k
In}, ’3‘-

{'3’

7} l
A P: r
531% .u.$..¥& _
.3 2.
:34...
5'31.

.A
, :2
as}... he». I.
.wndfiwiz: _

2 up»... .
$.fin. uhuwmuw

{A}: _
km

...,.:.w..§...,
m. AAA,
:

tittailai.
. .213... 8-3.5:
. P.-
A A AA 3.1,}? )0!
bywzwflwunfihmrruh a
. {$39}: an"??? .
HQ... 5&5». «(.Aihnr‘.
A (Bail. Enthugii 23.13.52}! 3
.3?..,...H§..1..mhin.unr . dug?» . r. 1 h». SLthkfi;
. lfikwlittikié . . 5:4 1.: 51L: sun: v» {.«tbzagriifif A
. 3.... ail... H.5: . . . i Er... J .4... If «I. .33.. t
A fisaifl‘i‘r . axxmwva hf... Mi 3 writ”. "fruaalrvg,
. . A z A I. A ,A . A firauffififi manna»: A; w in.
. . . , A A A afirsfin. h... 2 En! A? n...
. . .A A A ; 5.3“...siiuuhwhiz.
A A . . 1 . .A wfirwu...»€9.:.._
A A. . . . A A 4 {DA A
an; A , A . A . M iii
A .. u {I ~ . A A . A A
3.1.4.29 34.; . A . AA . . . . A . .
Izedfifiwfisf» .A , .
«v .n‘mk?» 1.1... . . it A . .
A A r54 A . 6.11.”... , AA
X ‘ . . . V ‘ ‘
. b5,

«it.
:2.

 

.wniifl A A
”3 3.: 3!. Tbrwfifilfi K.
.nA A _ 3 a : u i
gawk, (an, .Ezfiflijg. .511 we.
I y 1.! .1 a . .(IR. \5l. .1. _
1 5e )7 . 4 . ‘ 3‘1 .
A . .33; L? $11.15!}: . if:
A. A . A . . A . 585.1}... .tI‘t-i i1!..n!:_
. . . A . A. A 3:32.11 . i: Sfmé11§142
J... A g A . y . :A I... A
43“.. ayfamarcysd. A . . , . A aak¢flfl§¥iimfimflx¢
9‘. Heanyvav A A A u . s’fiflsflvs.flx..miiflrn
r mm}? . . A . . A . : auxlaiafiflxhflwffd
. A A . 123,!”th 252.0%) . .2)!
4 A ‘ ‘5... l‘ .fv yv‘ 14. “A“; ‘
K . A ugh’ufilipzfl : 131.1 55:91
pflfihflqu .. A oat... 33M“ «Loans. 4.
.flhnflnfia‘. ,

tr

Y . 3 1.
333A) 73-11? 1....
\,......I\ n A\\ 3: “

...
ttkh.’ o’5131 fl (.5: v.
.5 iJf-,2,¢J\..In.))hvr;imi|p
LL} 3.. :3. A
A A I {Mira
3'11. 35.: ,1: A 5)..
v.... :13: .r. $2,761.

“2:5... l. ,
A DPS}?
2. I

\‘(i‘é
ur.11tur!>ic\|::
5% L153}! Dig!
:1 )\9.7!«:v;\;.
.1.)
? ;
:1 .53., as!
‘15)): v.\
1:

fr}
{5:

l9}, .1- .
9:5; I \F?(\
. A . :2: .A

Mllé.»

i459
.\l.~;l./Ii 9..
9.31.. it...
. .7. 1

~ ,

 

A ....A.,.A.A‘..y_u}
I...» «in.

.1.

 

A : . A . Law“. :_ .A #— ‘ ... A . { . . . ..
.2 a? nufiwfiutkfius fiéflka».
.mvmwmmwfifiawég 3%..» $3

$5

LIIRAR Y

m' m
M

 

This is to certify that the

thesis entitled

AN INVESTIGATION OF FACTORS CONTROLLING THE FORMATION

OF METAL COMPLEXES CONTAINING MACROCYCLIC LIGANDS
presented by

Lawrence A. Funke

has been accepted towards fulfillment
of the requirements for

Ph.D. degree in Chemistry

Wm

[ fljor ProfeSSOI/

Date June 10, 1976 /

 

0-7 539

 

 

69/29;2£&j//

ABSTRACT
AN INVESTIGATION OF FACTORS CONTROLLING THE FORMATION
OF METAL COMPLEXES CONTAINING MACROCYCLIC LIGANDS
By

Lawrence A. Funke

Factors controlling the condensation reaction between [3,3'-
(ethylenebis(nitrilomethylidyne)]di-2,4-pentanedionato(2-)Jnlckel

abbreviated Ni(enp), and diamines to form compounds containing

macrocyclic ligands have been investigated.

 

Ni(enp)

The first part of this investigation was the study of the dynamic

linkage isomerization, i.e., the interchange between the coordinated

and uncoordinated "acetyl" groups, of Ni(enp). The interchange was

followed by proton magnetic resonance spectroscopy with some of the

hydrogens of the uncoordinated acetyl groups replaced by deuteriums

Lawrence A. Funke

‘finrthe purpose of "labeling." The process follows first order kinetics,

with the rate constant k at 35° equal to 0.8l :_0.05 x 10'4 sec-1. The

energy of activation, Ea, for the isomerization is 17.4 :_0.7 kcal mol"1
and the entropy of activation, ASI298, is -23 :_2 eu. It is concluded
that a C0 group has a significant lifetime coordinated to the nickel(II)
ion. The implications of this conclusion which relate to a possible
mechanism for reaction of Ni(enp) with diamines to give a macrocyclic
complex are discussed.

The second part of this investigation was the study of the kinetics
of the hydroxide ion catalyzed ring closure reaction of Ni(enp) with
ethylenediamine to form a compound containing a macrocyclic ligand.

The reaction is first order in Ni(enp) and hydroxide ion and second
order in ethylenediamine. Also, the rate of reaction decreases as the
polarity of the solvent increases. The rate constant k at 68.2° is

3 M'3 sec-1. The activation energy, Ea’

1

equal to 2.22 :_0.09 x lo-
and the entropy of activation, ASIZQB, are l0.4 i 0.8 kcal mol' and
-42 :_3 eu. respectively. Mechanisms consistent with these results

are discussed.

AN INVESTIGATION OF FACTORS CONTROLLING THE FORMATION
OF METAL COMPLEXES CONTAINING MACROCYCLIC LIGANDS

By

‘ _ GCCJ>

Lawrence A? Funke

A DISSERTATION

Submitted to
Michigan State Univeristy
in partial fulfillment of the requirements

for the degree of

DOCTOR OF PHILOSOPHY

Department of Chemistry

1976

ACKNOWLEDGMENTS

I would like to express my sincere thanks to Dr. Gordon A. Melson
for his guidance, assistance and friendship in the course of my graduate
4 study. I would also like to thank Dr. Carl H. Brubaker, Jr., for his
assistance, especially in my last year.

I am grateful to the Petroleum Research Fund and to the Department
of Chemistry, Michigan State University, for the financial aid which
has made my graduate education possible.

I would like to thank my friends, colleagues, and fellow golfers
for the many enjoyable hours spent together. Their friendship has made
my life as a graduate student much more tolerable.

Finally, I would like to thank my parents for their encouragement
and for the sacrifices that they have made so that I might have an

education.

ii

TABLE OF CONTENTS

Chapter

I.
II.

III.

GENERAL INTRODUCTION ...................

DYNAMIC LINKAGE ISOMERIZATION OF [3,3'-[ETHYLENE—
BIS(NITRILOMETHYLIDYNE)]-DI—2,4-PENTANEDIONATO(2-)]NICKEL.

INTRODUCTION ......................
EXPERIMENTAL SECTION ..................
Preparation of Materials ..............
Procedure for Obtaining Kinetics Data ........
Treatment of Kinetics Data .............
RESULTS AND DISCUSSION .................
KINETICS AND MECHANISM OF THE RING CLOSURE OF [3,3'-

[ETHYLENEBIS(NITRILOMETHYLIDYNE)]-DI-2,4-PENTANEDIONATO-

(2-)]NICKEL TO FORM A COMPOUND CONTAINING A MACROCYCLIC
LIGAND ..........................
INTRODUCTION ......................
EXPERIMENTAL ......................
Preparation of Materials ..............
General Procedure for Obtaining Kinetics Data. . . .
A. Reactions at Reflux Temperature .......

B. Reactions at Temperatures Below Reflux

Temperature .................
C. Physical Measurements ............
Treatment of the Kinetics Data ...........
RESULTS ........................

Concentration of Ni(enp) ..............

Concentration of Ethylenediamine ..........

Page

11
II
II
12
T6

3O
31
36
36
37
37

4O
41

42
43
48
49

Chapter
Concentration of Hydroxide Ion ............
Solvent .......................
Temperature .....................
DISCUSSION .......................
SUGGESTIONS FOR FUTURE WORK ...............
APPENDICES
A. ATTEMPTS TO PREPARE SOME COMPOUNDS CONTAINING MACROCYCLIC
LIGANDS ..........................
INTRODUCTION ......................
RING CLOSURE OF Ni(baen) ................
Refluxing Ethylenediamine ..............
Pressure Tube ....................
A l50° .....................
B 200° .....................
Solution .......................
RING CLOSURE OF Ni(baenBrz) ...............
Refluxing Ethylenediamine ..............
Solution .......................
RING CLOSURE OF Ni(btfaen) ...............
Refluxing Ethylenediamine ..............
Pressure Tube ....................
Solution .......................
CONCLUSIONS .......................
B. ALTERNATE METHODS OF SYNTHESIS OF COMPOUNDS CONTAINING
MACROCYCLIC LIGANDS ....................
INTRODUCTION ......................
RING CLOSURE OF Ni(enp) .................

iv

Page
5l
54
56
58
68

7T
7T
72
72
73

74
75
75
75
76
76
76
76
77

79
79
79

Chapter Page

Ethylenediamine ................... 79
l,2-Propanediamine .................. 80
l,3-Propanediamine .................. 80

RING CLOSURE OF Cu(enp) ................. 8O
Ethylenediamine ................... 8O
l,2-Propanediamine .................. 8l
l,3-Propanediamine .................. 81

LIST OF REFERENCES ........................ 83

LIST OF TABLES

Table Page

l Temperature Dependence of the Rate Constants for

Linkage Isomerization ................. 25
2 Experimental Conditions ................ 39
3 Dependence of the Rate on the Concentration of Ni(enp). 48
4 Dependence of the Rate on the Concentration of

Ethylenediamine .................... 49
5 Dependence of the Rate on the Concentration of

Hydroxide Ion ..................... 5l
6 Effect of Solvent on the Rate ............. 54
7 Dependence of the Rate on Temperature ......... 56

vi

LIST OF FIGURES

Figure Page
1 Heme a and derivatives ................. 3
2 Structure of coenzyme 81g, 5'deoxyadenosylcobalamin, III,
and vitamin 312, cyanoco alamin, IV ........... 4
3 Plot from KINFIT of data for linkage isomerization
at 60° ......................... 15

4 The TOO-MHz pmr spectrum of Ni(enp) at room
temperature ....................... l8

5 Methyl region of the lOO-MHz spectrum for Ni(enp-d) as
linkage isomerization takes place at 40° ........ 22

6 Plot of -log k v§_l/T°K for linkage isomerization. . . . 27

7 Electronic spectra of Ni(ean (l), and Ni(enp-M-en),

(2) ........................... 35
8 Sketch of the experimental apparatus .......... 38
9 KINFIT plot of absorbance data measured at 296 nm as

a function of time ................... 45
TO KINFIT plot of absorbance data measured at 342 nm as

a function of time ................... 47
ll Dependence of the rate on the concentration of

ethylenediamine ..................... 50
12 Dependence of the rate on the concentration of hydroxide

ion ........................... 53
l3 Effect of solvent on the rate .............. 55
l4 Dependence of the rate on temperature .......... 57
15 Series of Spectra obtained by following the reaction

of Ni(enp) with ethyTenediamine ............. 59
16 A schematic representation of a proposed mechanism

for the formation of Ni(enp-M-en) ............ 66

vii

CHAPTER I
GENERAL INTRODUCTION

GENERAL INTRODUCTION

 

Metal complexes of organic macrocyclic ligands have been known
for many years. Early examples were found in the natural products,
such as porphyrin and corrin ring derivatives, and in the phthalo-
cyanines. One of the principal reasons for the current interest
in complexes containing macrocyclic ligands is the possibility of
relating the results of this research to biologically important
molecules. Metal ions are essential for many biological systems,]'4
and a large number of these metal ions are coordinated by macrocyclic
ligands. Figure 1 shows the structure of heme a, the site of oxygen
reduction in cytochrome c, and is an example of an iron(II)-porphyrin
complex. Figure 2 shows the structures of vitamin B12 and a coenzyme,
examples of complexes of corrin ring derivatives. Vitamin B12 func-
tions in the liver and is used in the treatment of pernicious anemia
and other metabolic disorders. While the therapeutic value of
vitamin B12 is known, it is not understood why it functions as it
does.

Because of their structural similarity, some coordination
compounds of synthetic macrocyclic ligands are being proposed as
possible model systems for the study of naturally occurring compounds?"9
Some of the most productive research concerning model systems has been
the study of synthetic dioxygen carriers. Recently, Basolo, Hoffman

10 summarized their work in this area and Colean, £5.El;211

and Ibers
reported the synthesis, characterization and properties of an iron(II)-

porphyrin complex which reversibly binds dioxygen. These and similar

, cu, (H,
H, | H, I H
c C c C'
o c/ \c/ \c/ \c/ \c/ \c/ \m
./\/m " m N W " '
R CH
cu,
'c‘
a
‘c ‘H’ I
(la) R=R=H
(lb) R: R'= cH,CH3
O\\ ‘ (1c) R=R'=CH3
c cu, , .
H, (Id) R-H.R =cocu3
cg. cw,
u, u,
Ro-c/ :C-OR
0
CH CH cu
H; I ) “I I ) "1 l ,
c C C C C C"°\
c/\/\/\/\/\/\ CH
“1 c c c c c c ’
/ \ / H H, u u, H, H)
"1C CH
cu,
H
ch C
‘\CH,
(II)
°\\
c ‘ H:
x
H
CH, CH,
/
cu, cu,
/ \
H,co-C\\ 'C—ocu ,
O 0

Figure l. Heme a (Ia) and derivatives.

 

IV R - CN "”3

Figure 2. Structure of coenzyme 812, 5'deoxyadenosylcobalamin,
III, and vitamin Blz’ cyanocobalamin, IV.

studies have greatly increased our insight into the naturally
occurring dioxygen carriers.

It is only in the last two decades that a large number of new
macrocyclic-ligand complexes have been prepared and investigated.
Many excellent reviews of the field of macrocyclic ligands and their

‘2 in 1964, described the

metal complexes are available. Busch,
general characteristics of compounds containing macrocyclic ligands.
The "kinetic template effect," the sterically orienting influence
of the metal ion on the course of the formation of a coordination

‘3 reviewed

compound, was also discussed. Several years later, Busch
the synthesis, electronic structure, magnetic properties and stereo-
chemistry of the transition metal complexes of synthetic macrocyclic

ligands. At about the same time, Curtis14

described the synthesis,
physical properties and chemical reactivity of macrocyclic coordination
compounds formed by the condensation of metal-amine complexes with
aliphatic carbonyl compounds. Later reviews by Lindoy and Busch9 and
by Black and Hartshorn15 discuss extensively the general considera-
tions important in the design and synthesis of macrocyclic ligands

and their complexes. The relative advantages and disadvantages

of the two general synthetic approaches, namely, direct synthesis

of the ligand followed by coordination to the metal ion, and "in situ"
synthesis of the complex, i.e., with the metal ion already present
during formation of the ligand, are discussed. Recently, Christensen,

16

Eatough and Izatt compiled a list of 22l different macrocycles and

the metals bound by these ligands. The list includes macrocycles with

oxygen, nitrogen, sulfur and mixed donor atoms. The size of this
list indicates the wide variety of macrocyclic ligands that are
now available.

Some recent studies have concentrated on the effect that various
systematic changes in the macrocyclic ring structure have on the
properties of the metal complexes. By the use of oxidative dehydro-
genation, Hipp, gt_gl;,]7 synthesized metal complexes with macrocycles
having prescribed patterns of unsaturation. Watkins, gt_al;,]8
investigated the effect of varying ring size on the Mossbauer and
electronic spectra and the ligand field strength of saturated tetraaza
macrocyclic ligands coordinated to iron(II). Wagner and Barefield19
studied the N-alkylation of macrocyclic secondary amine complexes of
nickel(II). Chemical reactions of the new complexes and the effect
of N-alkylation on the kinetic stability of the complexes and the
ligand field strength of a given ligand system were discussed. Other

20'24 has resulted in the development of new macrocyclic

recent work
systems and of improved synthetic techniques.

A property of macrocyclic ligands that has attracted considerable
attention is their ability to stabilize unusual oxidation states of
metals. This ability is ascribed to the enhanced stability of the
complexes of macrocyclic ligands compared to similar non-cyclic

25'27 thus enabling the macrocycles to trap metal ions in

ligands,
uncommon oxidation states. This enhanced stability has been termed
the macrocyclic effect or multiple juxtapositional fixedness.

Macrocyclic complexes of nickel(I),28 nickel(III),28 cobalt(I),29

30’3] and silver(III),3] have been isolated and characterized,

silver(II),
while complexes of copper(III)32 have been formed but are spontaneously
reduced in acetonitrile. Possible applications of these complexes
could include their use as oxidizing or reducing agents.

Most of the research concerned with metal complexes of macro-
cyclic ligands has concentrated on the study of their synthesis,
properties and reactivity. Very little work to determine mechanisms
of formation of these compounds has been performed. In this thesis,

I will describe efforts to determine the factors controlling the

formation of compounds containing macrocyclic ligands.

CHAPTER II
DYNAMIC LINKAGE ISOMERIZATION
OF [3,3'-[ETHYLENEBIS(NITRILOMETHYLIDYNE)]-
DI-2,4-PENTANEDIONATO(2-)]NICKEL

INTRODUCTION

 

Jager has reported that reactions between complexes containing
tetradentate ligands, V, and some diamines result in the formation

33

of complexes with macrocyclic ligands, VI. However, compounds

similar to V, but with different substituents on the six-membered
chelate ring, VII, do not undergo reaction with diamines.5’20’34
One of the goals of this research was to determine the role of the
uncoordinated CO group in the ring-closure of V to form the macro-
cyclic complex, VI.

In these systems, the ring-closure may involve reaction of the
amine at the coordinated CO group or at the uncoordinated CO group
followed by rearrangement. Alternatively, the rate of interchange
between the coordinated and uncoordinated CO groups may be sufficiently
rapid that they are equivalent and thus it is not possible to determine

the position of attack. This interchange process or linkage isomeri-

zation may be represented as

éo

 

TO

AHHveu .AHHvez

H>

>.

AHHveu .AHHVez

 

II

In an attempt to draw some conclusions concerning the possible
mechanism for the ring-closure process, we have measured the rate
of interchange of the "acetyl" groups for V, X = C2H4 and M = Ni(II),

over a range of temperatures from 35 to 60°.35

EXPERIMENTAL SECTION

 

Preparation of Materials

 

[3,3'-[Ethylenebis(nitrilomethylidyne)]di-2,4-pentanedionato(2-)]-
nickel, abbreviated Ni(enp), was prepared as previously reported.36
The deuterated species, Ni(enp-d), was prepared by adding 0.75 g
of Ni(enp) to 40 ml of 99% MeOD (Aldrich) and adjusting the pH of
the mixture to 12 with NaOH in 020. The solution was stirred at
5° for 20 days and filtered, and the residue was dried overnight
at room temperature under vacuum over P4010. Use of longer times
and/or higher temperature resulted in a decrease in the selectivity
of the deuteration, i.e. significant deuteration of the downfield
methyl group as well as the upfield group occurs. If allowed

to stir over extended periods of time (several months), virtually
100% deuteration of both methyl groups will occur. The conditions
given above were found to yield the maximum selective deuteration,

approximately 30% of the upfield methyl protons, with no other

hydrogens of the ligand being exchanged for deuteriums.

Procedure for Obtaining Kinetics Data

 

All proton magnetic resonance (pmr) spectra were obtained with
a Varian HA-lOO spectrometer. TMS was used as an internal standard

and chloroform-d used as solvent for recording the complete spectrum

12

at room temperature (see Figure 4). Above room temperature, the
spectra were obtained with bromoform as solvent. (No significant
solvent effect on the spectrum was observed at room temperature.)

37 before

Temperature was measured by using an ethylene glycol standard
and after each run. The maximum temperature variation throughout a
run is estimated to be :_O.5°.

A weighed amount of Ni(enp-d) was dissolved in a known quantity
of bromoform (typical concentrations approximately 5 x lo"2 M) and
the solution was then transferred to an nmr tube. The tube was capped
and placed into the probe which had previously been adjusted to the
selected temperature. Approximately 5 minutes were allowed for the
sample to reach temperature. The timer was started and repeated
scans of the methyl region of the spectrum from low to high field
(6 ~2.5 --2.25 ppm) performed. These scans began m 5 Hz before the
low-field peak and continued to m 5 Hz after the high-field peak.
Total scan time was approximately l2 seconds. For a particular scan,
the time was recorded when the recorder pen reached the top of the

low-field peak. Repeated scans of the spectrum were obtained until

the two peaks were of equal height (see Figure 5).

Treatment of Kinetics Data
The heights of the two CH3 peaks were measured directly from
the spectra and the decrease to zero of the peak height difference
as a function of time was used for calculation of the kinetics of
the interchange process. The data obtained at a particular temperature

were treated by KINFIT, a nonlinear curve-fitting program,38 with the

13

Michigan State University CDC 6500 computer. The equation for a first

k t
obs was fitted to the data. A is

order process, viz., A = Aoe
the peak height difference in millimeters at time t (seconds), A0

is the peak height difference at t = 0 and kobs is the observed
first-order rate constant. Input data consisted of the experimental
values of A and t and the variances in those values. Variance in t

was estimated to be :_(l sec)2, and in A :_(l mm)2. (A good estimate
of variance is considered to be the error in measurement squared.)

A0 and kobs are the adjustable parameters fitted to the data. The
output includes both the value and the standard deviation of the
parameters. After the treatment of a set of data at a particular
temperature by the KINFIT program, those points which deviated
considerably from the calculated A v§.t curve were dropped from

the data and the program was rerun. In no case was there a significant
change in the value of kobs although the standard deviation of kobs
decreased markedly. Figure 3 shows a computer plot from KINFIT

illustrating the fit of a set of data.

Due to convergence problems, the Arrhenius Equation,

k = A e
could not be used directly with KINFIT to determine the activation
parameters. Instead, the rearranged equation

-E

a l
—- )
k=k eR—(T ref (2)

-N-a

14

Figure 3. Plot from KINFIT of data for linkage isomerization

at 60°C.

15

 

 

hwv «#40 JJZG .517 can .muuocmo cc onh\m~\c or w4c><v occaetoacoooooooaoveoooo L?u
I . ‘ m \I

Hmullim w m w u u wIIIIm w u multimillimoillviasorlIIIUIIIIUIIIIVIIIIH
v x u
H x H
H x x H
H an ace H
H c H
v w
H x C C x H
H x H
H x u n x H
H x CC x H
w cc w
H oc H
H x cc x H
H x x H
H x C x .H
.u C u
H o n
H x o n
H H
H x C x H
m- Cx U
H c H
m "x H
H x c H
H x H
m m
H c H
H H
H x x H
H c H
m. C x u
H c H
H c H
H c H
H x C H
V nu u
H y Cw x H
H C H
H H
H x H
m C u
H H
H x H
H c H
H H
m yc w
H C n
H H
H :H
H ex»
HmoIIIrIIIImIIIIMIIOIuIIIum:IIIwIcIIVIIIIWIIoIUIIIIrIIIImIIIIrIIIIwaunIUIIIIUIIIIuncluuuultmoaatvuoilH

> «hum: >a x (hqmn wzav uIH 2H mac yzHca CMquzugac crc Jq»rw:Hounyu xa wrawr u
HomHHCJn mH chwa 2w!) cum: vH u >47c. HzHoo empaqzuaqu a m2<mz c .HzHcc J<H7G:Hamnru (c arena .

Figure 3.

16

was used. Tref IS an arbitrary reference temperature and kref is

the rate constant at Tref’ The value of Ea is independent of the
chOIce of Tref’ as calculations using both 333 K and 308°K as Tref
have shown. For the calculation of Ea’ k and T were the input
variables and kref and Ea were the fitted parameters. Variance in

T was set at :(l°)2 and variance in k was the square of the standard
deviation of k. Calculation of the Arrhenius pre-exponential was
made by use of the equation

-E /RT
f
k =Aeal re (3)

where kref and Ea were the parameters calculated with equation (2).

Calculation of ASI298 was then made by use of the equation

+
EkT) +* AS (4)

log A = log ( 2 303 R

RESULTS AND DISCUSSION

 

The lOO-MHz pmr spectrum of Ni(enp), structure V, X = C2H4’ in
chloroform-d at room temperature is shown in Figure 4. The resonance
at 5 7.49 ppm (relative intensity 2) downfield from the internal TMS
standard is assigned to the hydrogens attached to the six-membered
chelate rings of the ligand in Ni(enp) and the resonance at 6 3.30 ppm
(relative intensity 4) is assigned to the methylene hydrogens. The
two remaining resonances (each of relative intensity 6) at 6 2.29 and

2.44 ppm are assigned to the hydrogens of the uncoordinated and

17

Figure 4. The lOO-MHz pmr spectrum of Ni(enp) at room

temperature.

 

18

onzm

 

 

Humzm

 

 

Egan
onvm

 

 

00$

_¢

\

omum

onxm

Figure 4.

19

coordinated CH3C0 groups, respectively. These latter two assignments
are made on the basis of assignments for similar compounds. In both
the nickel(II) complex VIII and the macrocyclic complex derived from

it by condensation with ethylenediamine, the methyl hydrogens are

HCO

 

5 2 0
\c7
CH3
CH3
VIII
39

observed as a singlet at 6 2.40 ppm, close to the position assigned

to the hydrogens of the coordinated CH3C0 group in Ni(enp). In

40

addition, Hipp and Busch have reported pmr spectra of some nickel(II)

20

complexes which contain macrocyclic ligands having uncoordinated
CH3CO groups in the-y(or 3) position of the six-membered chelate
ring; VI, M = Ni(II). Resonances in the range 6 2.27--2.30 ppm were
assigned to the hydrogens of the uncoordinated CH3C0 groups whereas
those in the region of 5 2.4 ppm were assigned to the hydrogens of
the methyl groups attached to the macrocyclic ring. The assignments
for Ni(enp) are consistent with these assignments.

The pmr spectrum of Ni(enp) remains unchanged in CHBr3 up
to llO°, above which temperature decomposition takes plaCe. Since
there is no evidence for line broadening and the methyl hydrogens
remain distinguishable in the two different environments, even at
elevated temperatures, linkage isomerization is not occurring
rapidly, i.e., on a nmr time scale. This observation correlates
with the findings of Holm, gt_al;,41 for M(triac)3 complexes (M = Al, Ga;
triac = 3-acetylpentane-2,4-dionato). In these complexes, which
contain coordinated and uncoordinated CH3C0 groups, line broadening
was not detected up to l80° and it was concluded that linkage
isomerization is not a rapid process.

In order to measure the rate of linkage isomerization for Ni(enp),
the methyl groups in the different environments were made identifiable
by means of selective deuteration (see Experimental Section). The
linkage isomerization was followed by monitoring the difference in
peak heights for the methyl hydrogens in the partially deuterated
Ni(enp-d) as it decreased to zero. Figure 5 shows the methyl region
of the pmr spectrum for Ni(enp-d) at various times at 40°. A

complete set of data normally consisted of between 40 and l00 such

21

Figure 5. Methyl region of the lOO-MHz pmr spectrum for Ni(enp-d)

as linkage isomerization takes place at 40°.

 

 

 

 

l
O

 

) J

1690

 

 

l

1

 

 

1

N

3180

 

J

 

W

 

 

N11

6000

t secs

 

 

l

1

 

 

1* N

 

 

N

i
00

 

 

23

scans of the methyl region. It should be noted that at no temperature
was the total amount of deuteration decreased, thus ruling out any
exchange with the solvent. In addition, there was no evidence of
exchange between added, undeuterated, ligand and deuterated coordinated
ligand.

For the linkage isomerization, represented by the following

scheme (where the asterisk signifies labeling by deuterium),

     

CH3
: E k
N o v—3
’J::' \\. // k
N1
/ \
L M
we can write
9§%= -k[(L) - (Mn (5)
and
19% km) - (Mn (6)

dt

24

Now, A, the experimental peak height difference, is pr0portional

to [(L) - (M)]. Thus,

 

dA d L - M
31?: U )(yc ( )J (7)
fl = d L) _ d(M) (8)
dt dt dt

Substituting (5) and (6) into (8),

alga -k[<L) - (Mn - kw.) - (Mn (9)
dA
a= -2k[(L) - (M)] = -2kA (10)
Therefore,
dA
T = -2k dt (11)
ln A = ln A0 - 2kt (12)
A = A0 e'2kt (13)

As noted in the Experimental Section, the expression

25

was used for treatment of the data. Equating this expression to

(l3), we observe that

k = kobS/Z (14)

The data obtained for the isomerization were treated as described
in the Experimental Section. The isomerization is a first-order
process and values of the first-order rate constants, k, obtained
over the temperature range 35~~60°C are listed in Table l. A plot

of log k y§_l/T°K is shown in Figure 6. From these data, the energy

Table 1. Temperature Dependence of the Rate Constants for Linkage

 

 

 

Isomerization

Temp, °K 103/1, °k" k x 104, sec"
308 3.24 0.81 :_0.05
313 3.20 1.20 :_0.08
317 3.15 l.64 :_0.09
320 3.12 2.47 :_0.17
325 3.08 3.42 :_0.38
329 3.04 4.99 1.0.44
333 3.00 5.0 :_0.5

 

 

of activation for the isomerization, Ea’ is 17.4 :_.7 kcal mol'],

and log A, 8.24. From the log A value, 85*298 is 23 :_2 eu.

26

Figure 6. Plot of -log k v_s_ l/T°K for linkage isomerization.

27

4.2 -

3.8 -'

-log k

3.6 -

 

 

 

300 3.08 3.16 3.24

1 3
T°K x 10

Figure 6.

 

28

The fairly large negative value of AST298 indicates that a
symmetric intermediate is produced in the linkage isomerization
process, which involves rotation about a C41C bond of the ligand.
We suggest that, in this symmetric intermediate, "coordination" of

both oxygens to the nickel(II) occurs. Fay42

has observed linkage
isomerization of triac in Co(triac)3 and Co(acac)2(triac). He
obtained Ea and AS1298 values of 39.6 : l.6 kcal mol'1 and 28 :_4 eu

and 39.9 i 1.8 kcal m1"

and 28 :_5 eu, respectively.‘ In these
cobalt complexes, the linkage isomerization is proposed to proceed
via a five-coordinate intermediate produced by Co——0 bond rupture
followed by rotation about a C3=C bond within the resulting
unidentate ligand and Co——0 formation. This mechanism is consistent
with the observation of a large positive value for ASTZQB. The
cobalt(III) ion is six-coordinated in these complexes and a
mechanism involving reduction of the coordination number of the
cobalt(III) ion is reasonable for the linkage isomerization. However,
for Ni(enp), the nickel(II) ion is four-coordinated and a mechanism
invoking association of both C0 groups with the nickel(II) ion
during the isomerization is both plausible and in agreement with
the observed negative value for AS1298.

The preexchange lifetime, 1, i.e., the lifetime of a particular
configuration prior to the isomerization taking place, can be obtained
from the rate constant by r = l/k. For Ni(enp-d) in bromoform, T

3 3

varies from l2.4 x l0 sec at 60°. From

sec at 35° to 1.66 x 10
this information, it is possible to draw some conclusions concerning

the ring-closure reaction of Ni(enp) with amines. Reactions of C0

29

groups with amines proceed via nucleophilic attack of the amine

at the carbon atom of the C0 group. As noted previously, there is

no evidence for exchange with free ligand. Therefore, a process

in which dissociation of the complex occurs, followed by ring closure
of the free ligand and then coordination of the macrocycle to the
nickel(II) ion, is unlikely. In Ni(enp), there are two potential
reaction sites, the carbon of the coordinated acetyl group and that
of the uncoordinated acetyl group. Since the linkage isomerization
is slow and a particular configuration of the ligand has a significant
preexchange lifetime, the nickel(II) ion will have a significant
effect on the electron density within the ligand. Coordination will
result in decreased electron density at the carbon of the coordinated
C0 group due to its electron-withdrawing nature. Thus the carbon
atom of the coordinated CD will be the preferred site for nucleo-
philic attack by amines in the ring closure process. Attempts to
confirm this conclusion by reaction of Ni(enp-d) with amines were

not successful however, since very rapid exchange of the deuteriums
of the uncoordinated acetyl groups occurred in the presence of amines

and the macrocycle obtained was nondeuterated.

CHAPTER III
KINETICS AND MECHANISM OF THE RING CLOSURE
OF [3,3'-[ETHYLENEBIS(NITRILOMETHYLIDYNE)]-
DI-2,4-PENTANEDIONATO(2-)]NICKEL
TO FORM A COMPOUND CONTAINING A MACROCYCLIC LIGAND

3O

 

INTRODUCTION

 

In the General Introduction, it was noted that the synthesis,
properties and reactivity of a large number of macrocyclic complexes
have been studied, while little work has been performed to deter-
mine mechanisms of their formation. In one of the few reported

studies of this type, Hurley, Robinson and Trotz43

investigated the
stepwise formation of nickel phthalocyanine by the isolation of
several intermediate products that are formed on the way to the
final complex. However, no kinetics data or activation parameters
were determined.

Blinn and Busch44

studied the reaction of 2,3-pentanedi0nebis-
(mercaptoethylimin0)nickel(II), IX, with a,a'-dibromo—o-xylene and
other alkyl dibromides to form a macrocyclic complex, XI. They
determined that the reaction proceeds via a two-step process. The
first step, the combination of one mercapto group with one end of
the dibromoxylene, is the slow, rate determining step. The second
step, the reaction of the remaining mercapto and bromo groups to
effect ring closure, is too rapid to follow. Blinn and Busch
ascribe this difference in rates to the fact that the reactive
sites for the second step are already sterically oriented for the
cyclization reaction. This last observation is the essence of what
is called the "kinetic coordination template effect." It is also
probable that precoordination of the dibromoxylene to the nickel ion

occurs, as implied by the proposed structure of X. In other studies

involving similar reactions, precoordination is proposed in the

31

32

reaction of benzyl bromide with various mercaptoaminenickel(II)

45

complexes and in the reaction of alkyl halides with substituted

B-mercaptoquinoline complexes of nickel(II).46’47

/—_———_\S

/N BF 5- CH2
Cfizi\c/N \N./: B'CH O CH3\C/ \ILI/
l .__.. 1
,CH
/c / \S 8rCH2 /c\\ /|/\ 2
CH N CH N M' S
3 \ /S 3 '\ /
IX Nr—W / x
CH /N Br 5" CH2
CZ
,z' ‘x ,/’|
CH \ N Bri\ 5-0-12
3 \___/
XI
The ability of a metal ion to orient reactants so as to facilitate
a certain reaction has been amply demonstrated.]2'16 In many instances

the products of a particular reaction carried out in the presence

48-50 Leussing and

and absence of a metal ion are quite different.
coworkers, who prefer the term promnastic (matchmaker) to template,
have studied the formation of Schiff bases from salicylaldehyde and

5] 52 They propose the

glycine, and salicylaldehyde and diamines.
formation of a ternary complex followed by condensation of the
complexed ligands within the primary coordination sphere of the
metal ion.

Because of the lack of mechanistic information concerning the

formation of macrocyclic complexes, the investigation of the ring

33

closure of Ni(enp), XII, with ethylenediamine to f0rm a compound

containing a macrocyclic ligand, Ni(enp—M-en), XIII, was undertaken.

H3C \c éo

 

N. .0
\ /
N1 + NHCHCHNH
2 2 2
/ \ 2
N

   

XII

In order to determine the factors controlling the formation of the
macrocyclic complex, the dependence of the rate of reaction on the
concentrations of Ni(enp), ethylenediamine and hydroxide ion and on
the solvent and temperature were studied. The course of the reaction
was followed by monitoring the electronic spectra of the starting
material and the product, shown in Figure 7, where l = Ni(enp) and

2 = Ni(enp-M-en). The decrease in the concentration of Ni(enp) and

the increase in the concentration of Ni(enp-M-en) were followed by

 

34

 

Figure 7. Electronic Spectra of Ni(enp), (l), and Ni(enp-M-en) (2)

 

35

.H wcamHu

E: zhozm4m><3

cow omm mwm

com

mum
o.o

 

 

I ¢.H

 

 

BONVBUOSBV

36

monitoring the absorbances at 296 nm and 342 nm, respectively, as
a function of time. The calculation of the rate constants is
described in the Experimental section. A mechanism consistent with

the results of this study will be presented in the Discussion.

EXPERIMENTAL

 

Preparation of Materials

 

[3,3'-[Ethylenebis(nitrilomethylidyne)]di—2,4-pentanedionato(2-)]-
nickel was prepared as previously reported.36
Ethylenediamine (en) was twice distilled from sodium hydroxide

under dry nitrogen and stored in a dry nitrogen atm05phere.

Tetrahydrofuran (THF) was dried by refluxing over calcium hydride,
distilled from calcium hydride and then stored over molecular sieves
("Linde,“ type 5A).

Absolute ethanol was used as provided.

The ethanolic sodium hydroxide solution was prepared by placing
the desired volume of absolute ethanol, previously degassed with dry
nitrogen, in an erlenmeyer flask fitted with a connecting tube
(gas adapter) and a sidearm with stopcock. The flask was placed in
an ice-water bath and a flow of dry nitrogen was maintained for several
minutes over the solvent. The weighed amount of reagent grade sodium
hydroxide was added slowly, a few pellets at a time, over a period of
several hours. The solution was constantly stirred by means of a

magnetic stirrer and a continuous flow of dry nitrogen was maintained

over the solution. The ice bath was kept near 0°. After complete

37

dissolution of the sodium hydroxide the stopcocks on the sidearm and
connecting tube were closed and the flask immediately taken into an
inert atmosphere dry box where the solution was filtered through a
medium fritted funnel. The filtrate was then placed in a clean plastic
bottle, the outside of which had previously been covered with black
electrical tape. The solution was removed from the dry box and

stored in a refrigerator. The solution was standardized by titra-

tion with potassium hydrogen phthalate, which had previously been

dried for one hour at ll0°. Phenolphthalein was used as the indicator.
The ethanolic sodium hydroxide was kept cold and in an inert atmosphere
as much as possible in order to prevent the decomposition of the
solution. It was found that if these precautions were not taken, a
yellowish-brown color appeared in a few days and that the solution
became progressively darker as time passed. The solution was

considered unusable at the first appearance of the color.

General Procedure for Obtaining Kinetics Data

 

A. Reactions at Reflux Temperature

All reactions were carried out in a 100 ml round bottom flask
fitted with the standard 24/40 ground glass joint, a l0/30 ground
glass thermometer port and a sidearm with stopcock (see Figure 8).
The flask was placed in a heating jacket attached to a variable
voltage transformer. Stirring was provided by a magnetic stirrer
and a teflon coated magnetic stir bar.

ll weighed quantity of Ni(enp) was placed into the flask.

(See Table 2 for the quantities of all reagents and solvents.)

 

 

-4>1-10

 

 

 

 

 

Figure 8. Sketch of the experimental apparatus.

 

39

 

 

 

0 0H000 0 00000 0 0H0eH 0 0H000 0 0H0aH "000

mszmmm Lou

0.0 0.0 0.0 000> 0.0 He .00>

0.00 0e0> 0.0H 0.0_ 0.0, He .2000>

0.00 00000 - 0.00 0.00 00> - 0.00 0.00 H5 .0000

0.00 0.00 0.00 0.00 0.00 He .HepoH>

0... 0.0 4.0mm.“ a. $00.0. 0.0 H 0.8 0.0 .xmmw. 8. 83.8....

0-00 x 00.0 0-0_ x H0.0 0L0> 0-0H x 00.0 0-0_ x 00.0 2 .H-:00

0.0 0._ 00.0 000> 0.0 2 .0000

0-0H x 00.0 0.0_ x 00.0 0.0_ x 00.0 0.0_ x 00.0 000> z .HHaemvHZH
mczpmcmasoH pco>Hom :oH muoncux: mcHsmHumcmecum HacmVHz

"co mama Ho mucmucmamo

 

 

mcoHpHucou HmucwsHLmaxm

.N mHan

40

THF was then added to the flask followed by the absolute ethanol

and then by the standardized stock sodium hydroxide solution. All

of these liquids were added via a pipet. A thennometer was inserted
into the thermometer port and a reflux condenser fitted to the flask.
A drying tube containing Drierite, to absorb water, and Mallcosorb,
to absorb carbon dioxide, was placed at the top of the condenser.

The stirred solution was then heated to reflux. After the solution
came to reflux and all of the Ni(enp) had dissolved, the ethylene-
diamine was pipetted into the solution. (The pipet was put through
the top of the condenser and placed so that the tip was just below
the mouth of the condenser when the amine drained into the solution.)
When the solution came to reflux again, usually after less than

30 seconds, the timer was started. Samples were periodically
withdrawn through the sidearm, which had been fitted with a serum
cap, by means of a microliter syringe, fitted with a Chaney adapter,
to assure consistent aliquot sizes. The sample aliquot was then
injected into a 25 ml volumetric flask containing absolute ethanol,
diluted to the mark with ethanol and stored in the dark until the
spectrum was recorded or the absorbances measured. Aliquot sizes were
chosen so that the total concentration of nickel complexes after

dilution was about 3 x 10'5 M.

B. Reactions at Temperatures Below Reflux Temperature
All reactions carried out at temperatures below reflux
temperature were done so by placing the flask and contents in a

thermostatically controlled oil bath. The temperature of the bath

41

varied less than :_0.05° from the set temperature. Stirring within

the reaction flask was provided by an air driven, immersible magnetic
stirrer and a teflon coated stir bar. The general procedure used for
obtaining the kinetics data at these temperatures was the same as for
those at reflux temperature up to just before the time at which the
ethylenediamine was added. After the Ni(enp) had dissolved, the solu-
tion was allowed to cool to just below reflux, at which time the
condenser was removed and the flask closed with a ground glass stopper.
The flask was then transferred to the oil bath where it was again
fitted with a reflux condenser. The condenser and flask were flushed
with nitrogen for about 30 seconds and then a slow flow of nitrogen
was maintained over the reaction mixture by means of adapters placed

at the top of the condenser (see Figure 8). After the solution reached
the desired temperature, ethylenediamine, preheated to 60° :_l°,

was injected into the solution through the sidearm via a l0 ml
graduated syringe, and the timer was started. The amine was preheated
because it was found that if amine at room temperature was injected
into the solution, lO-l5 minutes would elapse before thermal equili-
brium was reestablished. With the preheated amine, a constant
temperature was usually reached in less than 5 minutes. The temperature
of the reaction mixture varied less than :_0.2° once thermal equilibrium
was established. The procedure for obtaining aliquots and absorbance

measurements is the same as described in part A above.

C. Physical Measurements
All spectra were obtained by use of a Unicam SPBOO B spectro-

photometer. Absorbances used for the calculations were obtained

42

with a Beckman Model DB-GT Grating spectrophotometer. Both instru-
ments were calibrated for wavelength with a holmium oxide filter.
Matched quartz cells were used and absolute ethanol was used as
the reference solution.

Dielectric constants were obtained at 25° with a Wissenshaftlich-
Technische Werkstétten Dipolmeter Type DM—Ol. A calibration curve
of instrument readings as a function of dielectric constant was
obtained from the experimental readings and the literature values
of the dielectric constants of ethanol, tetrahydrofuran, methylene

chloride and l-propanol.

Treatment of the Kinetics Data

 

Reactions were followed by measuring the absorbance of the
diluted aliquot solutions at two wavelengths, 296 nm and 342 nm,
as a function of time. The data obtained correspond to the rate
of loss of starting material, Ni(enp), and the rate of formation of
product, Ni(enp-M-en), respectively. The data obtained were treated
by KINFIT, a non—linear curve fitting program,38 with the Michigan
State University CDC 6500 computer. Since the ratio of amine to
Ni(enp) was always very large (at least 200:1) all reactions were
treated as occurring under pseudo-first order conditions. Consequently,

the equation which was fitted to the experimental data was the pseudo—

first order equation

 

43

where At’ A0 and A00 are the absorbances at time equal to t, zero

and infinity respectively, t is the elapsed time in seconds and kobs

is the pseudo-first order rate constant. Input data consisted of

the experimental values of At and t and the variances in those values.
Variance in t was estimated to be :.(1 sec)2 and in At to be :_(.003)2.
A

A0° and k0 are the adjustable parameters fitted to the data.

0’ bs
The output includes both the value and the standard deviation of the
parameters. Figures 9 and 10 show plots, from KINFIT, of the absorbance
as a function of time data for a kinetics experiment, and illustrate
the agreement between experimental and calculated points.

For the calculation of the activation parameters, the Arrhenius
Equation
Ea

- _1_
1n k — 1n A - R (T)

was used, where k = kobs/[OH—Hen]2 (see Discussion). A plot of
ln k 1§_1/T was made and a least squares analysis of the data was
used for the calculation of the slope (-Ea/R) and intercept (1n A).
The energy of activation, Ea was then calculated from the slope.

The equation

1
1nA=1n(-°—,'j—t)+%

was then used to calculate the entropy of activation, AS+298‘

RESULTS

The rate constants, k , reported in this section are the

obs
averaged values of the rate constants calculated from the absorbance

 

44

 

Figure 9. KINFIT plot of absorbance data measured at 296 nm

as a function of time.

45

osom 7H Own orxa ow>c m2, Hemcup r mom Iioawo ozw w—IwHH 73a .09000000000000.000000000 07w

m-..'OVOOUVWOOOOWUOUOV'-OBVOU'OVIIOUDV---OV-3-3V00--VOOOOm'UOOV'---VOOOOVII---r30--."-OOUWOUOUWUUOOVOUOOW

I
x
C

r .3wa

KO

OX
CI
XC‘

 

o
x

H—ho—H‘f u.— u-n—U‘ mane->- U‘ nuwwunhrqrwwwwwo—h—h-U‘ u—nnu—U‘u—ur—o—U tut—Hwy p-

urchaser; wag—o-u-qr

WOWOOWVOOVU-O.IOw-OOOWUOO-v-3-OVDOOOWIOOOUOOOUUIOVUV'---W"'Om----m'l--V3-03“--‘UWOUOOTIIOOUVOOOUVUOO

> «04m: >8 0 «H409 usam 0:» 2— mac HzHon cuhcqzuaqu nzc 4<kzwsHamnsm 20 «20w: u
mcwHchn mH chua 7w13 cmm: vH u >42c. HzHco owpqqru;<u a vzcm: a .Hzmeo Jcpruanuaxu ‘0 mzaux 0

Figure 9.

46

Figure 10. KINFIT plot of absorbance data measured at 342 nm

as a function of time.

47

- Hru;H 0r myu
.ocno 7H oLFH orxc 30:0 n? Hovrur 5 men Ilonmb .Zu IHIHHH 72m oceanooooooooooooomocooo C70

VOWOUWOOOOVOUOUWUBOUWOUO'VOUVOVOOIUUOOUIVIU--U'---“O'OI-m-IOUUUOOUOUOO"v-0--.”'O'IU-|"W'--'V'---VOUOOb

"C

t- .- u—Lf B—D—nb—HU v—r-o-n- u‘ u—o— pa-nU‘o—t—ao-o—U >0— boo-mu“ o—cwu—t—U u—Hu—nHU‘u—owu—nv-Lrp—o-n—o-U 5-
C
II
”we..." . «Hp-0H” u—u QHHU .O'O—OD- ‘U ”O‘H‘u O—qu‘boh-CHHU bob-to 05-0“ boNb-nb—cu tau—pom;

Hx

. > 0040: >8 x «eguc 000m my» 7H mac m7Hco c0004:VJ0u are unhrwzHomnxw (0 vrrm: u
Homhhcgo wH CHVuo 7n!) (kw: V— n >J7CH HZch ocha:c_cc a vrnuz p .thcn .ququanVu Ar vrrua I

Figure 10.

48

as a function of time data obtained at the two wavelengths, 296 nm and
342 nm (see Introduction). Normally, the values of k296 and k342

deviated from k by less than 5%. Table 3, dependence of the rate

obs

on the concentration of Ni(enp), lists k296’ k and kobs to illustrate

342
the agreement between the rate constants calculated from the data
obtained at the different wavelengths. All other tables list only

the values of kobs'

The slopes and intercepts of the straight lines obtained from the
plots of the data were calculated by means of a least squares treatment
of the data.

Concentration of Ni(enp)

The results of the experiments used to determine the dependence

of the rate of reaction on the concentration of Ni(enp)are shown in

Table 3.

Table 3. Dependence of the Rate on the Concentration of Ni(enp)

 

 

 

. 4 4 4

[Ni(enp)] k296 x 10 , k342 x 10 , kobs x 10 ,

x 103, M sec-1 sec.1 sec-1
2.51 1.42 _+_ .04 1.43 i .02 1.43 1.04
5.02 1.40 i .05 1.35 i .03 1.38 i .05
7.50 1.55 i .07 1.41 1.03 1.49 1.07

 

 

49

The pseudo-first order expression

[Ni(enp)]t
1n [Ni(enp)]o = -kt

 

was used to calculate kobs'

the initial concentration of Ni(enp). the species being monitored, a

Since the value of kobs does not vary with

first-order reaction with respect to Ni(enp) is indicated.

Concentration of Ethylenediamine

The results of the experiments used to determine the dependence
of the rate of reaction on the concentration of ethylenediamine are
shown in Table 4. A plot of kobs y§_the square of the ethylene-

diamine concentration is shown in Figure 11. The slope of the straight

Table 4. Dependence of the Rate 0n the Concentration of Ethylenediamine

 

 

 

Volume kobs x 104.
of en, m1 [en], M [en]2, M2 sec"1
4.00 1.00 1.00 0.96 :_.03
6.00 1.50 2.25 1.74 :_.03
7.00 1.75 3.06 2.74 i .05
8.00 2.00 4.00 2.97 :_.06
9.00 2.25 5.06 3.57 :_.12

10.00 2.50 6.25 5.12 :_.11

 

 

 

kobs x IO4 (sec")

 

 

 

b

O l I
0 2 4 6

lien]2 (moles/liier)2

Figure 11. Dependence of the rate on the concentration of

ethylenediamine.

51

line is 7.5 1.5 x 10'5 11'2 sec“ and the intercept is 1.5 11.9

x 10'5 sec-1

, i.e., zero within experimental error. Hence, the rate
of the reaction is proportional to the square of the concentration of

ethylenediamine.

Concentration of Hydroxide Ion

 

The results of the experiments used to determine the dependence
of the rate on the concentration of hydroxide ion are shown in

Table 5. A plot of kobs !§_the concentration of hydroxide ion is

Table 5. Dependence of the Rate on the Concentration of Hydroxide Ion

 

 

 

[OH'] x 102, M kobs x 104, sec-1
1.88 0.73 i .08
3.71 1.25 1.08
5.55 1.85 i .09
5.54 2.27 i .05
7.52 2.41 i .08
8.43 2.97 i .08
9.30 3.25 1.09

 

 

shown in Figure 12. The slope of the straight line is 3.41 :_.11 x 10'3

M'1 sec"1 and the intercept is 0.77 :_7.5 x 10"6 sec-1

, i.e., zero within
experimental error. Thus, the rate is proportional to the concentration
of hydroxide ion and no reaction should occur in the absence of hydroxide

ion.

52

Figure 12. Dependence of the rate on the concentration of

hydroxide ion.

53

 

 

50:05.06 00. x P18
0.0 Qm 0.¢ QN O .
_ . . . 00
J QC
I 0.. on.
«.0
VA
.0.
.v
4 VN
aa
a
O
....
I No.

 

 

Figure 12.

 

 

54

Solvent
It has been shown53 that for a reaction between an ion and a

neutral molecule the following expression may be written

2 2
_ ' NZ 8 1 1
lnk"‘"ko “25151573

where k0' is the rate constant in a medium of infinite dielectric
constant, N is Avagadro's Number, Ze is the charge on the ion, r and

r1F are the radii of the reactant ion and the activated complex,
respectively, and D is the dielectric constant. To test if this
expression is applicable to the ring closure of Ni(enp), the dependence
of the rate of reaction on the dielectric constant of the solvent was
studied. The dielectric constant of the mixed solvent in a particular
experiment was measured as described in the Experimental section.

The experimental results are shown in Table 6 and a plot of 1n kObs

v§_l/D is shown in Figure 13. The slope of the straight line is

Table 6. Effect of Solvent on the Rate

 

 

 

 

Volume of .

Ethanol, m1 0 1/0 kobs x 104. sec“ 1n kobs
8.0 9.449 .1058 1.71 :_.09 - 8.577
9.0 9.548 .1035 1.43 :_.05 - 8.855
10.0 9.895 .1011 1.05 :_.07 - 9.150
11.5 10.197 .0981 0.92 :_.04 - 9.295
13.0 10.520 .0951 0.73 i..05 - 9.522
14.5 10.811 .0925 0.57 :_.04 - 9.755

 

 

55

 

 

 

 

- 8.6 -
-9.0 -
In k obs
-94—
-9.8 —
l L l l
.090 .095 .100 .105
l / D

Figure 13. Effect of solvent on the rate.

 

 

56

71.6 :_4.4 and the intercept is -15.2 :_.4. The data show that as the
polarity of the solvent increases the rate of the reaction decreases.

These results are consistent with the equation of Frost and Pearson.

Temperature

 

The results of the experiments used to determine the dependence
of the rate of reaction on temperature are shown in Table 7, and a
plot of 1n k v§_1/T is shown in Figure 14. The value of the rate
constant k was obtained from the pseudo-first order rate constant kob

s
by use of the equation

= kobs
[enJZIOH']

 

(see Discussion). The slope of the straight line is -5.25 :_.41

x 103 °1< and the intercept is 9.2 i 1.2. The energy of activation,

Ea’ and the entropy of activation, ASI298, were calculated as described
1

 

 

 

 

in the Experimental section, and are 10.4 :_.8 kcal mol' and
-42 :.3 eu, respectively.
Table 7. Dependence of the Rate 0n Temperature
3

~19— 4 —1 4 -3 -1
T, °C T, °K kobs x 10 , sec k x 10 , M sec 1n k
55.7 3.041 1.98 :_.06 12.3 :_.4 - 6.702
60.2 3.000 2.36 :_.09 14.6 :_.6 - 6.527
65.4 2.954 2.97 i .06 18.4 :_.4 - 6.297
68.2 2.930 3.59 :_.14 22.2 :_.9 - 6.108

 

 

 

 

 

 

-6.2 —
-6.4 -—
In R
-6.6 -
-6.8 ' ' ‘
2.92 2.96 3.00 3.04

l/T °K x 103

Figure 14. Dependence of the rate on temperature.

DISCUSSION

 

Figure 15 shows a series of spectra obtained by following the
reaction of Ni(enp) and ethylenediamine to form Ni(enp-M-en) as
described in the Experimental section. These spectra illustrate that
the decrease in the amount of starting material and the increase in
the amount of product may be monitored as a fUnction of time by
measuring the absorbances of the sample solutions at 296 nm and 342 nm.
respectively. The presence of an isosbestic point at 316 nm, as expected
from the spectra of Ni(enp) and Ni(enp-M-en), Figure 7, should also
be noted. The presence of an isosbestic point is normally considered
to show the existence of only two uniquely absorbing species, but

54 If an absorbing

this interpretation is not necessarily true.
species is present only in low concentrations, its presence will not
affect the appearance of an isosbestic point. Therefore, the simple
series of spectra obtained from the ring closure reaction of Ni(enp)
does not disprove the possibility of short-lived intermediates, but
only shows that appreciable concentrations of intermediates do not
build up.

The results of this investigation show there are several factors
controlling the ring-closure reaction of Ni(enp) with ethylenediamine
in solution. Since ring closure of the free ligand does not occur under
comparable conditions, the presence of the metal ion is essential to
the formation of the macrocyclic ligand. The dependence of the rate

on the concentration of hydroxide ion (see Table 5 and Figure 12), as

well as unsuccessful attempts to prepare Ni(enp-M-en) in its absence,

58

59

.mcHsmmnmcmesuw

squ HaemVHz mo 0000606; on» mcwonHo$ 00 06000000 mcpomam $0 mchmm .mH meamwm

E: 2.65.06?

 

 

000 on». 000 com 00.0 .
. 0 H _ 00
.. 0.0

v

.. 0.0 m

90 Av

I

q

.. 6.0 w

«a

«a
I md

 

 

 

 

60

show that hydroxide ion is necessary for the ring closure reaction.

The effect of the solvent on the reaction (see Table 6 and Figure 13),
i.e., the decrease in the rate as the polarity of the solvent increases,
is also an indication of an ion-molecule reaction as the rate deter-

53 55 however, warns that

mining step of the reaction. Laidler,
equations relating rate to dielectric constant are best used as semi-
quantitative formulations which allow only rough predictions to be

made as to the effect of changing the dielectric constant. Therefore,
no further quantitative analysis of the effect of the solvent will

be presented.

As might be expected, the reaction is first order in Ni(enp).
However, it is second order in ethylenediamine and these results
indicate that an intermediate which involves one Ni(enp) molecule
and two ethylenediamine molecules is formed. From the temperature
study data, activation parameters were calculated and the energy of
activation, Ea’ and the entropy of activation, 851298, are
10.4 :_.8 kcal mol"1 and -42 :_3 eu, respectively. The large negative
entropy change indicates the formation of a highly ordered intermediate.
These values are close to the values determined by Blinn and Busch44
for the reactions between 2,3-pentanedionebis(mercaptoethylimino)-
nickel(II), IX, and alkyl bromides, described in the Introduction to
this chapter. For these reactions, energies of activation between 10

and 13 kcal mol'] and entropies of activation between -26 and -38 eu

were determined.

61

A mechanism consistent with these observations may be represented

by the following scheme:

 

 

 

1 2
Ni(enp) + 2 en ‘\ A Step (1)
k-l
k2[0H']
A €> Ni(dang) Step (2)
slow
Ni(dang) 9? Ni(enp-M-en) Step (3)
fast

Step (1) is a pre-equilibrium process in which an intermediate, A,
probably an adduct between Ni(enp) and two ethylenediamine molecules,
is formed. Step (2), a slow step, is the hydroxide ion catalyzed
condensation of one of the amine groups of an ethylenediamine with
one of the coordinated C0 groups of Ni(enp) to form the intermediate
Ni(dang). Step (3) is a fast step in which the remaining, "dangling,"
amine group condenses with the remaining coordinated CD of Ni(dang) to
complete the ring closure and form the macrocyclic product, Ni(enp-M-en).
Since Step (3) is too fast to observe, it is not possible to determine
if hydroxide ion participates in the final step of the ring closure
reaction.

The initial condensation of the amine with Ni(enp) is proposed
to occur at the coordinated CO rather than the uncoordinated CD for
two reasons. From the data reported in Chapter Two, it was noted

that the electron density at the carbon atom of the coordinated C0

 

 

62

should be decreased due to the electron withdrawing nature of the
nickel(II) ion, and thus be the preferred site of nucleophilic attack.
Second, ring closure of the nickel(II) complex, VIII, occurs exclusively
at the coordinated C0 groups, as is shown by the positions of the

methyl peaks in the pmr spectra of VIII and the macrocyclic complex

derived from it. (See Chapter Two, Results and Discussion for details).

H5020 \c ¢0

   

VIII

The rate law for the reaction scheme proposed above, if one
assumes that the concentration of A is in a steady state, may be

derived as follows:

954:1 = k1 [Ni(enp)][en]2 - k_][A] - kZIAJEOH'] = 0 (15)

 

63

By rearranging (15) and solving for [A],

k [Ni(enp)][en]2
- 1 — (16)
k_] + kgfmi ‘J

 

Since Step (3) of the scheme is a fast step,

d N' -M- d N d
[ 1(engt en]: L 1(.tang)]- —k 2[A][OH ] (17)

By substituting (16) into (17),

 

 

d[Ni(enp—M-en)] = k2k1[N1(60P)][en]2[0H']
dt

 

 

- (18)
k_1 + k2[0H ]
_11/k
By multiplying (18) by l/kJ and substituting K = k 1_1/k
into the resulting equation,
. 2 -
d[Ni(enp-M-en)]j= k2K[N1(enp)][en] [0H ] (‘9)
dt k

1 + Eg-[OH']
-l
where K is the equilibrium constant for Step (1), the formation of A
from Ni(enp) and two ethylenediamine molecules.
At low hydroxide ion concentrations and where k2<<k_],
the rate law, (19) reduces to

 

d["i(e"3£"'e"’1 = k[N1(enp)][en12[0H’] (20>

 

 

64

which is the experimental rate law, where k = sz and kobs = k[en]2[0H'].

Values of k at various temperatures were shown in Table 7 and vary from

3 M'3 sec-1 at 68.2° to 1.23 x 10"3 M'3 sec-1 at 55.7°.

2.22 x 10'
The values of the individual rate constants, k], k_1 and k2

would be helpful in obtaining information about the nature of the

intermediate A. The value of k1 should be obtainable, since at high

hydroxide ion concentrations the rate law, (19), reduces to

 

d[Ni(en§;M-en] = k][Ni(enp)][en]2 (21)

 

i.e., the rate becomes independent of hydroxide ion concentration and

the values of k2 and k_]. Over the range of hydroxide ion concentra-

tions studied, however, this leveling was not observed. Experiments at

higher concentrations could not be performed because of the occurrence

of precipitate, which appeared to be nickel hydroxide. However, in the

study of the reaction of NiMMK, XIV, with l,3-propanediamine to form a
56

mono-condensed species, Shalhoub did observe zero order dependence

on the concentration of hydroxide ion at high concentrations.

 

65

If the value of the equilibrium constant K was determined the
rate constant k2 could also be calculated, since k = k2K. However,
no spectral changes which would indicate the formation of A were noted.
Therefore, the data needed to calculate K could not be obtained.
Despite the inability to calculate the specific rate constants
k], k_1 and k2
nature of the intermediate A and to propose specific mechanisms for the

, enough information is available to consider the

formation of Ni(dang). One possible mechanism is shown in Figure 16.
The intermediate A is Ni(enp) with an ethylenediamine molecule coor-
dinated at each axial position. Two amines coordinate due to the
greater stability of the six-coordinated, octahedral complex compared
to the five—coordinated, square-pyramidal complex which would result
if only one amine was coordinated to the nickel(II) ion. The next
step would be the nucleophilic attack of the uncoordinated end of
one ethylenediamine on the carbon atom of a coordinated C0, followed
by abstraction of a proton from the amine by a hydroxide ion.
Completion of this initial condensation results in the formation of
an imine, with release of a hydroxide ion, to yield Ni(dang). The
final step is the condensation of the unreacted end of the ethylene-
diamine with the remaining coordinated CO to form the macrocyclic
complex, Ni(enp-M-en). As with the macrocycle formation reactions

44 the last step is a fast step, since the

studied by Blinn and Busch,
reactive sites are sterically oriented for the cyclization reaction.
Thus, the ring closure reaction of Ni(enp) with ethylenediamine is

one of only a few reported examples of the "kinetic template effect,"

 

66

H.03050 00: 000 mmcHL
000H000 0:0 00 00000000 000000 .quL0H0 00 0000 0:0 000V .Hc0wziac0VHz
00 :0H00ELoH 0:0 00» 500000005 00000000 0 00 00000000000000 000050000 < .oH 00:0H0

Ac0iziac0sz A0000VHz <

K/ 1N: 1 IO” I/7~_\/7._ NI

> :\ .
z/.z\z . z/\;_.VzU oAflVz
zOz . 0 0A... 1V2 _ OAIIVZ

 

67

i.e., the sterically orienting influence of the metal ion on the course
of the formation of a coordination compound, applied to the formation
of a compound containing a macrocyclic ligand.

An alternative mechanism is similar to the mechanism just discussed
in that the intermediate A is the same, i.e., Ni(enp) with ethylene-
diamine molecules coordinated at the axial positions. In this second
mechanism, however, a hydroxide ion would abstract a proton from the
coordinated end of the amine, and then nucleophilic attack of the
resultant amide on the carbon atom of a coordinated CD would occur.
Condensation to yield Ni(dang) followed by the fast ring closure step
would complete the reaction.

Both of the proposed mechanisms are consistent with the
experimental entropy of activation, AST298 = -42 i 3 eu. The large
negative entropy change indicates the formation of an intermediate
that is much more ordered than the starting material. The proposed
intermediate A, which is formed by the coordination of two ethylene-
diamine molecules to Ni(enp), is such an intermediate.

The first mechanism postulates the breaking of the nickel-nitrogen
bond after the condensation of one end of the amine with a coordinated
C0 to form an imine linkage. The second mechanism requires the breaking
of the nickel-nitrogen bond before the ethylenediamine condenses to
become permanently attached to the coordinated ligand. It seems
more likely that condensation would occur prior to the breaking of the
bond through which the ethylenediamine molecule is attached to the nickel
complex. Therefore, the first mechanism might be favored over the

second mechanism.

SUGGESTIONS FOR FUTURE WORK

 

More information on the role of the metal ion is needed to help to
understand the mechanism of the ring closure reaction. This information
can best be obtained by studying the ring closure of other compounds
of the formula M(enp). The metal ions should be chosen so that a
variety of electronic configurations and preferred stereochemistries
are investigated. The study of the rate of ring closure of complexes
containing metal ions which exhibit Jahn-Teller distortion should
prove interesting, since the axial positions in these metal ions are

not as available for coordination as the axial positions in metals which

 

do not exhibit Jahn-Teller distortion. If ethylenediamine molecules
coordinate at the axial positions, as proposed for the intermediate A,
then the rate of reaction should be affected by this distortion.
Preliminary experiments indicate that the ring closure of Cu(enp) may
be studied by techniques similar to those used in the study of the
ring closure of Ni(enp).

A more detailed study of the role of the amine in the ring
closure reaction should be undertaken. The effect of varying the
basicity of the amine on the rate of reaction could yield interesting
information, although it might be difficult to distinguish between
steric effects and the effects of basicity. By substituting other
amines for ethylenediamine it may be possible to detect and/or isolate
one or both of the intermediates proposed in the reaction scheme.
Preliminary experiments show that the ring closure of Ni(enp) and Cu(enp)
with 1,2-propanediamine and l,3-propanediamine and of Ni(enp) with

1,4-butanediamine will occur.

68

69

The rates of condensation reactions, if any, between Ni(enp) and
monoamines should also be studied. The entropies of activation of
these reactions could prove especially useful. Since monoamines
contain only one functional group, mechanisms in which one end of
the amine coordinates to the nickel ion and the free end undergoes the
initial condensation with a coordinated C0 are not possible. There-
fore, on the basis of these studies, some of the possible mechanisms

for the formation of Ni(dang) may be eliminated from consideration.

 

APPENDICES

 

 

APPENDIX A
ATTEMPTS TO PREPARE SOME COMPOUNDS
CONTAINING MACROCYCLIC LIGANDS

1

7O

 

ATTEMPTS TO PREPARE SOME COMPOUNDS
CONTAINING MACROCYCLIC LIGANDS

INTRODUCTION

 

The ring closure reactions of Ni(enp), V, X = C2H4, and Ni(enda),

VIII, to form macrocyclic compounds, proceed quite readily.33 However,
it has been reported that the ring closure of Ni(baen), VII a, does

not occur.5’20

The structures of Ni(enp), Ni(enda) and Ni(baen) are
very similar, except for the substituents in the y (or 3) position
of the six-membered chelate rings. Therefore, in order to investigate

the effect of various substituents attached to the six-membered chelate

 

VII
X = C2H4
a) R1 = H; R2 = CH3
b) R1 = Br; R2 = CH3
c) R1 = H; R2 = CF3

71

 

72

rings on the ring closure reaction, attempts were made to prepare
compounds containing macrocyclic ligands from VII. Since the groups
in the y positions of Ni(enp) and Ni(enda) (CH3C0 and CZHSOCO,
respectively) are electron withdrawing, compounds containing electron
withdrawing substituents attached to the chelate rings were chosen

for study.

RING CLOSURE OF Ni(baen)

 

Bisacetylacetoneethylenediiminonickel(II), Ni(baen), VII a,

"35 Prepared as previously reported.57’58

 

Refluxing Ethylenediamine

 

Attempts were made to close Ni(baen) by using the methods
employed by Jager to form Ni(enp-M—en).33 About 0.5 g of Ni(baen
was added to 10 ml of anhydrous ethylenediamine and the solution was
heated to reflux and maintained at reflux temperature for 7 days. A
pink product, probably Ni(en)3X2, was filtered from the solution. The
mass spectrum of this solid showed no peaks above m/e = 60 (ethylene-
diamine). The ethylenediamine was removed from the filtrate in vaggg,
and a brown, tar-like substance remained in the bottom of the flask.
The mass spectrum of this substance indicated the presence of Ni(baen)
and ethyl enediamine but no peaks corresponding to the macrocyclic
complex were present.

The above procedure was repeated several times, and the length
of time the solution was heated was varied between two hours and two
weeks. No evidence for the formation of the macrocyclic complex was

obtained from any of the experiments.

Pressure Tube

 

Attempts to perfonn the ring closure reaction of Ni(baen) with
ethylenediamine at temperatures greater than the reflux temperature
of ethylenediamine were carried out with the reactants placed in a
glass pressure tube. For the experiments described in this section,
0.5 g of Ni(baen) and 10 m1 of anhydrous ethylenediamine were placed
in a pressure tube. The general procedure for preparing the pressure

tube and contents for an experiment are described elsewhere.59

A. 150°

 

For the first of these experiments, the sealed tube was placed
in an oil bath at 150° for 18 hours. After the tube and its contents
were cooled to room temperature, they were frozen in liquid nitrogen
and the tube was opened. After the contents of the tube melted, the
mixture was filtered through a medium fritted funnel. A very small
amount of solid was obtained and dried under vacuum over P4010. The
mass spectrum of this solid indicated the presence of a small amount
of macrocyclic complex but all attempts to isolate the compound failed.
The ethylenediamine was removed from the filtrate ig.yagug, and the
tar-like substance which remained was dissolved in xylene. Upon removal
of the xylene, jg_yaggg, red-brown needles were obtained. The mass

spectrum of this solid indicated the presence of Ni(baen) only.

B. 200°
The above procedure, in part A, was repeated, except the tube

was placed in a Wood's Metal bath at 200° for 40 hours. However, the

73

74

solution refluxed during the entire time period because the cooler,
unimmersed part of the tube acted as a condenser. The procedure in

part A, was again repeated except the tube was placed in a tube

furnace at a temperature of 200° for 24 hours. In each experiment, a
pink precipitate, assumed to be Ni(en)3X2, was obtained after filtration
of the mixture. The filtrate was treated as described above in part A,
and red-brown needles were obtained. The mass spectrum of this solid
indicated the presence of Ni(baen) only. No evidence for the formation

of the macrocyclic compound was obtained.

Solution

Attempts were made to effect the ring closure of Ni(baen) through
the use of conditions similar to those employed in the study of the
ring-closure reaction of Ni(enp) (see Chapter III). To a 50 ml round-
bottomed flask fitted with a sidearm and stopcock were added 0.28 g
(0.001 mol) of Ni(baen), 25 m1 of THF, 5 ml of 0.25 M ethanolic sodium
hydroxide and 6.7 ml (0.1 mol) of ethylenediamine. The solution was
heated to reflux and maintained at reflux temperature for 7 days.
Periodically, about twice a day, a sample was withdrawn from the
solution and the UV-visible spectrum recorded. The spectra indicated
a slight decrease in the concentration of Ni(baen), but no new
absorption peaks, which would indicate the formation of the macro-
cyclic complex, appeared. After 7 days, the solution was cooled and
then filtered. A pink solid, assumed to be Ni(en)3(0H)2, was

obtained. The solvent was removed from the filtrate ifl_vacuo. The

 

 

75

resultant tar-like substance was treated with xylene as described
above under Pressure Tube, part A, and red—brown needles were obtained.

The mass spectrum of the solid indicated the presence of Ni(baen) only.

RING CLOSURE 0F Ni(baenBrz)

 

Bis-3-bromoacetylacetoneethylenediiminonickel(II), Ni(baenBrz),

VII b, was prepared as previously reported.60

Refluxing Ethylenediamine
To a 50 ml round-bottomed flask were added 0.2 g of Ni(baenBrz),

and 10 ml of anhydrous ethylenediamine. The stirred solution was
heated to reflux. Within 45 minutes, the dark solution had lightened
considerably and a pink-violet precipitate had formed. The solution
was filtered and the precipitate dried. The mass spectrum of the
solid showed no peaks above m/e = 60 (ethylenediamine). The solid

was assumed to be Ni(en)3X2.

Solution

Several attempts to effect the ring closure of Ni(baenBrz) with
ethylenediamine were made by following the procedure previously
described for the ring closure of Ni(baen) in solution (see above).
The amounts of ethylenediamine and hydroxide ion were varied, but
for each experiment the results were similar. In each case, the
solution became considerably lighter in color and a pink solid,
Ni(en)3(0H)2, formed. No evidence for the formation of the macrocyclic

complex was obtained.

 

RING CLOSURE OF Ni(btfaen)

 

Bistrifluoroacetylacetoneethylenediiminonickel(II), Ni(btfaen),

VII c, was prepared as previously reported.58

Refluxing Ethylenediamine

 

To a 50 ml round-bottomed flask was added 0.5 g of Ni(btfaen) and
10 m1 of anhydrous ethylenediamine. The solution was heated to reflux
and maintained at reflux temperature for 24 hours. The solution was
cooled and filtered but no solid was obtained. With the addition of
50 ml of distilled water, a tan precipitate formed. The mixture was
filtered and the residue dried. The mass spectrum of the solid

indicated the presence of Ni(btfaen) only.

Pressure Tube

 

A pressure tube containing 0.3 g of Ni(btfaen) and 10 ml of

59 The tube was

ethylenediamine was prepared as previously described.
placed in a tube furnace, which had been preheated to 200°, for two
hours. The tube was removed from the furnace, cooled to 0° and
opened. The mixture was filtered and a pink solid, Ni(en)3X2,

recovered. No other solid could be isolated from the filtrate.

Solution

A solution containing 0.1 g of Ni(btfaen), 25 ml of THF, 2 ml of
0.15 M ethanolic sodium hydroxide and 3.8 m1 of ethylenediamine was
prepared. The solution was heated to reflux and maintained at reflux
temperature for 48 hours. Samples were withdrawn from the solution

periodically and the UV-visible spectrum recorded. The spectra

76

77

indicated some decomposition of Ni(btfaen) had occurred but no evidence
for the formation of the macrocyclic complex was obtained. The mixture
was cooled and was filtered. A pink solid was obtained. The solvent
was removed from the filtrate in vagug and 50 m1 of distilled water
was added. A tan precipitate formed. The mass spectrum of the dried

solid indicated the presence of Ni(btfaen) only.

CONCLUSIONS

Attempts to prepare macrocyclic compounds by the ring closure
reaction of VII with ethylenediamine were not successful. Ni(baen)
and Ni(btfaen) are not as reactive as Ni(enp) and decompose to
Ni(en)3X2 under severe conditions. Ni(baenBrz) is less stable than
Ni(enp) since decomposition occurs even under mild conditions. It
is evident that substituents attached to the chelate rings of VII
affect the stability of the metal complexes, but the role of these

substituents in the ring closure reaction of VII is uncertain.

 

 

 

APPENDIX B
ALTERNATE METHODS OF SYNTHESIS
OF COMPOUNDS CONTAINING MACROCYCLIC LIGANDS

78

 

ALTERNATE METHODS OF SYNTHESIS
OF COMPOUNDS CONTAINING MACROCYCLIC LIGANDS

INTRODUCTION

 

The inf0rmation obtained from the study of the factors controlling
macrocyclic ligand formation has resulted in the development of alter-
nate methods of synthesis of several metal complexes containing macro-
cyclic ligands. In the previously reported method of synthesis,33 a
mixture of Ni(enp) or Cu(enp) and a diamine is heated to reflux and
maintained at reflux temperature for several hours. Separation of
the macrocyclic product from the diamine is difficult and yields are
between 20 and 40%.

In solution, the ring closure reactions of Ni(enp) and Cu(enp)
occur quite readily. Isolation of the macrocyclic product from the
solution is easily managed and greater yields are obtained than by

the previous method of synthesis. These yields are between 45 and

76%. Details of the syntheses are given below.

RING CLOSURE OF Ni(enp)

 

 

Ethylenediamine
Into a 250 ml round-bottomed flask were added 0.5 g (1.5 x 10‘3 mol)
of Ni(enp). 100 ml of acetonitrile, 10 ml of 0.3 M (3 x 10‘3 mol) of

ethanolic sodium hydroxide, and 10 m1 (9 g, 0.15 mol) of ethylenediamine.
The solution was heated to reflux temperature and was maintained at that
temperature for 90 minutes. The reaction was assumed to be complete

when the absorbance peak for Ni(enp) was no longer observable in

79

 

 

 

 

80

the UV-visible spectrum. The solution was cooled to room temperature
and then added to 40 m1 of distilled water. A red precipitate was
produced. After the water mixture was cooled to near 0°, it was
filtered through a fine fritted funnel. The solid was dried overnight
at 100° under vacuum and identified by its mass spectrum as the
macrocyclic product, Ni(enp-M-en). Yield: 76%. (The mass spectrum

also indicated the absence of Ni(enp) in the product.)

l,2-Propanediamine

 

The procedure for performing the ring closure reaction of Ni(enp)
with 1,2-propanediamine and the identification of the product is
the same as that described with ethylenediamine, above, except that
13 ml (11 g, 0.15 mol) of l,2-propanediamine was used in place of the
ethylenediamine and the solution was maintained at reflux temperature

for 4 hours. Yield: 53%.

l,3-Propanediamine

 

The procedure for performing the ring closure of Ni(enp) with
1,3-pr0panediamine andthe identification of the product is the same
as with ethylenediamine, above, except that 13 ml (11 g, 0.15 mol)
of l,3-propanediamine was used in place of ethylenediamine and the

solution was maintained at reflux temperature for 4 hours. Yield: 58%.

RING CLOSURE 0F Cu(enp)_

 

Ethylenediamine

 

3

Into a 100 ml round-bottomed flask were added 0.5 g (1.5 x 10' mol)

3

of Cu(enp), 25 m1 of THF, 5 ml of 0.3 M (1.5 x 10- mol) of ethanolic

81

sodium hydroxide and 6 ml (5.4 g, 0.09 mol) of ethylenediamine. The
solution was heated to reflux and maintained at reflux temperature

for 15 minutes. The reaction was assumed to be complete when the
absorbance peak for Cu(enp) was no longer observable in the UV-visible
spectrum. The solution was cooled to room temperature and then added

to 200 m1 of distilled water. A purple precipitate was produced. After
the water mixture was cooled to near 0°, it was filtered through a fine
fritted funnel. The solid was dried overnight at 110° under vacuum

and identified as the macrocyclic product, Cu(enp-M-en), by its mass
spectrum. Yield: 72%. (The mass spectrum also indicated the absence

of Cu(enp) in the product.)

l,2-Propanediamine

 

The procedure for performing the ring closure reaction of Cu(enp)
with l,2-propanediamine and the identification of the product is the
same as with ethylenediamine, above, except that 6.5 ml (5.6 g, 0.075 mol)
of 1,2-propanediamine was used in place of ethylenediamine, 15 ml of
0.3 M ethanolic sodium hydroxide was used in place of 5 ml and the

solution was maintained at reflux temperature for 20 minutes. Yield: 53%.

l,3-Propanediamine

 

The procedure for performing the ring closure reaction of Cu(enp)
with 1,3-propanediamine and the identification of the product is the
same as with ethylenediamine, above, except that 6.5 ml (5.6 g, 0.075 mol)
of l,3-propanediamine was used in place of ethylenediamine, 15 m1 of
0.3 M ethanolic sodium hydroxide was used in place of 5 ml and the solu-

tion was maintained at reflux temperature for 75 minutes. Yield: 45%.

82

The assistance of William Stobby in the preparation of the

macrocyclic complexes derived from Cu(enp) is gratefully acknowledged.

 

 

LIST OF REFERENCES

 

(JON

10.

11.

12.
13.

14.
15.

16.

LIST OF REFERENCES

R. Bonnet, Chem. Rev., 63, 573 (1963).
R. F. Gould, (ed.), Adv. Chem. Ser., 0800492 (1971).

 

M. N. Hughes, The Inorganic Chemistry of Biological Processes,
Wiley, New York, 1972.

 

G. L. Eichorn, (ed.), Inorganic Biochemistry, Vols. 1 and 2,
Elsevier, New York, 1973.

 

T. J. Truex and R. H. Holm, J. Amer. Chem. Soc., 94, 4529 (1972).

 

C. L. Honeybourne and P. Burchill, Inorg. Nucl. Chem. Lett., 10,
715 (1974).

 

D. P. Riley, P. H. Merrell, J. A. Stone and D. H. Busch, Inorg.
Chem.,_14, 490 (1975).

J. J. Watkins and A. L. Balch, Inorg. Chem., 14, 2720 (1975).

 

L. F. Lindoy and D. H. Busch in "Preparative Inorganic Reactions,"
Vol. 6, W. L. Jolly, ed., Wiley-Interscience, New York, 1971,
pp. 1-61.

F. Basolo, B. M. Hoffman and J. A. Ibers, Accounts Chem. Res.,
8, 384 (1975).

 

J. P. Collman, R. R. Gagne, C. A. Reed, T. R. Halbert, G. Lang
and W. T. Robinson, J. Amer. Chem. Soc., 27, 1427 (1975).

D. H. Busch, Rec. Chem. Progr., 25, 107 (1964).

 

 

D. H. Busch, Helv. Chim. Acta, Fasc. Extraord. A. Werner, 50,
174 (1967).

N. F. Curtis, Coord. Chem. Rev., 3, 3 (1968).

 

0. St. 0. Black and A. J. Hartshorn, Coord. Chem. Revs., g,
219 (1972-73).

 

J. J. Christensen, 0. J. Eatough and R. M. Izatt, Chem. Rev.,
74, 351 (1974).

 

83

 

 

 

17.

18.

19.
20.

21.

22.
23.
24.
25.

26.
27.
28.
29.
30.

31.
32.
33.

34.

35.
36.

84

C. J. Hipp. L. F. Linday and D. H. Busch, Inorg. Chem., 11, 1989
(1972).

D. D. Watkins, Jr., D. P. Riley, J. A. Stone and D. H. Busch,
Inorg. Chem., 15, 387 (1976).

F. Wagner and E. K. Barefield, Inorg. Chem., 15, 408 (1976).

 

S. C. Tang, 5. Koch, G. N. Weinstein, R. W. Lane and R. H. Holm,
Inorg, Chem., 46’ 2589 (1973).

 

R. H. Prince, 0. A. Stotter and P. R. Woolley, Inorg. Chem. Acta,
2, 51 (1974).

 

. L. Honeybourne, Inorg. Nucl. Chem. Lett., 11, 191 (1975).

 

. H. Elfring, Jr. and N. J. Rose, Inorg. Chem., 14, 2759 (1976).

 

 

. M. Tait and D. H. Busch, Inorg, Chem., 15, 197 (1976).

 

C

W

A

D. K. Cabbiness and D. W. Margerum, J. Amer. Chem. Soc., 91,
6540 (1969).
F

M

D

J

M

 

. P. Hinz and D. W. Margerum, Inorg. Chem., 48’ 2941 (1974).
. Kodama and E. Kimura, J. Chem. Soc. Chem. Commun., 326 (1975).

 

. C. Olson and J. Vasilevskis, Inorg. Chem., 8, 1611 (1969).
. Vasilevskis and D. C. Olson, Inorg. Chem., 19, 1228 (1971).

 

(. O.)Kestner and A. L. Alfred, J. Amer. Chem. Soc., 94, 7189
1972 .

 

E. K. Barefield and M. T. Mocella, Inorg. Chem., 12, 2829 (1973).
D. C. Olson and J. Vasilevskis, Inorg. Chem., 10, 463 (1971).

 

E. G. Jfiger, 2. Chem., 8, 30, 392, 470 (1968); Z. Anorg. Allg.
Chem., 364, 177 (1969).

For details on attempts in this laboratory to form macrocyclic
compounds from VII, see Appendix A.

 

L. A. Funke and G. A. Melson, Inorg. Chem., 14, 306 (1975).

 

L. Wolf and E. G. Jfiger, Z. Anorg. Allg. Chem., 346, 76 (1966).

 

37.

38.
39.
40.
41.

42.
43.

44.
45.

46.

47.
48.
49.
50.
51.

52.

53.

54.
55.

56.

85

Varian A56/D Technical Manual, Publication #87-202-006 8168,
"V-4341/V-6057 Variable Temperature Accessory," pp. 15-19,
Varian Analytical Instrument Division, Park Ridge, Illinois.

J. L. Dye and V. A. Nicely, J. Chem. Educ., 44, 443 (1971).
Spectrum was obtained as described in the Experimental Section.

C. J. Hipp and D. H. Busch, Inorg. Chem., 74, 894 (1973).

 

J. R. Hutchinson, J. G. Gordon, II, and R. H. Holm, Inorg. Chem.,
74, 1004 (1971).

R. C. Fay, private communication.

T. J. Hurley, M. A. Robinson and S. I. Trotz, Inorg. Chem., 4,
389 (1967).

 

E. L. Blinn and D. H. Busch, Inorg. Chem., 7, 821 (1968).

E. L.)Blinn and D. H. Busch, J. Amer. Chem. Soc., 44, 4280
1968 .

J. A. Burke, Jr. and S. E. Campbell, J. Inorg. Nucl. Chem., 44,
1163 (1971).

 

 

J. C. Shoup and J. A. Burke, Jr., Inorg. Chem., 44, 1851 (1973).
G. A. Melson and D. H. Busch, J. Amer. Chem. Soc., 47, 1706 (1965).

 

S. G. McGeachin, Canadian J. Chem., 44, 2323 (1966).

 

C. M. Kerwin and G. A. Melson, Inorg. Chem., 47, 726 (1972).

 

 

?. Hoggood and D. L. Leussing, J. Amer. Chem. Soc., 27, 3740
1969 .

R. S. McQuate and D. L. Leussing, J. Amer. Chem. Soc., 27,
5117 (1975).

 

A. A. Frost and R. G. Pearson, Kinetics and Mechanism, John W.
Wiley and Sons, Inc., New York, 1961. PP. 147-150.

D. V. Stynes, Inorg. Chem., 44, 453 (1975).

 

K. J. Laidler, Chemical Kinetics, McGraw-Hill Book Company,
1965, PP. 225-228.

 

G. M. Shalhoub, Ph.D. Thesis, Michigan State University, 1976.

 

 

 

57.

58.

59.
60.

86

E. J. Olszewski, J. J. Boucher, R. W. Oehmke, J. C. Bailar, Jr.
and D. F. Martin, Inorg. Chem., 4, 661 (1963).

P. J. McCarthy, R. J. Hovey, K. Ueno and A. E. Martell, J. Amer.
Chem. Soc., 77, 5820 (1955).

C. M. Kerwin, Ph.D. Thesis, Michigan State University, 1972, p. 24.

L. F. Lindoy, H. C. Lip and W. E. Moody, J. Chem. Soc. (Dalton),
44 (1974).