IIHHHH

—
_
—
—
—

1W!!!

'1
.1

_'
muo l

I
‘h

 

THE PREPARATION OF SQME
“ANSI?! ON MEYAL ALKQX IDES

“12515 for fine 09(3ch of M. S.
MECHEGEN S’MTE UNE‘JEESETY
Marie: Wichoias
1964

 

THE-“‘3

(3.1

 

 

 

unortz'm STATE UWERSITY
DEPAC‘U: 15“."...LH'T or: {.35-23aviiSTfiY
EAST W'iSiNG, MICHIGAN

ABSTRACT
THE PREPARATION OF SOME TRANSITION METAL ALKOXIDES
by Mark Nicholas

Four alkoxides and chloroalkoxides of copper (II) have been pre-
pared and characterized. The alkoxides, Cu(OCH3)2 and Cu(OCZH5)2 were
prepared by reaction of stoichiometric quantities of the appropriate
lithium alkoxides and copper (II) chloride in alcohol. The chloride
alkoxides were also prepared by reaction of stoichiometric quantities
of the appropriate lithium alkoxide and copper (II) alkoxide in alcohol
and by reaction of copper (II) chloride and the copper (II) alkoxides
in alcohol.

The alkoxides are insoluble in all common solvents whereas Cu(OCH3)Cl
is soluble in acetone and Cu(OC2H5)C1 in 3-pentanone--both with pos-
sible solvoLysis. The infrared spectra of these and other alkoxides
show that bridging alkoxide groups, terminal alkoxide groups orlxxh<xu1be
detected by the position of the C-0 absorption band. For the copper (II)
alkoxides both bridging and terminal groups are judged to be present,
whereas only bridging groups are present in the chloride alkoxides.

Cu(OCH3)2 and Cu(OC2H5)2 and Cu(OC2H5)Cl each exhibit two bands
in their visible spectra whereas Cu(OCH3)Cl shows only one band. The
number of bands are taken as evidence that there is a reduction of
symmetry from octahedral to tetragonal. The position of the bands is
consistent with the spectrochemical series in that chloride is a weaker
ligand than ligands bound through oxygen, since absorption in the alkox-

ides is at higher energy than in the chloride alkoxides.

Mark Nicholas
The magnetic susceptibilities of the alkoxides and chloride alkox—
ides are found to be much lower than 1.72 B.M., the spin—only moment.
The magnetic moments of the chloride alkoxides are found to be higher
than those of the alkoxides. The spin pairing can be caused by either
a superexchange mechanism or by direct overlap of d orbitals of two cop-

per (II) ions.

The structure suggested for both the alkoxides and chloride alk-
oxides is a three dimensional polymer consisting of copper atoms
bonded in an tetragonal environment.

Investigation of mixed metal alkoxides resulted in the synthesis
of iron (III) phenoxide. Attempts to produce double alkoxides of thal-
lium (I) and iron (III), cobalt (II) and titanium (IV), and iron (II)

and titanium (IV) were all unsuccessful.

THE PREPARATION OF SOME TRANSITION METAL ALKOXIDES

By

Mark Nicholas

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Chemistry

l96b

ACKNOWLEDGMENT

The author wishes to express his sincere gratitude to Professor
Carl H. Brubaker, Jr. for his guidance and assistance during the course

of this investigation.

ii

TABLE OF CONTENTS
Page

I. INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . 1

II. EXPERIMENTAL . . . . . . . . . . . . . . . . . . . . . 6
Materials . . . . . . . . . . . . . . . . . . . 6
Preparation of Metal Alkoxides . . . . . . . . 6
Preparation of Metal Halides . . . . . . . . . 7

Analytical Methods . . . . . . . . . . . . . . 7

Preparation of Compounds . . . . . . . . . . . 11

Mixed Metal Alkoxides . . . . . . . . . . . . . 22

Spectrosc0pic Measurements . . . . . . . . . . 28

Magnetic Susceptibility Measurements . . . . . 28

III. RESULTS AND DISCUSSION . . . . . . . . . . . . . . . . . 31
Copper (II) Alkoxides and Chloride Alkoxides . . . . 31

Mixed Metal Alkoxides . . . . . . . . . . . . . . . 39

REFERENC$ o o o o o o o o o c o o o o o o o o o o o o o ’41

iii

Table

II.
III.

IV.

LIST OF TABLES

Magnetic susceptibilities and moments . . . . . .

C—O absorption bands of methoxides . . . . . . . .

C‘O absorption bands or ethOXidQS o o o o o o o o

Visible absorption maxima of copper (II) alkoxides '

and chloride alkoxides

iv

Page

. . 3h

LIST OF FIGURES

Figure Page
1. Infrared spectrum of TlOC2H5 . . . . . . . . . . . . . . . 8
2. Infrared spectrum of TlOC5H5 (Nujol mull) . . . . . . . . 9
3. Infrared spectrum of Fe(OC2H5)3 (Nujol mull) . . . . . . . 10
h. Infrared spectrum of Fe(OC6H5)3 (Nujol mull) . . . . . . . 13
5. Infrared spectrum of Cu(OCH3)2 (Nujol mull) . . . . . . . l6
6

. Visible absorption spectra of Cu(OCH3)2 and
Cu(OCH3)C1 (Nujol mull) . . . . . . . . . . . . . . 17

7. Infrared spectrum of Cu(OCH3)Cl (Nujol mull) . . . . . . . 18
8. Infrared spectrum of Cu(OC2H5)2 (Nujol mull) . . . . . . . l9

9. Visible absorption spectra of Cu(OC2H5)2 (Nujol mull)
and Cu(OC2H5)Cl (3 -pentanone) . . . . . . . . 2O

10. Infrared spectrum of Cu(OC2H5)C1 (Nujol mull) . . . . . . 21
11. Infrared spectrum of UA14(OC3H71)15 . . . . . . . . . . . 23
12. Infrared spectrum of Na2[Co(OC3H71)2C12] (Nujol mull) . . 26

13. Visible absorption spectrum of Na2[Co(OC3H7i)2C12]
(isopropanol) . . . . . . . . . . . . . . . . 27

INTRODUCTION

Metal alkoxide compounds have been known for over a century; however,
very little has been ascertained about their properties and structures.

Much investigation has been done concerning the reaction of alkox-
ides, particularly those of Na, K, Mg, Ca,_A1, and Ti in organic systems.
Review articles of this and other alkoxide chemistry have been written
by Bradley (1,2,3).

Some transition metals have been reported to form alkoxides. Most
investigated have been found to be covalent, volatile, polymeric substances
where the degree of polymerization varies from two to eight. There have
been no complete x—ray structure determinations for any of the transi-
tion metal alkoxides. Only one structure determination has been reported
-- that of 110C113 (h), which is tetrameric and is of a distorted cubic

symmetry as shown. A structural theory has been advanced by Bradley (5)

'T1

$11

T -———————__O

 

Thallous Ethoxide O = oxygen in ethoxide

and is based on hydrolysis experiments with titanium alkoxides and from
molecular weight studies (ebulliometry and cryoscopY).

The transition metal alkoxides of groups AB and SB have been most
extensively studied because of their covalency, volatility, and solubility

l

2

in organic solvents. Tetraalkoxides of titanium were first prepared in
1875 (6), and since then the following physical constants have been de-
termined: degree of polymerization, volatility, heat and entropy of
vaporization, density, viscosity, parachor, dipole moments, and refrac-
tivities. Reference to these can be found in Bradley's review artic1e(3).
The alkoxides of the congeners zirconium (7) and hafnium (8) have also
been synthesized, but have been less thoroughLy studied. Similar research
has been done with the group SB pentaalkoxides of niobium (9) and tanta-
lum (10), which were investigated by Bradley. Vanadium pentaalkoxides
have not yet been synthesized probably because of steric factors.

Caughlan (11) has found that the group AB tetraalkoxides have a max—
imum degree of polymerization equal to three and proposes that the mole-

cules consist of three interpenetrating octahedra as shown below:

«TN

Ti(OC2H5)4 O = oxygen in ethoxide

For group SB, the maximum degree of polymerization is two for the penta-
alkoxides, and he proposes a structure of two alkoxide bridged octahedra(lO)

as shown below:

O

of 7I O
/

Ta(OC2H5)5 O = oxygen in ethoxide

 

 

In the latter group, alkoxides have been synthesized where the metal
is of a lower oxidation state. More specifically, Bradley has reported
the existence of vanadium tetraalkoxides (l2) and trialkoxides (13).

The degree of polymerization of V(OCH3)4 is 2.79 or almost three. The
trialkoxides are solids insoluble in organic sorwafis,and hence no
physical properties were reported. Brubaker and Wentworth (1h) have
prepared the extremely unstable Nb(OC2H5)4 which is diamagnetic and so
possibly dimeric.

Cr(OC2H5)3 (15) and Mn(OC2H5)2 (l6) and their alcoholates are also
known. Unique alkoxides of tetravalent chromium have only recently been
reported (17), and these d2 compounds are paramagnetic, “eff = 2.79 -
2.82 B.M.

,Ferric alkoxides were first made in 1930 (18) and are known to be
trimeric (l9) and volatile under reduced pressure.

Barraclough (20) has recorded the infrared spectra of many of the
transition metal alkoxides and hasmade assignments to the metal-oxygen and

carbon-oxygen stretching frequencies.

11

Some lanthanide metal alkoxides have been synthesized.Mehotra (21)
has prepared simple trialkoxides of La, Pr, and Nd, which are all mono-
meric. Bradley (22) has investigated the covalent alkoxides of Ce(IV).
The primary alkoxides are polymeric solids which are volatile at high
temperatures and reduced pressures, whereas the tertiary alkoxides are
liquid monomers.

Both Bradley and H. Gilman have investigated actinide metal alkoxides.
Bradley has synthesized the covalent polymeric thorium tetraalkoxides,
of which the maximum degree of polymerization is six, and plutonium tetra-
alkoxides (2h). Alkoxides for the following uranium Species are known:
U(IV) (2S), U(V) (26), and U(VI) (both U(OR)6 and UOZ(OR)2 (27,28). As
most metal alkoxides these show considerable volatility and covalency.
Many novel synthetic routes were used for the preparation of the U(V)
and U(VI) alkoxides.

Of interest also are the double or mixed metal alkoxides, which in-
volve transition, actinide, or lanthanide elements. The compounds can
be represented as MM'(OR)X where M, M', or both are either transition,
lanthanide, or actinide elements. These compounds should contain bridg-
ing alkoxide groups between the different metals, which might then allow
electron super-exchange across the bridge.

Meerwein first synthesized double alkoxides in 1929 (29), but of
the more than SO reported by him, only six contained transition elements,
and half of these compounds have since been disproven (30). The remain—
ing three were formed by the reaction of NaAl(OC2H5)4 with CuClz, NiClZ,
and CoClz respectively. Bartley and Wardlaw (30) have prepared a series.
of double alkoxides of zirconium, MZr2(OR)9, where M is Li, Na, or K.

Most were found to be monomeric.

5
Albers (31) reported the preparation of the uranium aluminum iso-
propoxide UA14(OC3H;)15. Further work by Gilman (32), showed that
U(OC2H5)5 would readily form double alkoxides as NaU(OC2H5)6, but that

U(OC2H5)4 would not.

EXPERIMENTAL

Materials

All chemicals used in this investigation were of reagent grade un-
less otherwise stated. All solvents were dehydrated before using. Meth-
anol and ethanol were refluxed with magnesium shavings and were then
distilled, whereas, isopropanol was refluxed with calcium turnings and
swas then distilled. Benzene was refluxed with phosphorous pentoxide

and distilled.

Preparation of Metal Alkoxides

 

Aluminum Isopropoxide was prepared from aluminum foil and dry iso-

 

propanol according to the method of Wilds (33).

Thallium (I) Ethoxide was prepared according to West (A) from

 

thallium metal (K & K Laboratories - 99.8%), absolute ethanol, and oxygen.
The infrared spectrum is presented in Figure l and agrees with that

published by West.

Thallium (I) Phenoxide was prepared by the procedure of Menzies and

 

Christie (3h),.from thallium. Cf) ethoxide and phenol. The infrared

,spectnnnis presented in Figure 2.

Iron (III) Ethoxide was prepared by a method similar to that of

 

Thiessen (18). Anhydrous ferric chloride was added to an ethanol solu-
tion of sodium ethoxide. After refluxing for an hour, the solvent was
removed under reduced pressure, and the residue was extracted with ben-

zene. Iron (III) ethoxide was isolated from this solution. The

6

's

7

infrared spectrum is presented in Figure 3.

Titanium Isgpropoxide was kindly donated by P. Dupree and was Ander-

 

son, practical grade.

Preparation of Metal Halides

 

Uranium (IV) Chloride. UO4'xHZO was first prepared from uranyl

 

nitrate hexahydrate and 3% hydrogen peroxide (35). The dried product
was then heated in an oven at 2500C for four hours to yield uranium

trioxide. Following Hermann's procedure (36), uranium (IV) chloride
was made from the trioxide and hexachloropropylene (Aldrich Chemical

Co.).

Cobalt (II) Chloride. The anhydrous salt was prepared by dehydrat-

 

ing the hexahydrate by refluxing with thionyl chloride as described by

Prey (37) -

Copper (II) Chloride was prepared similarly to CoClz.

 

Iron (II) Chloride was synthesized from anhydrous ferric chloride

 

and chlorobenzene (38).

Analytical Methods

 

Iron Analysis. Fe(III) in Fe(OC6H5)5 was determined by ignition

 

of the sample to Fezos. In the Fe(III), Tl(I) mixtures, the samples
were first hydrolyzed in dilute HCl, and the TlCl was filtered from the
Fe(III) filtrate. After addition of potassium iodide, the liberated

iodine was titrated with standard thiosulfate solution. Fe(II) was

 

 

$100.2. Ho 55.302? vuuthH .H 833m
25qu 5 59325:
H: ma 3 2.. 0H m w m a
4— - _ _ H a . a a a

 

 

 

>

 

 

 

 

 

 

 

 

 

‘ A

__4_‘

 

 

 

 

 

 

nave: Mo 6530on panama; .H 0.26:
«conga cm 593 28:

 

 

 

 

a m a t e m a m
fl d u .—

>. . _

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

4H

mH

NH

HH

OH
a

See 3.35

mcouoma cw npmcmflm>m3

m m

_ fi

 

 

.Anxwoovfie mo sapwooam commumcH

 

 

 

.N SSE

A

 

 

 

 

 

 

lO

ma

A338 Hons—6 .nfinmwdouom mo 35.8on UupmpmcH .m 6.33m
2838 cm 593ng

 

 

NH 2 2 a , m a. e m 4
~ _ _ q u _ d . .

 

 

 

 

 

 

 

 

 

 

ll
determined by titration with standard Ce(SO4)2 after the samples were

hydrolyzed in dilute H2804.

Copper Analysis. Cu(II) samples were hydrolyzed in dilute HCl, and

 

after the addition of potassium iodide, the liberated iodine was titrated

with standard thiosulfate solution (39).

Uranium and Aluminum Analysis. Uranium was separated from aluminum

 

by formation of the peroxide (AC), and was then analyzed by ignition to
U303. Aluminum was determined by precipitation of the hydrous oxide

and subsequent ignition to the oxide.

Thallium Analysis. Thallium (I) was separated from iron in all

 

samples by precipitation of the thallium (I) thiourea complex (bl).
The complex was then dissolved in hot water and thallous chromate was
precipitated by addition of potassium chromate. This was dried at 110°

and weighed as TlZCrO4 (A2).

Phenoxide Analysis. After hydrolysis of Fe(OC5H5)3 in basic solu-

 

tion, phenoxide was determined by the bromination of phenol, reaction
of the excess bromine with KI, and titration of the liberated iodine

with thiosulfate (113) .

Carbon and Hydrogen Analysis. These were performed by Spang Micro-

 

analytical Laboratory, Ann Arbor, Michigan.

Preparation of Compounds

 

In all procedures involving the use or synthesis of moisture sensi-

tive materials, the utmost care was used to avoid contact with moisture.

12
Flasks were always flushed with nitrogen before use, and all critical
operations such as filtration, transfer of solutions, etc., were always

done in a nitrogen atmosphere dry-box.

Iron (III) Phenoxide. To 200 ml. of ethanol, 2.91 grams (0.127

 

moles) of sodium was added. After the sodium metal had completely
reacted, 11.29 grams (0.127 moles) of phenol was added, and the flask
was heated gently until the phenol dissolved. The following equilibrium

then exists in solution:

c2H5o‘ + C6H5OH

 

> C2H50H + c6H5o' (1)

6.85 grams (0.0h22 moles) of ferric chloride was added and an exothermic
reaction ensued. The solution became blood red and a cream colored pre—
cipitate formed. After refluxing for two hours, the ethanol was dis-
tilled under reduced pressure thus forcing the equilibrium (1) to the
right. The red solid residue was extracted with three 50 m1. portions
of benzene. Partial evaporation of the benzene solution plus addition
of 100 m1. of petroleum ether caused the precipitation of a black solid
which was filtered and washed with petroleum ether. The product (8.52
grams) was obtained in a 60.2% yield although more could have been iso—
lated by continued extraction. Upon grinding the black solid became
dark red. The infrared spectrum is presented in Figure A. However, no
visible spectrum could be obtained because a charge transfer band com-
pletely obscured any details of this region.

Analysis: Calc'd for Fe(0C5H5)3: Fe, 16.66; 0C6H5, 83.3h.

Found: Fe, 16.373 OC6H5, 83.60.

A32: flownzv .nAnmmuov om Mo 550.0QO woquMCH .4 6.33““
98.82: cm 53:39pm:

2 m w a e m a m

T F b

 

13

 

 

 

q _ E m H

‘-
‘

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1h

CopperjII) Methoxide. Lithium methoxide was prepared in situ by

 

adding 0.5h2 grams (0.078 moles) of lithium to 100 m1. of methanol.
Fifty milliliters of a methanolic 0.78 M. CuClZ solution was slowly
added to this in an erlenmeyer flask, and a light silky green precip—
itate formed. Agitation of this mixture by rapid shaking, changed the
color of the precipitate from green to deep blue. The coarse blue
solid was then filtered on a sintered glass frit, washed with methanol
and then dried in yagug, 3.00 grams was recovered in an 89.5 % yield.
The infrared and visible spectra are presented in Figures 5 and 6
respectively.

Analysis: Calc'd for Cu(OCH3)2: Cu, 50.59; C, 19.12; H, b.82.

Found: Cu, 50.57; C, 18.99; H, h.67.

Copper (II) Chloridemethoxide. Investigation of the previous re-

 

action revealed that the green precipitate was a different species,
Cu(OCH3)Cl. The green solid can be prepared in three different ways.

From the previously mentioned reaction, when the supernatant liquid
was carefully decanted from the green precipitate without any agitation,
the green solid was isolated.

Addition of CuClZ to a suspension of Cu(OCH3)2 in methanol, caused
the color of the solid to turn from blue to green. The reaction was
complete when excess CuClZ colored the previously clear methanol.

In contrast to the previous two methods, when a solution of
LiOCHs in methanol was added to a solution of CuClz in methanol, green

Cu(OCH3)C1 formed as long as CuClZ was present in excess.

15
In all three methods the solid was filtered on a sintered glass
frit, washed with methanol, and dried ip‘yagpp. The infrared and visible
Spectra are presented in Figures 7 and 6 respectively.
Analysis: Calc'd for Cu(OCH3)Cl: Cu, h8.88; C, 9.2h; H, 2.32.

Found: Cu, h9.l6; C, 9.31; H, 2.30.

Cppper (II) Ethoxide. This was made exactly as the corresponding

 

methoxide using lithium ethoxide and ethanol. The reaction is quantita—
tive and the product which precipitates is blue-green. The carbon and
hydrogen analysis of this compound,however, is reported as being quite
low although the metal analysis is very satisfactory. It is suspected
that hydrolysis occurred in the process of analysis. In the synthesis
and handling of this compound, very strict measures were taken to ex-
clude moisture. The infrared spectrum of this compound shows no bands
due to the hydroxyl group or water as is usually found when hydrolysis
has occurred. The infrared and visible absorption spectra are presented
in Figures 8 and 9 respectively.

Analysis: Calc'd for Cu(0C2H5)2: Cu, u1.3h; c, 31.26; H, 6.59.

Found: Cu, h1.0h; C, 27.55; H, 6.01.

Copper (II) Chloride Ethoxide. This was synthesized as the analagous

 

copper (II) chloride methoxide, using lithium ethoxide and ethanol. The
green product was quantitatively precipitated and recovered. The infra-
red and visible absorption spectra are presented in Figures 10 and 9
respectively.

Analysis: Calc'd for Cu(OC2H5)Cl: Cu, hb.09; C, 16.67; H, 3.50.

Found: Cu, h3.38; C, 16.58; H, 3.51.

16

 

 

 

 

 

 

A258 Howszv .NAnmUOvao .Ho 8593QO vammmcun .m 6.3m;
95935 5 cpmcoflgm:
fl 2 NH 2 2 a w a. e m a
4 J a m _ w a a A ~ .

 

 

 

 

l6

 

 

AHHDE Homszv .mfinmoovso mo espuomam UopmumcH .m updmwm
mcopowe ca cumcmao>m3
s: 2 S 2 2 a m a. e m a
. _, a a ._ a. 1 a w _ .

 

 

 

 

 

 

 

 

17

00m

Aaase Hemszv. .Hoflnmuovso use NAnzuovso mo mtpueam compatOmam madamw> .e tundra

mcouowemafims cw npucoao>m3

00a ome ooe 0mm com om:

 

 

 

_ . a . _ _ . \_

 

HoAnzuovso

NAn260vso ..... /

l
.3

Na.

m
Koueqaosqv aAiquaH

l
\o

L
[T

 

 

18

:H

mH

NH

OH

aflaas Howazv .HUAnmUOVsD mo Evapommm woumpwcH

mcopowe cw camcoflm>m3

m m N. o m

+1 1 m

«h
-p

.5 mudmmm

 

 

q q

 

 

 

 

 

 

 

 

 

 

 

1?

 

 

.AHHSE acmvzv .NAnmNUovso mo Edmuouam wouwnMCH .m madman
mcouowe cm camcoao>m3
fl 2 NH 2 2 a m a e m a m
. H1 . _ n _ _ _ _ a .

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

20

00w

0mm

0mm

000

.aofimmwoovso dew Nahmwoovso no threads coaaaaomam madama>
mcouomswfifims cm numcmflm>m3

oom

.m oudmmm

 

 

Aococmpcoaumv

Aaasa HohszV

aoxnzwoovso

Nflhxmoowso

*1

 

_

/
/
I l I
U\ 4: «\
Aoueqhosqv afiiqetag '

f’
“1

l
N

 

 

AHHDE HOmSZV .HoAnmwoovso mo evapomam woummHCH .OH mudmwm

mcouowz CH npmcon>m3

21

JH

MH

NH

HH

OH

P

H

m
_

m

d

A,

‘

 

 

 

 

 

 

 

 

 

 

 

 

 

22

Mixed Metal Alkoxides

 

Uranium (IV) tetrakis (tetraisoprppoxoaluminate): This compound

 

was made according to the method of Albers (31). Sodium isopropoxide
was added to 100 ml. of a 1.6 M aluminum isopropoxide solution until all
the NaAl(OC3H7i)4 had precipitated. To this was added, 169 ml of a
0.180 M solution of UCl4 in isopropanol. From this dirty green mixture,
the viscous green double alkoxide could be isolated by vacuum distilla-
tion. The liquid is subject to rapid hydrolysis and oxidation and
showed noticeable decomposition within a week, even when stored under
nitrogen in a dry-box.

The infrared spectrum is presented in Figure 11. The visible
spectrum showed only a broad charge transfer band. No E.P.R. signal
could be detected for the sample when measured pure or as a solution in
benzene.

An lysis: Calc'd for UA14(oc3H7i)16: U, 18.h3; A1, 8.36.

Found: U, 18.82; Al, 8.89.

Attempted Preparation of a Thallium (I), Iron (III) Double Alkoxide.
Two procedures were tried without success for isolating a double alkoxide.

The reactants were added according to the following stoichiometry

To a solution of 0.0033 moles of Fe(0C2H5)3 in CC14, 0.01 moles of
T10CZH5 were added; 110 m1 of absolute ethanol was added, and a very
small amount of light orange precipitate formed. This was filtered and
analysis showed less than 5% iron. The infrared spectrum showed no

alkoxide absorption bands, and the substance did not burn in a flame.

,23

 

4H

mH

NH

HH

.wHAHhmnoovaab mo euppoomm punmpHCH .HH oudmfim

mcopome CH npmcmHo>m3

OH m m m b m a m

 

 

 

. a _ _ H . . \

 

 

 

 

 

 

 

 

 

 

2h

In the second method, benzene was used as the solvent. To 25 m1.
of a 0.0133 M Fe(OCZH5)3 benzene solution, 0.01 mole of T10C2H5 was
then added. No precipitate was noticed. Removal of the solvent under
reduced pressure yielded a tan solid. This was washed with ethanol and
benzene until the filtrate was colorless. The resultant tan solid
showed no alkoxide absorption in the infrared and also did not burn in
a flame. It is assumed that decomposition took place, and the solid
residue is mainly thallous ferrate (III).

Analysis: Calc'd for TlFeOZ: Tl, 69.933 Fe, 19.11.

Found: Tl, 67.06; Fe, 19.02.

Attempted Preparation of a Thallium (I),Iron.(III) Double Phenoxide.
Forty milliliters of a saturated benzene solution of thallium (I) phen-
oxide was added to NO m1 of a saturated benzene solution of iron (III)
phenoxide. The solution was refluxed for five hours but no precipitate
formed. Upon slow evaporation of the solvent, the white thallium (I)
phenoxide began to crystallize from the solution. Hence it was concluded

that no reaction occured.

Attempted Preparation of a Cobalt (II), Titanium (IV) Double Alkoxide.

 

A different type of reaction was attempted as compared to that used for
iron and thallium. Since cobalt (II) alkoxides are unknown, it was
thought that cobalt (II) chloride might react with an equimolar amount
of sodium isopropoxide and titanium isopropoxide (NaTi(OC3H7i)5) in
isopropanol. Accordingly 0.0532 moles of both Na0C3H7i and Ti(OC3H7i)4
were added to 200 ml of iSOpropanol. To this 0.0262 moles of CoClZ

was added, and the resultant blue solution was refluxed for eight hours.

25
After the removal of the solvent under reduced pressure, the residue
was extracted with benzene and filtered. A blue solid was then pre-
cipitated by subsequent addition of petroleum ether. This solid con-
tained chlorine but not titanium. Bartley (30) by the similar reaction
of CoClZ with NaZr2(OC3H7i)9 isolated a similar blue solid which contained
no zirconium and was found to be Na2[Co(0C3H7i)2C12]. The infrared and

visible spectra are presented in Figures 12 and 13.

Attempted Preparation of an Iron (II), Titanium (IV) Double Alkoxide.
The reaction was done similarly to that with CoClZ, using the same
stoichiometry. However, upon addition of the iron (II) chloride, a
large quantity of brown precipitate formed. This was of a very muddy
texture and could be filtered only with the greatest of difficulty. It
was filtered and dried under vacuum to yield a light brown powder. This
solid did not burn, did not absorb in the infrared, was contaminated
with sodium chloride, and was not soluble in common organic solvents.

Analagously when FeClz was added to solutions of sodium iSOprop-
oxide in iSOprOpanol, sodium ethoxide or lithium ethoxide in ethanol,
and sodium methoxide or lithium methoxide in methanol, the same muddy
type of precipitate resulted. The precipitates from the reactions in-
volving the lithium alkoxides were, however, uncontaminated with lithium
chloride, and hence the precipitate should have been a pure alkoxide.
Analysis for Fe(II) showed, however, less than 10% in both cases. It
seems then most probable that oxidation had occurred and that regardless
of whether NaTi(0C3H7i)5, NaOCZHs, or LiOC2H5 are used, no definite

product can be isolated.

26

 

4H

mH

AHHSE HOmszv .meUNAHAInUOVOUHNmzmo mupooam UmpmnHCH .NH mudmmm

mcouoHs CH cpmcmHo>m3
NH AH OH a w a e m a m

P

_ H _ _ . a _ _ _

 

 

 

 

 

 

 

 

 

 

27

00w

 

0mm
A

AHocmaouaomHg, .H~H0~Aahmnuovoogwmz Ho sapwooam cowpauomnm mHnHmH>. .mH musmfim

com

1!

mcouoHaHHHme cm cumcon>m3

ea

 

co:

m (\I

L:
KOUBQJOSQV.3AIQEIZH

in

 

 

27

00w

 

0mm
4

AHocmaouaomHv .HNHUNAHsmnoovooammz mo suppooam coHQQMOmnm anmmH> .MH unamwm

oom

mcouuHaHHHHE CH camcoHo>w3

as as
_

 

l
N

.l
‘2

ll
1::
Abusqzosqv aAiquaH

L
in

 

 

28

Spectroscopic Measurements

 

All infrared spectra were obtained with the Perkin—Elmer Model 21
Spectrophotometer. All solid samples were mulled in Nujol.

The visible spectra have been obtained with a Beckman Model DB
Spectrophotometer. Most of the solids were mulled in Nujol since no
solvent was available. All mulls and solutions necessary for the visible
and infrared spectra of moisture sensitive compounds were prepared in a

dry-box.
Magnetic Susceptibility Measurements

Magnetic susceptibilities were determined by the Gouy method, employ-
ing a semi-micro balance and the electromagnet described by Vander
Vennan (AA). The susceptibility was calculated by use of the following

equation

F' + a
X 06 =
l W

where X is the gram susceptibility; w is the weight in grams of the
sample; B is the tube constant; F' is the corrected force in mg. and is
equal to the force, F, experienced by the sample plus the force,<§ , ex-
perienced by the glass tube; a is a constant allowing for the suscepti-
bility of the displaced air.

The mole susceptibility, Xm’ can be calculated by multiplying the
gram susceptibility, X, of the sample by the molecular weight. Xm', the

corrected mole susceptibility, is obtained from Xm by adding to it the

appropriate Pascal constants—-the corrections for the diamagnetic ligands

(115) .

29
The magnetic moment, u, is then calculated from the following equa-
tion:

1 = 2.811xmv(1+e>11/2

The Curie—Weiss temperature, 0, was not calculated for any of the com-
pounds and was assumed to be small.

The tube which has a standard tapered glass joint to prevent hydrol-
ysis of samples was calibrated with both Hg[Co(SCN)4] and water (A6).
Hg[Co(SCN)4] was used for the calibration at low field strength and was
suitable for determining the susceptibility of samples having two or more
unpaired electrons. The calibrant has a gram susceptibility of l6.AA
x 10-6 at 20°. To calculate the gram susceptibility of the calibrant at

any given temperature, the following equation is used

_ 1 2.8A 2
X - NIZJTTLfi-C]

M is the molecular weight, A8l.9, C, 177 x 10—6, is the diamagnetic cor-
rection for Hg[Co(SCN)4], 6 equals 10°, and u equals A.AA Bohr Magnetons

(A7). Substitution yields the following:

—3
.O 2 O '-
X = 5 3 +x18 _ - 0.37 x 10 6

 

The tube was calibrated at 23° and 28°, where the gram susceptibility
was respectively 16.27 x 10'6 and 16.00 x 10‘6.

For water which is useful as a calibrant when the samples are dia-
magnetic or have one unpaired electron, the gram susceptibility is —0.720
x 10'6 at 200 and dX/dT = 0.0012/degree at 200 (A6).

The results for the compounds measured are presented in Table I.

30

Table I. Magnetic susceptibilities and moments

 

 

 

Compound Temp. 0K Xm' x 106 u (B.M.)
Cu(OCH3)Z 297 528 1.12
296 A80 1.07
Cu(OCH3)Cl 297 1,160 1.67
298 1,180 1.68
Cu(0C2H5)2 296 579 1.18
296 636 1.23
Cu(OC2H5)Cl 299 828 1.A1
296 810 1.39
Fe(0C6H5)3 301 3A,A00 9.1h
301 36,000 9.28
Fe(OCzH5)3 301 9,890 A.91
301 9,720 b.85
Na2[Co(0C3H7i)2C12] 298 8,860 A.60

UA14(OC3H71)15 300 1,230 3.21

 

RESULTS AND DISCUSSION

Copper (II) Alkoxides and Chloride Alkoxides

 

The alkoxides synthesized, Cu(OCH3)Z and Cu(0C2H5)2 are seemingly
quite different from the typical transition metal alkoxides in that
they are not volatile or soluble iri inert solvents. However, they do
resemble a second type of alkoxide—-that typified by V(0CH3)3, U(OCH3)4,
or La(0CH3)3. The compounds are generally insoluble and have the metal
ions in a low oxidation state and with small ionic radii. These alk-
oxides are probably covalent polymers as contrasted to the usual small
telemeric metal alkoxides.

The chloride alkoxides, too, have different pr0perties from the
known chloride alkoxides, e.g., titanium (IV) and zirconium (IV) (A8).
The chloride alkoxides of the aforementioned elements are considered to
be covalent and similar to the parent alkoxides, whereas the chloride
alkoxides of copper (II) resemble the copper (II) alkoxides in chemical
properties. 7

Before postulating possible structures for these new compound, it
is first necessary to consider the spectra, both visible and infrared,
magnetic properties, and chemical properties of the compounds.

The infrared spectra of methoxides or ethoxides are relatively
simple in the region where the C-0 stretching vibration band appears.
For the methoxides a single, strong wide band occurs between 1020-1100
cm_1, and for the ethoxide this band appears at 1025-1070 cm—l. Barra-
clough (20) in a study of the infrared spectrum of metal alkoxides,

postulated that two C-O stretching vibration bands are possible because

31

32
of the difference between terminal and bridging alkoxide groups in co-

valent metal alkoxides.

CR 3
T
M<:\/M 11—00113

Y
CR3

bridging terminal

Such differences between bridging and terminal groups have been
noted in compounds having CO, CN-, SCN-, and NO as ligands. Two or
more absorption bands are attributable to the stretching vibrations of
these groups (A9). However, alkoxide ligands do not Show two such dis-
tinct bands in the region of the C—0 stretching vibration. Because the
C—0 band is very broad and the region of absorption is at best double
the half-width of the C—0 band, one would expect superimposition of
terminal and bridging group absorptions, or possibly the presence of a
shoulder when both types of alkoxide groups are present simultaneously
in a molecule. As expected, then, only one band is found for the
ethoxide or methoxide ligand not only in the compounds synthesized in
this research, but also in the Spectra reported in the literature
(A,1A,20,50,51). It is thus the position of the single band (and in
few cases an accompanying shoulder) which could be indicative of the
type of alkoxide linkage present.

Inspection of the position of the C—0 stretching band reveals that
there are three rather distinct regions of absorption. The alkoxides,
listed in Tables II and III with the frequencies of the bands, are class-

ified according to the region in which the C—0 band appears. Compounds

Table II. C-O absorption bands of methoxides.

33

 

 

Compound

Region of Absorption (cm-1)

 

I 11 III
[(CH3)4N]2[Nb(OCH3)C15] (50) 1130 --- __-
(CsHeN)2[Nb(OCH3)C15] (50) 1095 --- ___
(CQHBN)2[Nb(OCH3)C15] (SO) 1110 ___ -__
CU(OCH3)2 --- 1056 -__
Cu(OCH3)C1 --- --— 101A
TIOCH3 (A) _-- --- 1030

 

3A

Table III. C-O absorption bands of ethoxides.

 

 

Compound Region of Absorption (cm-1)
I II III

 

[(CH3)2NH2]2[Nb(OC2H5)C15] (50) 1070 -—- ---

(C5H6N)2[Nb(OC2H5)C15] (50) 1070 ___ ___
(CQHBN)2[Nb(OC2H5)C15] (50) 1070 --- ___
Ti<oc2H5). (20) 106A --- 10A2(Sh)*
Ta(OCZH5)5 (20) 1072 --- 1030(sh)
Nb(OC2H5)5 (20) 1063 --- 1029(sh)
Nb(OC2H5)4 (1A) --- 1060 _--
111(002115)3 (51) ___ 1059 ___
Fe(OC2H5)3 —-- 1060 _-_
Cu(OC2H5)2 --- lohg ___
Cu(OC2H5)C1 --— ___ 1027
TlOCsz --- --- 1030

 

*(sh) = shoulder.

35
absorbing in region (I) are believed to have terminal alkoxide groups
only. Absorptions in region (II) are believed to be due to both term-
inal and bridging alkoxide groups, while absorption in region (III) is
indicative of bridging alkoxides only. Thus by this classification,
the copper (II) alkoxides would contain both terminal and bridging
groups whereas the copper (II) chloride alkoxides would contain bridging
groups only.

This distinction is very marked in the methoxides but less so in
the ethoxides because of the smaller variation in the C—0 stretching
frequency. It is probable that in the higher analog alkoxides (prop-
oxides, butoxides, etc.) the systematic variation in C—O stretching
frequencies will not be noticeable because the strength of the inter-
molecular bridging bonds will be greatly reduced, hence causing negligible
shifts in the C-0 stretching frequency.

The visible absorption spectra of solutions and Nujol mulls of the
copper (II) alkoxide and chloride alkoxide compounds have been measured,

and the absorption maxima are listed in Table IV.

Table IV. Visible absorption maxima of copper (II) alkoxides and
chloride alkoxides

 

 

 

Compound Frequencies ef Absorption Medium
(cm 1)

Cu(0C2H5)é 19,AOO 15,500 Nujol

Cu(OC2H5)C1 20,800 lA,7OO 3-pentanone

Cu(OCH3)2 19,100 15,700 Nujol

Cu(OCH3)C1 --- 1A,500 Nujol

 

36

Of these spectra only that of Cu(OCZH5)C1 is doubtful. Since the
solvent is 3-pentanone, it is possible that some solvolysis has occurred
and that the band at 20,800 cm-1 is due to coordinated solvent. (The
Spectrum of Cu(OCZH5)Cl in Nujol contains no absorption bands.) This band
is also present in the spectrum of CuClZ in 3-pentanone. In acetone,
for example, the spectrum of Cu(OCH3)Cl corresponds exactly to that of
CuClz, hence a Nujol mull was used with this compound.

Copper (II) in an octahedral environment, should exhibit one symmetric
band in the visible region corresponding to the 2Eg-——-:> Zng transition.
Most spectra of Cu(II) complexes (for example aquo or ammine complexes),
however, Show one asymmetric band which can be resolved into two compo-
nent bands (52). This asymmetry results from a tetragonal distortion
as caused by the Jahn-Teller effect. In the limiting case of tetragonal
distortion (a square planar configuration), three separate absorption
bands are expected, and have been observed in copper (II) ethylacetyl-
acetonate, for example (53).

The splitting of the d orbitals can be represented with respect to

the amount of tetragonal distortion.

”””””,,..——*‘-—‘—“ bl (dxz_y2)
d

 

 

(1::"3’2 :
b2 (dxy)
gxy -———- --—- a1 (d 2)
1 8(dyz’dxy)
Oh I D4h

> Increased Tetragonal Distortion -————-—e>

37

The following three transitions would then be expected: e -—H> b1,

 

al > b1, and b2 -—-—> b1 (A). In all four compounds, a band is

seen at lA,500—15,7OO cm_l. This is either the b2——+>b1 transition or
both the b2———>bl and a1———>b1 transitions. Because of this uncertainty,
A cannot be measured. It should be noted that the band appears at lower
energy in the chloride alkoxides than in the alkoxides, in agreement
with the fact that chloride is lower in the spectrochemical series than
oxygen bonded ligands.

The magnetic data (Table I) for these compounds yields further in-
formation as to their structure. C0pper (II), having a (19 configuration,
should have a spin-only magnetic moment of 1.72 B.M. However, because
of spin-orbit coupling, the experimental moments are usually higher
(1.70-2.20 B.M.). Thus in the alkoxides and chloride alkoxides, extensive
spin-pairing is indicated by their low magnetic moments.

The spin pairing can be caused by either a super exchange mechanism
or by direct overlap of d orbitals of two copper (II) ions. The latter
effect is found in copper (II) acetate monohydrate which has a magnetic
moment of l.A B.M. at room temperature (5A). Without x-ray data or
knowledge of the variation of the molar susceptibility with temperature,
the cause of the low moments cannot be ascertained. It should be noted,-
however, that the chloride alkoxides have higher magnetic moments than
the alkoxides. Oxygen bridges facilitate electron-transfer more than
chloride bridges.

Three different structures for the alkoxides and chloride alkoxides
can be proposed, and these will be discussed with respect to the spectra

and magnetic properties.

38
One structure is that of a one-dimensional polymer consisting of

staggered hexamer units.

 

// /’ 0’__H_____ d/

..___.__ o _._____.

0 represents the ligand.

The copper atoms are in a tetragonal environment, but, there is the
inherent difficulty that certain of the alkoxide ligands will formally
contain seven coordinated oxygen atoms. On that basis, the above struc.
ture does not seem likely. Two other structures seem to be more reason-

able, of which the Ofirst is basically the CuClZ structure.

/\ /
./ <> \.

\
\

/

0 represents the ligand

Each copper atom has an immediate square planar environment, but
also interacts electrostatically with the ligands above and below the
plane. In this way there is tetragonal distortion, but in this structure

all the ligands are bridging.

39
Finally, and perhaps the most plausible structure, is that of a

three dimensional polymer.

Oé O‘—'“'
|

[

——- ~—

 

If
0
, q
I
0 represents the ligand

Each copper atom has a tetragonal environment, and there are non-
equivalent ligands. The ligands in the xy-plane are all bridging, whereas
the lIgands on the z—axis (the axis of distortion) are essentially term-
inal because of their extended distances from the copper. Hence the cop-
per alkoxides will have both bridging and "terminal" alkoxide groups if

the chlorine atoms are on the z-axis. This interpretation would readily

explain the infrared data.

Mixed Metal Alkoxides

 

Due to the paucity of information about mixed metal alkoxides it is
very difficult to offer reasons for the failure to form double alkoxides
between thallium (I) and iron (III), cobalt (II) and titanium (IV), and
iron (II) and titanium (IV). Because Meerwein (29) had prepared many
aluminum (III) or zinc (II) double alkoxides, Similar double alkoxides
should be easily synthesized using metal alkoxides of "amphoteric" char-
acter. Hence titanium (IV) and iron (III) were hopefully substituted

for aluminum (III), but the reactions were unsuccessful.

A0

In the attempt to prepare these double alkoxides, the new compound
iron (III) phenoxide was synthesized in order to study its reaction.with
thallium (I) phenoxide. This compound, Fe(OC6H5)3, has an abnormally
high magnetic moment, 9.21 B.M., which indicates there is a ferromagnetic
interaction between the iron (III) nuclei or that some ferromagnetic
impurity is present.

In attempting to prepare a cobalt (II), titanium (IV) double alkoxide,
the complex Na2[Co(0C3H7i)ZC12], first reported by Bartley (30), was
synthesized. Bartley, while trying to prepare the analogous cobalt (II),

zirconium (IV) double alkoxide, inadvertently isolated this.

10.

ll.

12.
13.
1h.

15.
16.

17.
18.

19.

REFERENCES

Bradley, D. C., Record Chem. Progr. 21, 179 (1960).

Bradley, D. C., "Metal Alkoxides" in "Advances in Chemistry Series"
V. 23, American Chemical Society, Washington, 1959.

Bradley, D. C., "Progress in Inorganic Chemistry", F. A. Cotton,
Ed., V. 2, Interscience Publishers Inc., New York, 1960.

Dahl, L. F., G. L. Davis, D. L. Wampler, and R. West, J. Inorg.
Nucl. Chem. 2A, 357 (1962).

Bradley, D. C., Nature $82, 1211 (1958).
Demarcay, E., Compt. rend. 8Q, 51 (1875).

Bradley, D. C., F. M. El-Halim, E. A. Sadek, and W. Wardlaw, J. Chem.
Soc., 2032 (1952).

Bradley, D. C., R. C. Mehrotra, and W. Wardlaw, J. Chem. Soc., 163A
(1953).

Bradley, D. C., B. N. Chakravarti, and W. Wardlaw, J. Chem. Soc.,
2381 (1956).

Bradley, D. C., W. Wardlaw, and A. Whitley, J. Chem. Soc., 726 (1955).

Caughlan, C. N., H. S. Smith, W. Katz, W. Hodgson, and R. W. Crowe,
J. Am. Chem. Soc. 73, 5652 (1951).

Bradley, D. C., and M. L. Mehta, Can. J. Chem. A9, 1183 (1962).
Bradley, D. C., and M. L. Mehta, Can. J. Chem. A9, 1710 (1962).

Wentworth, R. A. D., and C. H. Brubaker, Jr., Inorg. Chem. 3, A7
(1961). '

Thiessen, P. A., and B. Kandelaky, Z. anorg. Chem. 181, 285 (1929).
Druce, J. P., J. Chem. Soc., 1A07 (1937).

Hagihara, N., and H. Yamazaki, J. Am. Chem. Soc. 81, 3160 (1959).
Thiessen, P. A., and O. Koerner, Z. anorg. Chem. lég, 65 (1929).

Bradley, D. C., R. K. Multani, and W. Wardlaw, J. Chem. Soc.,
A153 (1958).

A1

20.

21.

22.

23.

2A.

25.

26.

27.

28.

29.

30.

31.

32.

33.

3A.
35.
36.

37.

A2

Barraclough, C. 0., D. C. Bradley, J. Lewis, and I. M. Thomas, J.
Chem. Soc., 2601 (1961).

Misra, S. N., T. N. Misra, R. N. Kapour, and R. C. Mehotra,
Chem. Ind., 120 (1963).

Bradley, D. C., A. K. Chatterjee, and W. Wardlaw, J. Chem. Soc.,
2260 (1956).

Bradley, D. C., M. A. Saad, and W. Wardlaw, J. Chem. Soc., 1091
(195A).

Bradley, D. C., B. Harder, and H. Hudswell, J. Chem. Soc., 3318
(1957).

Jones, R. 0., G. Karmas, G. A. Martin, Jr., and H. Gilman, J.

Am. Chem. Soc. 78, A285 (1956).

Jones, R. 0., E. Bindschadler, G. Karmas, F. A. Yeoman, and H.
0i1mah, J. Am. Chem. Soc. 78, A287 (1956).

Jones, R. 0., E. Bindschadler, G. Karmas, G. A. Martin, Jr., J. R.
Thirtle, F. A. Yeoman, and H. Gilman, J. Am. Chem. Soc.
18, 6030 (1956).

Bradley, D. C., and A. K. Chatterjee, J. Inorg. Nucl. Chem. lg,
71 (1959).

Meerwein, H., and T. Bersin, Ann, 828, 113 (1929).

Bartley, W. G. and W. Wardlaw, J. Chem. Soc., A21 (1958).

Albers, H., M. Deutsch, W. Krastinat, and H. von Osten, Chem. Ber.
85, 267 (1952).
Jones, R. 0., E. Bindschadler, D. Blume, G. Karmas, G. A. Martin

Jr., J. R. Thirtle, and H. Gilman, J. Am. Chem. Soc.
78, 6027 (1956).

Wilds, A. L., "Organic Reactions", Roger Adams Ed., V. 2, J. Wiley
and Sons, New York, N.Y., 19AA, p. 198.

Sidgwick, N. V., and L. E. Sutton, J. Chem. Soc., 1A6l (1930).
Quinn, L., Ph.D. Thesis, Michigan State University, 1961, p. 2A.

Hermann, J. A. and J. F. Suttle, "Inorganic Synthesis", T. Moeller
Ed., V. 5, McGraw—Hill Book Co., New York, N.Y., 1957,
p. 1A3.

3

Pray, A. R., ibid., p. 153.

38.

39.

A0.

A1.

A2.

A3.

AA.
A5.

A6.
A7.
A8.
A9.

50.
51.

52.
53.
5A.

A3

Kovacic, P. and N. 0. Brace, "Inorganic Synthesis", E. G. Rochow,
Ed., V. 6., McGraw-Hill Book Co., New York, N.Y., 1960
p. 172.

Willard, H. H., N. H. Furman, and C. E. Bricker, "Elements of
Quantitative Analysis", D. Van Nostrand Co. Princeton,

N. J., 1956, p. 266.

9

Rodden, C. J., and J. C. Warf, "Analytical Chemistry of the Man-
hatten Project", C. J. Rodden Ed., McGraw—Hill Book Co.,
New York, N. Y., 1950, p. 18.

Kolthoff, I. M., and R. Belcher, "Volumetric Analysis", V. 2,
Interscience Publishers Inc., New York, N. Y., 1957,

p. 371.

Hillebrand, W. F. and G. E. F. Lundell, "Applied Inorganic Analysis",

J. Wiley and Sons, New York, N. Y., 1955, p. A78.

Stone, K. 0., "Determination of Organic Compounds", McGraw—Hill
Book Co., New York, N.Y., 1956, p. 121.

Vander Vennan, R. E., Ph.D. Thesis, Michigan State University, 195A.

Figgis, B. N. and J. Lewis, "Modern Coordination Chemistry", J.
Lewis and R. G. Wilins Ed., Interscience Publishers Inc.,
New York, N. Y., 1960, p. A06.

Ibid., p. A15.

Figgis, B. N. and R. S. Nyholm, J. Chem. Soc., A190 (1958).

Brown, R. N. and 0. Winter, J. Chem. Soc., 29A7 (1961).

Cotton, F. A., "Modern Coordination Chemistry", J. Lewis and R. G.

Wilkins, Ed., Interscience Publishers Inc., New York, N. Y.,

1960) p0 353.
Wentworth, R. A. D. and C. H. Brubaker, Inorg. Chem. 2, 551 (1963).

Wilhoit, R. C., J. R. Burton, Fu—tien Kuo, Sui-rong Huang and A.
Viquesnel, J. Inorg. Nucl. Chem. 28, 851 (1962).

Jfirgensen, C. K., Acta Chem. scand. 8, 1A95 (195A).
Graddon, D. P., J. Inorg. Nucl. Chem. AH: 161 (1960).

Figgis, B. N. and R. L. Martin, J. Chem. Soc., 3837 (1956).

‘

mummy LIBRARY

- dud-£55..