close

Вход

Забыли?

вход по аккаунту

?

Metal-Olefin and -Acetylene Bonding in Complexes.

код для вставкиСкачать
Metal-Olefm and -Acetylene Bonding in Complexes
By Frank R. Hartley[*]
The molecular orbital (MO) and valence bond (VB) descriptions of the bonding in metalolefin and -acetylene complexes are examined in relation to their ability to explain the known
properties of these complexes. It is concluded that whereas there are difficulties with the valence bond description, the molecular orbital description can satisfactorily explain all the
known properties of these complexes and has the necessary versatility to enable it to be
readily applied to any series of metal-olefin or -acetylene complexes. It is further shown that
olefin and acetylene complexes can be usefully divided into two main classes: those that
resemble the platinum(1r)-complexes(class S), and are either square-planar or octahedral with
essentially free rotation about the metal-olefin bond; and those that resemble platinum(0)complexes (class T) and are either trigonal planar or trigonal bipyramidal and in which there
is no rotation about the metal-olefin or -acetylene bond.
1. Introduction
The metal-olefin bond is usually described in molecular
orbital terms using the theory which Dewad'] put forward
for silver(r)-olefin complexes and which Chatt and Duncansodz1subsequently extended to include platinum(I1)-olefin
complexes. The bonding, according to the Dewar-ChattDuncanson model essentially involves a o-bond, from the
filled prr-orbital of the olefin to an empty hybrid orbital on
the metal, complemented by a K-back bond, from a filled
hybrid orbital on the metal to the empty prr*-(antibonding)-orbital of the olefin. In this bonding scheme the olefin
is formally neutral and the platinum in the + 2 oxidation
state.
Gel'mud3' has put forward an alternative description of
the bonding within the framework of the VB theory. This
involves two platinum-carbon o-bonds giving a three
/c
membered Pt I ring. In this bonding scheme the olefin
C
'
carries a formal 2 - charge and the platinum is in the + 4
oxidation state. Although this description of the bonding
went out of favor after the Dewar-Chatt-Duncanson
description was put forward, it has recently been revived
to explain some of the properties of metal-olefin complexes
that were apparently hard to explain on the Dewar-ChattDuncanson theory.
The present paper effectively falls into three parts. In the
first part (Section 2) the more important properties of
platinum(I1)-olefin and -acetylene complexes, together,
where appropriate, with their palladium(I1) and nickel@)
analogs, are put forward. The molecular orbital and
valence bond descriptions are then examined to see how
they can explain these properties. It will then be shown
that a molecular orbital description, modified slightly from
the rather rigid scheme put forward initially, can explain
all the presently observed phenomena and further that this
['I
596
Dr. F. R. Hartley
Department of Chemistry
The University
Southampton SO 9 5NH (England)
description has greater flexibility than the valence bond
description.
In the second part (Section 3) the same treatment is accorded to platinum(0)-olefin and -acetylene complexes, together, where appropriate, with their palladium(o) and
nickel(0) analogs. Again it will be shown that all the
current experimental data can be understood in terms of
the molecular orbital approach, which because of this and
its greater flexibility is considered to be the more useful
description of the bonding.
In the third part (Section 4) the more important properties
of the olefin and acetylene complexes of the other transition metals are examined and it is shown that the complexes fall into one of two classes. One class, class S,
resemble the platinum(I1) complexes and the other, class T,
the platinum(0) complexes[*].
Throughout this paper only monoolefins and occasionally
non-conjugated diolefins are considered because the bonding in conjugated polyolefin-metal complexes is considerably more complex and less well understood owing to the
greater number of olefin orbitals that must be considered.
2. Platinum(II)-Olefm and -Acetylene Complexes
2.1. Experimental Observations
In this section are listed the more important experimental
observations on platinum(@-olefin and -acetylene complexes that any successful description of the bonding must
be able to account for. Where appropriate the properties
of the palladium(I1) and nickel(I1) analogs are also considered.
1. The structure of the olefin or acetylene is modified only
slightly on coordination. Although platinum(I1)-olefin and
-acetylene complexes have been widely investigated by
X-ray diffraction there are many difficulties, which have
been summarized by the present author14].These have led
[*I
These two classes and the significance of S and T are discussed in
Section 4.4.
Angew. Chem. internal. Edit. / Vol. 11 (1972) / N o . 7
surprising, since it would be anticipated that with such a
low energy barrier to free rotation crystal forces would
lead to a distortion from 90". The more accurate results
obtained recently indicate angles slightly different from
90" (see Table 1).
to a number of inaccurate results being published. However, a survey of the more accurate results (Table 1)suggests
that on coordination there is a very slight lengthening of
the multiple-bond and that the substituents are bent out
of the plane of the multiple-bond away from the metal.
Table 1. Bond lengths and angles in olefins and acetylenes coordinated to platinum(i1) and palladium(1r)
Multiple bond
length
Complex
Coord.
Free
Angle that alkyl
substituent is bent
out of the plane of
the multiple bond(")
1.37
1.335
-
(A)
Angle between C=C axis
and square plane(")
Ref.
84
[51
71
C61
85
[71
90
89
2. The energy barrier to free rotation of the olefin about
the platinum(I1)-olefinbond is low1'o-'21, generally lying
between 10 and 15 kcal/mol. This observation makes the
early reports that the plane of the multiple bond is perpendicular to the square-plane surrounding the platinum atom
3. The olefin can twist slightly about the carbon-carbon
double-bond axis. This is indicated by the lower 195Pt-'H
coupling constant in the NMR spectrum of the proton cis
to the methyl group in [Pt(CH,CH=CH,)Cl,]
than of
the trans-pr~tons['~!
~
Table 2. Platinum-ligand bond lengths trans to olefins.
Metal-ligand bond
Bond
length
(A)
Ref.
P
Pt-CI
Bond
length (A)
Ref.
2.05
in [Pt(dipentene)CI,]
HsC,
Pt-CI
Comparison
metal-ligand bond
''
,H
Pt -c1
11-
in
,C,
H
CH,
I
yH2
2.33
2.33
P11
2.33
2.32
[*'I
2.35
C6I
CIT
E3C' 'C&
Angew. Chem. internat. Edit. / Vol. I1 (1972) No. 7
597
4. The NMR spectra show that both the cis and the trans
proton-proton coupling constants of olefins decrease on
coordination to platinum suggesting that the sp2hybridization of the carbon atoms of the free olefin has altered slightly
towards sp3 hybridization["* 141.
5. The infrared spectra indicate. that the C=C stretching
frequency['5r'61 of olefins decreases by about 300 cm- '
and the C-C stretching frequency["] of acetylenes decreases by about 250 cm- on coordination to platinum(x1).
The value quoted here for olefins is that originally suggested
by Babushkin et aZ.["], but rejected by subsequent workers
in favor of a decrease of about 140 cm- on coordination
to platinum(I1) until the recent reinvestigation of the IR
and Raman spectra of Zeise's salt and its deuterated
ethylene analog by Hiraishi" 51.
'
6. Olefins exert a strong trans-effect as deduced from their
influence on the substitution of ligands in the trans-posit i ~ n " ~As~ .mentioned above, much of the data on platinum(II)-olefin complexes obtained by X-ray diffraction
is of doubtful accuracy, but the more accurate results (see
Table 2) do suggest that olefins have little or no trans-influence as deduced from the lengths of the platinum-ligand
bonds trans to olefins.
2.2. Description of Bonding
If the x, y , and z axes are those shown in Figure 1, then the
envisaged a ooriginal molecular orbital treatment".
(olefin to platinum) bond being formed by donation of
charge from the full pn-orbital on the olefin to the empty
5dZ25d,, - y 2 6s 6p, hybrid orbital on platinum and a x-(platinum to olefin) bond being formed by back-donation of
charge from the 5d,, 6p, hybrid orbital on platinum to the
empty pn*-(antibonding) orbital on the olefin. This bond-
Y (to the PtC1, plane)
Fig. 1. Axes used in discussing platinum(~~)-olefin
complexes.
ing scheme required the olefin to be bound perpendicularly
to the square-plane surrounding the platinum and the
early X-ray structures seemed to support this[231.However,
subsequent NMR evidence suggested that the energy
barrier to free rotation of the olefin about the platinumolefin bond was low1''- "]. This can be accounted for by
allowing the x-back donation to occur from either the
5d,,6py hybrid orbital or the 5d,, orbital on platinum (the
6p, orbital is not available for hybridization as it is involved
in bonding to the cis-chloride ligands). Since calculations
have shown[241that d,-p,
hybridization involving an
unoccupied p-orbital can lead to a strengthened metalligand x-bond by increasing the overlap power of the metal
598
d,-orbital, and that only a small amount of p character
need be added to the d-orbital for this effect to be noticed,
it is anticipated that the 5d,, orbital will give less overlap
with the olefin p*-(antibonding) orbital than the 5dy,6p,
hybrid orbital. Consequently use of the 5d,, orbital not
only accounts for the observation of rotation about the
metal-olefin bond but also explains why the perpendicular
arrangement of the C=C double-bond (with respect to the
square-plane through the Pt atom and the C1 atom) is the
most stable. This argument has recently[25,26]been put on
a semiquantitative basis by calculating the total energy of
the complex truns-[Pt(NH,)C1,(C,H4)] as a function of
the angle between the C=C double-bond axis and the
PtNCl, plane. The results show a deep energy minimum
when this angle is 90", which corresponds to the position
necessary for maximum x-back donation from the 5dy,6p,
orbital of the platinum and a shallow minimum when the
angle is O", which corresponds to the position necessary
for maximum x-back donation from the 5d,, orbital of the
platinum.
Recent molecular orbital calculations[27*
281 have indicated
that the platinum 6s orbital is not significantly involved in
the bonding. This has necessitated modification of the
empty acceptor orbital on the metal from d2sp to d2p2.
The molecular orbital treatment, as outlined above, accounts for the slight weakening of the olefinic doublebond on coordination, as required by the X-ray and IR
studies, since both loss of electron density from the olefin
p-(bonding) orbital and gain of electron density in the
olefin pn*-(antibonding) orbital will result in a weakening
of the C=C double-bond.
The bonding in platinum(I1)-acetylenecomplexes is formally
similar to that in platinum(II)-olefin complexes. The two
perpendicular pn-orbitals of the acetylene interact to give
a cylindrically symmetric bond[291which forms a o-bond
in a similar way to the olefin to platinum o-bond. The platinum to acetylene n-bond is similarly analogous to the
corresponding bond in the olefin complexes.
The high trans effect of olefins is readily explained, as described previously by the present authorf4j, by the above
bonding scheme in terms of the x-bonding mechanism for
the trans effectE3'].The negligible trans influence of olefins,
where the trans influence is defined as the tendency of a
ligand to weaken the bond trans to
arises
naturally from the bonding scheme described above, since
ligands which exert their trans effect by a x-acceptor mechanism will only cause weakening of the trans bond if this
bond has a x-component. If there is no x-component (e.g.
in Pt-N) or only weak x-bonding (e.g. in Pt-Cl) in the
trans bond then that bond may either not be weakened or
even be ~trengthened~~~!
The experimental material discussed so far has been that
traditionally used to support the molecular orbital description of the bonding. It is important, howeler, to consider
how the molecular orbital description agcounts for (i) the
bending back of substituents out of the plane of the multiple
bond away from the metal, (ii) the lowering of the doublebond stretching frequenr: from that of an olefin to a value
half-way between that of an olefin and an alkane and (iii)
the reduction that occurs on coordination of the cis and
Angew. Chem. internal. Edit. / Vol. 11 (1972) / N o . 7
trans proton-proton coupling 'constants in the NMR
spectra. All these experimental results have been used to
advocate a greater or lesser modification of the sp2 hybridization of the carbon atoms of free olefins towards sp3 hybridization on coordination. To include these results within
the molecular orbital description it is necessary to consider
what the bonding scheme would look like if the olefinic
carbon atoms were to be completely rehybridized to sp3.
In such a case the olefin is considered to be a dicarbanion
and the platinum atom to be formally in the +4 oxidation
state. Although superficiallythis appears to be very different
from the o-donationlx-back donation scheme discussed
above, in reality the two approaches merge. This can be
seen from Figure 2, which shows the effect on the distribu-
substituents R (where R = H or a more bulky group) on the
multiple bond being bent out of the plane of the multiple
bond away from the metal, since on going from the situation in Figure 2a to that in Figure 2b the electron density
is shifted towards the C-R bonds. This will repel the C-R
bond by bond-pairlbond-pair repulsion so forcing R away
from the metal[351.When this purely qualitative argument
was examined semi-quantitatively using the CNDO-MO
theory (Complete Neglect of Differential Overlap Molecular Orbital Theory) it was shown that for acetylenes addition of electron density to the p*-(antibonding) orbital of
the multiple bond must cause either cis- or trans-bending
of the substituents. Cis-bending is favored by both nonbonded interactions between the substituents and the central metal group, and by the fact that it enables the s-orbitals
of the acetylene to assume the correct symmetry to contribute to the bonding[36! Other explanations for the cisbending of the substituents, that are in effect equivalent to
the present one, consider the coordinated olefin or acetylene
to be in an excited ~ t a t e [ ~ ~However,
. ~ * ] . there is some disagreement as to the description of the excited state[37 391.
bI
ai
Region of high charge density
Region of Low charge density
ma
Fig. 2. Effect of varying the relative importance of the u- and x-components of the platinum(Ir)-olefin bond on the density of the charge
between the olefin and platinum.
(a) sp2 hybridized carbon atoms with a strong o-(olefin to platinum)component and a weak x-(platinum to olefin)-component in the platinum-olefin bond. (b) sp2 hybridized carbon atoms with a weak
a-component and a strong x-component in the platinum-olefin bond.
(c) sp3 hybridized carbon atoms.
tion of electron density in the region between the metal and
the olefin when the relative importance of the 0-and x-components of the metal-olefin bond are altered. The two extremes, (a) and (c), involve a migration of electron density
/c
from inside the Pt, I
triangle, where it is located for sp2
C
hybridized carbon atoms which give strong o-donation
and weak x-acceptance of electron density, to the edges of
the Pt'
L
[ triangle, where it is located for sp3 hybridized
\C
carbon atoms. A carbon hybridization scheme mid-way
between sp2 and sp3 would give an electron density distribution mid-way between these two extremes. This mid-way
point coincides with that for an sp2 hybridization scheme
in which the A-back donation of charge (from platinum to
olefin) is as important as or more important than the odonation of charge (from olefin to platinum). There is some
evidence, such as that from stability constant studies for
platinum(r1)-olefincomplexes[341,that the x-back donation
of charge is more important than the a-donation for the
formation of a stable platinum(i1)-olefin bond (i.e. the
situation shown in Figure 2 b).
The electron density distribution around the olefin shown
in Figure 2 b accounts for the reduced proton-proton
coupling constants observed in the NMR spectra of coordinated olefins, which have been quoted as evidence for
some sp3 hybridization on the olefinic carbon atoms['2,141.
The bonding scheme in Figure 2 b can also account for the
Angew. Chem. internat. Edit. 1 Vol. 11 (1972)
No. 7
2.3. Summary
In conclusion, the o-donationlx-back donation molecular
orbital description of the bonding in platinum(i1)-olefin
and -acetylene complexes can explain all the known properties of these complexes. There is no necessity to invoke
changes of hybridization of the olefinic carbon atoms from
sp2 to sp3, since the same effect can be obtained by altering
the relative importance of the 0-and x-components of the
metal-olefin bond.
3. Platinum(0)-Olefin and -Acetylene Complexes
3.1. Experimental Observations
The more important experimental observations that any
satisfactory description of the bonding in the zerovalent
olefin and acetylene complexes of platinum, palladium and
nickel must be able to explain are as follows :
1. The multiple bond is weakened and lengthened on coordination. This is indicated by the bond lengths of coordinated olefins and acetylenes as determined by X-ray diffraction (Table 3) and also by the lowering of the C x stretching frequency of acetylenes by about 5OOcm-' on coordination[47!
2. X-ray diffraction studies have shown that the axis of the
multiple bond lies close to the plane containing the platinum and the other ligands. Values of between 8 and 14"
have been reported for the dihedral angle a (Figure 3 and
Table 3). The angle p (Figure 3) which the multiple bond
subtends at the platinum atom is about 40" (Table 3).
3. X-ray diffraction studies have indicated that the substituents bound to the multiple bond are bent out of the
plane of this bond away from the platinum atom (Table 3).
599
the NMR time scale (i. e. less than 1.2 s- I). Further similar
NMR evidence on related complexes supports these obs e r v a t i o n ~ [571.~ ~ ,
7. The geminaI fluorine-fluorine coupling constant in the
NMR spectrum of [(PPh,),Pt(CF,-CF=CF,)]
is of the
same order as in a saturated fluorocarbon, which suggests
sp3 hybridization of the "olefinic" carbon
3.2. Molecular Orbital Treatment of the Bonding
Fig. 3. The structure of olefin and acetylene complexes of platinum(0).
The plane containing the points u, v, w and x together with the Pt atom
is perpendicular to both the PtCC and PtPP planes.
The observations firstly that olefins and acetylenes in
which strongly electron-withdrawing substituents are
bound direct to the multiple bond give more stable complexes than ligands with less electron-withdrawing substituents and secondly that percyano-olefins and -acetylenes,
which are known to be very good a-acceptor ligands but
Table 3. Results of X-ray diffraction studies of platinurn(0) complexes.
Complex
C-C or CZ?$
Bond length ( A )
Complex
[(PPh,), Pt(TCNE)l [c]
[(PPh,),Ni(C,H.,)I [dl
[(PPh,),Pt(NC-C=C-CN)]
[(PPh,),Pt(PhC-CPh)]
[('BuNC),Ni(TCNE)] [c]
[a]
[b]
[c]
[d]
1.59
1.46
1.40
1.32
1.48
r]
a [a1
Ref.
Y PI
Free
1.34
1.33
1.19
1.19
1.34
8.3
12
8
14
23.9
41.5
42
32.5
39
44.4
40
28.4
[40,411
r421
WI
[441
[451
See Figure 1 for definition of a and P.
y=angle by which substituents are bent out of the plane of the multiple bond away from the platinum atom
TCNE = tetracyanoethylene.
Isomorphous with platinum(0) analog [46].
4. Olefins and acetylenes in which strongly electron-withdrawing substituents are bound direct to the multiple bond
give more stable complexes than olefins with less electronwithdrawing substituents. This is indicated firstly by the
decrease in the tliermal stability of [(PPh,),Pt@,C=CR,)]
in the order R = CN (decomp. point 268-270°C[481)
> R = F (decomp.point 218-220°C[491)~9R=H (decomp.
point 122-125 0C[501),and secondly bi,gtudies of equilibrium (I), which showed that the position ofthe equilibrium
favors the formation of the platinum(0) complex with the
acetylene containing the most electron-withdrawing substituents (ac and ac' are substituted acetylenes)['', 521.
[(PPh,),Pt(ac)]
+ ac'
$
[(PPh,),Pt(ac')]
+ ac
(1)
5. The melting-points of the complexes [(PPh,),Pt(olefin)],
which vary from 122-125°C for the ethylene complex to
268-270 "C for the tetracyanoethylene (TCNE) comp l e ~ [ ~are
~ ] ,relatively high indicating the considerable
thermal stability of the complexes.
6. The observation of two separate coupling constants between the methyl group and the phosphorus atoms in the
NMR spectrum[541of [(PPh,),Pt(PhC=CCH,)] and the
inequivalence of the two phosphorus
in
[(PPh,),Pt(CF,==CFX)] (where X=Cl or CF,) both
indicate that rotation of the olefin or acetylene about the
platinum( 0)-olefin or -acetylene bond is slow relative to
600
Angle
P [a1
poor o-donors, form stable complexes with platinum(0)
suggest that if the bonding in platinum( 0)-olefincomplexes
is described in terms of the Dewar-Chatt-Duncanson
modeI['*21it would be expected that the a-bond (from platinum(0) to the unsaturated ligand) would be fairly strong
whereas the o-bond (from the unsaturated ligand to platinum(o)) would be almost non-existent. This would lead
to a substantial increase in the electron density in the pn*(antibonding) orbitals of the olefin or acetylene which
would weaken and lengthen the multiple-bond as is observed (cf. Section 3.1). Furthermore, molecular orbital calculations have shown that addition of electron density to the
pn*-(antibonding) orbital of TCNE will increase the bondorder of the C-CN
The decrease of the C-CN
bond length from 1.449 A in free TCNE[s91to 1.41 kO.04 A
on coordination to platinum( o ) [ ~ O , 411 is thus consistent
with the Dewar-Chatt-Duncanson model. The observation
that the substituents attached to the multiple bond are bent
out of the plane of this bond away from the platinum atom
is also consistent with a substantial electron density in the
p*-(antibonding) orbitals of the ligand, as discussed above
in connection with platinum(I1) complexes.
The near planagty of platinum( 0)- and nickel(0)-olefin and
-acetylene complexes (see Table 3 and Figure 3) as opposed
to the almost perpendicular arrangement of the multiplebond in relation to the square-plane around the platinum
atom in the divalent complexes is readily explained by the
Angew. Chem. internat. Edit. 1 Vol. 11 (1972) / N o . 7
molecular orbital treatment. Zerovalent platinum complexes are tetrahedral (e.g . [(PPh3),Pt(CO)]160~611
and
[Pt(PF3),][621)because for ad" atom there is no difference
in the ligand field stabilization energies between a squareplanar and a tetrahedral arrangement so that the ligands
arrange themselves as far apart as possible (i. e. tetrahedral)
to minimize bond-pairbond-pair repulsion energies[351.
However, in an olefm or acetylene complex, where the
n-back donation of charge from metal to olefin is large and
the a-donation from olefin to metal is small, there will be
rather less than ten d-electrons around the platinum atom.
This will lead to a Jahn-Teller distortion in which the tetrahedron I
flattened, thus increasing the energy of one of the
t,, orbitals, which is left partly empty, and decreasing the
energy of two of the t,, orbitals, which are left full, relative
to the energies in an undistorted tetrahedral environmen t I6 1' .
The molecular orbital explanation for the observation that
the geminal fluorine-fluorine coupling constant in the
the multiple bond lies in the PtP, plane. The fact that these
calculations predict a minimum energy when the dihedral
angle CL (see Figure 3) is zero is probably an indication of
the approximate nature of the calculations rather than an
indication of the importance of crystal packing effects in the
solid state, since crystal packing effects would be
expected to vary markedly over the range of complexes shown in Table 3, whereas, with the exception of
[(t-BuNC),Ni(TCNE)] the complexes show a remarkably
constant value of a.
Recent molecular orbital calculations[27*
281 have reinforced
the conclusion mentioned above that the strength of the
metal(0)-olefinor -acetylene interaction is strongly dependent on the amount of n-back donation from the metal.
These calculations also indicated that the metal s-orbital
was not involved to any significant extent in bonding. When
group theory was used to develop a new bonding model
that was consistent with this the bonding scheme in Table 4
was obtained.
Table 4. Orbitals involved in the bonding in metal(o)-olefin and -acetylene complexes
+z
(from plane of
i
Y
Type of bonding
Orbitals involved in olefin complexes
Orbitals involved in acetylene complexes
cr-donation to
the metal
From a-orbital of olefin to dp2
(dxy+ p, + p,) hybrid on metal
n-back donation
from the metal
From dxz-yzorbital on metal to
(antibonding) orbital on olefin
(1)From x,,-orbital of acetylene to dp2
(d,,+px+py) hybrid on metal.
(2) From %,-orbital of acetylene to d2
(d,,+d,,) hybrid on metal.
(IFrom
)
dxz-yzorbital on metal to
(antibonding) orbital on acetylene.
(2) From d Z(d,,+ dyz)hybrid on metal to
a:-(antibonding) on acetylene.
NMR spectrum of [(PPh,),Pt(CF,-CF=CF,)]
is of
the same order as in a saturated fluorocarbon is the same
as advanced above for a similar observation with platinum(r1) complexes (cf. Figure 2 and the related discussion).
XI-
<,-
3.3. Valence Bond Treatment of the Bonding
In the valence bond description the olefin or acetylene is
considered to be a dicarbanion which forms two a-bonds
to platinum, which is itself in the 2 oxidation state. This
description can readily account for the following experimental observations :
The absence of free rotation of the olefin or acetylene about
the platinum(0)-olefin or -acetylene bond arises because
the platinum 6pz and 5d,, orbitals are not readily available
for back-donation to the unsaturated ligands because they
are involved in back-donation to the triphenylphosphane
I.The substituents at the multiple bond are bent away
ligands. Thus the flattened tetrahedral structure, in which
from the platinum atom on coordination. This would be
the back-donation occurs from the dX2-,,2~ r b i t a l [ ~ ' ,is~ ~ ] , an immediate consequence of the rehybridization of the
considerably more stable than any other arrangement, so
carbon atoms from the sp" to the sp"+' arrangement
preventing the occurrence of free rotation about the planecessary for the formation of two platinum-carbon atinum(0)-olefin or -acetylene bond. This argument has
bonds (n= 1 for acetylenes and n =2 for olefins). However,
recentlyLz5, been put on a semi-quantitative basis by
if the carbon atoms of the olefin are rehybridized from sp2
calculating the total energy of the complexes
to sp3 on coordination, then the angle y in Table 3 should
[(PH,),Pt(CH,C-CCH,)]
and [(PH3),Pt(C2H,)] as a
be 54"44'. Since the observed values of y (Table 3) are subfunction of the angle between the multiple bond axis and
stantially less than 54"44', the argument in favor of sp3
the PtP, plane. The results show a single minimum when
hybridization does not appear to be warranted, particularly
+
Angew. Chem. internat. Edit.
Vol. I 1 (1972) 1 No. 7
601
in view of the ease with which the MO theory can interpret
the experimentally observed values of y.
and rehybridization of the carbon atoms of the multiple
bond[661,has recently been criticized on the grounds
2. The geminal fluorine-fluorine coupling constant in the
NMR spectrum of [(PPh,),Pt(CF,-CF=CF,)]
is of the
same order as in a saturated fluorocarbon, which suggests
sp3 hybridization of the “o1efinic”tarbon atoms and hence
platinum-carbon o - b o n d ~ [ ~ ~ !
that the
twisting is very
[(t-BuNC)2Ni(TCNE)1where the largest
Observed
was
even in
Of ci yet
2. The small values observed for the C-pt-c
angle (p in
Figure 3) together with the values of about 70” observed
for the C-C-Pt
angle, which are in contrast to the optimum C-C-Pt
angle of 109’28’ for the formation of a
3. There is no free rotation of the olefin or acetylene about
the metal-olefin or -acetylene bond.
4. The VB treatment is useful in accounting for the anal-
strong o-bond, suggest that the three-membered
ogies between the acetylene complexes and cyclopropenes
as shown in the preparation, bromination, and addition of
hydrogen halide reactions given
Pt/y
C
‘
ring would be very strained. However, since the‘melting
points of the [(PPh,),Pt(olefin)] complexes vary from
CN
Bromination
1
L‘B r
HX
(X=CI,Br)
PPh3 CN
I
I
X-Pt-C*
I
PPh,
Y-CN
Addition of
hydrogen halide
However, whilst the valence bond conception provides a
useful way of looking at these reactions, the mechanisms of
the bromination and hydrohalogenation reactions are not
yet sufficiently well understood to know if the structural
and mechanistic implications of the valence bond formulation are valid.
In spite of the apparent successes just listed, the valence
bond description is hard pressed to account for the following experimental observations :
1. The angle ci in Figure 3 generally lies between 8” and 14”.
To account for this it is necessary to assume that the eclipsed
position of the substituents on the multiple-bond is sufficiently unstable to cause the sp3 hybrid orbitals to rotate
towards the staggered position[651.However, these substituents are sufficiently remote from each other for this to
be rather unlikely. A suggestion,similar to the present one,
that the values of ci arise from a combination of twisting
602
A
122-125°C for the ethylene complex to 268--270°C for
the TCNE complex[531,there would not appear to be. as
much strain as is suggested by these small angles.
3.4. Summary
The MO and VB descriptions of the bonding in the zerovalent complexes are equivalent in that both predict that
the platinum atom has less than ten Sd-electrons (i. e. is
formally in a positive oxidation state) and that the coordinated unsaturated ligand has a higher electron density
around it than when it is free. In addition, both theories
suggest that the bulk of the electron density in the platinumolefin or -acetylene bond lies along the edges of the Pt,
/ cI
C
triangle rather than in the center of this triangle (cf. Fig. 2c).
Angew. Chem. internat. Edit. 1 Vol. I1 (1972) No. 7
been accorded to these complexes is slight relative to that
given to the complexes of nickel, palladium, and platinum.
The MO description has two important advantages over
the VB description. Firstly it can readily explain all the
experimental observations reported to date for these complexes whereas the valence bond description cannot, and
secondly its flexibility in terms of variations of the relative
strengths of the o-and x-components of the metal-olefin
bond allows essentially the same bonding description to be
used for the complexes of a whole series of metals.
4.1. X-Ray Diffraction Studies
The structural data obtained from X-ray diffraction, summarized in Tables 5 and 6, is similar in its consequences for
the bonding scheme to that in Tables 1 and 3. Thus in all
the complexes the multiple-bond increases in length on
coordination and where they have been observed the substituents attached to the multiple bond have been found to
be bent out of the plane of that bond away from the metal
atom. In the square-planar rhodium(1)complexes (Table 5 )
the olefin axis is approximately perpendicular to the ligand
plane as found in the square-planar platinum(I1)complexes
reported in Table 1, whereas in the trigonaI-bipyramidal
4. Olefin and Acetylene Complexes of Other Metals
In the present section the physical properties of olefin and
acetylene complexes of the transition metals other than
nickel, palladium, and platinum are considered. It is immediately apparent that the depth of study that has so far
Table 5. Structural data on square-planar [a] rhodium([)-olefincomplexes.
Complex
Structure
Multiple bond length
(A)
Angle (") between C=C axis
and square plane
Ref.
Complex Free
[a] [RhCl(PPh,),(C,F,)]
[b] Ref. [69].
[c] Ref. [70].
is substantially distorted from square-planar towards a tetrahedral arrangement
Table 6. Structural data on trigonal bipyramidal iridium(1) and iron(o) olefin complexes.
Complex
Structure
Multiple bond length (A)
Complex
Free
y La]
Ref.
(7
11
[IrH(CO#PPh,),(FUMN)]
[b]
[IrBr(CO)(PPh,),(TCNE)] [c]
PII,P,~" &CN
1r-11
Ph3P'A
'>If
CiY
Ph3P,yo
C/'Cv
Ir--li
Ph3P'Ar
'\\C;i
CN
-
-
~711
1.51
1.34
35
[71]
1.42 (racemic isomer)
1.35 [d]
-
[72]
1.35 [d]
26
[73]
1.35 [d]
27
[73]
11.35 [d]
37
[73]
graphically
I.'f--ll
oc'&
1.43
Y
,'
CcQIl
optically
.
. active
1.401 isomers
1.34
[a]
[b]
[c]
[d]
-
y=angle by which the substituents are bent out of the plane of the double-bond
FUMN = fumaronitrile.
TCNE = tetracyanoethylene.
Reference [75].
Angew. Chem. internat. Edit. 1 Vol. 11 (1972) 1 No. 7
603
complexes(Table 6) the olefins lie in or nearly in the trigonal
plane, which is similar to the arrangement in the zerovalent
complexes reported in Table 3.
The X-ray diffraction studies that have been reported so
far thus provide no information which cannot be accounted
for in the molecular orbital description of the bonding as
outlined above for divalent and zerovalent complexes of
platinum.
4.2.
of the values which are quoted in Table 7 may be open to
some doubt due to the problem ofcoupling ofthe >CH, deformation mode with the C=C stretching mode. What is
certain is that there is a decrease in the C=C stretching
frequency on coordination in agreement with the predictions of the MO bonding scheme.
NMR Studies
The temperature dependence of the 'H-NMR spectrum of
[(x-C,H,)Rh(C,H,),]
can best be understood in terms of
free rotation about the rhodium-ethylene a-bond axis with
a rotational energy barrier of about 6 k~al/mol~'~].
This
can be understood in an essentially similar way to the rotation of olefins about the metal-olefin bond in platinum(I1)
complexes discussed above, and therefore once again does
not introduce any new experimental data that the MO
bonding scheme cannot accommodate.
A study of the "F-NMR spectra of the fluoro-olefin complexes of rhodium(I),iridium(1)and nickel(0)has shown that
these fall into two main series[771.In the first series, typified
by [Rh(acac)(C,F,)] and [Ni(PPh,),(C,F,)],
all the
fluorine atoms absorb in the same region suggesting a
square-planar structure with free rotation about the metaltetrafluoroethylene bond. This structure is of course analogous to that found for platinum(r1) complexes. In the
second series, typified by [Rh(acac)(n-C,H, INC)(C,F,)],
and [IrCl(CO)(PPh,),(C,F,)], two fluorine resonances
were observed indicating the presence of two sets of fluorine
atoms in chemically non-equivalent environments. By
analogy with [IrCl(CO)(PPh,),(TCNE)] the structure of
the iridium(1) complex is probably trigonal bipyramidal :
When an olefin coordinates to [IrCl(CO)(PPh,),] there is
an increase in the CO stretching frequency of the coordinated carbon monoxide ligandEE6'.Furthermore, this increase is greatest for olefins with weak a-donor but strong
x-acceptor capacities (see Table 8). An increase in the CO
Table 8. Carbonyl stretching frequencies for iridium complexes (from
Ref. [86-881).
~
Complex
vco (cm-')
Ir"'
1955
2017
2018
2025
2029
2040
2060
2080
1937
1952
1958
2020
2040
2045
2074
c1
A
If the tetrafluoroethylene molecule lies in the trigonal plane
and if there is no rotation about the iridium@)-tetrafluoroethylene bond two I9F-NMR signals would be expected.
This structure, as already pointed out in Section 4.1, is
analogous to the structure found in zerovalent platinum
complexes. Thus these "F-NMR results substantiate the
conclusion reached above from the X-ray diffraction results
that olefin complexes fall into two main classes, those
formally analogous in structure and bonding to the square
planar platinum(I1)complexesand those formally analogous
to the platinum(0) complexes.
4.3. Infrared Studies
The IR spectra of olefin complexes indicate that the doublebond stretching frequency decreases on coordination,
although, as mentioned above, the absolute magnitudes
604
~_______
Oxidation State
Ir'
Ir"'
stretching frequency of a metal-carbonyl complex indicates
a reduction in the x-back donation of charge from the metal
to carbon monoxide. Such a reduction in x-back donation
accompanies the oxidation of a metal; for example
[IrCl(CO)(PPh,),] containing iridium(1f has vco equal to
1966 cm- ', whereas [IrCI,(CO)(PPh,),] containing iridium(rrr) has vco equal to 2080 cm- A series of complexes
such as that shown in Table 8 demonstrates the advantages
of the M O approach to metal-olefin bonding because of its
built-in flexibility.It is apparent that as Table 8 is descended
there is an increase in the relative importance of the x-component of the metal-olefin bond and a decrease in the relaAngew. Chem. internat. Edit. / Vol. I1 (1972) / N o . 7
tive importance of the o-component of that bond which
leads to an effective transition from iridium(1) to iridium(@
as the Table is descended.The strength of the MO approach
lies in its relative disinterest in the oxidation state of the
metal, whereas by contrast the VB approach can only draw
the two extreme formulations shown in Figure 4, and then
consider resonance between unequal amounts of these two
canonical forms.
a) Iridium(0
b) Iridium(II1)
Fig. 4. Extreme valence bond formulations for iridium-olefin complexes.
the experimental observations reported to date for these
two classes whereas the VB description cannot. The M O
scheme is simple to visualize and, since it is anticipated that
the relative importance of the 0-and n-components will
vary according to (i) the metal atom, (ii)the oxidation state
of the metal atom, (iii) the nature of the olefin, and (iv) the
nature of the other ligands present, this description of
the bonding can readily be applied to all metal-olefin complexes. It should be pointed out that this scheme precludes
any involvement of the carbon 25 orbital in the metal-olefin
bond. As yet no evidence that unequivocally supports
carbon 2s participation has been put forward.
1 should like to thank Imperial Chemical Industries Limited
for the award of a Research Fellowship during the tenure
of which part of this work was carried out.
Received: May 21, 1971 [A 891 IE]
German version: Angew. Chem. 84,657 (1972)
4.4. Summary
The physical properties of metal-olefin and -acetylene
complexes just discussed suggest that they fall into two
classes, class S and class T"].
Class S Complexes are formally analogous in their properties and bonding to the square-planar complexes of platinum(Ii), discussed in Section 2 of this paper. It is also
possible to have ligands above and below the square-plane
and still have essentially the same properties. For example
the room temperature proton NMR spectrum of the octahedral complex cis-[Re(CO),(C,H,),]
PF, exhibits only
the single proton signalt801that would be consistent with
the free rotation of ethylene about the metal-olefin bond
found for other class S complexes.
+
Class T Complexes are formally analogous in their properties and bonding to the trigonal complexes of platinum(0)
discussed in Section 3 of this paper. Again it is possible to
have ligands coordinated above and below the trigonal
plane giving trigonal bipyramidal complexes.
Thus class S complexes, as the name implies, involve three
other ligands coordinated to the metal in the same plane as
the unsaturated ligand, whereas class T complexes only
have two other ligands in addition to the unsaturated
ligand. One of the principal differences between the two
classes of complexes is that whereas class S complexes exhibit free rotation of the olefin or acetylene about the metalolefin or -acetylene bond, there is no such free rotation in
the class T complexes.
5. Conclusions
It has been shown that the transition metal complexes of
monoolefins and monoacetylenes fall into two main structural classes which are labeled S and T. An examination of
of the description of the bonding in each of these two classes
has indicated that the MO scheme is able to interpret all
[*I The letters S and T are taken from the square-planar and trigonal
structures of the model platinum complexes.
Angew. Chem. internat. Edit. / Vol. I 1 (1972) 1 NO. 7
[I] M . J . S . Dewar, Bull. SOC.Chim. Fr. 18, C79 (1951).
[2] J . Chart and L. A. Duncanson, J. Chem. SOC.1953,2939.
131 A . D: German, C. R. Acad. Sci. URSS 24, 549 (1939).
[4] F . R . Hartley, Chem. Rev. 69,799 (1969).
[5] J . A . J . Jarois, B. I: Kilbourn, and P . G . Owston, Acta Crystallogr.
827, 366 (1971).
[6] E . Benedetti, P . Corradini, and C. Pedone, J. Organometal. Chem.
18, 203 (1969).
[7] G . R. Daoies, W Hewertson, R . H . 8. Mais, P. G . Owston, and C . G .
Patel, J. Chem. SOC.A 1970, 1873.
[8] N . C. Baenziger, G . F. Richards, and J . R . Doyle, Acta Crystallogr.
IS, 924 (1965).
[9] J . R . Holden and N . C. Baenziger, J. Amer. Chem. SOC.77, 4987
(1955).
[lo] A . R . Brause, F . Kaplan, and M . Orchin, J. Amer. Chem. SOC.89,
2661 (1967).
[ill C. E. Holloway, G. H d l e y , B. F . G . Johnson, and J . Lewis, J. Chem.
SOC.
A 1969.53.
[12] C. E. Holloway, G . Hulley, B . F . G . Johnson, and J . Lewis, J. Chem.
SOC.A 1970, 1653.
[I31 H . P . Fritz, K . E . Schwarzhans, and D. Sellmann, J. Organometal.
Chem. 6, 551 (1966).
[14] I:Kinugasa, M . Nakamura, H . Yamada, and A . Saika, Inorg. Chem.
7, 2649 (1968).
[15] J . Hiraishi, Spectrochim. Acta 25A, 749 (1969).
[16] J . Hiraishi, D. Finseth, and F . A. Miller, Spectrochim. Acta 25A,
1657 (1969).
[17] E . 0. Greaues and P . M . Maitlis, J. Organometal. Chem. 6, 104
(1966).
[is] A . A . Babushkin, L. A . Griboo, and A . D.German, Russ. J. Inorg.
Chem., English Trans]. 4, 695 (1959).
[19] F. B a s d o and R . G . Pearson: Mechanisms of Inorganic Reactions.
2nd t d n . Wiley, New York 1967, p. 355.
[20] P . R. H . Alderman, P . G . Owston, and J . M . Rowe, Acta Crystallogr.
13, 149 (1960).
1217 G . H . W Milburn and M . R . P u t e r , J. Chem. SOC.A 1966, 1609.
[22] N . C. Baenziger, R. C. Medrud, and J . R. Doyle, Acta Crystallogr.
18, 237 (1965).
[23] J . A . Wunderlich and D. P . Mellor, Acta Crystallogr. 7, 130 (1954);
8, 57 (1955).
1241 D. P . Craig, A . Maccoll, R . S . Nyholm, L. E. Orgef,and L. E . Sutton,
J. Chem. SOC.1954, 332.
[25] K . S. Wheelock, J . H . Nelson, L. C. Cusachs, and H . B. Jonassen,
J. Amer. Chem. SOC. 92, 5110 (1970).
[26] J . H . Nelson and H . B. Jonassen, Coord. Chem. Rev. 6,27 (1971).
1271 J . H . Nelson, K . S. Wheelock, L . C. Cusachs, and H . B. Jonassen,
Chem. Commun. 1969,1019.
[28] J . H . Nelson, K . S. Wheelock, L. C. Cusachs, and H . B. Jonassen,
J. Amer. Chem. SOC.91, 7005 (1969).
1291 C. A . Coulson: Valence. Oxford University Press, London 1952,
p. 193.
605
[30] J . Chatt, L. A. Duncanson, and L. M . Venanzi, J . Chem. SOC.1955,
4456.
[31] A. Pidcock, R. E. Richards, and L. M . Venanzi, J. Chem. SOC.
A 1966, 1707.
[32] L. M . Venanzi, Chem. Britain 4, 162 (1968).
[33] D. M . Adams, J . Chatt, J . Gerratt, and A. D. Westland, J . Chem.
SOC.1964,734.
1341 R. G . Denning, F. R. Hartley, and L. M. Venanzi, J . Chem. SOC.
A 1967, 324.
[35] R. J . Gillespie and R. S. Nyholm, Quart. Rev. Chem. SOC.11,339
(1957).
[36] A . C. Blizzard and D. P. Santry, J . Amer. Chem. SOC.90, 5749
(1968).
1371 R. Mason, Nature 217, 543 (1968).
1381 R. McWeeny, R. Mason, and A. D. C.Towl, Discuss. Faraday SOC.
47, 20 (1969).
[39] J . N . Murrell, Discuss. Faraday SOC.47, 59 (1969).
[40] C . Panattoni, G. Bombieri, U . Belluco, and W H . Baddley, J . Amer.
Chem. SOC.90,798 (1968).
[41] G . Bombieri, E. Forsellini, C. Panattoni, R. Graziani, and G . Bandoli,
J . Chem. SOC.A 1970,1313.
[42] C. D. Cook, C.H . Koo, S. C. Nyburg, and M . 7: Shiomi, Chem.
Commun. 1967,426.
[43] C.Panattoni and R . Graziani in M . Cais: Progress in Coordination
Chemistry. Elsevier, London 1968, p. 310.
[44] J . 0. Glanuille, J . M . Stewart, and S. 0. Grim, J . Organometal.
Chem. 7, P9 (1967).
[45] J . K. Stalick and J. A. Ibers, J. Amer. Chem. SOC.92,5333 (1970).
[46] C.D. Cook and G. S. Jauhal, J . Amer. Chem. SOC.90,1464 (1968).
[47] J . L. Boston, S. 0. Grim, and G. Wilkinson, J. Chem. SOC.1963,3468.
1481 W H . Baddley and L. M . Venanzi, Inorg. Chem. 5, 33 (1966).
[49] M . Green, R. B. L. Osborn, A. J . Rest, and F. G. A. Stone, Chem.
Commun. 1966,502.
[50] C. D. Cook and G. S. Jauhal, Inorg. Nucl. Chem. Lett. 3,31(1967).
[Sl] A. D. Allen and C. D. Cook, Can. J . Chem. 41, 1235 (1963).
1521 A . D. Alkn and C . D. Cook, Can. J . Chem. 42,1063 (1964).
[53] W H.Baddley, Inorg. Chim. Acta Rev. 2, 7 (1968).
1541 E . 0. Greaves, R. Bruce, and P. M . Maitlis, Chem. Commun. 1967,
860.
[SS] M . Green, R. B. L. Osborn, A. J . Rest, and F. G. A. Stone, J. Chem.
SOC.A 1968,2525.
[56] J . H. Nelson, H . B. Jonassen, and D. M . Roundhill, Inorg. Chem. 8,
2591 (1969).
[57] E. 0. Greaves, C.J . L. Lock, and P. M . Maillis, Can. J . Chem. 46,
3879 (1968).
[ S S ] B. R. Penfold and W N. Lipscomb, Acta Crystallogr. 14,589 (1961).
606
[59] D. A. Bekoe and K . N. Trueblood, unpublished results (1967),
quoted in [45].
1601 !T G. Albano, G. M . Basso Ricci, and P. L. Bellon, Inorg. Chem. 8,
2109 (1969).
[61] K G. Albano, P. L. Bellon, and M . Sansoni, Chem. Commun. 1969,
899.
[62] J . C. Marriott, J . A. Salthouse, M . J . Ware, and J . M . Freeman,
Chem. Commun. 1970,595.
[63] L. E. Orgel: An Introduction to Transition Metal Chemistry.
Methuen, London 1960, p. 65.
[64] G. L. McClure and W H . Baddley, J . Organometal. Chem. 25,261
(1970).
[65] E . L. Eliel: Stereochemistry of Carbon Compounds. McGrawHill, New York 1962.
[66] P. Heimbach and R. Traunmiiller, Liebigs Ann. Chem. 727, 208
(1969).
[67] 7:Kashiwagi, N . Yasuoka, N . Kasai, and M. Kukudo, Chem. Commun. 1969, 317.
1681 P. B. Hitchcock, M . McPartlin, and R. Mason, Chem. Commun.
1969,1367.
[69] L. E. Sutton: Tables o f Interatomic Distances and Configurations
in Molecules and Ions. Chem. SOC.Special Publ. No. 18, London 1965,
p. M 95s.
[70] I . L. Karle and J . Karle, J. Chem. Phys (8. 963 (1950).
1711 L. Manojlouii-Muir, K. W Muir, and J . A. Ibers, Discuss. Faraday
SOC.47, 84 (1969).
[72] C. Pedone and A. Sirigu, Acta Crystallogr. 23, 759 (1967).
[73] C.Pedone and A. Sirigu, Inorg. Chem. 7,2614 (1968).
[74] A. R. Luxmooreand M . R. Truter,Acta Crystallogr. 15,1117(1962).
1751 C.J . Brown, Acta Crystallogr. 21, 1 (1966).
[76] R. Cramer, J . Amer. Chem. SOC.86,217 (1964).
[77] G. W Parshall and F . N . Jones, J . Amer. Chem. SOC.87,5356 (1965).
[78] M . J . Grogan and K . Nakamoto, J. Amer. Chem. SOC.88, 5454
(1966).
[79] E. 0. Fischer and K . Fichtel, Chem. Ber. 94, 1200 (1961).
[SO] E. 0. Fischer and K . Ofele, Angew. Chem. 74.76 (1962).
[81] E. 0. Fischer and M . Herberhold, Experientia, Suppl. 9,259 (1964);
Chem. Abstr. 62, 579f. (1965).
[82] H. D. Murdoch and E. Weiss, Helv. Chim. Acta 46, 1588 (1963).
[83] R. Cramer, Inorg. Chem. 1, 722 (1962).
[84] M . J . Grogan and K . Nakamoto, J. Amer. Chem. SOC.90,918 (1968).
[85]
[86]
[87]
[88]
H . W Quinn and D. N. Clew, Can. J. Chem. 40, 1103 (1962).
W H . Baddley, J. Amer. Chem. SOC.90,3705 (1968).
L. Vaska, Accounts Chem. Res. I, 335 (1968).
L. Vaska, private communication.
Angew. Chem. internat. Edit.
Vol. I 1 (1972) / NO.7
Документ
Категория
Без категории
Просмотров
1
Размер файла
943 Кб
Теги
bonding, metali, olefin, complexes, acetylene
1/--страниц
Пожаловаться на содержимое документа