr molecules leading to a more but negative AS* for Ph3CCI and

Ph2CHBr, but instead the reverse has been found.

A third type of transition state (type (III)) can be considered
for oxidative addition of the Pt(O) complex to Ph3CC1 or Ph2CHBr. This

consideration arises out of the different products of the reaction

which may be the result that a new mechanism is operative. Such a


L/ Pt ....C....CPh

(III)
transition state, which is assumed to have more product-like (Ph3C-
and L2PtC1) than reactant-like character, would be attractive for

three reasons. Firstly, it would explain that, for example, Ph3C- is
formed in the reaction of Pt(O) with Ph3CC1 but CH3. is not formed

with CH3I. The reason is that such a transition state, leading to the

relatively unstable CH3* is energetically prohibitive, whereas transition







state of type (I), leading to the formation-of the normal oxidative-

addition product is now preferred. On the other hand, type (II)

transition state would be unfavorable for Ph3CC1 because of steric

reasons. With Ph2CHBr, the relatively stable (cf CH3.) diphenylmethyl

radicals, once generated, are unstable with respect to the dimer and

hence Ph2CH-CHPh2 is obtained as the organic product. Secondly, the

proposed transition state is not at variance with the activation

parameters obtained. There is little difference in AS* for both Ph3CCl

and Ph2CHBr because the solvent molecules are not displaced from the

vicinity of the aromatic rings since there are farther from the reaction

site than they would be if type (I) transition state were operative.

Furthermore, the more negative AS* for CH31 can be rationalized on the

basis of less carbon-halogen bond breaking in the transition state

so that the decrease of entropy is greater. Finally, the proposed

transition state is a polar one and is consistent with the increase

in rate with increase in polarity of the solvent.

From the above discussion on the kinetics and transition state

of the reaction of Pt(PPh3)4 with Ph3CCl and Ph2CHBr, a mechanism for

the reaction consistent with the kinetics and activation parameters,

can now be proposed as follows:


PtL4 -+ PtL3 + L (38)

K
PtL3 PtL2 + L (39)

kl 6+ 6 I
PtL3 + RX [L3Pt ....X....R ] + L3Pt X + R. (40)

I fast
L3Pt X + RX cis- and/or trans-PtX2L2 + L + R- (41)








PtL2 + RX 2 L2Pt X + R- (42)
fast
L2Pt X + RX cis- and/or trans-PtX2L2 + R. (43)

PPh3
trans-PtX2L2 h cis-PtX2L2 (44)

The overall reaction is thus:

PtL4 + 2RX cis-PtX2L2 + 2L + 2R.(R-R) (45)

In spite of the activation parameter data, one might argue that the
reaction proceeds via a normal oxidative addition to form PtC1(Ph3C)-
(PPh3)2, followed by a second oxidative addition forming PtC12(Ph3C)2-
(PPh3)2 which then undergoes reductive elimination to give cis-PtCl2(PPh3)2
and Ph3C-CPh3, the latter dissociating to give Ph3C*. If this were so,
then, one should be able to isolate the first oxidative-addition product
PtC1(Ph3C)(PPh3)2 especially with less than the stoichiometric amount
of Ph3CC1. Attempts to isolate either PtCI(Ph3C)(PPh3)2 or PtC1(Ph3CH)-
(PPh3)2 were unsuccessful. Even though the relatively unstable
PtC1(Ph3C)(PPh3)2 were formed, it seems unreasonable to expect a further
molecule of the bulky Ph3CC1 to add more rapidly than the first. Thus,
the proposed free-radical mechanism appears to be quite sound.

In order to extend the free-radical reactions of tertiary and
secondary aromatic halides to d8 systems and perhaps also to find more
corroborating evidence for the radical nature of the mechanistic path-
way, the reactionsof trans-chlorocarbonylbis(triphenylphosphine)-
iridium(I) with Ph3CBr and Ph2CHBr were carried out. When Ph3CBr
was added to a benzene solution of IrC1(CO)(PPh3)2, the yellow solution







immediately turned orange and an esr spectrum of the solution showed
the presence of Ph3C-. On stirring for 3 hours, > 80% yield of
IrBr2Cl(CO)(PPh3)2 was obtained. In a similar manner, using Ph2CHBr
instead of Ph3CBr, yields IrBr2CI(CO)(PPh3)2 and Ph2CH-CHPh2 in nearly
quantitative yield after 12 hours' reaction. These two reactions
strongly reinforce the assertion of a free-radical mechanism since

Ph3C* and Ph2CHCHPh2 could not have arisen from two successive oxi-
dative additions of the halide to IrC1(CO)(PPh3)2. The first oxidative-
addition step would have formed the IrBrCI(Ph3C)(CO)(PPh3)2 which is
coordinately saturated and a second oxidative addition of Ph3CBr
would have yielded an extremely improbable octa-coordinated IrV species.
It should also be mentioned that the oxidative addition of methyl
halides, CH3X, to trans-IrCl(CO)(PPh3)2 to form IrXCI(CH3)(CO)(PPh3)2
is well documented18'47 and no further reaction of CH3I with IrC1I(CH3)-
(CO)(PPh3)2 has ever been found. Finally, when IrC1(CO)(PPh3) was
reacted with a mixture of Ph3CBr and Ph2CHBr in an equimolar ratio of
all three reactants, the products of the reaction were found to be
IrBr2C1(CO)(PPh3)2 and pentaphenylethane. The pentaphenylethane was
identified by its mass spectrum which showed intense peaks corresponding
to m/e 243 and 167 which were assigned to the Ph3C+ and Ph2CH+ ions.
Two peaks of considerably weaker intensity at m/e 410 and 333 correspond-
ing to the parent molecular ion and Ph4C2H+ were also observed. This
experiment provides further evidence for a free-radical reaction of
IrC1(CO)(PPh3)2 with Ph3CBr and Ph2CHBr. The pentaphenylethane is
presumably formed by combination of Ph3C- and Ph2CH- radicals. Thus,
a new free-radical mechanism for the reaction of IrC1(CO)(PPh3)2 with
tertiary and secondary aromatic halides can be proposed as follows:







IrC1(CO)(PPh3)2 + RX -+ IrC1X(CO)(PPh3)2 + R- rate determining (46)

IrC1X(CO)(PPh3)2 + RX IrClX2(CO)(PPh3)2 + R- fast (47)

IrC1(CO)(PPh3)2 + 2RX IrC1X2(CO)(PPh3)2 + 2R.(R-R) overall (48)
where R = tertiary or secondary aromatic group

X = halogen
It is not possible at this time to determine the precise nature of the
transition state. This information must await the results of experi-
ments for the determination of activation parameters and rate
dependence on the polarity of solvents. It would also be of interest
to study the effect of rate of reaction when one varies the halogen
on both the halide and the Ir(I) complex.

Reactions of Tetrakis(triphenylphosphine)platinum(0) With Interhalogens

The preceding discussion has focused on the reaction of Pt(PPh3)4
with organic halides which have been demonstrated to react in two ways.
A primary halide such as CH31, adds oxidatively to form trans-
PtICH3(PPh3)2,17 whereas, secondary and tertiary aromatic halides
yield cis-dihalogenobis(triphenylphosphine)platinum(II). In the midst
of the work just described, the interest to investigate the reaction
of Pt(PPh3)4 with interhalogens was aroused. Questions such as these

were asked: (1) Will the interhalogens add oxidatively to the Pt(0)
complex analogously to CH3I to form new, mixed dihalogenobis-
(triphenylphosphine)platinum(II)? Or (2) Will they react to form
dihalogenobis(triphenylphosphine)platinum(II) complexes containing

only one type of halogen in a reaction analogous to that with
triarylmethyl halides? (3) What is the stereochemistry of the
products and how do these products come about? Thus, in an







attempt to answer some of these questions, the investigation of the

reaction of Pt(PPh3)4 with interhalogens was initiated.

The choice of the interhalogens, namely, iodine monobromide and

iodine monochloride, was dictated by their stability with respect to

disproportionation and by their commercial availability. When Pt(PPh3)4

was reacted with IC1 in 1:6, 1:3 and 1:1.2 mole ratios, the pre-

dominant, halogen-containing Pt(II) species were found to be trans-

PtCl2(PPh3)2 (74%), trans-PtI2(PPh3)2 (> 80%) and cis-PtClI(PPh3)2,
respectively. The trans-PtI2(PPh3)2 and trans-PtCl2(PPh3)2 were

identified by decomposition point, elemental analysis and by comparison

of their ir spectra with those published. It is worthwhile to point

out at this time that Mastin36 has proposed that in complexes of the

type PtX2(PPh3)2, where X = C1, Br, I, the cis- or the trans-isomer

can be differentiated by examination of the ir spectra in the region

480-550 cm-1. In cis-isomers, the ir spectra contain a strong ab-

sorption at 550 + 5 cm-1 whereas, this same absorption is relatively

weak in the trans-isomer. This difference is clearly illustrated in

cis- and trans-dibromobis(triphenylphosphine)platinum(II) (figures 19 A

and B). This observation was utilized time and again to differentiate

the cis- from the trans-isomers. Furthermore, several additional

generalizations may be made concerning the differences between the

cis- and the trans-isomers: (1) Cis-isomers are usually less colored

than trans-isomers. (2) The solubility in non-polar solvents is

greater for the trans-isomers. (3) The decomposition point is higher
for the trans-isomer than for the cis-isomer. (4) For the two weak

absorptions just below 1600 cm-1 and which are present in the ir

spectra of both isomers, the more intense one appears at higher energy

















ii
:. 1 ?0 25










II |
















(A) I (B)

I I I i I I I
600 500 400 300 600 500 400 300

WAVENUMBER CM-1


Figure 12. Infrared Spectrum of (A) Cis- and (B) Trans-
PtBr2(PPh3)2







in the spectra of the trans-isomer. This last generalization was first
48
reported by Bland and Kemmitt .
The reaction of Pt(PPh3)4 with IC1 in a 1:6 mole ratio provides

the quickest synthetic route and the highest yield (74%) so far reported

for the preparation of trans-dichlorobis(triphenylphosphine)platinum(II).

Conventional methods of synthesis yield only the cis-isomer. The most

recently published preparation of the trans-isomer involves photochemical

isomerization of the cis-isomer with a reported yield of 40%49. Besides

the relatively low yield, the photochemical preparation takes a much

longer reaction time (4 hours). It also suffers from the undesirable

necessity of having to separate one desired isomer from a mixture con-

taining both isomers. This results in lower yields in the purification

process if pure isomer is required. The present method, using Pt(PPh3)4

and IC1, requires only a few minutes of reaction time and results in

the formation of the trans-isomer in high yield.

With a mole ratio of Pt(PPh3)4 to IC1 of 1:3 the product is trans-

PtI2(PPh3)2. This result is rather surprising in view of the product
got from the 1:6 addition, viz., trans-PtCl2(PPh3)2. Firstly, one

would expect the IC1 to serve in its usual capacity as a chlorinating

agent. Secondly, this oxidative addition is quite different from the

reaction of Pt(PPh3)4 with CH3I and Ph3CC1 discussed previously. The
"normal" oxidative product is not isolated as in the reaction with

CH3I and the reaction is different from that ofPh3CC1 in that the

halogen atoms coordinated to the Pt atom is the more electropositive

end of the molecule IC1, whereas, for Ph3CC1, the C1 atom is the more

electronegative portion of the molecule.







In an attempt to isolate any stable precursors of trans- PtI2-

(PPh3)2, the reaction of Pt(PPh3)4 with IC1 was again carried out using
a 1:1 mole ratio of the reactants. In this reaction, only about 60%
of the Pt metal was precipitated as halogeno-Pt(II) after the usual
work-up; the reason for this can be attributed to competition for IC1

by the platinum(0) and free PPh3, the latter of which rapidly forms
PPh3IC150. The precipitated product was found to contain predominantly

cis-PtC1I(PPh3)2 and some trans-PtI2(PPh3)2. This cis-isomer of
PtC1I(PPh3)2 was inferred from the ir spectrum which contained the
strong absorption at 550 cm-1 characteristic of the cis-isomer.

In the foregoing discussion, the discovery that the reaction
of Pt(PPh3)4 with IC1 gives rise to three distinct Pt(II) complexes

depending upon the relative amounts of the reactants employed, was
presented. But, how can these diverse products be rationalized?

Let us first consider the formation of trans-PtI2(PPh3). If it is
assumed that IC1 adds oxidatively to Pt(PPh3)4 to form first trans-

PtCII(PPh3)2, there remains the problem of accounting for the

subsequent transformation of trans-PtC1I(PPh3)2 to trans-Ptl2(PPh3)2.

Two plausible routes will be considered: (1) A second molecule of

IC1 can undergo a second oxidative addition to trans-PtClI(PPh3)2 to

form an octahedral Pt(IV) complex which subsequently undergoes a

reductive elimination of C12, or (2) a substitution reaction can
occur in which a chloride ion is replaced by an iodide ion upon attack
of IC1. The possibility of a second oxidative addition of IC1 to

trans-PtC1I(PPh3)2 yielding an octahedral complex, which subsequently

undergoes a reductive elimination, can be ruled out in this system

since, if the octahedral complex were formed, it would be expected to

be quite stable under the reaction conditions and time employed







(5 minutes). Furthermore, there is evidence that when a dihalogeno-

bis(triphenylphosphine)platinum(II) reacts with an interhalogen,

substitution reactions occur faster than addition reactions. In many

of these types of reactions to be discussed later, the substitution

products are isolated, but, if the reaction time is increased, Pt(IV)

complexes are formed. This observation suggests that in a system

containing dihalogenobis(triphenylphosphine)platinum(II) and an inter-

halogen, the thermodynamically stable species is tetrahalogenobis-

(triphenylphosphine)platinum(IV). This observation lends support to

the earlier argument against reductive elimination. Thus, based upon

the assigned premise trans-PtI2(PPh3)2 is produced as a consequence

of a substitution reaction of IC1 on trans-PtClI(PPh3)2.

A rather obvious question arises, however. Why should trans-

PtC1I(PPh3)2 be proposed as a precursor to trans-Ptl2(PPh3)2 when

cis-PtCll(PPh3)2 is the isolated product in the 1:1 addition? The

compelling reason is that substitution reactions involving square

planar Pt(II) complexes are stereospecific, i.e., substitution

reactions involving cis- and trans-complexes yield cis- and trans-
51-54
isomers, respectively No exceptions have been reported.

Furthermore, stereospecificity obtains in all the substitution reac-

tions of Pt(II) complexes studied during the course of this work.

If trans-PtI2(PPh3)2 must be formed from trans-PtC1I(PPh3)2, one must

account for the observation that the only mixed halogeno-complex
isolated was cis-PtC1I(PPh3)2. These seemingly contradictory "truths"

can be accounted for by assuming that, after the initial formation in

solution of trans-PtCllI(PPh3)2, the excess of free phosphine causes

rapid isomerization to the cis-PtCll(PPh3)2 which precipitates from







the reaction mixture, the low solubility of the cis-isomer in benzene
contributing significantly to the driving force of the reaction. It
is well known that free phosphine catalyses isomerization of this
kind. Allen and Baird reported that on addition of a trace of tri-
phenylphosphine to a chloroform solution of trans-PtC12(PPh3)2, the
yellow color of the solution is rapidly discharged resulting in the
cis-isomer55. This rapid isomerization was also demonstrated in this

laboratory to occur in benzene solution whereupon cis-PtCl2(PPh3)2

precipitated out.
The product obtained when Pt(PPh3)4 reacts with IC1 in a 1:6
mole ratio is trans-PtCl2(PPh3)2. Initially, one might be tempted to
assume that the trans-PtC1I(PPh3)2, which is formed first, would be
converted into trans-Ptl2(PPh3)2 which, subsequently, undergoes halogen

substitution to form trans-PtCl2(PPh3)2. This assumption would seem
to be substantiated by the observed conversion by IC1 of trans-
PtI2(PPh3)2 to trans-PtCl2(PPh3)2. However, upon examination of the

reaction sequence shown below, it can be seen that

I-- P I P I P C1- P
IC1 IC1 IC1 IC1
Pt(O) ----- Pt -- Pt -- Pt -- Pt
-Cl2 -2I -12
P---C P- P--C1 P-CI


there is a flaw, albeit a subtle one; namely, the formation of both
trans-PtI2(PPh3)2 and trans-PtCl2(PPh3)2 from the same species, trans-
PtC1I(PPh3)2. The final product, trans-PtCl(PPh3)2,which is more
thermodynamically stable than trans-PtI2(PPh3)2 in the presence of
IC1, must be formed directly from trans-PtC1I(PPh3)2. If this be so,

the question arises as to why trans-PtI2(PPh3)2 is the predominant







product when the ratio of Pt(0) to IC1 is 1:3. Accordingly, it is
proposed that when the 1:3 ratio obtains, trans-PtC1I(PPh3)2 is, as

usual, formed first. But the subsequent attacking species is not
IC1, but another species. If it were IC1, trans-PtC12(PPh3)2 would

result as discussed above. The most probable attacking species is

I2C1- which will be formed from the reaction of PPh3 and IC1 according
to equation (45)
PPh3 + 2IC1 PPh3C1+I2C1- (45)

This rapid reaction is analogous to the reaction of PPh3 with IBr to
50
form PPh3Br+I2Br Analogous ionic species, in which the halogen

bound to the Group VA element is the one of lower atomic number, are

also formed with the reaction of triphenylarsine with IBr5056. Thus,
when the mole ratio of Pt(PPh3)4 to IC1 is 1:3, there is very little

IC1 to react with the trans-PtC1I(PPh3)2. Instead, the resulting

attack by 12C1~ on trans-PtC1I(PPh3)2 yields trans-PtI2(PPh3)2. All

the products in the reaction of Pt(PPh3)4 with different mole ratios

are now successfully rationalized.

Although there is no direct evidence concerning the nature of
the transition states involved in the oxidative addition of IC1 to
Pt(PPh3)4 yielding trans-PtClI(PPh3)2, it would appear to be useful

to speculate on the probable mechanism (Figure 20) involved in the

reaction. In the solution, Pt(PPh3)3 is assumed to be trigonal
planar. Attack by ICl at the Pt atom is perpendicular to the trigonal

plane. This is the preferred direction of attack which is the direc-
tion of the filled dz2 orbital of the Pt atom, such that a a-bond

can be formed between the metal atom and the incoming atom. At this

time, electron transfer from the metal to IC1 occurs with subsequent










P
Id~
-43
P



P P

I p


P]
P _-_ -
<_.___


-P


-P


trans


-Cl-
Z-4


Cl1


[P] P

P

S(D)

-P


-P


I P


Cl P

cis


Figure 20.
to IC1 to


Proposed Mechanism for the Oxidative Addition
Pt(PPh3)3 in the Absence of Free Phosphine


C16--







cleavage to form a distorted tetrahedron which collapses to a square

planar configuration typical of Pt(II) complexes. A chloride ion then

attacks the Pt atom along pz perpendicular to the plane with subsequent

formation of trigonal-bipyramidal species from which the product is

obtained.

Pentacoordinated species have been detected spectrophotometrically

in substitution reactions of PtC12(NH3)2 by N02 57, and are implicated

in cis-trans isomerizations of Pt(II) complexes58,59. A characteristic

property of trigonal-bipyramidal species is pseudorotation, which is

defined as the intramolecular process whereby a trigonal-bipyramidal

molecule is transformed by deforming bond angles in such a way that it

appears to have been rotated by 90 about one of the trigonal-planar

interatomic bonds. Thus, in the diagram below, the substituent A

remains fixed while the apical substituents are pushed backward and

the equatorial substituents pulled forward until the process leads

to the second trigonal bipyramid, which appears to have been produced

by rotating the first about the bond from the fixed substituent A

(the "pivot") to the central atom. The symbol [A] is used to denote



[A]





pseudorotation with A as a pivot. By means of pseudorotation, sub-

stituents in apical positions are placed in equatorial positions and

vice versa. From nmr studies on stabilities of possible positional

isomers in a trigonal-bipyramidal structure containing different

ligands, some empirical rules have been formulated with regards to the







preference of different ligands for either apical or equatorial

sites 6061. Ligands which are strong i-acceptors (CO, SnCl3-) prefer

equatorial sites, whereas, strong a-donors (H-, CH3-) and electro-

negative ligands favor apical positions. Those ligands which do not

show a propensity to be a T-acceptor or a o-donor (PR3) have no

particular preference for either site. Moreover, the stability of

positional isomers is dependent on steric interaction between ligands,

more bulky groups favoring equatorial sites. On the bases of these

rules, it can be seen in Figure 20 that, the square pyramidal structure

(E) can only give rise to structure (B) which pseudorotates to (A)

and from which the product trans-PtC1I(PPh3)2 finally results. The

product can also be obtained, though less likely, directly from (B).

(C) and (D),which can give rise to cis-PtCI(PPh3)2, are not formed

from (E), (A) or (B) by fluxional changes because they are the higher

energy forms of the possible position isomers on two counts: (1) Only

one electronegative halide ligand is apical and (2) The three PPh3

ligands experience most steric repulsions with two of them in equatorial

and the other at an apical position. Thus, on the basis of both elec-

tronegativity and steric effects, (C) and (D) are of higher energy

than (A) and, probably, are also of higher energy than (B) on the basis

of the overriding steric effect. Hence, no cis-isomers are formed.

When the Pt(O)species is Pt(PPh3)2, the mechanism is again the

same if it is assumed that it is trigonal planar in solution as a

consequence of solvation. It will be recalled that the rate of reac-

tion of this species with CH3I is several times faster than the rate

of reaction of Pt(PPh3)3 with the same halide. This can be accounted







for by the fact that PPh3 is a better w-acceptor than is the solvent

molecule so that the greater electron density on the Pt atom in Pt-

(PPh3)2 (solvent) makes it more nucleophilic and, hence, more reactive.
Alternatively if, in solution, the Pt(PPh3)2 species is linear,

the trans-product can be obtained if the IC1 attacks and forms the

trigonal planar PtI(PPh3)2+ with a subsequent attack of Cl- from above

the trigonal plane and finally, the collapse of the distorted tetra-

hedron into trans-PtCII(PPh3)2.

The mecahnism for the formation of trans-PtICH3(PPh3)2 from

Pt(PPh3)4 and CH3I can be postulated to follow a similar pathway.

CH3- (a good o-donor) and IF (an electronegative ligand) both prefer

apical positions and all the arguments advanced above obtain.

The reaction of Pt(PPh3)4 with IBr is analogous to that with

ICl. Using a Pt(O) to IBr ratio of 1:1 and 1:6, cis-PtBrI(PPh3)2

and trans-PtBr2(PPh3)2 (86%) are obtained, respectively. With the

ratio of reactants intermediate between the above two mole ratios,

e.g., 1:2, the product is an approximate mixture of cis-PtBrI(PPh3)2

and trans-PtI2(PPh3)2. Infrared spectroscopy of a reaction mixture

containing an initial mole ratio of Pt(O) to IBr of 1:3 still indicates

the presence of some cis-complex, cis-PtBrI(PPh3)2. In terms of %

yield and expediency, the reaction of Pt(PPh3)4 with IC1 in a 1:6

mole ratio, again provides a superior method of synthesis of trans-
PtBr2(PPh3)2 over any other published methods. The mechanism of the

reaction is proposed to be similar to that of Pt(PPh3)3 and IC1 and

will not be discussed further.







Oxidative Addition of Halogens to Tetrakis(triphenylphosphine)platinum(0)

Customary reactions involving the oxidative addition of Br2 and

12 to Pt(PPh3)4 have been reported either explicitly or implicitly to
yield invariably cis-PtBr2(PPh3)2 and PtI2(PPh3)2, respectively.1'5'62'63
Although no report of the reaction of Pt(PPh3)4 with Cl2 has been

reported, the general consensus among chemists in this research area
is that cis-PtCl2(PPh3)2 would result. This is evidenced by the fact

that, hitherto, synthetic methods for the synthesis of trans-PtX2(PPh3)2
(X = C1, Br or I) involve either photochemical49 or thermal isomeriza-
tion36 of the cis-isomers or from the reaction55 of PtHC1(PPh3)2 with

HgCl2 (which gives a very low yield) instead of the obvious reaction
of Pt(PPh3)4 directly with the halogens. The first inkling that the
addition of the halogens to Pt(PPh3)4 might yield trans-dihalogenobis-
(triphenylphosphine)platinum(II) came when the mechanism for the
oxidative addition of IC1 to Pt(PPh3)4 (discussed earlier) was con-

sidered. It is recalled that the latter reaction yielded trans-
PtClI(PPh3)2 as a consequence of the proposed energetically favored
trigonal bipyramidal transition state, PtCII(PPh3)3, in which the

electronegative halogen atoms occupy apical positions and the bulky
triphenylphosphine ligands occupy equatorial positions. This more
stable transition state can only yield trans-PtC1I(PPh3)2 by elimina-

tion of PPh3 from an equatorial position. It became apparent that,
if halogen is used instead of IC1, the most stable transition state in
the oxidative addition to Pt(PPh3)4 would also have to be one in which

the electronegative halogens are apical and the three PPh3 ligands
are at equatorial sites and that trans-isomers would result. The







fact that only cis-somers have been reported must be because of
isomerization, in the presence of triphenylphosphine, of the initially
formed trans-isomer to the cis-isomer as in the case of trans-PtClI-

(PPh3)2. This trans-cis isomerization has already been reported to
occur rapidly for PtCl2(PPh3)2 in chloroform This author has
demonstrated that the rate of trans-cis isomerization in benzene for

PtX2(PPh3)2 (X = C1, Br or I) follows the order:
C1 > Br >> I

Thus, in order to demonstrate that the addition of halogens to Pt(PPh3)4

is a trans-addition by isolating trans-isomers, the experimental

conditions must be such that no free triphenylphosphine is present

after the formation of trans-PtX2(PPh3)2. The phosphine-catalyzed

trans-cis isomerization reaction can then be prevented.
One way to prevent isomerization of the trans-products is to

"tie up" the free triphenylphosphine present in the reaction mixture
by using more than the stoichiometric amount of halogen required for

the oxidative-addition reaction. The excess halogen reacts with PPh3

to form [PPh3X]X (and/or [PPh3X]X3 depending on the amount of X2

used). Previous investigators, who obtained cis-isomers, were

reluctant to use an excess of the halogen, presumably for fear of
further oxidation of the initially formed PtX2(PPh3)2 to PtX4(PPh3)2.

However, it was found that the first oxidative addition of X2 to'
Pt(PPh3)4 proceeds faster than the second oxidative addition of X2 to

PtX2(PPh3)2 and by limiting the reaction time to 3 minutes or
less, the formation of Pt(IV) complexes can be eliminated. For 12

and Br2, the reaction with Pt(PPh3)4 for 3 minutes using a 4:1 mole








ratio of halogen to Pt(O) complex was found to be sufficient to yield

exclusively trans-PtI2(PPh3)2 and trans-PtBr2(PPh3)2, respectively.
The conditions were modified for the reaction of chlorine with Pt(PPh3)4.
Since trans-PtCl2(PPh3)2 is more rapidly isomerized in the presence
of PPh3 than trans-PtBr2(PPh3)2 or trans-Ptl2(PPh3)2, it was found

necessary to add the Pt(PPh3)4 solution in a fast dropwise fashion to
an excess of a stirred benzene solution of C12 so that at any given
time during the addition, no free phosphine can be present in the
solution. The reaction time was also limited to 1.0 minute to prevent

oxidation to PtC14(PPh3)2 since C12 is more reactive with both Pt(PPh3)4
and PtCl2(PPh3)2 than Br2 or 12 with the corresponding reactants.
With all three halogens, no cis-isomers could be detected by ir
spectroscopy in the products. Thus, the addition of the halogens to
Pt(PPh3)4 is demonstrated to be a trans-addition reaction. It is of
interest to note that this demonstration also provides very convincing
confirmation to the validity of the mechanism of the oxidative addition
of IC1 to Pt(PPh3)4 proposed earlier.
The reaction of 12 with Pt(PPh3)4 was also repeated using the

stoichiometric amount of reactants (1:1) for the formation of PtI2(PPh3)2.
The reaction time was extended to 15 minutes. It was found that
trans-PtI2(PPh3)2 was formed in spite of the presence of free tri-
phenylphosphine being present under these conditions. This demonstrates
that trans-PtI2(PPh3)2 isomerizes rather slowly in the presence of

PPh3 and suggest strongly that earlier investigators had indeed obtained
the trans-PtI2(PPh3)2 but thought that they had cis-PtI2(PPh3)2 instead.
Their mistaken notion might have been the result of an analogy with







the result of C12 and Br2 with Pt(PPh3)4 where cis-isomers are indeed

formed under their conditions of a 1:1 mole ratio of the reactants.

With trans-Ptl2(PPh3)2 cis-trans isomerization appear to occur

more readily instead of trans-cis isomerization. Mastin has reported36

that the thermal isomerization of the cis-isomer occurs in a refluxing

chloroform solution containing 2% ethanol. This author has found

that the isomerization also proceeds in a refluxing solution of

benzene and even in the solid state by heating at 2000. Thus, trans-

PtI2(PPh3)2 appears to be relatively more thermodynamically stable
with respect to cis-PtI2(PPh3)2 than trans-PtBr2(PPh3)2 and trans-

PtC12(PPh3)2 with respect to their respective isomers.

Substitution Reactions of Dihalogenobis(triphenylphosphine)platinum(II)
Complexes

In previous discussion on the mechanism of the oxidative-addition

reaction of Pt(PPh3)4 with IC1 (1:6 mole ratio), it was proposed that
initial oxidative addition produced trans-PtClI(PPh3)2 which then

reacted with more IC1 to form the final, isolated product trans-

PtC12(PPh3)2. It became necessary then, to perform the experiment
involving the reaction of cis-PtCll(PPh3)2 with IC1 in an attempt to

demonstrate if indeed, this reaction would yield cis-PtC12(PPh3)2.
If this result obtains, then by analogy,the reaction of trans-PtClI-

PPh3)2 with IC1 would have to yield trans-PtC12(PPh3)2. It is noted
here that cis-PtC1I(PPh3)2 was used out of necessity, because the
trans-isomer was not isolated. Subsequent to this experiment, a

number of reactions of cis- and trans-PtXY(PPh3)2 (X or Y = C1, Br

or I) with IC1, IBr and Br2 were studied.








The complexes, cis-PtXY(PPh3)2, are only sparingly soluble in

benzene and hence the reactions were carried out in either CHC13 or

CH2Cl2 in which they are soluble. For the complexes, trans-PtX2(PPh3)2,

benzene was used as the solvent. The general procedure which was

followed involved the mixing of an ethereal solution of the interhalogen

or halogen with a solution (C6H6, CH2C12 or CHC13, where appropriate)

of PtXY(PPh3)2 and stirring the mixture for 5-15 minutes. The reaction

mixture was then concentrated to a small volume followed by addition

of MeOH to effect more complete precipitation of the product. On

subsequent recrystallization from benzene (for trans-products) or

from chloroform-methanol (for cis-products), the pure products were

obtained.

From the reactions of cis- and trans-PtXY(PPh3)2 with inter-

halogens and halogens, the following observation are made: (1) The

reactions are substitution reactions in which a less electronegative

element is always replaced by a more electronegative one. For example,

the reaction of cis-PtCII(PPh3)2 and trans-PtBr2(PPh3)2 with IC1 yields

cis- and trans-PtCl2(PPh3)2, respectively. (2) The reactions are

stereospecific, i.e., cis- and trans-reactants yield cis- and trans-

products, respectively. This observed stereospecificity is in accord

with all other substitution reactions of square-planar Pt(II) complexes.

(3) Most of the reactions are completed in about 5 minutes at room

temperature. The occurrence of such facile substitution reactions

of Pt(II) complexes, which are well-known for their inertness, is

surprising. (4) The yields are high (> 80%). (5) Under the reaction

conditions which were employed, no oxidative addition of the interhalogen