होम Transition Metal Chemistry Reductiveversuscoupling pathways in the reactions of nickel and copper vapours with the...

Reductiveversuscoupling pathways in the reactions of nickel and copper vapours with the mono-halobenzenes

, , ,
यह पुस्तक आपको कितनी अच्छी लगी?
फ़ाइल की गुणवत्ता क्या है?
पुस्तक की गुणवत्ता का मूल्यांकन करने के लिए यह पुस्तक डाउनलोड करें
डाउनलोड की गई फ़ाइलों की गुणवत्ता क्या है?
खंड:
15
भाषा:
english
पृष्ठ:
8
DOI:
10.1007/bf01061941
Date:
August, 1990
फ़ाइल:
PDF, 924 KB
0 comments
 

To post a review, please sign in or sign up
आप पुस्तक समीक्षा लिख सकते हैं और अपना अनुभव साझा कर सकते हैं. पढ़ूी हुई पुस्तकों के बारे में आपकी राय जानने में अन्य पाठकों को दिलचस्पी होगी. भले ही आपको किताब पसंद हो या न हो, अगर आप इसके बारे में ईमानदारी से और विस्तार से बताएँगे, तो लोग अपने लिए नई रुचिकर पुस्तकें खोज पाएँगे.
Transition Met. Chem., 15, 317-324 (1990)

Reaction of nickel and copper vapours with halobenzenes

317

Reductive v e r s u s coupling pathways in the reactions of nickel and
copper vapours with the mono-halobenzenes
Scott E. DougLass, Scott T. Massey, Susan G. Woolard and Robert W. Zoellner*

Department of Chemistry, Northern Arizona University, Flagstaff, Arizona 86011-5698, USA
Summary
The reactions of nickel and copper vapours with
mono-halobenzenes to form benzene and biphenyl are
reported. Benzene formation apparently proceeds through
the intermediacy of the phenyl radical, which is produced
when metal clusters abstract halogen from the halobenzene. Biphenyl formation proceeds via an oxidative
insertion/disproportionation/reductive elimination sequence, and requires that metal atoms, rather than metal
clusters, initiate the reaction sequence. The concentration
of the metal in the matrix determines the pathway of the
reaction. Cluster formation and benzene production are
favoured by high metal concentrations, while low metal
concentrations in the matrix favour biphenyl production.
Copper vapours are anomalously less reactive than nickel
vapours, even though copper reagents are used in the
Ullmann reaction to produce biaryls through the
coupling of haloarenes. This result appears to be due to
the preferential formation of atoms or very small clusters
over large clusters or crystallites under the conditions of
the metal vapour synthesis technique.
Introduction
The heterogeneous Utlmann reaction (1) is a widely used
method for the coupling of haloarenes to produce many
symmetrical and unsymmetrical biaryls through the use
of copper powder as a reagent ~2). Relatively high
temperatures are required for the traditional Ullmann
reaction (up to 300~ in refluxing solvent) ~3). However,
milder conditions (as low as 0 ~C) are possible with a
copper reagent prepared by the reduction of copper(I)
salts with lithium or potassium naphthalenide ~4).
The mechanism for the heterogeneous Ullmann
reaction is thought to ; be that depicted in Figure la ~3-6).
There are apparently no free radical intermediates
involved in the process (7' 8, 9). The Ullmann reaction can
also be carried out under homogeneous conditions using
soluble copper(I) salts; the mechanism of the homogeneous
reaction is similar (Figure lb) (8' lo).
The coupling of haloarenes to biaryls can also be
accomplished using activated nickel powders (11) or
soluble nickel complexes (12'13). The mechanism of the
heterogeneous reaction (11) is thought to be that depicted
in Figure 2a, where a classical oxidative addition/disproportionation/reductive elimination sequence occurs.
In the homogeneous reactions, however, detailed studies
led to two alternative mechanisms, each with significant
differences from what is thought to occur in the
heterogeneous reaction.
Beginning with isolated and characterized arylnickel(II)
halide complexes of phosphines, Tsou and Kochi (12)
concluded that phosphine complexes of nickel(I) and
arylnickel(III) halides were reactive intermediates in a

* Author to whom all correspondenceshould be directed.
0340-4285/90 $03.00 +.12

(a)
ArX 4- 2 Cu ~

>ArCu ~+ Cu~X

2ArCu r

>Ar-Ar+2Cu

ArCu ~+ ArX

(b)

~

>Ar - Ar + Cu~X

k1

ArX + C u t ~ A r C u H ~ X
k-1

ArCuHX + ArX

k2 >Ar -- Ar + culilx2

ArCu~lIX+H +

k3>Ar H + C u i l I X

Figure 1. Proposed mechanisms for the coupling of haloarenes
in the Ullmann reaction using (a) heterogeneous copper(0)
reagents and (b) soluble copper(I) triflate salts. In (a), X = CI,
Br, or I; in (b), X = Br. The soluble copper species present [such
as ArCu~in (a) and ArCumX in (b)] almost certainly have other
ligands, including solvent, associated with the complex. The
exact nature and number of these associated species are
unknown.

radical chain process which produced biaryls from
haloarenes (Figure 2b). Free radicals were conclusively
eliminated as possible intermediates in their reaction
system. Conversely, in the work of Semmelhack and
co.workers~l 3), a second oxidative insertion (following an
oxidative insertion on the nickel(0) starting material to
form a nickel(II) intermediate) to form a nickel(IV)
intermediate is proposed. This nickel(IV) species then
reductively eliminates the biaryl (Figure 2c).
The technique of Metal Vapour Synthesis (MVS) (14' 15)
can provide a method for the preparation of activated
metal powders or slurries (aluminium, indium, zinc,
cadmium, tin, lead, and nickel). These slurries are useful
in a variety of reactions with haloalkanes and haloarenes (16). In some cases, aryl- and alkyl-complexes of
metals may be isolated from these reactions. Unfortunately, reactions of haloarenes with nickel slurries do not
result in high yields of biaryls.
In contrast with the apparently low reactivity of the
nickel slurries produced by MVS toward haloarenes,
certain haloarenes do react readily with nickel vapours
after co-condensation at - 1 9 6 ~
to yield both the
insertion product ArNinX and the disproportionation
product ArNiHAr. When the haloarene is chloro-,
bromo- or iodo-pentafluorobenzene, both of these
9 1990 Chapman and Hall Ltd

318

S. E. Douglass et al.

Transition Met. Chem., 15, 317-324 (1990)

(a)

Experimental
NiO+Ar X

L >ArNi[tXLn

2ArNillXLn

>Ar2Nil%~ + NiHX2

Ar2NiJlL n

) Ar - Ar + Ni ~

(b)
NiJXL3 + ArX

ArNilHX2 L + 2 L

ArNiItIX2L 4- ArNiEIXL2
Ar2Nill~XL

Metal vapour co-condensation/addition ( CCA ) reactions
A schematic diagram of the metal vapour co-condensation/addition (CCA) reactor is found in Figure 3. The
metal (nickel, 99.9%, copper, 99.999%, Cerac, Inc.) to be
vaporized is placed in an integral alumina-coated
tungsten crucible (GTE Sylvania Emissive Products)
connected to water-cooled copper electrodes. The reactor
is evacuated under dynamic vacuum to approximately
1 x 10 . 4 torr, and the external wall of the reactor bottom
is cooled to - 196 ~ C with liquid nitrogen. At this point,
previously degassed (by repeated freeze/pump/thaw
cycles), dry, benzene-free toluene (Fisher ACS-certified)
is inlet as a vapour through the shower-head, and the

) Ar2NiltIXL + NiI"X2L2
k

K

>Ar -- Ar -I- NilXL3

(c)

B IG
Ni~

+ ArX

) ArNiiIXL2 + L2
V^

ArNilIXL2 + ArX

Ar2NiwX2

>Ar2Nilvx2 + L 2

L.

C

G

VA

C

> A r - - A r + Nillx2

Figure 2. Proposed mechanisms for the coupling of haloarenes
using (a) heterogeneous nickel(0) reagents, (b) soluble,
pre-formed arylnickel(II) halide complexes, and (c) soluble,
nickel(0) complexes of cyclooctadiene or phosphines. In (a),
L = solvent (the exact nature and number of associated solvent
molecules are unknown), while in (b), L = phosphine ligand,
and in (c), L2 = cyclooctadiene or two phosphine ligands.
Except for the explicitly isolated and characterized complexes
in (b) and (c), i.e., ArNiI~XL2 and Ni~
respectively, the
numbers of ligands have been arbitrarily assigned assuming
four-coordinate complexes.

intermediates can be trapped as the phosphine adducts,
and the latter disproportionation product can be isolated
as the t/6-toluene adduct (17).
The focal point of the reactions of metal vapours with
haloarenes has always been the production of isolable
metal complexes of the haloarenes. When no such
complexes could be isolated, the organic products of the
reactions, if any existed, were ignored (18). We have begun
a systematic investigation of MVS reactions with
haloarenes which focusses specifically on the organic
products of the reactions. In so doing, we hope to better
delineate the intrinsic reactivity of the metal vapours with
haloarenes and to study the reaction mechanisms
involved when metal vapours interact with haloarenes.
Further, we shall compare our results with those for the
classical Ullmann reaction so as to determine how metal
vapours are similar to or different than classical methods
for the coupling of haloarenes. Herein we report our
results for nickel and copper vapour reactions with the
mono-halobenzenes.

D
F
E

~oo~
~oo~

eg.3

U
Figure3. Schematic drawing of the metal vapour co-condensation/addition (CCA) reactor (not to scale). Letters on the
figure refer to the following: A, flask containing degassed
toluene; B, outlet to vacuum system; C, water-cooled copper
electrodes; D, reactor top; E, 3000 ml reactor bottom; F, external
liquid nitrogen coolant level; G, high vacuum valves; H,
shower-head for dispersal of toluene; I, crucible containing
nickel; J, magnetic stir bar; K, flask containing toluene solution
of the halobenzene with L, inlet isolated fi'om contact with
matrix and crucible.

Transition Met. Chem., 15, 317-324 (1990)

Reaction of nickel and copper vapours with halobenzenes

toluene begins to condense and freeze on the internal
wall of the reactor bottom.
During the initial inlet of toluene, electric power
(220 V/50 A input to a variable transformer, output to a
100~ duty cycle step-down transformer at 0 10 V/50 A)
to the crucible is slowly (0.1-0.2 V/min) increased until
the metal in the crucible melts. Slight adjustments in
the power supply during the time the metal is molten are
performed to permit an appropriate rate of vaporization
of the metal from the crucible. These adjustments are
made so as to allow the co-condensation of the metal
with the toluene to produce a matrix which appears to
contain a toluene to metal ratio appropriate to the
reaction being carried out. (Such visual observations of
the matrix can only be approximate but, with experience,
matrices with relatively high or tow concentrations of
metal can be produced with fair reproducibility.) The
co-condensation is allowed to proceed for up to three
hours under dynamic vacuum. The frozen matrix of
toluene and the metal is initially orange to red-brown,
but darkens as the co-condensation proceeds.
At the end of the co-condensation time, the electric
power to the crucible is slowly (0.3-0.5 V/min) decreased
until zero power is reached, at which time the toluene
inlet is stopped. The reactor is then isolated from the
dynamic vacuum, and the previously degassed solution
of the halobenzene (chlorobenzene, Baker analysed;
bromobenzene, Kodak, 99~) dissolved in toluene is inlet,
as a liquid, into the reactor through a tube so as to
prevent contact with the frozen matrix. The dewar of
liquid nitrogen is then removed, and the reactor is allowed
to warm to ambient temperature under static vacuum.
The solution of halobenzene in toluene is intimately
mixed with the frozen matrix as the matrix melts using
a magnetic stir-bar placed in the bottom of the reactor
before the reaction was begun.
When the reactor and the toluene solution (which now
contains metal and metal halide particles as a slurry)
containing any unreacted halobenzene and any organic
products reach ambient temperature (20 to 30 rain), the
solution/slurry is siphoned out of the reactor under an
atmosphere of dry, deoxygenated nitrogen or argon, and
is filtered through a "D" porosity glass frit into a standard
Schlenk tube to remove metal and metal halide particles.
[CAUTION: The finely divided metal and metal
halide particles produced in the CCA reactions described
above and in the DCC reactions described below are

319

often extremely pyrophoric. Rapid exposure to air or
oxygen may cause spontaneous ignition. It is advisable
to handle any finely divided metal or metal halide
particles under an atmosphere of inert gas and, if
exposure to air is required, to allow such exposure only
under well-controlled conditions.]
Table 1 gives details of the specific reaction parameters
for each of the CCA reactions, which were carried out
only for nickel with chlorobenzene and bromobenzene.

Metal vapour direct co-condensation (DCC) reactions
The procedure for a metal vapour direct co-condensation
(DCC) reaction uses a similar apparatus (Figure 4) and
procedure to that used in the CCA reactions. However,
instead of co-condensing toluene with the metal, the
previously degassed halobenzene (fluorobenzene, Aldrich
99~; chlorobenzene and bromobenzene, as described
above; and iodobenzene, Aldrich 99~o) is co-condensed
directly with the metal. In these cases, the matrix is dark
brown almost from the beginning of metal vaporization.
At the end of the reaction time, after the power to the
crucible is at zero, the reactor is allowed to warm to
ambient temperature under static vacuum. When the
reactor has reached ambient temperature, dry, oxygenfree toluene is added to the reactor, by syringe, under an
atmosphere of dry, deoxygenated nitrogen or argon. The
resulting toluene solution of halobenzene and organic
products, containing the metal and metal halide particles
as a slurry, is siphoned out of the reactor and filtered, as
described above.
Tables 2 (nickel) and 3 (copper) give the specific
reaction parameters for the D D C reactions, which were
carried out with each of the halobenzenes.

Gas chromatography
The toluene solutions isolated from the metal vapour
reactions are analytically diluted to a standard volume
(either 50 ml, 100 ml, 200 ml, or 250 ml, depending upon
the volume of solution isolated from the reaction) with
benzene-free toluene. These solutions are analysed using
a Beckman Model 990 Flame-Ionization gas chromatograph, He carrier gas, and either a 10m SE-30 (for
benzene) or a 10m OV-17 (for biphenyl) column.
Standard curves to quantify the instrument response were
developed using analytically-prepared solutions of

Table 1. Experimental parameters for nickel vapour reactions with the mono-halobenzenes via co-condensation/addition.
Halobenzene

Amount of PhX added"

Nickel metal vaporizedb

Raft&
PhX:Ni

Reaction
time (min)

CCA-1

PhC1
PhC1

CCA-3

PhBr

CCA-4

PhBr

0.5049 g (8.603 mmol)
[into 20.0ml toluene]
0.8060g (13.73 mmol)
[into 25.0ml toluene]
0.4653 g (7.928 mmol)
[into 20.0 ml toluene]
0.6719 g (11.45 mmol)
[into 25.0ml toluene]

5.7:1
[21.9:1]
3.6:1
[17.1:1]
6.0:1
[23.7:1]
4.2:1
[20.5:1]

110

CCA-2

5.00 ml (49.1 mmol)
[in 25.0ml toluene]
5.00ml (49.1mmol)
[in 25.0ml toluene]
5.00ml (47.6 mmol)
[in 25.0ml toluene]
5.00 ml (47.6 mmol)
[in 25.0ml toluene]

140
120
135

"The halobenzene is added to the reactor as a solution in toluene after the co-condensation of nickel with toluene is complete (see text for
experimentaldetails), bUncorrectedfor metal vapour which becomesdeposited on the shower head or the electrodesduring the reaction and does
not reach the matrix. Value in square brackets is the volume of toluene co-condensedwith the nickel. Walue in square brackets is the ratio of
toluene to nickel.

S. E. Douglass et al.

320

Transition M e t . Chem., 15, 317-324 (1990)

G

B

I

J

v^
c

I\

G
ViC

J

, \
t
= =G

! fH

00o

\

'/

Figure 4. Schematic diagram of the metal vapour direct
co-condensation (DCC) reactor (not to scale). Letters on the
figure refer to the following: A, flask containing degassed
halobenzene; B, outlet to vacuum system; C, water-cooled
copper electrodes; D, reactor top; E, 3000 ml reactor bottom;
F, external liquid nitrogen coolant level; G, high vacuum valves;
H, shower-head for dispersal of halobenzene; I, crucible
containing nickel or copper; J, magnetic stir bar. Stippled areas
denote ground glass joints.

benzene and biphenyl in benzene-free toluene, and these
were always run before and after the analysis of a reaction
product to ensure that the instrument response was
reproducible from one analysis to the next.
Results and discussion
The only major organic products detected from the metal
vapour reactions of nickel and copper with the
mono-halobenzenes were benzene and biphenyt. The
results for the nickel vapour reactions, both for the CCA
and DCC reactions, are listed in Table 4, while the results
for the copper vapour reactions are listed in Table 5.
Co-condensation~addition reactions

Upon examination of the data in Table 4 for the nickel
vapour reactions, there is evidently a significant difference
between the CCA reactions and the DCC reactions: the

CCA reactions produced no biphenyl. This result
indicates that metal atoms (or extremely small metal
clusters) are responsible for the production of biphenyl,
and that complexation phenomena may also be
important, as is discussed below.
In a CCA reaction, nickel vapours are co-condensed
with toluene so as to presumably form the metastable
(decomposition temperature ca. - 70 ~ C (19)) bis(toluene)nickel* complex. This complex then acts as a "carrier"
for the nickel atoms, bringing the atoms, in a reactive
form, to the halobenzene (such nickel atoms, after
co-condensation with toluene, have been referred to as
"solvated nickel atoms" because of their enhanced
reactivity(22)). However, since toluene melts at
- 95 ~ C (23a), which is only 25 ~ C below the decomposition
temperature of the nickel-toluene complex, nickel cluster
formation would be expected to begin at the melting
point of toluene, or even below,** This cluster formation
may simply occur too rapidly to allow any reaction
between the nickel and the haloarene.
Even if the bis(toluene)nickel complex reaches the
solution of the haloarene in toluene at the bottom of the
reactor, competitive complexation phenomena may also
inhibit reactivity. Arene complexes of nickel (24' 25, 26) and
of cobalt (25'26) are highly labile, and readily exchange
the bound arene ligand for other arenes or a-donor
ligands. However, arenes containing one or more
fluoro- or trifluoromethyl-substituents will not exchange
with complexed benzene or complexed arenes containing
electron-donating substituents. Further, complexed
benzene or complexed arenes containing electrondonating substituents, in exchange reactions with arenes
containing electron-donating substituents, appear to
reach an equilibrium with the exchanging arene such that
the complex which contains the arene with the substituent
with the greater electron-donating capability is favoured
and found to be present in excess. Thus, even if the bis(toluene)nickel complex reaches the halobenzene intact,
the halobenzene may not exchange with the (presumably
loosely bound) toluene, and reaction may not occur to
produce biphenyl.
Since the CCA reactions do produce benzene, the
above analysis apparently precludes the possibility of
benzene formation through complexation of the
halobenzene to nickel. Benzene formation in the CCA
reactions may be occurring by some other mechanism
whereby nickel clusters or crystallites (of unknown size)
react with the halobenzene to abstract halogen. This may
or may not occur with formation of a phenyt radical, but
the phenyl group would, after abstraction of the halogen
by the nickel clusters or crystallites, pick up a proton
adventitiously. The source of this proton, at present, is
unknown.
Direct co-condensation reactions

Figure 5 depicts the ratios of mmol(benzene or biphenyl)
to mmol(nickel vaporized) for the nickel and copper
* This is formally a twenty-two electron complex if both toluene
ligands are bound in an ~6-fashion.Such a formulationis extremely
doubtful, however;the exactmode of binding of the tolueneligands,
and eventheirexactnumber,is the subjectofmuchcontroversy(2~z1).
** Such clusteringand decompositionat the melting point of toluene
is also indicated by the appearance of the matrix as melting occurs.
The matrix rapidlyturns black and exhibitsmetal particleformation
as melting occurs.

Transition Met. Chem., 15, 3 1 7 - 3 2 4 (1990)

Reaction of nickel a n d copper v a p o u r s with h a l o b e n z e n e s

321

Table 2. Experimental parameters for nickel vapour reactions with the mono-halobenzenes via direct co-condensation.
Halobenzene

Amount of PhX inlet

Nickel metal vaporized a

Ratio
PhX: Ni

Reaction
time (rain)

DCC-01
DCC-02
DCC-03
DCC-04
DCC-05
DCC-06
DCC-07
DCC-08
DCC-09
DCC-10
DCC-11
DCC-12
DCC- 13
DCC-14
DCC-15
DCC- 16

3.12ml (33.2mmol)
3.3~7ml (35.8 mmol)
3.62ml (38.5mmol)
25.7ml (253mmol)
25.7 ml (253 mmol)
37.8 ml (372 retool)
19.3 ml (184 mmol)
22.0ml (210 mmol)
23.1ml (220mmol)
24.0ml (228 mmol)
24.9 ml (237 retool)
26.8 ml (256 mmol)
29.4 ml (280 mmol)
7.85 ml (70.5 mmol)
10.1 ml (90.6mmol)
12.6 ml (113 mmol)

0.0768g
0.4243 g
0.3011g
0.0173g
0.6293 g
0.8587 g
0.2761 g
0.5730 g
0.5857 g
0.1978 g
0.1562 g
0.3669 g
0.2015 g
0.2497 g
0.0506g
0.2856 g

25.3:1
5.0:1
7.5:1
861:1
23.6:1
25.4:1
39.1:1
21.5:1
22.0:1
67.7:1
89.1 : 1
41.0:1
81.6:1
16.6:1
105 :I
23.3:1

80
95
110
190
170
120
80
140
135
125
110
120
110
135
65
110

PhF
PhF
PhF
PhCI
PhC1
PhC1
PhBr
PhBr
PhBr
PhBr
PhBr
PhBr
PhBr
Phi
Phi
Phi

(1.31mmol)
(7.230 mmol)
(5.130mmol)
(0.294mmol)
(10.72 mmol)
(14.63 mmol)
(4.704 mmol)
(9.763 mmol)
(9.980mmot)
(3.370mmol)
(2.661 mmol)
(6.251mmol)
(3.433 mmol)
(4.255 mmol)
(0.862mmol)
(4.866 mmol)

aUncorrected for metal vapour which becomes deposited on the shower head or the electrodes during the reaction and does not
reach the matrix.

Table 3. Experimental parameters for copper vapour reactions with the mono-halobenzenes via direct co-condensation.
Halobenzene

Amount of PhX inlet

Copper metal vaporize&

Ratio
PhX: Cu

Reaction
time (rain)

DCC-17
DCC-18
DCC-19
DCC-20
DCC-21
DCC-22
DCC-23
DCC-24

9.19 ml (97.9 mmol)
11.3 ml (120 mmol)
3.50 ml (34.4 mmol)
24.7 ml (243 mmol)
29.7 ml (292 mmol)
22.5ml (214mmol)
25.1 ml (239 mmol)
10.6ml (94.7 mmol)

0.0789 g
0.0257 g
0.1136 g
0.026 t g
0.1696 g
0.1124g
0.5457 g
0.0414g

79.0:1
29.7:1
19.2:1
591 : 1
109 : I
121:1
27.8:1
145:1

80
90
85
80
80
70
80
80

PhF
PhF
PhCI
PhC1
PhC1
PhBr
PhBr
Phi

(1.24 mmoI)
(0.404 mmol)
(1.788 mmol)
(0.411 mmol)
(2.669 mmoI)
(1.769mmol)
(8.587 mmoI)
(0.651mmoi)

aUncorrected for metal vapour which becomes deposited on the shower head or the electrodes during the reaction and does not
reach the matrix.
Table 4. Products from the nickel vapour reactions with mono-halobenzenes.
Halobenzene

CCA-1
CCA-2
CCA-3
CCA-4
DCC-01
DCC-02
DCC-03
DCC-04
DCC-05
DCC-06
DCC-07
DCC-08
DCC-09
DCC-10
DCC-11
DCC-12
DCC-13
DCC-14
DCC-15
DCC-16

PhCI
PhC1
PhBr
PhBr
PhF
PhF
PhF
PhCI
PhCl
PhC1
PhBr
PhBr
PhBr
PhBr
PhBr
PhBr
PhBr
Phi
Phi
Phi

Products (mmol)

Ratio

benzene

biphenyl

Phil: Ni

Phil: PhX

PhPh: Ni

PhPh:PhX

1.46
0.844
0.375
1.76
b
b
b
0.358
0.640
4.80
3.83
1.42
2.04
0.987
0.648
0.800
0.288
b
0.405
0.568

a
a
a
a
a
.
a
1.25
"
0.756
0.566
1.93
1.84
3.43
~
c
4.28
a
0.632
0.639

0.170:1
0.0615:1
0.0437:1
0.154:1
_
_
_
1.22:1
0.0597:1
0.328:1
0.814:1
0.145:1
0.204:1
0.293:1
0.243:1
0.128:1
0.0839:1
_
0.470:1
0.117:1

0.0297: !
0.0172:1
0.00788: t
0.0370:1
_

-

_
0.00142:1
0.00253:1
0.0129:1
0.0208:1
0.00676:1
0.00927:1
0.00433:1
0.00273:1
0.00312:1
0.00103:1

_
_
_
4.25:1
0.0517:I
0.120:1
0.198:1
0.184:1
1.02:1
~
c
1.25:1

0.00447:i
0.00503:1

0.733:1
0.131:!

0.00698:1
0.00565:1

_
_
_
0.00494:1
0.00203:1
0.00308:1
0.00919:1
0.00836: I
0.0150:1
~
0.0153:1

aNo biphenyl detected at detection limits of less than 0.0005mmol total biphenyl in reaction product mixture, bNo benzene detected at detection
limits of less than 0.001 mmol total benzene in reaction product mixture. ~
for these two determinations of biphenyl were considered unreliable
due to instrument malfunction and are not reported.

322

S.E. Douglass et al.

Transition Met. Chem., 15, 317-324 (1990)

Table 5. Products from the copper vapour reactions with mono-halobenzenes.
Halobenzene
DCC-17 PhF
DCC-18 PhF
DCC-19 PhC1
DCC-20 PhC1
DCC-21 PhC1
DCC-22 PhBr
DCC-23 PhBr
DCC-24 Phi

Products (retool)
benzene
biphenyl

Ratio
Phil: Cu

a
a
0.347
0.1 i6
0.678
1.94
3.06
0.303

_

b
b
b
b
b
0.285
1.18
0.0822

Phil: PhX
_
0.0101:1
0.00048 : 1
0.00232:1
0.00907:1
0.0128:1
0.00320:1

0.194:1
0.282:1
0.254:1
1.10:1
0.356: I
0.465:1

PhPh: Cu

PhPh: PhX

_
_
--

_
_
---0.00133:1
0.00494:1
0.00087:1

-0.161:1
0.134:1
0.126:1

aNO benzene detected at detection limits of less than 0.001 mmol total benzene in reaction product mixture, bNo biphenyl detected at detection
limits of less than 0.0005 mmol total biphenyl in reaction product mixture.

vapour reactions with the mono-halobenzenes. For both
nickel and copper, no reaction was observed with
fluorobenzene. With regard to the production of benzene,
the order of reactivity was PhCI > PhBr ~ Phi for nickel,
and PhBr > Phi > PhC1 for copper. Nickel reacted with
the halobenzenes to produce biphenyl in the order
PhC1 > PhBr > Phi, while copper reacted with PhBr and
Phi at about the same (relatively low) level and did not
react with PhC1 to produce any biphenyl. Generally,
copper was more reactive than nickel toward the
production of benzene, while nickel was more reactive
than copper toward the production of biphenyl.
During these reactions, since P h - - X bonds must be
broken and (presumably) nickel-halide bonds must be
formed, a comparison of bond dissociation energies with
product formation was required. Table 6 lists selected
bond dissociation energies for the relevant bonds.
Apparently, the strength of the P h - - F bond is too great,
and the bond energy difference between the P h - - F bond
and either the N i - - F or the C u - - F bond is too large,
for any reaction to occur with fluorobenzene.
With regard to benzene production by nickel, the bond
energy differences A(AHf) are in the order PhC1 >
PhBr = Phi, which is the opposite trend expected for a
reactivity increase with a decrease in bond energy. In the
case of copper, however, bromobenzene is the most
reactive halobenzene, and bromobenzene is the halobenzene with the lowest bond energy difference.

Biphenyl production by nickel is again opposite to that
predicted by bond dissociation energies and bond energy
differences. Production of biphenyl by copper did not
exhibit any marked trends, nor was production of
biphenyl very high in comparison to nickel. The data
suggest that, at least for nickel and possibly for copper,
neither the breaking of the P h - - X bond nor the
formation of the M - - X bond is intrinsically related to
the reactivity of the halobenzene with metal vapours
under these conditions.
In order to obtain a more detailed understanding of
the reactions occurring in the D C C reactions, a study
was carried out in which the concentration of nickel
co-condensed with a matrix of bromobenzene was varied.
The results of this study are depicted in Figure 6, where
the ratio of benzene produced to nickel vaporized is
plotted, and in Figure 7, where a similar plot for biphenyl
production is shown.
Figures 6 and 7 offer a distinct contrast in the
production of benzene as compared to the production of
biphenyl. As is readily apparent, there is no correlation
to be found in Figure 6, but a high degree of correlation
is noted in Figure 7. Thus, Figure 7 indicates that as the
ratio of bromobenzene to nickel increases, more biphenyl
is produced per nickel atom. As the ratio of
bromobenzene to nickel increases, the excess bromobenzene may prevent nickel atoms from recombining to
form nickel clusters or crystallites and increase the chance

2.5

2.0

-c
1,oo

075

2

=o

1.o
o

0.5

0.50

0.25

.9

.9

*~

O.0

PhH

PhPh
PhF

PhH

PhPh
PhCI

PhH

PhPh
PhBr

PhH

PhPh
Ph i

ooo

10

20

30

40

50

60

(0

80

ratio of bromobenzene t o nickel vaporized

Figure 5. Comparison of benzene (Phil) and biphenyl (PhPh)

Figure 6. Relationship of benzene production to the concen-

production in nickel and copper vapour DCC reactions with the
mono-halobenzenes (PhF, PhC1, PhBr, Phi). Open bars refer
to nickel vapour reactions, solid bars refer to copper vapour
reactions, and an 'x' indicates that no product was detected.

tration of nickel in a bromobenzene matrix. The best-fit linear
regression line through these points (not shown) has a coefficient
of determination, r 2, of 0.0196, indicating that the points cannot
be fitted well to a line.

Reaction of nickel and copper vapours with halobenzenes

Transition Met. Chem., 15, 317-324 (1990)

323

13

~

~,_

~,

~

o5

o

~))_

o.c
20

30

40

50

60

70

BO

90

100

X

"P

<~:>-x
Ma t o m s ~176176176176
I
>
M

ratioofbromobenzene vaporized
~o n~ckel

Figure 7. Relationship of biphenyl production to the concentration of nickel in a bromobenzene matrix. The best-fit linear
regression line through these points, as drawn in the figure, has
a coefficient of determination, r 2, of 0.901.
that nickel atoms will complex with the rc electrons of
the bromobenzene.
The co-condensation event is complex. Figure 8
illustrates a hypothetical cross section of the matrix as
it forms at - 1 9 6 ~
Freezing is not instantaneous:
during the time a nickel atom is in the "gas-liquid zone"
or the "liquid-solid zone", the atom can "skate" through
the semi-solid, solidifying material with ease, contacting
many other atoms or molecules as the atom does so. The
recombination of nickel atoms to nickel clusters or
crystallites is a low energy process (~v), and this
recombination reaction can only be minimized under
conditions in which the reactant ligand (in this case
bromobenzene) is present in sufficient excess.
This, Figure 7 indicates that nickel atoms are necessary
in order for biphenyl to be produced, and when the
number of nickel atoms is increased under conditions
where cluster formation is prevented (or the concentration
of nickel in the matrix is decreased, which amounts to
the same change of conditions), the amount of biphenyl
produced per atom of nickel will be increased.

F
R
0
Z
E
N
Z
0
N
E

LN 2

LIQUIDSOLID
ZONE

GAS-LIQUID
ZONE

- 196o0

~lass
wal!

Temperature

increases

Figure 8. Schematic diagram of a cross-section of the low
temperature matrix in the MVS reactor during metal
vaporization/ligand co-condensation. Because of the relatively
inefficient conduction of heat through the frozen zone and the
glass reactor wall, the ligand/metal co-condensate cannot freeze
instantaneously. Metal atoms are generally free to "skate"
through the gas-liquid and liquid-solid zones, contacting many
other molecules (and atoms or clusters of atoms, if there is too
small an amount of the co-condensing ligand present) as they
d o so,

tion

M ~

?rtion
~M--X

dispr~176176176
~, ~ M ~

red~ ~tit7io
+ MX2

Figure 9. Proposed reaction scheme for the interactions of
metal vapours and mono-halobenzenes to produce benzene and
biphenyl. See text for details.
Benzene production is quite different from that of
biphenyl. Since there is no correlation with the
concentration of nickel in the matrix, nickel atoms are
apparently not required for benzene production. Further,
the wide variation in the amount of benzene produced
is indicative of reactions in which free radicals are
involved(~8). Thus, it appears that the production of
benzene in DCC reactions, as was noted in CCA
reactions earlier, requires nickel clusters or crystallites
and appears to involve the intermediacy of phenyl
radicals.
Reaction scheme
From our data, and the preceding discussions, the
reaction scheme in Figure 9 appears to represent our
results with moderate precision. The first step apparently
involves complexation of the metal with the halobenzene.
When the halobenzene is present in an insufficient excess
(or when, as in a CCA reaction, the halobenzene contacts
the metal only after the matrix begins to warm), metal
clustering begins to occur because of the inefficient
complexation process in this system of limited halobenzene. These clusters are responsible for the abstraction
of the halogen from the haloarene, and the phenyl radical
thus formed picks up a proton adventitiously during the
warming process.
When the halobenzene is present in sufficient excess,
complexation is efficient and prevents metal clustering.
An oxidative addition of the metal to the P h - - X bond
then occurs, followed by a disproportionation step to
form the biaryl species. The biaryl then reductively
eliminates biphenyl.
Among the questions which remain involves the
problem of nickel apparently being more reactive than
copper in the coupling of the mono-halobenzenes to
produce biphenyl in the DCC reactions. The answer may
lie in the ratios of halobenzene to copper vaporized in
Table 3. In most cases, the ratios are quite high, favouring
copper complexation to the halobenzene and preventing
cluster or crystallite formation. In the Ullmann reaction,
highly dispersed, freshly prepared copper powder is the
reagent of choice (2,3'4). Thus, copper clusters or
crystallites, or the surface of bulk, polycrystalline copper,
may be necessary for the coupling reaction to occur, and
under the conditions of the DCC reactions, clusters are
simply not favoured.
Such a finding may also serve to explain the generally
higher production of benzene by copper vapours as

324

S.E. Douglass et al.

Transition Met. Chem., 15, 317-324 (1990)

Table 6. Selected bond dissociation energies.
Halobenzene

AHf (298 K)
k J/moP

Nickel
halide

AHf (298 K)
kJ/mol b

A(AHf)

Copper
halide

AHf (298 K)
kJ/mol b

A(AHf)

Ph--F
Ph--C1
Ph--Br
Ph--I

523
397
335
268

Ni--F
Ni--C1
Ni--Br
Ni--I

435
372
360
293

88
25
-25
-25

Cu--F
Cu--C1
Cu--Br
Cu--I

413
383
331
197

110
14
4
71

aReference (23b). bReference(23c).
compared to nickel vapours. According to the proposed
mechanisms in Figure 1, copper(0) does not form an
insertion product with a haloarene. Rather, at least two
copper atoms appear to be necessary in order for a
"double abstraction" of the haloarene to form ArCu I and
CuIX. If copper atoms are the only major species present
in a copper DCC reaction, only abstraction of the halogen
from the halobenzene may occur, forming Cu~X and the
phenyl radical. Although the possibility exists that the
phenyl radical can dimerize to form biphenyl, concentration effects may make the production of benzene
significantly more favourable.
Thus, a generalized reaction scheme for the reductive
versus coupling pathways in the reactions of metal
vapours with halobenzenes has been developed. However,
questions remain concerning (a) the source of hydrogen
in the reaction in which phenyl radicals are converted to
benzene, (b) the role of metal-containing intermediates
(such as A r - M n - X and Ar-M~I-Ar) in the reaction
scheme, and the relationship of these intermediates to,
among other factors, bond dissociation energies and the
rate-determining step(s) in the reaction, (c) the role of
clusters, and cluster formation processes, in copper vapour
reactions with halobenzenes and (d) the effect of electronic
factors and substituent effects on the rc-complexation step
of the reaction. Investigations are currently underway in
these regards through (a) deuterium labelling studies, (b)
E H T calculations, (c) concentration studies of copper in
halobenzene matrices and (d) studies using appropriatelysubstituted haloarenes in DCC reactions.

Acknowledgements
This work was supported in part by a Flinn Foundation
Grant of Research Corporation (to R.W.Z.), and aided
by a Grant-in-Aid of Research from Sigma Xi, the
Scientific Research Society (to S.T.M.). We wish to thank
the Office of Sponsored Research Administration and the
Organized Research Committee of Northern Arizona
University for their support.

References
~I)B.P. Mundy and M.G. Ellerd, Name Reactions and
Reagents in Organic Synthesis, Wiley-Interscience, New
York, 1988, pp. 212-213.
(2)j. March, Advanced Organic Chemistry: Reactions,
Mechanisms, and Structure, Third Edition, Wiley-Interscience, New York, 1985, pp. 597-598.
(3)M. Goshaev, O. S. Otroshchenko and A. S. Sadykov, Russ.
Chem. Rev. (Engl. Trans.), 41, 1046 (1972).

~4)G. W. Ebert and R. D. Rieke, J. Org. Chem., 53, 4482 (1988);
J. Or9. Chem., 49, 5280 (1984).
/5)p. E. Fanta, Synthesis, 9 (1974); Chem. Rev., 64, 613 (1964);
Chem. Rev., 38, 139 (1946).
f6)R. G. R. Bacon and H. A. O. Hill, Q. Rev. Chem. Soc., 19,
95 (1965).
(7)y. Mugnier and E. Laviron, J. Chem. Soc., Perkin Trans. 2,
1264 (1979).
(8)T. Cohen and I. Cristea, J. Am. Chem. Soc., 98, 748 (1976).
(9)T. Cohen and T. Poeth, J. Am. Chem. Soc., 94, 4363 (1972).
(l~ Cohen and J. G. Tirpak, Tetrahedron Lett., 143 (1975).
(11)H. Matsumoto, S.-i. Inaba and R. D. Rieke, J. Org. Chem.,
48, 840 (1983).
(12)T. T. Tsou and J. K. Kochi, J. Am. Chem. Soc., 101, 7547
(1979).
(13)M. F. Semmelhack, P. Helquist, L. D. Jones, L. Keller, L.
Mendelson, L. S. Ryono, J. G. Smith and R. D. Stauffer, J.
Am. Chem. Soc., 103, 6460 (1981).
/~4)j. R. Blackborow and D. Young, Reactivity and Structure
Concepts in Organic Chemistry. 9. Metal Vapour Synthesis
in Oroanometallic Chemistry, Springer-Verlag, Berlin, 1979.
/aS)K. J. Klabunde, Chemistry of Free Atoms and Particles,
Academic Press, New York, 1979.
(16)K. J. Klabunde and T. O. Murdock, J. Org. Chem., 44, 3901
(1979).
/17)K.J. Klabunde, B.B. Anderson, M. Bader and L.J.
Radonovich, J. Am. Chem. Soc., 100, 1313 (1979).
/18)K. J. Klabunde, private communication.
/19)Reference (15), page 105.
(2~ F. Efner, D.E. Tevault, W.B. Fox and R.R.
Smardzewski, J. Organomet. Chem., 146, 45 (1978).
12~)S. C. Davis, Ph.D. Dissertation, "Clustering of metal atoms
in alkane media", The University of North Dakota, Grand
Forks, North Dakota, 1980.
(22)K. J. Klabunde, H. F. Efner, T. O. Murdock and R. Ropple,
J. Am. Chem. Soc., 98, 1021 (1976).
(23)R. C. Weast, M.J. Astle and W.H. Beyer (Eds.), CRC
Handbook of Chemistry and Physics, 66th Edn. CRC Press,
Boca Raton, Florida, 1985. la)Page C-518. (b)Page F-193.
(~ F-174 to F-184.
(24)R. G. Gastinger, B. B. Anderson and K. J. Klabunde, J. Am.
Chem. Soc., 102, 4959 (1980).
(2S)M. M. Brezinski, M.S. Thesis, "The chemistry of bis(pentafluorophenyl) (tl6-toluene)cobalt(II): Arene displacements and exchange reactions", Kansas State University,
Manhattan, Kansas, 1982.
(26)M. M. Brezinski and K. J. Klabunde, Organometallics, 2,
1116 (1983).
(27)Reference (15), page 68.
/2S)D. Katakis and G. Gordon, Mechanisms of Inorganic
Reactions, Wiley-Interscienee, New York, 1987.

(Received 19 July 1989)

T M C 2129