A.

 

This is to certify that the

thesis entitled

PHOTOLYSIS OF TITANOCENE DICHLORIDE

presented by

— Zei-Tsan Tsai

has been accepted towards fulfillment
of the requirements for

,_£hm_ degree in W

H .

Major professor

Date 7LW- ‘5" “17%

 

0-7 639

 

 

 

 

PHOTOLYSIS OF TITANOCENE DICHLORIDE

By

Zei—Tsan Tsai

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements

for the degree of

DOCTOR OF PHILOSOPHY

Department of Chemistry

1978

ABSTRACT
PHOTOLYSIS OF TITANOCENE DICHLORIDE
By
Zei-Tsan Tsai

The photochemistry of titanocene dichloride, Cp2TiCl2,
2, (CD = nS-CSHS), and its analogs is reported. The
photolysis of Thin the presence of Ph3CCI or nitrosodurene
at 20° gives Ph3é radical or a b-line esr signal attribut-
ed to 2,3,5,6-tetramethylphenyl cyclopentadienyl nitroxide,
respectively. Both the CpTiCl2 fragment and the spin
adduct of the Cp radical and nitrosodurene were observed
when irradiation was carried out at —85°. Irradiation
of l in a benzene matrice (-196°) produced detectable Cp
radical and Cp'I‘iCl2 signals. The photogenerated CpTiCl2
species can abstract a halogen atomfrom either 1 or CHBr3
to produce a titanocene trihalide. The photogenerated Cp
radicals can couple with each other in pure THF or benzene
to give 1,5-dihydrofulvalene. The photolysis of l in the
presence of alcohols or acetone yields Cp(alkoxyl)TiC12.
All of the observed photochemistry can be interpreted as
arising from homolytic Ti-Cp n-bond cleavage in the excited
states derived from ligand + metal transitions associated

with the Ti-Cp fl-bond.

To my parents who have given me so much.

11

ACKNOWLEDGMENTS

I would like to express my sincere appreciation to my
advisor, Professor Carl H. Brubaker, Jr., for his patient
guidance, encouragement and assistance throughout my re-
search.

I would also like to thank Professors T. J. Pinnavaia,
P. J. Wagner and G. E. Leroi for serving on my committee.

Financial support from the Institute of Nuclear Energy
Research, Republic of China, is greatly appreciated.

Finally, I wish to express my deepest appreciation
and thanks to my wife, Suing, for her love, inspiration

and understanding.

iii

TABLE OF CONTENTS

Chapter Page

LIST OF TABLES. . . . . . . . . . . . . . . . . . v
LIST OF FIGURES . . . . . . . . . . . . . . . . . vi
INTRODUCTION. . . . . . . . . . . . . . . . . . . l
RESULTS AND DISCUSSION. . . . . . . . . . . . . . 8

l. Spectra. . . . . . . . . . . . . . . . . 8
2. Esr Studies. . . . . . . . . . l2
3. Photolysis in the presence

of bromoform . . . . . . . . . 27
A. Photolysis in pure THF and

benzene. . . . . . . . . . . . . . . . 30
5. Photolysis in the presence

of alcohols and acetone. . . . . . . . . 37
CONCLUSIONS . . . . . . . . . . . . . . . . . . . U0

EXPERIMENTAL. . . . . . . . . . . . . . . . . . . U1

1. General. . . . . . . . . . Al
2. Preparation of (05D5)2—

T1012. . . . . . . . . . . . M2
3 Preparation of (Mecp)2-

T1012. . . . . . . . . . . . . . N3
A. Preparation of nitro-

sodurene . . . . . . . . . Au
5. Preparation of Ph3CCl .and.

Ph3C radical . . . . . . . . . . . an
6. Esr experiments. . . . . . . . . . . AA
7. Photolysis in the presence

of bromoform, alcohols and

acetones . . . . . . . . . . . . . . A5
8. Photolysis in pure THF and

benzene. . . . . . . . . . . . . . . N8

BIBLIOGRAPHY. . . . . . . . . . . . . . . . . . . 49

iv

Table

LIST OF TABLES

Esr parameters of the Cp radical.

Electronic spectral band maxima for

Cp2TiCl2 and (MeCp)2TiCl2

Products obtained from photoreactions

of I with substrates. . . . .

Mass spectral patterns of photo-
products of I in the presence of

substrates. . . . .

Page

38

39

LIST OF FIGURES

Figure Page

1 Comparison of absorption spectra
of Cp2T1012 (---—) and (MeCp)2TiCl2
(————) in CH3CN at the identical
concentration of 3.3 x 10"3 fl_. . . . . 10
2 Comparison of absorption spectra of
Cp2T1C12 (----) and (MeCp)2T1012
(-———) in THF at the identical con-
centration of 3.3 x 10'3 n. . . . . . . 11

3 Esr spectrum of the photolyzed
Ph3CCl in benzene at 20°. . . . . . . . 13
h Esr spectrum of Ph3C radicals

obtained from.a photoreaction of I

and Ph3CCl in benzene at 20°. . . . . . l3
5 Esr spectrum of the photolyzed THF

solution of I_at 20° for a 30 min

irradiation . . . . . . .3. . . . . . . 15
6 Esr spectrum of the nitroxide radi-

cal obtained from a photoreaction

of I with ArNO in CH013 at 20° for

a 1 min. irradiation. . . . . . . . . . l7
7 Esr spectrum of the nitroxide radi-

cal obtained from a photoreaction

of (CSD5)2T1C12 with ArNO in THF

vi

Figure

10.

ll

12

Page

at 20° for a 6 min. irradiation . . . . 19
The decay of the esr signal, obtained

from a photolyzed CHC13 solution of I

in the presence of ArNO at 20°, after
extinguishing the lamp for (a) im—

mediate (b) 2 (c) 3 (d) A (e) 5

min . . . . . . . . . . . . . . . . . . 22
Esr spectra of the photolyzed

benzene solution of I in the

presence of ArNO at 20° (a) for

a 20 min. irradiation (b) after

extinguishing the lamp. . . . . . . . . 23
Esr spectrum of the nitroxide

radical obtained from a photo—

reaction of (MeCp)2TiCl2 with

ArNO in chloroform at 20° for a

2 min. irradiation. . . . . . . . . . . 25
Esr spectrum of the photolyzed

THF solution of I.at -78° for a

9 min. irradiation. . . . . . . . . . . 26
Esr spectrum of the photolyzed THF

solution of.I in the presence of

ArNO at -85°. Irradiation time

(receiver gain): (a) 6 min. (1.6 x

102) (b) 8 (5 x 102) (c) 10 (10 x 102)

vii

Figure

13

1A

15

l6

17

18

19

Page

(d) 12 (8 x 102) (e) 15 (8 x 102)

(r) 17 (8 x 102). . . . . . . . . . . . 28
Esr spectrum of the photolyzed

benzene solution of I_at -196°

for a 185 min. irradiation. . . . . . . 29
Uv absorption spectrum of the

separated THF solution from the

photolyzed THF solution of I. . . . . . 31
Uv-visible absorption spectrum of

the filtrate obtained from the

treated residue of the photolyzed

THF solution of I . . . . . . . . . . . 33
Uv absorption spectrum of the

separated benzene solution from

the photolyzed benzene solution

of I, . . . . . . . . . . . . . . . . . 3A
Uv absorption spectrum of the

filtrate obtained from the treated

residue of the photolyzed benzene

solution of‘I . . . . . . . . . . . . . 35
Uv absorption spectrum of the

separated benzene solution from

the photolyzed benzene solution

of (MeCp)2T1012 . . . . . . . . . . . . 36

The photolysis apparatus. . . . . . . . A6

viii

Figure

20

Page
The changes in the visible absorp-
tion spectrum of a photolyzed
benzene solution of I in the pres-
ence of l—propanol for (a) 0 (b) 10
(c) 15 and (d) 20 min. irradiations . . A7

ix

INTRODUCTION

Titanocene dichloride, Cp2Ti012, ;, (Cp = nS-CSHS),
has been widely studied, mainly in connection with re—
actions yielding the disubstituted o-alkyl and n-aryl deriva—
tives, and often used in homogeneous catalysisl. The Ti-Cp
n-bond in structure I is thermally inert in most chemical
environments. For instance, I will not yield ferrocene
by reaction with ferrous chloride2, or nickelocene by
reaction with nickel carbony13. Electrochemical reduction
of I destroys the Ti-Cl bonds, but leaves the Ti-Cp n—bonds
intact“. Notable thermal reactions involving Ti-Cp n-bond
rupture in I are cleavage by amines5 and redistribution

reactions with T101“, as in reaction 16.

T101“ + Cp2TiCl2 + 2 CpTiC13 (1)

Also, heating a saturated solution of I in dimethyl sul-
foxide at 70° effects an irreversible n-c rearrangement of
the Cp ligands7.

Photochemically-induced reactions of organotransition
metal compounds have been known for a long time8, but they
.have undergone detailed exploration only in the past 20

lyears. Most investigation concentrated on the binuclear

carbonyls of cyclopentadienyl manganese, rhenium, molyb-
denum and tungsten. They undergo homolysis of metal-metal
bonds (reaction 2), as evidenced by

hv
R(CO)nM-M(C0)nR -> 2 M(C0)nR (3)

(R = nS-CSHS), co; M Mn, Re, Mo, w; n = 3, 5)

the following observations: (1) the clean stoichiometric

formation of CpM(CO)3Cl (M Mo, w>9 and M(C0)501 (M =

Mn, Re)10 when C01“ or CHCl3 is the chlorine donor, (2)

the observation of Ph3c- radicals when Ph3CCl is the chlor-
ine donor9b and (3) the formation of the spin adducts of
M(C0)nR with nitrosodurenell. In addition, several review
articles have appeared in recent yearslz, but very few
examples involved study of titanium complexes.

Hunter and Winter13 found that titanium haloakoxides
were suitable for photoreduction to provide a very simple
route to Ti(III) complexes. Photoreduction was also
found with an oxalato-Ti(IV) complex in aqueous oxalic
acid solutionlu. The photocleavage of the Ti-Cp n-bonds
in Cp2TiMe2 was not observed. The reaction instead pro-
ceeds vIa homolytic cleavage of Ti-C o—bond, with the
formation of the black titanocene and methyl radicals, as
in reaction 3. The titanocene was characterized by the

hv

formation of a metallocycle derivative in the presence
of diphenyl acetylene or Cp2Ti(C0)2 in the presence of

0015’16. The methyl radical abstracted a hydrogen atom

from either a methyl group or a Cp ring to form methanel6

or formed stable spin adducts with the appropriate spin

trapping agentsl7.

18

However, Harrigan and coworkers observed that the

photolysis of I in chlorinated hydrocarbons (CHCl or C01“)

3
gave CpTiCl3 and they proposed that light-induced cleavage

of the Ti-Cp n-bond occurs, as in reactions A and 5.

hv
;.-+ CpT1c12 + °Cp (H)
CpTiC12 + CHc13 + CpTiC13 + -CH012 (5)

Furthermore no evidence for Ti-Br bond rupture was found
on photolysis of Cp2TiBr2 in chloroform. At about the
same time, Vitz and Brubaker19 observed that exchange of
the Cp ligands between molecules of 2.1“ benzene and the
alcoholysis of I in a methanol-benzene solutions occur
solely by photochemical processes, as in reactions 6 and
7. Studies of photoexchange of Cp ligands have been

hv
(05H5)2TiC12 + (C5D5)2TiCl2 -+

2 (C5H5)(C5D5)T1012 (6)

hv
I + MeOH -> Cp(Me0)TiC12 + C5H6 (7)

20

extended to ziconium analogs and to systems with d1

metal ion complexes, including CpZVCl2/Cp2VCl2—dlo and
21. I photoelectron spectra of

10
Cp2MCl2 (M = Ti, Zr or Hf) have been reported22, in which

Cp2TiCl/Cp2TiCl-d The He
transitions involving electron transfer from the Cp ring
to the empty d—orbitals of the metal occur at lower energy
than those arising from the Cl ligands to the same orbitals.
The direct detection and identification of short-lived
free radicals by esr is possible only if the radicals are
produced in relatively high concentration in the esr cavity
by intense In slip irradiation or by rapid-mixing flow
systems. Recently, an indirect technique for the identi-
fication of free radicals in reacting systems has been de-
veloped23. This method relies on the idea that the reactive
free radicals add easily to diamagnetic scavengers (spin
traps) to form stable radicals (spin adducts), whose esr
spectra afford information about the structure of the initial
free radicals trapped. The two kinds of spin traps, most
commonly used, are nitroso-compounds and nitrones. They
react with various kinds of free radicals to give relatively
stable nitroxides. Nitrosodurene (ArNO, Ar = 2,3,5,6-

MeuC6H) has proved2u

to be a versatile spin trapping re-
agent, because of the simplicity of the spectra of the

Spin adducts and of its stability towards photolysis

comparable to others of its class, such as 2-methy1—2-

nitrosopropane and 2,A,6-tri—t-butyl nitrosobenzene. For
example, nitrosodurene has been employed successfully in
the formation of spin adducts with photogenerated Mn(C0)5

l and with photogenerated methyl

radicals from Mn2(CO)lOl
radicals from Cp2MMe2 (M = Ti, Zr or Hf)16.

It has been reported25 that vacuum flash pyrolysis of
nickelocene generates 9,10-dihydrofu1va1ene which readily
undergoes rearrangement to another isomer, 1,5-dihydroful-
valene, in dilute solution, with a tl/2 of 52.3 min. at
30.0°. Both isomers can be characterized by uv absorption

spectrometry; the former shows a Ama of 2A0 nm (537.5x103,

x
heptane, 0°), while the latter gives Amax of 336 nm (e z

1.1lx1014

, heptane). The pathway leading to the formation
of 9,10-dihydrofulvalene is the coupling of two Cp radicals.
Cyclopentadiene is thermally unstable as well as photo-
sensitive. It couples with itself in the Diels-Alder di-
merization to yield tricyclo(5,2,1,02’6)-deca-3,8-diene

'6 '12-]. S-1

with a second-order rate constant of 2-2.5x10
at 25° 26. Ultraviolet irradiation of cyclopentadiene
produced bicyclo(2,l,0)-pent-2-ene27. Neither compound
shows a discernible uv absorption spectrum.

The esr study of the Cp radical has been extensively
investigated by several authorsza. The spectrum consists
of six equally-spaced lines with relative intensity of

l:5:10:10:5:l and has no measurable anisotropy. The

hyperfine splitting due to five equivalent hydrogens de-
pends somewhat on the generation method of Cp radical
(Table 1).

We have attempted to determine the primary photoin-
duced species of I. Ph3C01, nitrosodurene, and CHBr3
were used to react with the photolytically generated species,
providing information about the photobehavior of I.

Dry acetone has been used as a medium for kinetics
studies of the substitution reaction of I with halides29.
The photolability of I towards acetone as well as extended

alcoholysis studies have been also explored.

 

 

 

 

 

xpoz moan QEMH mm mnzmmmha swan 3xIH m spas .mhs m pom
mace NHOHBNQO mo scandaom ocoucmp m no COHpmHumppH m.m
mmm om: 0p om- msmmn cossomfim spa: mmmo co cospwfismssa o.o
n .mmm moat mcmaopaoaozo CH
memo was mnosmusv co cofipsflom s so mamsHOQOnm o.m
mmm mama on final dEwH mm madmmmma swan m new: mason
: pom mmmo we Hmummpo mawcfim m no coapwfioman H.Mm.m
pmm moan .mcoooppcm no Emma amazomHoE mo mHmzHOme mm.m
owm wmfiu on omau msmmp cospomflm spa: mmmo mo cofipmfiemssH H.Hm.m
9mm mm mofippma manpcmsmcm
CH mp5H0m mm cams .mmmo no mam»a0aumnl> H.m
.mom Acov .QEmB awofipwp no mo nonpme soapwhocmo Amy mm
.HMoprm do on» we mpmmempma 9mm .H magma

RESULTS AND DISCUSSION

1. Spectra

The band maxima and molar absorptivities are summarized
in Table 2 for I and (MeCp)2T1012. The electronic spectra
of both complexes have been measured in CH3CN and THF and
are shown in Figures 1 and 2. The absorption bands of I
are in good agreement with the reported data18’3o, except
that a band is observed at about 255 nm which was not

300 303 were used as solvent.

reported when acetone and benzene
The underlying purpose of the methyl substitution on the
Cp ring is to elucidate the electronic spectra, in which
both d-d and charge-transfer transitions may be revealed.
For the latter excitation, the observed bands, correspond-
ing to ligand + metal or metal + ligand transitions, depend-
ing on whether the band, observed on methyl substitution,
will shift to lower or higher energies3l.

The spectra are dominated by an intense (e 2 20,000

M'1 cm'l) absorption with a maximum in the uv region.

The band indicative of methyl substitution on a Cp ring
shifts to lower energies compared to that of I. Each
complex also exhibits a lower intensity band (a z 200 Mfl

cm‘l) in the visible region. The band maxima and molar

Table 2. Electronic spectral band maxima for Cp2TiCl2
and (MeCp)2T1012.

 

 

 

Complex Solvent Bands, nm ( , M51 cm—l)
Cp2T1C12 CH3CN 522 (170), 392 (1800), 250 (19800)
THF 515 (150), 385 (2730), 255 (17700)

(MeCp)2T1012 CH3CN

THF

528 (220), 395
520 (230), 388

(1680),
(19A0),

255 (25300)
258 (22500)

 

 

10

 

.a muoaxm.m to
COHumppcooCOo HmoHpCmcH map um mommo EH A O NHOHBmfiaoozv
can AIIIIO NHOHBNQO mo whpowmm cofiuahomnm no somHLmQEoo

 

E: .prcoHo>m3
omm com 0mm oom om: oo: omm oom Omm

 

OOH " H
UmpSHHU

OH . H
OmdeHU

 

.H wadem

Aqrsueq reorndo

11

.a mIOme.m
Oo eOHpmspewocoo HOOHscmOH we» as ems OH H V NHOHONHOOOEV
vcm Aulnuv NHOHBNQO mo mpuomaw SOHpQHownm mo coanquoo

 

Em .npwcmHm>w3

 

omw OOO Omm OOm cm: 00: omm oom 0mm

 

 

             

     
 

i O.O
OOH " H
OmOsHHO m
m - m.O
OH H H m - .
OOHOHHO i O O
... . O
M a; . O.O

 

.m onszm

KQISUGG tearqdo

l2

absorptivities depend somewhat on the solvents used. Aceto-
nitrile causes a red shift compared to THF. Both I and
(MeCp)2T1012 are d° complexes. The shifts of the absorp-
tion bands of the methyl substituted species to lower
energies, as well as the dependence of absorption band
positions on solvents, confirms that the corresponding
excitations do not involve electrons occupying d-orbitals.
The three transitions are all of ligand + metal character.
The M9 calculation for I has been carried out by several

authors. Some authors22b’30b’32

22a

regarded I as a C mole-

2V
cule, others consider that I belongs to CS symmetry.
Nevertheless, the three excitations occur in such a way

that the electron transfers from the Cp ring to the empty

d-orbitals of the Ti metal.

2. Esr Studies

 

Photolysis of a benzene solution of I in the pres-
ence of Ph3CC1 produced the esr detectable Ph3C free radi-
cal (Figure U), while the solution containing Ph3CCI alone
did not give one (Figure 3). This information means that
photolysis of a solution of I undergoes free radical for-

mationgb, with formation of either a Cp radical and CpTiCl2

or a Cp2T101 fragment and a Cl atom. In either case, the

CpT1012 or Cp2TiC1 is a d1

complex which can be detected
directly by its esr signal. Photolysis of a THF solution

of I_at 20° for 2 min. yielded a singlet with a g value of

13

3360 G

\L

hx-pVMr-‘mtyLN-MWW-”st/«ue M»

10 G

A
I

Figure 3. Esr spectrum of the photolyzed Ph30C1 in
benzene at 20°.

3360 G

M

“ 10G

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure A. Esr spectrum of Ph3C radicals obtained from
a photoreaction of I_and Ph3CC1 in benzene at
20°.

1A

1.976. At increasing irradiation time, a second singlet
with a g value of 1.953 gradually appeared (Figure 5).
The singlet with g = 1.976 and the presence of weak satel-
lites, attributable to hyperfine interaction with Ti iso-
topes (u7Ti and “9T1, nuclear spin of 5/2 and 7/2 present
in natural abundance of 7.75 and 5.51%, respectively),
allow one to identify this signal; it is certainly due to
a Ti(III) species with the unpaired electron residing largely
on the meta133. The second singlet was undetermined.
After 2.5 hrs. of irradiation, the solution turned green.
Both signals last for at least 18 hours, after the end
of irradiation, in darkness at room temperature.

In order to obtain more constructive information from
a spin trapping experiment, an appropriate spin trapping
agent was selected. Because esr spectroscopy is such an
extremely sensitive probe for free radicals that a spin
trapping experiment may yield a "positive" result on a
minor side reaction, while the main reaction is overlooked,
if it is nonradical, or even when it does involve a radical,
if the radicals are not readily trapped or yield non-per-

3“ that nitroso-

sistent spin adducts. Kinetics have shown
compounds are effective spin trapping agents towards various
of n-alkyl radicals; the rate constant for their formation
is as fast as 3.9 x 107 Mfl s'l.

A photolyzed CHCl3 solution of I in the presence of

nitrosodurene at 20° displayed a four equally-spaced

15

336A G 3u07 G

1.978
1.953

09
H
II

 

20 G

 

NV
09

m
II

 

 

 

Figure 5. Esr spectrum of the photolyzed THF solution
of I_at 20° for a 30 min. irradiation.

16

lines (Figure 6) with relative intensity of 1:2:2:1. The
distance between two adjacent lines is 13.0 G. A rational
assignment to this signal is 2,3,5,6-tetramethy1phenyl
cyclopentadienyl nitroxide, as expected for the radical
resulting from the 0p radical adding to nitrosodurene

(reaction 8), with hyperfine splittings: aH = 13.0 G and

(343 (”43 [4
H H
ArNO + Cp -> H N (8)
a e H
“3 CH3 H

aN = 13.0 G. The signal corresponding to the unpaired
electron at oxygen (discussed later) was first split by
the 1”N nucleus (I = 1) into a triplet, then each of trip-
lets was split by the 6 hydrogen on the Cp ring, into a
doublet, giving six lines. Since the value of aN equals
that of aH, the resulting esr spectrum exhibits a quartet
with a relative intensity of 1:2:2:1. The reasons are
that (1) the values of 3N (13.7 G) and aH (12.17 G) on the
previous reported spin adduct of ArN(0)CH3 are close and
(2) none of the hyperfine splitting arising from the 'Y
hydrogens on ArN(0)CH2CH3 was foundzu. Consistent with

this assignment, when the experiment was repeated with

17

338“ G

O' H H

CH3 CH3
m g = 2.005
aN = 13.0 G
aH = 13.0 G

 

 

 

 

 

 

 

10 G

 

V/

 

 

 

 

Ffiigure 6. Esr spectrum of the nitroxide radical obtained
from a photoreaction of I with ArNO in CHC13
at 20° for a l min. irradiation.

l8

(C5D5)2T1012, a Spectrum having a triplet of triplets
pattern was obtained, as in Figure 7. The smaller 1:1:1
triplet is attributed to the B deuteron (I = l).

The g value of 2.005 and the hyperfine coupling constants
of the spin adduct of nitrosodurene and the Cp radical was
assessed from the observed spectrum. The g value and the
1“N coupling constant of a nitroxide radical can be best

discussed in terms of the two valence structures A and E.

Due to the large spin-orbit coupling constant of oxygen and

R
../R \. /R

T Z T
O- ..

:O
A B

to the presence of nonbonding, lone-pair electrons on oxygen,
the structure AIis mostly responsible for the observed
positive shift of the g value (g > ge = 2.0023).

It has been shown that the coupling constant of the
8 hydrogen of a nitroxide radical, aH(B)’ is related to
the dihedral angle, 035, i.e., the angle between the
B hydrogen-carbon-nitrogen plane and the carbon-nitrogen-
p(N)-orbital plane. An equation relating the coupling

constant and 0 was proposed initially by Heller and

l9

CH3 CH3 D
33516 3‘0
N C
\ \l/ ‘ H (J D 0
l ”3 “3
g = 2.005
aN = 13.0 G
2.0 G

 

 

 

L
36::

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

10 G

A
I

Figure 7. Esr spectrum of the nitroxide radical obtained
from a photoreaction of (C D ) T1012 with ArNO
in THF at 20° for a 6 min Irgagiation.

2O

McConnell36 and is given by Equation 1, where B0 and B2
a = (B + B cos2e)p (1)
H(B) o 2 N

are empirical constants and 0N is the spin density at the
nitrogen pTr orbital. A good approximation with B0 = 0

and B2 = 50 G has been obtained for neutral hydrocarbon
radicals350. The value of pN can be estimated35b by Equa-

tion 2, where p0 is the spin density at the oxygen atom.
aN = 35.61 0N + 0.93 00 (2)

Substitution of the values of pN, BO and B2 into Equation
1 gives Equation 337 (neglecting the po term).

aH(B) = 50(c0520)pN = 50(cos20)(aN/36) G (3)

Substitution of the values of aH (13.0 G) and aN (13.0 G)
into Equation 3 gives 6 = 35.5°. As expected, for steric
reasons, the nitroxide radical is forced to adopt the
conformation Q; the 8 hydrogen has twisted out of the

nodal plane of the pfl(N) orbital about 35.5° (only if the

B hydrogens are not located in the nodal plane of the

pfl(N) orbital, i.e., 0 # 90°, is there an observable hyper-
fine splitting by the 8 hydrogen). The narrow line width

of the spectrum of 9 (AH = 0.5 G) suggests

peak-to-peak

21

 

that the conformation 9‘15 locked.

The signal of the photolyzed IHArNO solution decayed
in the absence of light to a much weaker six line signal
(Figure 8). No attempt was made to assign it. The an-
nihilation of the signal can probably be attributed to the
untrapped Cp radicals adding to the nitroxyl group of ArN(0)Cp
to form ArN(0Cp)Cp38, since the Cp radical resembles the
cyclopentyl radical and is neither a good oxidant nor a
reductant. The rate constant for this reaction is likely
to be greater than that for the reaction of ArNO with Cp
radical.

The same esr studies were repeated with benzene as
the solvent. The esr spectrum of a photolyzed benzene
solution centaining I and ArNO gave a quartet but the signal
lacked a discernible relative intensity of l:2:2:l (Figure
9), compared with that of the case of the CHCl3 solution
(Figure 6).

The spin trapping experiment with (MeCp)2TiCl2 was

 

 

 

 

V

Figure 8. The decay of the esr signal, obtained from a
PhOtOlyzed CHCl3 solution of I in the presence
of ArNO at 20°, after extinguIShing the lamp
for (a) immediate (b) 2 (c) 3 (d) A (e) 5 min.

 

 

 

 

 

 

 

 

 

 

 

 

H H
3321: G H N j
I
‘L ‘ l 0' H H
l‘ 1 CH3 CH3
W (J
(a) ryJ Wm
g = 2.005
aN = 13.0 G
i N ' aH = 13.0 c
20 G
g

(b) M

Figure 9. Esr spectra of the photolyzed benzene solution
of I.in the presence of ArNO at 20° (a) for a
20 min. irradiation (b) after extinguishing the
lamp.

2“

carried out in order to characterize the primary photo-
induced species of I further. The system gave two signals:
a triplet and a quartet (Figure 10), indicating that the
MeCp radical can bind to nitrogen in two sites. The quartet
is the same one as mentioned before, while the triplet is
caused by species that do not involve the 6 hydrogen.

If the photolysis of I results in the homolytic cleav—
age of the Ti-Cp w-bond, one expects to observe the nitroxide
radical from a Cp radical as well as from the CpTiCl2
species. Failure to form a stable spin adduct of ArNO and
CpT1012 might be due to this adduct's readiness to undergo
photolysis to produce a diamagnetic Ti(IV) species, as
evidenced by a red precipitate found in the esr tube after
5 min. of irradiation.

Low temperature esr studies were attempted in order
to identify both the Cp radical and the CpTiCl2 species
simultaneously. The photogenerated CpTiCl2 species from I
in THF at -78° was easily characterized by esr spectros-
copy (Figure ll). Since the melting point of CSH6 is -80°,
it was reasonable that the rate of annihilation of the photo-
generated Cp radicals is faster than the esr time scale,
consequently it could not be observed in this system.
However, photolysis of a THF solution of I in the pres-
ence of ArNO at -85° yielded a quartet, which can be
assigned to the spin adduct of the Cp radical and ArNO, in

addition to a singlet assignable to the CpT1012 species,

25

3388 G

H
CH3 H
"O?
OH H
33926 "3%
g =2.005
(L aN = 13.0 G
a = 13.0 G

 

 

 

 

 

 

 

 

 

“ v
100 ”08””

 

 

 

 

 

 

 

Figure 10. Esr spectrum of the nitroxide radical obtained
from a photoreaction of (MeCp) T1012 with ArNO
in chloroform at 20° for a 2 m n. irradiation.

26

3273 G

20 g g = 1.976

 

 

 

 

Figure 11. Esr spectrum of the photolyzed THF solution
of I_at -78° for a 9 min. irradiation.

27

as in Figure 12. The another singlet with a g = 1.953
was undetermined and appeared in previous system in the
absence of ArNO (Figure 5).

Photolysis of I in benzene for 185 min. at —196° gave
a singlet with g = 1.983, CpTiClZ, an apparent quartet
with g = 2.003 and a spacing of 5.5 G, and a doublet with
g = 2.01A and a spacing of A73 G (Figure 13). The apparent
quartet has a relative intensity of l:2:2:l and there are
two hidden peaks indicated by star marks: that is, the
actual spectrum is a sextet and due to the Cp radical
(Table l). The doublet signal is unidentified and probably
cannot be attributed to the hydrogen atom whose esr spectrum
has a splitting of 506 G39. The same esr studies were
repeated with (C5D5)2TiC12 in benzene and with I in deuter-
ated benzene. Those two systems also gave the identical

doublet with spacing of A73 G.

3. Photolysis in thegpresence of bromoform

In order to gain insight into the source of chlorine
for the formation of CpTiCl3 during the photolysis of I
in CHC13, CHBr3 was used as a substrate. The photolysis of
a benzene solution containing 8 x 10"3 M'of I and 8 x 10"2
M CHBr3 gave CpTiClB, CpTiBrZCl and CpTiBrCla. They were
identified by mass spectrometry. The photopathway t0

CpTiBr012 is expected to be analogous to reactions A and

Figure 12.

 

(a)

(b)

 

(c
20 G
——>
d
( ) 3271 c 3312 G
(e)
$1 = 1.977
3226 G 82 = 1'953
(f) ‘1’ Z81 zg2

g = 2.005

 

Esr spectrum of the photolyzed THF solution of
I,in the presence of ArNO at -85°. Irradia-
tion time (receiver gain): (a) 6 min. (1.6
x 102) (b) 8 (5 x 10 ) (c) 10 (10 x 102) (d)
12 (8 x 102) (e) 15 (8 x 102) (r) 17 (8 x 102).

29

I .20HpmHomHnH .CHE mmH m pom
omeI pm H mo :0HpsHom mcmuson uman0po£Q on» we Esnpomam Hmm .mH mmstm

 

 

 

 

 

 

O mmmm

 

 

 

 

 

 

 

 

O mmmm ‘

 

 

 

 

 

O mmmm

30

5. There are two possible ways to obtain CpTiClB: (1)
the exchange of halides between molecules of CpTiBrClz,
as in reaction 9 and (2) the abstraction of chlorine atom

from I by CpTiClZ, as in reaction 10.
2 CpTiBrC12 + CpTiC13 + CpTiBr2Cl (9)

CpTiCl + l + CpTiCl3 + Cp2TiCl (10)

2

The coresponding unirradiated solution was stirred for
2 days under darkness and gave Cp2TiClBr, determined by

mass spectrometry, instead of CpTiCl3 and CpTiBrCl2.

A. Photonsis in pure THF and benzene

 

The uv absorption spectrum of the separated THF solu-
tion from the photolyzed THF solution of I displayed two
absorption peaks: 336 nm and 2A3 nm (Figure 1A). The
peak at 336 nm is probably due to 1,5-dihydrofulvalene25
-- the coupling of two photogenerated Cp radicals. In
addition to the coupling process, the photogenerated Cp
radical may abstract a hydrogen atom from either solvent
or I_to form CSH6 which has a uv absorption at about
2A0 nm (superimposed on 9,10—dihydrofulvalene25). Although

THF is not a good hydrogen donor towards a radical, we

cannot ascertain that the peak at 2A3 nm is completely due

31

.H mo COHpSHom mus UmnzHouona

 

 

 

m Eon
n» w COHpsHom mme vopmpmdmm on» p0 Esnpomam COHpQHomnm >2 .zH mnsw
E: .spwcmHo>m3 Hm
omm ozm com com 0mm
. a 0.0

1

I. N00

l :00

 

KATSUGG reorndo

32

to 9,10-dihydrofulvalene. But 1,5-dihydrofulvalene is
derived from 9,10-dihydrofulvalene25, 9,10-dihydroful-
valene may be attributed in part to the 2A3 nm peak. The
photoproduct of 9,10-dihydrofu1valene only amounted to

about 8 x 10-“

M (approximately a yield of 10%) which is
insufficient to give an 1H nmr spectrum).

Figure 15 is the uv absorption spectrum of the fil-
trate obtained from the treated residue of the photolyzed
THF solution of I. The filtrate is yellow, indicating that
the residue contains CpTiCl3 -- I is almost completely in-
soluble in hexane. The peak at 375 nm is probably due to
both CpTiCl3 and 1,5-dihydrofulvalene, while the peak at
235 nm may be due to the combined absorption of 05H6 and
9,10-dihydrofulva1ene.

For benzene solutions, the uv absorption spectra were
less informative (Figures 16 and 17). These spectra at—
tributed to benzene which possibly undergoes photo or
thermal reaction with the Cp species or dihydrofulvalene
isomers in Diels-Alder reactions._

Additional information was obtained from the photolysis
of a benzene solution of (MeCp)2TiCl2 (Figure 18). The
peak at 329 nm is probably due to the isomeric species
resulting in the coupling of two photogenerated MeCp radi-

cals.

33
Optical Density

.M mo COszHom mme cmumHouosa can mo osvama cmummnp on»
soap UmcHMpno mpmeHHm map mo Esppomam COHuQHOmnm mHnHmH>I>D .mH mpstm

E: anpwcmHo>m3

 

OOH OOH OOH oom ONm OON Osm
c H a q u q H u - 4 u - 00H
0.0
H.O . . H.H
1 4 MW
2
NO . . NH m.
9
.. 4 T.
G
m.o H 1 m.H m
S
. I.
1 1.
K
eO . I eH
O O . .mH

 

 

3A

.M mo COHpSHow mcmncmn UonHouona
on» Eopm COHpSHom mummcon Umumpmamm on» we Euppomam GOHpQHomnm >9

E: .npwcmHo>m3

on: 00: com 0mm owm

D .

.mH mmstm

o.o
Hu
.0
q
T.
O
m.

m.o
ad
a
u
S
Tr
AI?
nA

3.0

 

35

.M mo COHHSHom mcmncmn ooumHOHOnQ on» no mschmH
pounce» on» Eopm Umchuno mpmeHHm 0:» mo Euppomam COHpromnm >9

E: .nuowHm>m3

OOm Oem oom OON

1)
‘
ID

 

D
V
‘ C ‘

 

 

 

 

.OH mpstm
0mm
o.o
. m.o
. 2.0

 

Karsuec IPOIQdO

36

 

j l

 

0.7

0.5
>.
.p
H
U)
c
(D
D
H
m
0
.r—1
4.)
a
c:

0.3

0.1

Figure 18.

I L A

290 330 370 A10
Wavelength, nm

Iv absorption spectrum of the separated benzene
solution from the photolyzed benzene solution of

(MeCp)2TiC12.

37

5. Photolysis in the presence of alcohols and acetone

The photolysis of‘I (8 x 10'3 M_in benzene) in the pres-
ence of l-propanol or 2-propanol (8 x 10-2 M) produced
Cp(OCH20H2CH3)TiC12 and Cp(OCH(CH3)2)TiC12, respectively
(Table 3). Both isomers can be distinguished by their mass
spectra: the former has a fragment at m/e of 213, corres-
ponding to Cp(OCH2)TiCl2+, while thelatter has one at 227,

representing Cp(OCHCH3)TiCl +, as in Table A.

2

The photoproducts were predicted in a previous reportlg,
in which it was suggested that the photogenerated Cp radical
might abstract a hydrogen atom from an alcohol molecule to
give an alkoxyl radical which consequently combines with
CpTiCl2 to produce Cp(alkoxyl)TiCl2.

The photolysis of a benzene solution of I_in the pres—
ence of acetone led to Cp(OCH(CH3)2)TiCl2 which was iden-
tified by a comparison of its mass spectral pattern with
that of a known Cp(OCH(CH3)2)T1012 sample obtained from
thelight-induced reaction of I with the 2-propanol, Table
A.

When acetone is excited to its triplet state (n + n*),
the nonbonding electron, which is promoted into the anti-
bonding n* orbital, is concentrated largely on the carbon
atom, leaving the oxygen atom with a reactivity similar
to that of an alkoxyl radicallZd. The triplet carbonyl
compound, therefore, reacts with CpTiCl2 to generate

Cp(OC(CH3)2)TiC12, which subsequently abstract a hydrogen

atom from the solvent to give final product.

 

LI? .1. “IFS-s];
h

38

Table 3. Products obtained from photoreactions of I
with substrates.

 

 

 

Irradiation m.p. Yield
Substrates time (min.) Photoproducts (°) (%)a
1-pr0panol 30 Cp(0C3H7-n)TiC12 33-3A 3
2-propanol 35 Cp(OC3H7-i)TiC12 105-110b 30
acetone 50 Cp(OC3H7-i)TiC12 105-110b 22

 

 

a. Based on I- b. lit. 113-11A.5”°.

39

Table A. Mass spectral patterns of photoproducts of I
in the presence of substrates.

 

 

Relative Intensities

 

 

m/e Ions (M+) 1-propanol 2-propanol acetone
2H2 CpTiC12(OC3H7) 0.21 0.28 0.2“
227 CpTiC12(OCHCH3) ---- 1.A1 1.A0
213 CpTiCl2(OCH2) 0.72 ---- -—--
183 CpTiC12 0.76 1.15 1.16
1A8 CpTiCl 0.AA 0.66 0.60
65 C5H5 1.00 1.00 1.00

 

 

CONCLUSIONS

The photochemistry of

I is consistent with the

excitation of the charge-transfer (Cp ligand—)Ti metal),

which leads to a homolytic cleavage of Ti—Cp n-bond.

Scheme 1 is a summary of the fate of the photogenerated

Cp radical and CpTiC12 spec

ies in different chemical en-

 

 

vironments.
”“3001; Ph30
”NO : ArN(0)Cp
hv CHBr

Cp2TiCl2 I 0p + TpTiC12 _.‘

3-’ CpTiBrCl

_

2

—CP_2Ii—Cl2——» CpTiCl3

acetone

 

21¢ Cp(OCH(CH3)2TiC12

 

Scheme 1. Possible pathway

generated Cp rad
ROH* l-propano

s
ROH

‘
r

Cp(0R)T1012

s of the photolytically
ical and CpTiCl species.
1 and 2-propan01.

A0

EXPERIMENTAL

1. General

 

Manipulations of air-sensitive complexes were carried
out under an argon atmosphere, by using either Schlenk
or vacuum manifold techniques.

Ir spectra were taken by use of Perkin Elmer Model
A57 grating spectrophotometer in NuJol, Mass spectra by
use of a Hitachi-Perkin-Elmer Model RMU-6 mass spectrometer,
esr spectra by use of a Varian Model E-A esr spectrometer
with a temperature variable controller, uv-visible absorp-
tion spectra by means of a Unican Model SP-800 spectro-
photometer or by use of a Cary Model 17 spectrophotometer,

and l

H nmr by use of a Varian Model T-60 spectrometer.
Reagent or spectroscopic grade solvents were used
throughout. All solvents and substrates were dried or
refluxed over appropriate drying agents and distilled under
argon prior to use. The drying agents were the molecular
sieve AA for acetone, anhydrous CaSOu for alcohols, sodium
benzophenone ketyl for benzene and THF (tetrahydrofuran),
sodium metal for hexane, and calcium hydride for aceto-
nitrile. Bromoform was used as commercially available

reagent grade chemical without further purification.

Titanocene dichloride was purchased from Alfa Product and

A1

 

 

 

 

A2

purified by either recrystalizing it twice from toluene
or Soxhlet extraction from chloroform saturated with HCl
gas and argon. D20 was obtained from Mallinckrodt Chemi-
cal Works. Cyclopentadiene was obtained by cracking com-
mercial dicyclopentadiene (Matheson Coleman & Bell) over
BaO through a 30-cm Vigreux column. Similarly, methyl-
cyclopentadiene was obtained by cracking the methylcyclo-
pentadiene dimer (Aldrich Chemical Company Inc.). C5D6
was prepared from the method of Switzer and Rettigul,

beginning with 05H6, KOH, and D2O-dioxane cosolvent.

2. Preparation of (C52512T1C12

Perdeuterotitanocene dichloride was prepared from TiClu
(A.3 ml, 3.9 mmol) and sodium cyclopentadienide-d5 (9 g,
9.7 mmol) in THF in an argon atmosphere. The mixture was
refluxed for 2 hours. Then the THF was evaporated and the
residue was extracted, by using a Soxhlet apparatus with
chloroform under HCl atmosphere, followed by cooling and
filtration; yield 5 g (ca. 50% based on T101“); mass
spectra [e/m (relative intensitY)]: 258 (57), 223 (17),
118(100), 153(A0), 118(1A), 70(21); no peaks at e/m of 2A8
and 65, corresponding to Cp2T1012 and CSHS’ respectively,

were found.

A3

3. Preparation of (MeCp)2TiC12

Bi (methylcyclopentadienyl)titanium dichloride was

A2

prepared by the established method with some modifi-

cationsu3. Sodium methylcyclopentadienide was prepared
by a reaction of freshly distilled MeCpH (100 ml, about

1 mol) with sodium metal (23 g, 1 mol) in 300 ml THF

under an argon atmosphere. The resulting solution (from
colorless to deep violet, depending on the extent of
oxidation) was added dr0pwise under argon to a 200 ml

dry benzene solution of TiClu (A9.5 ml, about 0.A5 mmol)
which was precooled to 0°. Once the reaction mixture had
cooled, the solvent was removed into a dry-ice trap.

The residue was placed in a large cellulose thimble of a
Soxhlet extractor and the reaction product extracted with

a mixed solvent of benzene and chloroform (1:1). Cool-

ing of the mixed solvent solution led to a red solid.

This solid was collected by filtration, washed with pentane,
and dried in vacuo to give 23 g of moderately pure product
(19% yield based on TiClu). Further purification was
achieved by recrystalization from toluene to give brilliant
red crystalline leaflets; mp 219—22l° (decomp.) (lit.

mp 217-2l8°, decomp.); ir spectrum (nuJol, cm'l) 3100 (m),
1500(m), 1055(8), 10AO(s), 939(w), 859(VS), 829(m), 701(w);

1H nmr (00013) 62.3 (s,6), 6.25(s,8).

AA

A. Preparation of nitrosodurene

Nitrosodurene was synthesized by the established
procedureuu, beginning with durene, mercury acetate and
ethyl nitrite; mp 159° (decomp.) (lit. mp 160°, decomp.);
ir spectrum (nuJol, cm-l) y(N-0): 1510 (monomer), 1265

(dimer, trans).

5. Preparation of Ph3001 and Ph3C radical

Both triphenylchloromethane and triphenylmethyl radi-
cal were prepared by the known procedureus, beginning with
benzene and 0C1“. The measured g value of the Ph3C
radical was 2.0027, compared with the published value of

2.002uu5.

6. Esr experiment

The reaction solutions (8 x 10'3 M for Ti(IV) com-

plexes, 8 x 10-2

M for either Ph3CCl or ArNO) were care-
fully introduced into 2 mm o.d. Pyrex tubes by a syringe
under argon atmosphere and then degassed. They were then
mounted in the cavity of a Varian Model E-A spectrometer
equipped with a variable temperature controller. Ir-
radiation was carried out with a Hanovia 287, 1000—W com-

pact arc lamp in a Model LH 15 1N Schoeffel Lamp Housing.

The light was focused through a lens and filtered through

A5

a A cm water filter. DPPH (2,2-diphenyl-l-picryhydrazyl,
g = 2.003710.0002) was used for magnetic field calibra-
tion and measurement of g values. The estimated accuracy

of hyperfine coupling constants is 10.1 G.

7. Photolysis in thegpresence of bromoform, alcohols and

acetone

The preparative, photolysis apparatus is shown in Fig-
ure l9. Argon gas was bubbled through the solution in
the apparatus (set in a 20% alcohol-water bath) during
irradiation and samples were withdrawn by a syringe per-
iodically. The samples were transferred into capped
l—cm quartz cells in order to monitor the extent of
photolysis by promptly measuring the optical density at
520 nm. Figure 20 shows one example and shows the changes
in the visible spectrum of a reaction mixture during irradia-
tion. If the irradiated solution was exposed to air for
a short time (less than A min.), no effect on the visible
absorption measurement was observed. With prolonged ex-
posure, precipitation occurred. When the peak at 520 nm
disappeared, the photoreaction was stopped. Then the
irradiated solution was dried under vacuum at room tempera-
ture and the residue was sublimed under vacuum at A5°-50°.
The photoproduct was collected from the cold finger in a
dry box and identified by melting point measurement and

mass spectrometry. Thermal nonlability of these systems

A6

water outlet <:::J~V fiij; water inlet

3 cm
F

cooling Jacket

 

 

 

 

argon inlet
argon outletrll fléé

.- --—--~u.\ v

 

 

 

 

 

 

 

a
/l’
//
Hanovia r’T‘
A50-W lamp
9 cm
/0 1
reaction K\\::::~»H.._xu :j::L’U __
vessel

 

 

 

 

 

 

 

 

 

 

\A/
| 5cm '
I 9 cm I

l 12 cm I

Figure 19. The photolysis apparatus.

Optical Density

A7

2.0

r (a)

1.2

(b)

(e)

0'“ ’ (d)

 

000 l l j

 

A50 500 550 600
Wavelength, nm

Figure 20. The changes in the visible absorption spectrum
of a photolyzed benzene solution of I'in the
presence of l-propanol for (a) 0 (b) 10 (c)
15 and (d) 20 min. irradiations.

A8

was confirmed by the determination of the optical density
at 520 nm after the same treated solution was stirred in
darkness for 3 days. No significant decrease in the

optical density at 520 nm was found.

8. Photolysis in pure THF and benzene

A 100 ml THF or benzene solution of 8 x 10"3 M_of I
or (MeCp)2TiCl2 was irradiated for 2.5 hrs. at room
temperature with an unfiltered A50-W Hanovia medium-pres-
sure mercury lamp. After irradiation the solvent was
removed to a dry-ice trap and subsequently analyzed by uv
spectrometry. The residue was extracted with hexane and

filtered. The filtrate was analyzed by uv spectrometry.

BIBLIOGRAPHY

1.

10.

BIBLIOGRAPHY

R. C. Wailes, R. S. P. Coutts and H. Weigold, Organo-
metallic Chemistry of Titanium, Zirconium and Haf-
nium, Academic Press, New York (197A).

A. N. Nesmeyanov, O. V. Nogina, and V. A. Dubovitskii,
Izv. Akad. Nauk SSSR, Otd. Khim. Nauk, lA81 (1962),

C. A. 58: lA87f.

P. M. Maitlis, A. Efraty, and M. L. Games, J. Organo-
metal. Chem., 2, 2A8 (1962).

(a) S. P. Cubin and S. A. Smirnova, J. Organometal.
Chem., 20, 229 (1969).

(b) H. Hsiung and G. H. Brown, J. Electrochem. Soc.,
110, 1035 (1963).

(c) R. E. Dessey and R. B. King, J. Amer. Chem. Soc.,
88, 5112 (1966).

(a) L. J. Baye, Syn. Inorg. Metal-Org. Chem., 2,
A7 (1972), C.A. lg: 12103As.

(b) P. C. Bharara, J. Organometal. Chem., 21, 199
(1975). --

R. D. Gorsich, J. Amer. Chem. Soc., 80, 57AA (1958).
M. Tsutsui and C. E. Hudman, Chem. Lett., 777 (1972).

(a) L. Mond and C. Langer, J. Chem. Soc., 59, 1090
(1891)

(b) J. Dewar and H. 0. Jones, Proc. pr. Soc. Ser.
A, 19, 558 (1905)-

(a) R. M. Laine and P. C. Ford, Inorg. Chem. , I6,
388 (1977).

(b) M. S. Wrighton and D. S. Ginley, J. Amer. Chem.
Soc., 31, A2u6 (1975).

M. S. Wrighton and D. S. Ginley,ibid., 21, 2065 (1975).

A9

11.

12.

13.

1A.

15.

l6.

l7.

18.

190
20.

21.

22.

50

(a) A. Hudson, M. F. Lappert and B. K. Nicholson,
J. c. s., Dalton, 552 (1977).

(b) A. Hudson, M. F. Lappert and B. K. Nicholson,
J. C. 8., Chem. Comm., 966 (197A).

(a) M. S. Wrighton, Chem. Rev., IA, A01 (197A).

(b) M. S. Wrighton, Top. Curr. Chem., 65, 37 (1976).
(c) E. A. Koerner von Gustorf, L. H. G. Leenders, I.
Fishler and B. N. Perutz, Adv. Inorg. Chem. and
Radiochem., I9, 65 (1976).

(d) M. F. Lappert and P. W. Lednor, Adv. Organometal.
Chem., IA, 3A5 (1976)-

M. R. Hunt and G. Winter, Inorg. Nucl. Chem. Lett.,

J. Shiokawa and A. Matsumoto, Asahi Garasu Kogyo
GiJutsu Shoreikai Kenkyu Hokoku, 207 (197A), C. A.

ii: 977% (1976).

(a) H. Alt and M. D. Rausch, J. Amer. Chem. Soc.,
29. 5936 (197A).

(b) H. Alt and M. D. Rausch, Z. Naturforsch, 30b,
813 (1975).

M. D. Rausch, W. H. Boon and H. G. Alt, J. Organo-
metal. Chem., 1A1, 299 (1977).

E. Samuel, P. Maillard and C. Ciannotti, ibid., A2,
289 (1977).

R. W. Harrigan, G. 8. Hammond and H. B. Gray, ibid.,
£1. 79 (197A).

Ed Vitz and C. H. Brubaker, Jr., ibid., 82, 016 (197A).

M. H. Peng and C. H. Brubaker, Jr., ibid., 135, 333
(1977).

(a) Ed Vitz, P. J. Wagner and c. H. Brubaker, Jr.,
ibid., 10A, c33 (1976).

(b) 6Ed Vitz and C. H. Brubaker, Jr., ibid., 107, 301
(197 ).

G. Condorelli, I. Fragala and A. Centineo, ibid., 81,
311 (1975).

23.
2A.

25.

26.
27.

28.

29.

30.

31.

32.

51

E. G. Janzen, Accounts Chem. Res., A, 31 (1971).

S. Terabe, K. Kuruma and R. Konaka, J. C. S.
II, 1252 (1973).

Perkin

E. Hedaya, D. W. McNeil, Paul Schissel and D. J.

McAdoo, J. Amer. Chem. Soc., 98, 52A8 (1968)
A. Wasserman, J. C. S., 870 (1939).

J. I. Brauman, L. E.
Amer. Chem. Soc., 88, 8A6 (1966).

Ellis, E. E. van Tamelen, J.

(a) A. Faucitano and F. F. Martinotti, European

Polymer J., IQ, A89 (197A).

(b) R. E. Linder and A. C. Ling, Can. J. Chem., 88,

3982 (1972).

(c) S. I. Ohnishi and I. Nitta, J. Chem. Phys., 9

28A8 (1963).

(d) P. J. Zandstra, J. Chem. Phys., A8, 612 (196A).

(e) G. R. Liebling and H. M. McConnell, ibid., Ag,

3931 (1965).

(f) P. J. Krusic and J. K. Kochi, J. Amer.
22; 7155 (1968).

Chem. Soc.,

(g) R. W. Fessenden and S. Ogawa, ibid., 88, 3591

(196A).

(h) T. Kawamura and J. K. Kochi, J. Organometal. Chem.,

51, 79 (1973).

C. H. Langford and J. P. Aplington, ibid., A, 271

(1965).

(a) Ed Vitz, Ph.D. Thesis, Michigan State University,

(197“).

(b) J. c. w. Chien, J. Phys. Chem., 81, 3A77 (1963).

(c) J. L. Burmeister, E. A. Deardroff and C. E. van

Dyke, Inorg. Chem., 8, 180 (1969).

K. R. Gordon and K. D. Warren, Inorg. Chem.,

987 (1978).

J. W. Lauher and R. Hoffmann, J. Amer. Chem.
9_8_. 1729 (1976).

ll,

Soc.,

33.

3A.

35.

36.

37.

38.

39.

A0.

A1.

A2.

A3.

AA.

52

B. A. Goodman and J. B. Raynor, Adv. Inorg. Radio-
chem., I1, 135 (1970).

P. Schmid and K. U. Ingold, J. Amer. Chem. Soc.,
100, 2A93 (1978).

(a) E.G. Janzen, B. R. Knauer, L. T. Williams and
W. B. Harrison, J. Phys. Chem., 1A, 3025 (1970).

(b) E. G. Janzen and J. W. Happ, ibid., 13, 2335
(1969).

(c) M. D. Sevilla and G. Vincow, ibid., lg, 36A7
(1968).

(d) T. Nishikawa and K. Someno, Bull. Chem. Soc.,
Jap., A1, 2881 (197A).

0. Heller and H. M. McConnell, J. Chem. Phys., 3g,
1535 (1960).

P. H. Kasai and D. McLeod, J. Phys. Chem., 8g, 619
(1978).

S. Nigam, K. D. Asmus and R. L. Willson, J. Chem.
Soc. Faraday Trans. I, 72, 232A (1976).

(a) R. W. Fessenden and R. H. Schuler, J. Chem. Phys.

(b) G. Lehmann, Angew, Chem. Int. Ed. Engl., I1,
89 (1978).

A. N. Nesmeyanov, O. V. Nogina and V. A. Bubovitskii,
Izv. Akad. Nauk SSSR. Ser. Khim., 527 (1968).

M. E. Switzer and M. F. Retting, Inorg. Chem., I3,
1975 (197A).

L. T. Reynolds and G. Wilkinson, J. Inorg. Nucl.
Chem., A, 86 (1959).

(a) J. L. Petersen and L. F. Dahl, J. Amer. Chem.
Soc., 21, 6A22 (1975).

(b) D. G. Sekutowski and G. D. Stucky, Inorg. Chem.,
IA, 2192 (1975)-

(a) L. 1. Smith and F. L. Taylor, J. Amer. Chem. Soc.,

51; 2370 (1935).

AA.

A5.

A6.

53

(b) L. 1. Smith and F. L. Taylor, J. Amer. Chem. Soc.,
21, 2A60 (1935).

A. I. Vogel, Practical Organic Chemistry, John Wily
& Sons, Inc., New York, NY (1962), p. 815.

D. B. Chesnut and G. J. Sloan, J. Chem. Phys., _;,
637 (1960).