вход по аккаунту


Electron-Rich Half-Sandwich ComplexesЧMetal Bases par excellence.

код для вставкиСкачать
[237] H. Schmidbaur, C. E. Zybill, G. Miiller, C. Kriiger, Angew. Chem. 95
(1983) 753; Angew. Chem. Int. Ed. Engl. 22 (1983) 729.
[238] 3. Stein, J. P. Fackler, C. Paparizos, H. W. Chen, J . Am. Chem. Soc. 103
(1981) 2192.
[239] C . Kriiger, J. C. Sekutowski, H. J. Fuller, 0. Gasser, H. Schmidbaur,
Isr. 1. Chem. I S (1976/77) 149.
12401 D. Seyferth, S. 0. Grim, J. Am. Chem. Soc. 83 (1961) 1610.
[241] N. A. Nesmeyanov, V. M. Novikov, 0. A. Reutov, Izu. A h d . Nauk
SSSR Otd. Khim. Nauk 1964, 722.
12421 N. A. Nesmeyanov, V. M. Novikov, 0. A. Reutov, J. Organomet. Chem.
4 (1965) 202.
[243] H. Schmidbaur, K. H. RBthlein, Chem. Ber. 107 (1974) 102.
[244] Y. Yamamoto, H. Sugimoto, Bull. Chem. Soc. Jpn. 53 (1980) 3176.
[245] D. R Mathiason, N. E. Miller, Inorg. Chem. 7 (1968) 709.
[246] H. Schmidbaur, 0. Gasser, T. E. Fraser, E. A. V. Ebsworth, J. Chem.
Sac. Chem. Commun. 1977, 334.
12471 H. Schmidbaur, T. Costa, Chem. Ber. 114 (1981) 3063.
I2481 H. Schmidbaur, G. Muller, Monafsh. Chem. 111 (1980) 1233.
[2491 H. Schmidbaur, W. Wolf, Chem. Ber. 108 (1975) 2851.
12501 D. J. Burton, R. Takei, S. Shin-ya, J . fluorine Chem. 18 (1981) 197.
[251] N. C. Baenziger, R M. Flynn, D. C. Swenson, N. L. Holy, Actu Crystallogr. B 34 (1978) 2300.
Electron-Rich Half-Sandwich ComplexesMetal Bases par excellence
By Helmut Werner’
Dedicated to Professor Ernst Otto Fischer on the occasion of his 65th birthday
Electron-rich, half-sandwich complexes of the type C,R,ML2 or C,R,MLL‘ are built up of
an aromatic five- or six-membered ring, a d8-metal, and either a pair of two-electron donors
or an equivalent chelating ligand. Such complexes behave like Lewis bases and react with a
wide variety of electrophiles, El or ElX, to form products with a new metal-element bond.
According to their reactivity they are comparable to the Vaska-type compounds. Certain of
the products obtained after addition of the electrophile undergo interesting subsequent
reactions in which, for example, metal complexes containing molecules that are unstable in
the free state, such as CS, CSe, CH$, CH,Se, CH2Te, CH3CHS, CH3CHSe, CH2=C=S,
CH2=C=Se, and CH2=C=Te are formed. Moreover, cycloadditions as well as reactions
with coordinatively unsaturated transition-metal compounds which result in formation of
heterometal binuclear complexes demonstrate that the metal bases C,RnMLz and
C,R,MLL’ are valuable synthetic building blocks. Furthermore, very recent investigations
have indicated links between metal basicity and the problem of C-H activation.
1. Introduction
Half-sandwich complexes are almost as old as ferrocene-the progenitor of all sandwich compounds. Shortly
after the structure of dicyclopentadienyliron had been accurately determined and analogous metallocenes had
been synthesized, Fischer und Hafner”’ reported the
synthesis of tetracarbonyl(cyclopentadienyl)vanadium,
C5H5V(COk. Later work-involving highly competitive research efforts in Munich and Harvard-produced the homologous manganese and cobalt complexes C5H5Mn(C0)3
and C5HSCo(C0)~2*31.
Because the metal atom in these
compounds is enclosed on only one side by a C,H,-disk
(as in the metallocenes) but is bonded to conventional ligands on the other side, the structure resembles a “halfsandwich”. This term was later applied to similar compounds: instead of CO, ligands such as NO, halogenides,
isocyanides, phosphanes, and olefins were now coordinated to the metal, and instead of CSHs other ring systems
such as C4H4,
C7H7 or CsH8 were employedl4].
[‘I Prof. Dr. H. Werner
Institut fur Anorganische Chemie der Universitxt
Am Hubland, D-8700 Wiirzburg (Germany)
Angew. Chem. Int. Ed. Engl. 22 (1983) 927-949
There was no indication in early publication^[^.^^ that
half-sandwich complexes can behave as metal bases, i. e.
that they react with electrophiles, El or ElX, to form new
metal-element bondscs1.Such behavior was hardly to be expected since both the carbonyl(cyclopentadieny1)metals
mentioned above and most of the analogous compounds
with the general formula C,H,ML,
are 18-electron complexes. At least in a formal sense these possess no lone pair
of electrons capable of forming a covalent bond with a
Lewis acid. Vaska-type compounds trans-[MX(CO)L,]
(M=Rh, Ir; X=C1, Br, I, N,; L=PR3, P(OR),, AsR3 etc.)
do possess such a lone pair of electrons and react not only
with known electrophiles, e. g. HCl, CH31, CH,COCl, BX3,
SOz, etc. but also with H2 or R3SiH by oxidative addition@’. The Vaska-type compounds are thus typical metal
bases, and it is not only thanks to their chemistry that the
concept of metal basicity has become increasingly known
since the beginning of the seven tie^'^^.
The central atom in the trans-[MX(CO)L] complexes is
a d8-system, as is the case in the dicarbonyl(cyc1opentadieny1)metal compounds C5H5M(C0)2(M = Co, Rh, Ir). In
an article which indicated the great significance of the nucleophilicity of transition-metal atoms in complexes having full valence shells, Wilkinson et
mentioned that
0 Verlag Chemie GmbH, 6940 Weinheim, 1983
0570-0833/83/1212-0927 $ 02.50/0
CSH5Co(CO)zis rapidly decomposed in strong acids, such
as &SO, or BF,. H20, whereas in weakly acid media, such
as CF3C02H,it is stable for at least 10 minutes. It did not
prove possible using NMR spectroscopy to confirm the
proposal that a cation [CSHsCoH(CO),]’ was formed in
the process.
The first evidence for such an interaction between the
central atom in CSH5Co(CO), and an electrophile was provided by Kemmitt et aI.[sl.They isolated 1 : 1 adducts from
the reaction of the half-sandwich complex with HgXz
(X=CI, Br, I), which they then compared with known
compounds such as C4H8S.HgC12, and whose formation
they interpreted as a Lewis acid/Lewis base reaction. This
suggestion was confirmed shortly thereaftercg1by X-ray
crystallographic studies on [C5HsCo(C0)2HgClz].Similarly, with HgCI2 and HgBr,, C,H,Rh(CO), forms 1 : 1 adducts whose IR data indicate the presence of a direct metal-metal bond[”].
Further development of this field was stimulated by the
work of Graham et a!. which appeared in 1970/71c11!
They were the first to show that the compound
CSH5Rh(CO)(PMe2Ph),i. e. a derivative of CSH5Rh(C0),,
reacts with Clz, BrZ, CH,COBr, and CF,COCI at
[C5HSRhR(CO)PMezPh]X (R= C1, Br, COCH,, COCF,;
X =halogen) which can be isolated as stable salts after exchanging Xe for BPh?‘””]. The iridium complex
C5H51r(CO)PPh3 not only behaves similarly but is also
capable of forming a hydridometal cation with HBr in
CHzClzat - 70°C. This cation, [C,H51rH(CO)PPh3]+, can
again be isolated as the BPh,
Kinetic data for the
compounds C5HSM(CO)PPh3indicate that the metal basicity increases along the sequence Co 5 Rh < Ir (i. e.
within the group from top to bottom) and that there is
also an increase in basicity along the sequence CSHSCo(CO)PPh3 < C5HSCo(CO)PMePh2 <
C,H,Co(CO)PMe2Phf”c1. It thus became evident that a
better donor ligand would increase not only the electron
density but also the nucleophilicity of the metal-an observation which has since been confirmed in many different
ways in our own investigations.
2. Metal Bases having Five-Membered Ring Ligands
L = C0,PF3,P(OR)3,PMe3
iPrMgBr. The bis(triphenylphosphane)cobalt(r) complex
is formed without having to add PPh, to
the reaction mixture, although in low yield. The rhodium
compound CSHsRh(PPh3),is obtainable analogously[’’1. In
addition, starting from [CsMesRhC12]2and L (molar ratio
1 :4; L=PMe3, PMe2H, PMeZPh,P(OMe),) and using Na/
Hg as the reducing agent, some of the half-sandwiches can
be prepared almost quantitatively‘l6]. The reduction of
with Zn is highly unusual
in that it yields the tetrafluoroethylene complex
analogue of the compounds
CsHSRh(C2H3R’)PR3obtained in reactions (4) and (7).
The most direct route to metal bases having five-membered ring ligands is offered by the exchange of coordinated halides for C5H5,CSH4R,CsMeS,or C9H7.Thus, for
example, the cobalt complexes CsHsCoL2 (L= PMe,Ph,
PMePh,[I7], PPh3[181,PEt3[I9],and PMe,’’’]) may be prepared by allowing the appropriate paramagnetic cobalt(1)
compound CICO(PR~)~
to react with M‘C5Hs (M’= Na, Tl)
or LiC,Me,. C5HSCo(PPh3), can also be obtained from
C O H ( N ~ ) ( P Pand
~ ~ )CSHJZ1].
Both Vaska-type compounds and binuclear complexes
[L2MC1I2and [L(L’)MCI], (M = Rh, Ir) are suitable educts
for the preparation of the complexes C5H5MLz and
CsHSMLL’(M = Rh, Ir), respectivley, by substitution of a
halide- by a cyclopentadienyl-ligand.Several examples of
such syntheses are collected together in reactions (3) and
M=Rh; L=L‘=PPh3[zz]
M = R h ; L=L’=P(OR)3[261
; = P R ~ ,L’=P(OR?,[Z~I
M = R h ; L=PPh3, L’=C2H4‘IIb1
M=W, Ir; L=PPh,, L’=CO[’lb]
M = R h ; L=PMe3, L’=C2H3R”’*]
M=Rh, Ir; L=PPh3, L’=CS‘231
M = R h ; L=P(iPr),, L’=C2H4[281
M = R h , Ir; L=P(iPr),, L‘=CzRJ24.251 M = I r ; L=P(iPr),, L’=C8H14[Z51
2.1. Synthetic Methods
Both ligand exchange reactions and redox reactions
have been used for the preparation of metal bases having
five-membered ring ligands. Thus, the dicarbonyl complex
mentioned in the introduction was synthesized from cobaltocene and CO by displacing one of the
cyclopentadienyl ligands by two CO group^[^,^]. In this
case, the ligand exchange is associated with reduction of
cobalt(r1) to cobalt(r). In the same way, the compounds
and C,H5Co[P(OR),], (R= Me, Et, Ph)[I3]
are also accessible. Analogously, C,H,Co(PMe,), may be
prepared according to reaction (1)1141, although this
method is hardly recommended because of the amount of
trimethylphosphane required.
A reductive method starting from a cobalt(1rr) compound is given in reaction (2)[”]. The reducing agent is
According to reaction (3), the preparation of
C5HSRh(PPh3), involves a reaction type related to that
used for the synthesis of CSHSRh(PMe3)2[29~301
C9H7Rh(PMe3),[311from [Rh(PMe,),]Cl; in the cation of
this salt the rhodium atom is square-planar coordinatedr3”.
Similarly, the compounds (CsH4R)Rh(PMe3)Z (R= iPr,
~ B u ) ‘ and
~ ~ ] CSMeSRh(PMe3)$’61,
which are analogous to
C5HSRh(PMe3)2,are obtained by allowing [(PMe3)2RhC1]2
to react with TlC5H4Ror NaCsMe,.
Half-sandwich complexes of the type CSR5M(CO)Land
CsR5M(CZH4)L(M=Co, Rh; R = H , Me) are also frequently prepared by ligand displacements, as in reactions
( 5 ) and (7). Replacement of CO by a phosphane or phosphite generally takes place under milder conditions than
the displacement of ethylene. Kinetic studies by SchusterAngew. Chem. Int. Ed. Engi. 22 (1983) 927-949
Woldan and BusoZO[~~'
confirmed an A- or SN2-typemechanism for the formation of C,H,Rh(CO)L from
C,H,Rh(CO), and L [see reaction (511. Disubstitution as
shown in reactions (6) and (8) has been observed only in a
few instances up till now. Since the n-acceptor ligand CO
or C2H4 is bonded much more firmly in the compounds
C,H,M(CO)L and C,H,M(C,H4)L than in the educts, it is
evident that reactions (6) and (8) involve a very high activation energy.
Reaction (5):
R = H ; M = C o ; L=PPh3[341, PMePh2, PMeZPh, P(C6H11)3f11C1,
R = H; M = Rh; L= PPh3, P(nBu),, P(OMe),'341, PMezPh"'al,
R = M e ; M = C o ; L=PMe3[201
R = M e ; M = R h ; L=PMe3, PzMe4'381
Reaction (4):
R = H ; M = R h ; L=P(OMe)3f341
Reaction (7):
R = H ; M = C o ; L=PMe3, P(OMe)3[391
R = H ; M = R h ; L=PMe3*, P ( ~ B u ) ~ * [ ~ ' )
R = M e ; M = R h ; L=PMe3,P,Me,"61
Reaction (8):
R = H ; M = R h ; L=PMe3*, PEt3*, P(OEt),*, P(OC6H4Me)3*i401
* not isolated, characterized by NMR spectroscopy.
The displacement of a phosphane ligand in
C5H5Co(PMe3), by CO'4'1 or CNR (R= Me, tBu, Ph)[4*1is
also possible. Interestingly, the corresponding reaction of
C , M ~ , C O ( P M ~with
~ ) ~CO yields the monocarbonyl compound, while with CNMe the bis(isocyanide) complex
C , M ~ , C O ( C N M ~is) formed[201.
2.2. Complexes of the Type CsRsML,
2.2. I . Dicarbonyl- and Bis(o1efin)metal Complexes
C5H5Co(C0)2is not stable in acids['], and nothing is
known about the reactions of C5H5Rh(C0), with HX
to form [CSH5RhH(C0),]X. Dicarbonyl(cyc1opentadieny1)metal compounds are thought to be inert toward methyl iodide[431,which is also to be expected in view of the
good acceptor properties of CO and the resulting relatively
low electron density at the metal. However, if C5H5is replaced by C5Me5and the central atom thus becomes more
electron-rich, both C5Me5Rh(C0), and C,Me,Ir(CO), react
with Me1 [reaction (9)][""'. Whereas for M = Ir the reaction
does not go beyond the stage of the salt-like complex 3,
the corresponding rhodium compound cannot be isolated
and very rapidly yields the product 4 with CO insertion (or
CH3 migrationr45a1).More recently, evidence has been presented that protonation of CsMe,Ir(C0)2 gives the corresponding hydridometal cation[45b1.
Angew. Chem. Int. Ed. Engl. 22 (1983) 927-949
1 , 4 : M = Rh; 2 , 3 : M = Ir
With HCI the compound C5H,Rh(C,H4), forms the
ethylrhodium(rr1) compound C5H5RhCZH5(C2H4)C1[461.
conjectured intermediate hydrido(o1efin)metal cation
could not be detected either here or in the reactions
of mixed bis(o1efin) complexes of the type
C,HsRh(C2H4)(C2H3X)(X = F, CN, C0,Me) with HCl[471.
That the rate of the reaction of C5H5Rh(C2H3X),with HCI
increases along the sequence X =C N < C0,Me <CH3'471,
supports the postulated protonation at the metal. Since the
donor character of the olefin increases along the same sequence, the resulting increase in electron density at the metal should favor electrophilic addition.
2.2.2. Bis(ph0sphite)metal Complexes
In addition to the carbonyl(phosphane) derivatives, the
bis(phosphite) compounds have also played a major role
in establishing the metal basicity of cyclopentadienylmetal
complexes of the type C5RsML2 or C5R5MLL'. Although
the first representative of this group, the complex
C5H,Rh[P(OMe),12 had already been described in 1966[341,
its reactivity was not investigated in detail at that time.
We were closely associated with compounds of this type
in our endeavors to find new synthetic routes to trinuclear
complexes of the "super-sandwich" type[481,[C5H5Co(pP(OR)20)3Co(p-P(OR)20)3CoC5H5].These compounds
were initially prepared by pyrolysis of C,H,CO[P(OR)~]~
(R,=Me, Et)j4'l, although-as Klaui and Dehnicke1501,and
more recently Brill et aZ.["l have demonstrated-they are
more readily produced starting from cobalt(Ir1)-trialkylphosphite or cobalt(rI1)-dialkylphosphonate complexes.
Attempts to obtain analogous trinuclear rhodium complexes starting from C,H,Rh[P(OR),], proved to be unsucc e s s f ~ l ~We
~ ~discovered
' ~ ~ ~ . from these studies, however,
that the bis(phosphite) compounds C,H,Rh[P(0R),I2 are
good nucleophiles and that they react (at least for R = Me)
even under very mild conditions with Br~nstedacids, methyl iodide, or Meerwein's reagent, [OMe3]BF4,to form the
corresponding rhodium(1rr) c o r n p l e x e ~ [ For
~ ~ ~the
~ ~ me~.
thylation reaction the sequence given in Scheme 1 has
been confirmed[541.
Scheme 1. The bar symbolizes a cyclopentadienyl ligand.
The iodide salt 6 obtained by electrophilic addition at
-30°C in acetone eliminates methyl iodide when heated
to form the phosphonate compound 8, which is an isomer
of the starting complex 5. Evidently, the rupture of an
0-CH3 bond is highly favored kinetically compared to
that of a Rh-CH3 bond, an observation in accord with the
high metal basicity of 5. The conversion of 5 into 8 illustrates that trialkylphosphite complexes frequently behave
like the intermediate moiety [R'P(OR),]+ (R=alkyl) in the
Michaelis-Arbuzov reaction and thereby offer a convenient route to metal phosphonate compounds.
Scheme 2. IRh]=CsH5RhCH,; R=OMe.
Proof of this is also to be seen in the subsequent reactions undergone by some of the products obtained from
the metal bases C5HSM[P(OMe),],. Thus, 8 reacts with alkali metal iodides, for example, to yield rhodium(phosphonate) complexes 9, which react with HCl to give 10;
the latter presumably contains a six-membered ring having a symmetric O H 0 bridge1561.Starting from 10 or 11,it
is possible to prepare polynuclear complexes 12 (see
Scheme 2) which are similar to the "super-sandwich" comp o u n d ~ ' ~ ~The
, ~ ~marked
ability of metallabis(phosphonates) to complex alkali metal ions is also illustrated by
the fact that the compounds 15 prepared from
(C,H,R)Co(CH,),PR; (see Section 3.2) act as strong complexing agents toward Li+, Na', and K + . The complexes
formed release the alkali metal ion M + from the coordination sphere only when HCI is added, but not on addition
of water (see Scheme 3)[581.
crystallizes with one equivalent of NaI[591.In this salt there
exist intermolecular interactions between the sodium ions
and a pair of C6H60sI[P(0)(OMe)2]~anions, with the result that the O=P-0s-P=O
unit is no longer planar
and the 0 . . .O distance relative to that in
C6H60sI[P(0)(OMe)2]2H(a compound readily comparable
with 10) is substantially increased.
2.2.3. Bis(phosphane)metal Complexes
Since phosphanes are better donor ligands than phosphites, the C5H5M(PR3), complexes should also be
stronger metal bases than their phosphite analogues
C,H,M[P(OR),],. This prediction proved to be true, in particular when PMe, was used as a phosphane ligand.
Whereas the compounds C,H,M[P(OMe),12 (M = Co, Rh),
for example, can be protonated only with strong acids such
as CF3C0,H or HBFJs5],the conversion of the phosphane
complexes 18 and 19 into 16 and 17, respectively, can be
carried out using NH4PFg117*291.
The high basicity of 18 is
evidenced very clearly by the fact that the compound
reacts slowly even with methanol to form the hydridometal
cation [CSH5CoH(PMe3)2]
Some typical examples illustrating the reactivity of the
complexes 18 and 19 with electrophiles are summarized in
In addition to the Lewis acids SnC14,
Scheme 41*7,29,60J.
Me,SnCl, and Me3GeC1 also ZnCl,, HgCl,, and CuCl (as a
PMe3 adduct) were allowed to react with 18 and compounds having Co-Zn, Co-Hg and Co-Cu bonds isolatedIm1.Adducts of boron halides BX3 (X = F, C1, Br) with
14 are similarly accessibler6'!
20,21: R = M e
2 2 , 2 3 : R = Et
31:E = Ge
32,33:E = Sn
24,25:R = Me
26,27:R = P h
Li, ria, K
2 8 : X = C1
29:X = I
R, /R
R \R
16, 18, 20, 22, 24, 26, 30, 32: M.=Co
17, 19, 21, 23, 25, 27, 28, 29, 31, 33: M = Rh
Scheme 4. L = PMe,.
Scheme 3. [Col=(C5H4R')CoPMe3; R'=H, iPr, tBu; R=OMe.
The type of coordination of the M f cations by the oxygen atoms of the phosphonate groups was established by
X-ray analysis of the complex C6H60S1[P(O)(oMe)2]2Na,
which is related to 9 and to the cation of 13, and which
The differing behavior of methyl and ethyl iodides, on
the one hand, and isopropyl and tert-butyl halides, on the
other, toward bis(trimethy1phosphane) compounds of the
type C5H5M(PMe3), is quite remarkable. Under the same
conditions that 18 reacts with Me1 and EtI to form 20 and
Angew. Chem. Int. Ed. Engl. 22 (1983) 927-949
22, respectively (see Scheme 4), it forms the alkylcyclopentadienyl(hydrid0)metal complexes [(C5H4R)CoH(PMe3)2]Br
with iPrBr and ~ B u B ~ ~The
’ ~corresponding
hexafluorophosphates 34 and 35 (Scheme 5) can readily be converted
by NaH into the neutral compounds 36 and 37 which,
after further reaction with iPrBr or tBuBr, afford the doubly ring-substituted products 38, 39, or 40. Subsequent deprotonation with NaH yields the complexes 41-43. These
also react very rapidly with methyl iodide, i. e. an electrophilic addition involving the formation of a Co-CH, bond
is possible even when two bulky substituents are present
on the ring. The introduction of a third alkyl group to the
cyclopentadienyl ligand, however, is not possible; 41-43
are inert toward iPrBr and tBuBr[”].
- LL’ I.1) R B r
38 -40
34,36:R = i P r
35,37:R = t%u
41 -43
= iPr,R =
= R = tBu
Scheme 5. L= PMe,.
It is surprising that in 40 and 43 (R= R’=tBu) rotation
about the metal-ring bond is so strongly hindered that rigid
conformers are present in solution up to a temperature of
ca. 100°C. It is likely here that the size of the phosphane
ligands is an important factor. Even for 42 a frozen conformation can be detected below - 10°C. The fact that the
rotational barrier is lower than in 43, and that for 41 there
is absolutely no spectroscopic evidence of any hindrance
to free rotation can be explained by means of a model
based on a “gear wheel m e ~ h a n i s m ” l ’ ~ ~For
~ * ~a. 1,3(tBu)&H3 ring it is clearly not possible for the tert-butyl
groups to pass the PMe, ligand as they rotate about the
ring-metal axis (despite free rotation about the C-tBu
bond). However, if in the tBu groups a CH, group replaces
an H atom, i. e. tBu is substituted for iPr, the two ring substituents can probably arrange themselves in such a way
that at the moment of passing a phosphane the H atom of
the isopropyl group is directed “inwards” and-like the
teeth on a gear wheel-fits in-between the methyl groups
on the phosphorus (Fig. I).
MO calculations1631indicate that the barrier to rotation
about the ring-metal bond in half-sandwich complexes of
the type C5H5ML2and C5H5ML3is extremely low. Thus,
for purely electronic reasons the formation of preferred
conformers is not to be expected. Steric effects should thus
play the determining role. Recently, using NMR spectroscopy Pomeroy and Harrison‘641succeeded in demonstrating rotational hindrance in a similar compound
Angew. Chem. Int. Ed. Engl. 22 (1983) 927-949
Fig. 1. Proposal based on a model of the space-filling capacity of 1.3(iPr)2CsH,Co(PMe3)2 41 1621.
A proposal for the mechanism of the highly unusual ring
substitution of 18 is shown in Scheme 6II9]. It draws on
concepts which help to understand the aromatic nature of
cyclopentadienylmetal complexes such as Fe(C5H& or
C,H,Mn(C0),’41. We assume that in the initial step-completely analogously to that for the reactions of 18 with methyl and ethyl iodides-an electrophilic addition takes
place and the ionic compound A is formed; intermediate
B is produced from A by intramolecular rearrangement.
RReaction of B with the metal base D, which acts autocatalytically as indicated in Scheme 6, leads to E (34 and 35 in
Scheme 5), which after completing the exchange can be
isolated (along with small amounts of [C,H,CoH(L),]X) in
more than 90% yield.
Evidence that the hydride hydrogen atom in E originates from the cyclopentadienyl ring in 18 (and not from
the alkyl halide) is provided by the reaction of
C,D,CO(PM~,)~ with
data[”’ also support the formation of the postulated intermediate B as do our recently published results on a
[C5Me5Co(PMe3)(CNR)2]z+and CD3N02 in the presence
of traces of NEt,. In this case it is assumed[201that the intermediate formation of a cobalt(1) complex with an (q4C5Me5-exo-CD2N02)ligand occurs.
Scheme 6. L=PMe,; R=iPr, tBu.
The alternative to the proposal outlined in Scheme 6
would be a radical reaction. A one-electron transfer from
18 to the electrophile RX would then initially occur to
93 1
form a “contact radical pair”, similar to that postulated for
the intermediate in the reaction of carbonylmetalate ions
[M(CO),]- (M = Mn, Re) with the corresponding halo
c o m p l e x e ~ I ~Co-R
~ ~ . or C5H5-R bond formation could
then occur from the radical pair (C,HSCoLeO, RXe0}.
One reason we believe this alternative is possible, although for RX = MeI, EtI unlikely, is the result of the reaction of 19 with iPrBr or tBuBr which produces a mixture
of [C5H5RhBr(PMe3),]Br, [(C5H4R)RhBr(PMe3),1Br and
other compounds which have not yet been identified[331.
We assume that the cations involved are formed via free
radicals, whose appearance as intermediates seems probable on the basis of CIDNP measurements[331.A radical
cation, [CSHsRh(PPh3)2]e”, is also allegedly produced
as an intermediate in the formation of [(q’OC1~H~)Rh,(PPh3),l(BF4), from C ~ H S R ~ ( P P ~and
AgBF,‘66a1. McKinneyrWb’was also able to synthesize the
BF, salts of the corresponding cobalt complexes,
[CsHSCoLz]+ (L= PPh3, PEt3, PEtPhz, P(OiPr),), both by
oxidizing C5H5CoL2and by reaction of C5H5CoXLwith
TlBF, in the presence of L. Finally, the observation that
C5MesCo(PMe3), reacts with methyl iodide to form
[C5Me5CoI(PMe3)2]Iand not [C5MeSCoCH3(PMe3)2JIfits
well into the overall picture of conceivable radical reactions involving electron-rich half-sandwich compounds*201.
in organic solvents and thus precipitate. The heteroallene
complexes 46 and 48 react with free trimethyl- or triphenylphosphane to give the corresponding thio- or selenocarbonylcobalt compounds 50 or 51L71-731.
Brief mention will
also be made of the reactivity of 50 and 51 toward electrophilic substrates in Section 2.3.1.
c 5H sCo( CS)L
Scheme 7. L = PMe,
2.2.4. CO, Homologues QS Electroplriles
Carbon dioxide, a well-known electrophile, has recently
been allowed to react with electron-rich metal compounds,
especially in connection with attempts to “activate” CO,
by transition metals. However, there are very few notable
results of such attempts[671;CO, is evidently too inert toward “soft” bases, among which are the half-sandwich
compounds discussed here.
Rather more reactive than CO, itself are its homologues
(often designated as heteroallenes), some of which react
very readily with metal bases. The pioneering work in this
field was carried out by Baird and Wilkinson[681,who, as far
back as 1966, studied the reactivity of the compounds
RhC1(PPhs)3, Pd(PPh3),, and Pt(PPh,), toward CS,.
Among others, they isolated the first thiocarbonylmetal
complex, tran~-[RhCl(CS)(PPh~),]‘~~~.
That half-sandwich compounds behave in a similar way
was primarily established by Yamazaki et aI.r’s,221.By allowing CSH5M(PPh3),(M = Co, Rh) to react with CS, they
in modest yields-the
C5H5M(q2-CS2)PPh3in which, as in 44,the carbon disulfide is probably bonded to the metal via the C atom and
one S atom[”].
Because CsH5C~(PMe3)2
18 and C,H,Rh(PMe3)z 19 are
stronger metal bases than the analogous triphenylphosphane derivatives, they react significantly faster with
CS, and other COz homologues and yield-as shown in
Scheme 7 using 18 as e ~ a m p l e [ ~ ~ - ~whole
~ ] - a series of
novel metal@)and metal(II1) complexes. The initial step in
these reactions involves attack of the nucleophilic metal
center at the electrophilic C atom of the heteroallene. The
subsequent elimination of a phosphane ligand will probably be favored because PMe3 forms-at least with CS, and
CSe,-zwitterionic 1 : 1 adducts which are rather insoluble
X-ray structural analyses have been carried out on 44
and 451707721.
These have revealed very similar distances for
the three-membered CoCS ring and also an almost identical SCX bond angle (X = S or C) for the no longer linear
CS, or thioketene molecules in the complex. It thus appears that both heteroallenes possess rather comparable ligand characteristics. An X-ray analysis of 48 is now underway, and we are very optimistic that the result will confirm the structural proposal presented in Scheme 7. Recently, we were able to confirm for the palladium comthat the thiocarpound [1,2-C6H4(CH2PPh2),IPd(q2-CSSe)
bony1 selenide is indeed coordinated to the metal via C
and Se and not via C and Sr7,].
CsHsCo(PMezPh)2 53 behaves quite similarly to 18 towards CO, homol~gues~~’~.
Using CSz, compound 52, i. e.
the analogue of 44, is formed practically quantitatively,
while with CSez the five-membered ring complex 55 is obtained in addition to 54 (and small amounts of
C,H5Co(PMe2Ph)CSe3).The formation of 55 can be interpreted according to a 1,3-dipolar cycloaddition of 54 to a
second CSe, molecule. A compound having the composition C5H5Rh(PMe3)C2S4,which is structurally very closely
related to 551751, is formed in the reaction of
with CS2. As confirmed in this particular instance by X-ray analysis, the atoms of the metallaheterocycle are roughly coplanar. Moreover, the C-S distances in the ring and to the exocyclic sulfur atoms are almost identical, which suggests the presence of a highly delocalized n-electron system17s1.The reaction of 53 with
CSSe yields compounds analogous to 49 and 501731.It is
possible here also that an qz-CSSe complex is initially
formed which immediately reacts with PMezPh or CSSe to
give the products indicated.
Angew. Chem. Int. Ed. Engl. 22 (1983) 927-949
52 C & sCo(1)'-C S 2) P M e F h
(10) CS,
i 9 2-CSe
P M e Zp h 54
The results summarized in Scheme 7 and in equations
(9) and (10) generally support that-in contrast to the reactions of the half-sandwich complexes 18 and 53 with
CS2-use of the more reactive heteroallenes CSSe and
CSe, yields via the initially formed C,HSCo(q2-CSSe)PR3
or C5H5Co(q2-CSeZ)PR3compounds further, secondary
products, some of which are structurally novel. Presumably both the lower dissociation energy of the C-Se bond
(as compared to the C-S bond) and the greater reactivity
of the CoCSe three-membered ring in the dihapto-bonded
CSSe and CSe, complexes are responsible for their formation.
The CoCS three-membered ring in the compounds F
and G postulated as the primary intermediate in the reactions of 18 with COS and SCNR is probably even more labile. These intermediates react very rapidly with the PMe3
released in the primary step by SPMe3 elimination to form
the carbonyl or isocyanide complexes 56 or 57, and 58
which-as shown below-are themselves metal bases (see
Section 2.3.2). The course of the reaction of 18 with SCNR
outlined in Scheme 8 is supported by the fact that under
certain conditions (especially with excess isothiocyanate)
the metallaheterocycles 59 and 60 are isolated in addition
to 57 and 58. These heterocycles structurally resemble the
compounds 55 and CSH5Rh(PMe3)CzS4'42J.
PMe,Ph), CsHSRh(CO)PPh3, CSHSIr(CO)PPh3,etc." 'I, they
react with alkyl halides (especially with methyl iodide) to
form the acyl complexes.
Protonation to afford the hydridometal cations
[C,R,MH(CO)PR;]+ has proved to be successful in only a
few cases so far. Starting from the carbonyl(tripheny[phosphane) compounds CSHsM(CO)PPh3only the iridium derivative forms a stable cation of this type[11b1,
which merely
confirms the tendency to increasing basicity according to
the scheme 3d- <4d- < Sd-element (in the same triad) put
forward by Shri~er[~].
When we turn to the carbonyl(trimethylphosphane) complexes C5H5M(CO)PMe3a stable salt,
[C5HSRhH(CO)PMe3JX(X = BF,, CF3S03),can be isolated
for M = Rh1371.The corresponding C,Me,Rh compound
can be prepared similarly'381.
Only very little is known about the reactivity of phosphane(thiocarbony1) complexes, CsH5M(CS)PR3, towards
electrophiles. Faraone et al.[231
have studied the reaction of
CSHsIr(CS)PPh3 61 with excess methyl iodide and isolated
the thiocarbene compound 62.The reaction pathway postulated in reaction (12) is based on the plausible assumption that 61 behaves similarly to C5H51r(CO)PPhJ1lbl, and
that the intermediate H is produced primarily by oxidative
addition of a methylcarbenium ion to the metal. It is quite
possible that there is a further intermediate between H
and I involving a dihapto-bonded thioacyl group, as indicated by Roper et al. for the corresponding osmium comple~es[~~".
57,59: R
58,60: R
= Me
= Ph
Scheme 8. [Co]= CsH5CoPMe3
2.3. Complexes of the Type CSR,MLL'
2.3.1. Carbonyl(phosphane)- and
lXwcarbonyl(ph0sphane)metal Complexes
In terms of their nucleophilicity, the CsRsM(CO)PR3
(R=H, Me; M=Co, Rh, Ir; R'=alkyl, aryl) half-sandwich
complexes-numerous examples of which are known-occupy an intermediate position between the bis(phosphane)- and dicarbonylmetal compounds. As Graham et al.
have shown for C5H5Co(CO)PR; (PR; = PPh3, PMePh?,
Angew. Chem. Inf. Ed. Engl. 22 (1983) 927-949
In reacting with methyl iodide, the trimethylphosphane
compounds 50 and 64 yield-in contrast to 61-no products analogous to H, I, or 62.Even electrophilic addition
at the sulfur atom with formation of a methylthiocarbyne
complex (which occurs in the reaction of [W(CS)(CO)(diphos),] with CF3S03Me[771,cannot be proven1731.50 reacts
with CH212 to give a mixed product containing both
[C5H5CoI(PMe3),]I and, rather surprisingly, the binuclear
compound C5HsCo(p-CS)2Co(PMe3)CsH5,whose structure, with unsymmetric CS bridges as a characteristic feature, has been established by X-ray analysis[781.
The carbonyl- and thiocarbonyl(trimethy1phosphane)
complexes 56 and 63, and 50 and 64 behave very similarly toward the 16-electron species [(C,H,R)Mn(CO),]
(R- H, Me). The latter are typical electr~philes~~'~,
and accordingly react with the half-sandwich metal bases to give
the binuclear compounds 65-68 [reaction (13)] in high
We assume that an interaction initially
takes place between the nucleophilic metal center in 50,
56, 63,or 64 and the electronically and coordinatively unsaturated manganese atom, followed by stabilization re933
sulting from the formation of the CO or CS bridges. X-ray
indicate that a
crystallographic studies of 65[’*l and 66r831
cis-arrangement of the cyclopentadienyl ligands is preferred in each case. The CS group in 66 and 68 is exclusively
bridging (and is not in a terminal site), a fact which
emerges particularly from the IR data173J.In the synthesis
of 66 and 68 a second product is also formed containing a
C,H,Mn(CO), group as the ligand on the sulfur atom[811.
The yields obtained vary according to the amount of
50, 56,63,64
56,65: M = C o ; E
50,66: M = C o ; E
63.67: M = Rh;E
64.68:M = R h ; E
65,67:R = M e
66,68:R = H ( 1 3 )
2.3.2. Isocyanide@hosphane)metalComplexes
The isocyanide complexes C5H5M(CNR)PMe3(M = Co,
Rh) behave as nucleophiles toward both methyl iodide and
heteroallenes. Interestingly, the course of the reactions of
the cobalt compounds 57, 58 and 71 (Scheme 9) with Me1
is utterly dependent upon the group R of the isocyanide ligand. With R = M e and tBu, in pentane as solvent the
[C5H5CoCH3(CNR)PMe3]+ cations are obtained which
can be characterized readily as the PF, salts 69, 70. With
R = Ph, the corresponding cationic complex is rather unstable and rapidly rearranges to form the neutral imidoylcobalt(Ir1) compound 72 when the attempt is made to dissolve and recrystallize it from acetone/ether14,’. This subsequent reaction is to be expected in view of the similar ligand characteristics of CO and CNR and the comparable metal basicity of C5H5Co(CO)PMe3 56 and
C5H,Co(CNR)PMe3 (57, 58, 71).
However, if the methyl isocyanide and tert-butyl isocyanide complexes [C5HsCoCH3(CNR)PMe3]I are dissolved in acetone, no product analogous to 72 is formed.
When R=tBu another cation (73 with iodide as the anion)
having a dihapto-bonded imidoyl group is obtained after
migration of the CH3 group from the metal to the isocyanide
It is conceivable that an intermediate
analogous to 72 or 73 is also formed when
[C,H,CoCH3(CNMe)PMe3]I is dissolved in acetone; however, this is extremely labile and thus reacts with the solvent via a [3 + 21-cycloaddition to form the metallaheterocycle 74 (Scheme 9)Is4]. This product has been characterized by X-ray structure analysis[s5J. The reaction of
[C5H,CoCH3(CNMe)PMe3]I with acetonitrile gives 7 5 , a
five-membered ring compound containing Co; this is
probably formed via intermediate J. The structure of 75
has also been established by X-ray analysis1841.
The reaction of 57 with methyl and phenyl isocyanates
and of 58 with methyl isocyanate probably proceeds via
attack of the nucleophilic cobalt atom at the electrophilic
center of the heteroallene to yield the four-membered ring
compounds 76-78 (Scheme
Methyl isothiocyanate
reacts analogously with 57, 58, and 82. Surprisingly, the
- A,?,
57, 58,71
Scheme9. 57, 69, 74, 75: R=Me; 70, 71, 73: R=tBu; 58, 72: R=Ph.
metallaheterocycles 79-81 prepared in this way are also
formed from C5H5Co(CNMe)PMep and SCNR, which
suggests the intermediate formation of a four-membered
ring containing a sulfur atom. Cleavage of a C-S bond
and renewed ring closure leads to the thermodynamically
stable product bearing a methyl group on the ring nitrogen
atom. As indicated by X-ray crystallographic studies of the
complexes 79 and 80, planar CoCNC four-membered
76 - 78
57: R = M e
58:R = P h
76:R = R’ = M e
77: R = M e , R ’ = P h
7 8 : R = Ph,R’ = M e
79 - 81
79:R = Me
80:R = P h
81,82: R = C6Hi1
C f i & o ( CNR)PMeS
Scheme 10.
Angew. Chem. Int. Ed. Engl. 22 (1983) 927-949
rings are present whose C and N atoms are, as expected,
coordinated in a strictly trigonal-planar manner[861.Protonation and methylation of the heterocycles 76-81 lead to
formation of cyclic aminocarbenecobalt compounds,
which also opens up a route to metallaindole derivative~'~'].
In connection with the preparation of 80 from 57 it
should also be mentioned that the isocyanide compound
C,H,Co(CO)CNMe is formed when the complex
C5H,Co(CO)PMe3 56 -which is comparable with 57reacts with SCNPhLgol. In contrast, the reaction of
CSH,Co(CO)PPh3 with SCNPh leads to C5H,Co(q2SCNPh)PPh3, among others, and not to a four-membered
ring species related to 80L8*].
2.3.3. Olefin(phosphane)metalComplexes
The lack of evidence of attack by a proton at the metal
in the reaction of bis(o1efin)rhodium complexes with HCl
is certainly not mirrored by the mixed olefin(phosphane)
compounds 83-85. Using HBF, in ether, one obtains the
BF, salts 86-88, which are also relatively stable in solution (e.g. CH3N02)'281.The temperature dependence of the
NMR spectra and deuteration experiments indicate that in
solution a rapid proton exchange takes place involving the
olefinic protons and the hydridic proton. The positions of
the equilibria in reactions (15) and (16) are certainly
shifted to the side of the hydrido(o1efin) complexes, since
it has proved impossible either by 'H- or 13C-NMR spectroscopy to establish the presence of an alkylrhodium cation'*''.
83 - 85
86 - 88
83, 86: R = M e , R' = H
84, 87: R = R' = M e
85,88:R=iPr,R'= H
The reaction of the styrene complex 89 with HBF, or
CF3COZH/NH4PF6leads to a product corresponding analytically to a hydrido(o1efin) compound analogous to 8688. The I R and NMR spectra of this product, however,
show no evidence of an Rh-H bond. As proved by X-ray
structure analysis (X = PF,), in the crystal exclusively 90 is
present, i. e. the methylbenzyl group formed from the proton and the styrene is linked to the metal via a n-ally1 type
bond[281.The formation of 90 probably takes place via the
intermediates K and L which-as demonstrated by labeling experiments-are in equilibrium with M and 90
(Scheme 11). The fact that the complex 90 -in contrast to
other q3-methylbenzyl metal compounds-possesses a rigid rather than a fluctuating structure in solution may arise
because the rhodium(n1) has the preferred octahedral
[[ R h 1-C H2C H2P h
Scheme 11. [Rh]=CsHsRhPMe3; X=BF,, PF,.
The methylation of the rhodium atom in 83 takes place
similarly [reaction (17)]. The ethylene(methy1) complex
formed, 91,is stable at room temperature and reacts only
upon warming in tetrahydrofuran (THF) to eliminate C2H4
and yield 92.The analogous compound obtained by allowing 89 to react with Me1 (PhCHCH2 instead of CzH4) is
much more labile than 91 and rearranges to 92 when dissolved in CH3N0J2']. It would appear that in cations of
composition [CsRsRhCH3(C2H3R?PR3]+ the positive
charge is stabilized by increasing donor character in the ligands. It is thus easy to understand why the compound
[C5H5RhCH3(CzH4)PPh3]Idescribed by Graham et
rapidly eliminates ethylene even at room temperature,
whereas 91 does not. However, what does not fit into this
picture is the fact that the cations [CSMeSRhCH3(C2H,)L]
(L= PMe3,
CsMesRh(C2H,)L are very labile, and in solution in the
presence of iodide they are spontaneously converted into
the neutral complexes CSMeSRhCH3(L)IL'61.
[Rh] = C5H&hPMe3
It is interesting to note that protonation of the propenerhodium compound 84 yields the two diastereomeric pairs
of 87,not in equal amounts but roughly in the ratio 70 :30,
i. e. an asymmetric induction occurs[891.Since the configuration at the two chiral centers (the metal and the methylsubstituted carbon atom of propene) is constantly changing as a result of equilibrium (16), one diastereomer is rapidly transformed into another. This is evidenced in the
NMR spectra by the line broadening of the signals from
the ring and phosphane protons or carbons atoms1281.
Angew. Chem. Int. Ed. Engl. 22 (1983) 927-949
A number of subsequent reactions involving the ethylene(hydrid0) cation 86 should also be mentioned. With
C1-, Br- and I - as reagents, even at room temperature in-
sertion of the olefin into the Rh-H bond occurs to form
the ethyl(ha1o) complexes C5H5RhC2H5(PMe3)X(X =C1,
Br, I)f371.In the reaction of 86 with ethylene an insertion
also occurs, although the product formed, 93, is stable
only under a C2H4 atmosphere and transforms into 86
when any attempt at isolation is made. However, PMe3 undergoes nucleophilic addition to the ethylene to yield the
0-phosphonioethyl compound 94[371.Insertion of C'H4
into the Rh-CZH5 bond in 93 could not be detected, even
with ethylene under pressure.
- 1
Displacement of olefin occurs when 86 reacts with
P(iPr)3 and
BF4 salt
[C5H5RhH(PMe3)P(iPr)3]+is obtained. This can readily be
deprotonated, providing access to the mixed bis(phosphane) complex C5H5Rh(PMe3)P(iPr)3[371.
Scheme 12. R=iF'r.
ligand with phenyl substituents in c i s - p o ~ i t i o n sAmong
the subsequent reactions of this compound, special mention must be made of the rearrangement to the metallaheterocycle 102, which is isomeric with 100. This rearrangement takes place in methanol upon addition of NH,PF, or
traces of CF3C0,H. We assume that the trifluoroacetate
group of 101 is initially displaced by the solvent, and that
the cationic intermediate formed reacts by ortho-metallation. The lability of the Rh-OCOCF, bond in 101 is evident from the smooth reaction with NaI or CH3MgI to
form 1031241.
The five-membered metallacycle in 102 possesses an approximate chair-like conformation which, together with the varying C-C bond distances, suggests that
there is no delocalized n-electron system pre~ent'"~.
2.3.4. Alkyne(ph0sphane)- and
finylidene(phosphane)rnetal Complexes
The reversibility of olefin insertion into the metal-hydride bond in hydrido(o1efin)metal complexes is well doc~mented['~],
as is evident, for example, in reactions (15)
and (16) and in Scheme 11. On the other hand, there is no
proof of a corresponding equiiibrium between alkyne(hydrido)- and vinylmetal compounds. All presently known
syntheses of vinyl complexes from alkynes have been carried out by allowing a metal hydride to react with CzRz or
RC2H, i.e. the M-CR=CHR(or M-CH=CHRand
linkages have not been produced by reactions between alkynemetal compounds and acids HXf9'].
The compounds 95 and 100 recently prepared by us,
which are analogues of the olefin complexes 83-85, react
with CF3COzH or HBF, even at room temperature["].
After addition of PF,, 95 forms the syn-1-methylallyl
complex 98, which is slowly converted into the thermodynamically more stable anti-isomer 99rg11.
Investigation of
the reaction pathway indicates that in the presence of a
weak acid such as H 2 0 (on A1203)an initial rearrangement
of 95 to the isomeric methylallene compound 96 is followed by protonation to yield the ally1 complex. The synconfiguration of the I-MeC,H, group in 98 has been confirmed by NMR data and by the formation of the cis-2-butene compound 97 upon addition of a hydride ion.
The reaction of 100 with CF3C02H yields the complex
101, which is known from X-ray analysis to contain a vinyl
Scheme 13. R=iPr.
Investigations are currently underway on the protonation of the cobalt compound C5H,Co(CzPhz)PMe,'931.The
work on rhodium species (Schemes 12 and 13) does not allow any definite statement on whether the attack of the
acid takes place at the metal or at the C=C triple bond.
The exclusive formation of 101 from 100, which can be
understood in terms of a cis-addition would appear to support a primary interaction between the metal and the acid.
However, in view of results on the stereochemical course
of reactions of metal hydride complexes with alkyne~["~,
caution seems to be indicated here.
Angew. Chem. Int. Ed. Engl. 22 (1983) 927-949
An attempt to obtain the acetylenerhodium compound
CSH5Rh(C2H2)PiPr3analogous to 95 or 100 by allowing
104 to react with NaCsH, led rather surprisingly to almost
quantitative formation of the vinylidene complex 1051941.
Depending on the reaction conditions, the phenylacetylene compound 106 reacts with NaC5H5 to form either the
alkyne complex 109 or the vinylidene complex l10[951.
conversion of this type of a I-alkyne to a vinylidene ligand
in the coordination sphere of a transition metal has already been observed in related cyclopentadienylmanganese, cyclopentadienyliron, and cyclopentadienylruthenium compounds. In general, it has been explained by assuming the formation of an alkynyl(hydrid0) intermediate’”’. Such an intermediate was first isolated during the
reaction of 106 to form l10[951.
The equilibrium between
106 and 107 in ether or toluene can be shifted completely
to the right by addition of pyridine to yield the chiral octahedral complex 108. The latter reacts with NaC5Hs in
THF at room temperature to give the phenylvinylidene
compound 110. If, however, a solution of 108 in THF is
very rapidly treated at 0°C with an equimolar amount of
NaC5HS and worked-up in the cold, the square-planar
complex 111 is is01ated‘~~I.
This complex reacts with cyclopentadiene virtually quantitatively to give 110; when C5D6
is used, CsD,Rh(=C=CDPh)PiPr, is
should be a high electron density at the metal center. Accordingly, the vinylidenerhodium complexes should also
behave as metal bases.
The first tentative steps in this direction seem to support
this suggestion. Thus, 105 reacts with HC1 (1 :1) in benzene to yield the chloro(viny1)rhodium compound 112,
whose formation may take place via a hydrido(viny1idene)rhodium cation[971.Using CHzNzin the presence of
Cuzc leads to the allene complex 113 but not to the isomeric carbene compound N which should result from CH,
addition to the vinylidene C=C double bond. The electrophile CH, thus clearly prefers the more electron-rich
Rh=C instead of the C=C bond in adding to the “metallaallene” moiety C,H,(L)Rh=C=CH,. It thereby provides
a new synthetic route to allenemetal complexes[971.
H. .H
114 -116
115: E = Se
116: E = Te
Scheme IS. R=iPr.
Recently, by making use of the electrophilic addition of
a sulfur, selenium, or tellurium atom to the Rh=C bond in
The reaction sequence shown in Scheme 14 illustrates
that although the conversion of a 1-alkyne into a vinylidene ligand actually involves an alkynyl(hydrid0) intermediate there is no intramolecular hydride shift from the metal to the carbon as originally supposedr961.In fact, a twostep elimination/addition is responsible for the final isomerization which occurs. Rearrangement of 109 into 110 (or
vice versa) is not observed either on irradiation or on after
leaving the compound to stand for a considerable time in
As established by X-ray structure analysis, there is a
fairly short bond in complex 110 between the metal and
the a-C atom of the phenylvinylidene which is some 1520 pm shorter than in carbenerhodium c o r n p ~ u n d s This
distance points to considerable overlap of the orbitals involved in the Rh-C-71 bond, which indicates that there
Angew. Chem.
Ed. Engl. 22 (1983) 927-949
105, it became possible for the first time to prepare transition-metal compounds with thio-, seleno- and telluroketenes, CH,=C=E, i.e . with molecules which are very
short-lived (E = S, Se) or unknown (E =Te) in the free state
(Scheme 15)[941.The complexes 114 and 115 are remarkably stable both in the solid state and in solution, whereas
the telluroketene compound 116 reacts slowly in solution
to form 105 and metallic tellurium. X-ray analysis of 114
has shown that in the crystal both enantiomers are present‘”]. Overall, there is a close analogy between the structural data of 114 and those of complex 45 obtained from
the half-sandwich complex 18 (see Scheme 7). This complex, 45, contains the thioketene 1,1,3,3-tetramethyl-2thiocarbonylcyclohexane as ligand, which in contrast to
CH2=C=S is stable in the free form. Behrens et UZ.[~*] have
isolated from this compound and, more recently, from di-tbutylthioketene, mono- and binuclear thioketene complexes by reaction with various carbonylmetal compounds.
In these thioketene complexes the appropriate thioketene
is coordinated either via the sulfur (i. e. monohapto) or as a
bridge between two metal atoms. Some iridium(1) and platinum(o) complexes with $-bonded bis(trifluoromethy1)thioketene described in 1970 by Stone et aZ.[991are
structurally related to 114 (and 45); they were synthesized,
however, not from (CF,),C=C=S but rather from heterocyclic, CF3-substituted sulfur compounds.
2.4. Reactions of C5R5ML2and C5R5MLL’with
Dihslogeno- and Trihalogenomethanes
In a brief review on the chemistry of carbodiphosphoranes which appeared in 1979, Schmidbaurl’ool advanced
the provocative suggestion that the bonding patterns in
C(PR3)2molecules could be described not only in terms of
the resonance structures a and b but also as an extreme
case in terms of c.
We took this as a challenge and attempted to prepare
binuclear metal complexes according to reaction (20). In
such complexes both of the metal atoms would be linked
only by a carbido bridge. The metal base, ML,, would then
act as a donor toward the “naked” C atom, just as the
phosphane group in the above-mentioned formula c for
To determine what factors influence the conversion of a
halogenmethyl(phosphane) complex into a halogen(methylenephosphorane) complex, we changed both the dihalomethane and the metal base. 19 also reacts smoothly
with CH2Cl2 (though somewhat less rapidly than with
CH,I,) to form the cation [C5H5RhCH2CI(PMe3),]+,
whose PF6 salt 120 is exceptionally stable[’091.When 19
reacts with CH2Br2a mixed product i s formed from which
after dissolving and reprecipitating with NH4PF6 the bromomethyl complex 121 and the isomeric methylenephosphorane 122 can be isolated; the latter is obtained quantitatively after addition of NEt,. Isomerization of 120 could
not be observed even after it was allowed to stand for some
considerable time in the presence of triethylamine[391.
Scheme 16. [Rh]=C,HsRhPMe,.
As a test molecule we first selected the bis(trimethy1phosphane)rhodium complex 19, which reacts with CH21Z
under exactly the same conditions as it does with methyl
iodide to form the iodomethylrhodium(I1I) compound 117
in 62% yieldl’O’l.The latter is stable for several days in nitromethane, although in dimethyl sulfoxide it rapidly loses
“CH,” to form [CSH5RhI(PMe3)2]I.
The presence of a reactive C-I bond in the cation of
117 is evidenced by its reaction with PMe3, which leads
quantitatively to 118. With NEt3, no nitrogen-ylide complex analogous to 118 is formed from 117, which rather
surprisingly isomerizes to 119[’0’1.The anion clearly is not
involved in this rearrangement since the BF4 salt also
reacts analogously to 117.
[Rh]= C,HSRhPMe,
The conversion of 117 to 119 is catalyzed not only by
NEt3 but also by NaOMe and-disappointingly in view of
our stated objective-by the complex 19. Even after
changing the reaction conditions and replacing 19
by the more reactive half-sandwich compounds
C5HSRh(CNMe)2,C S H 5 F e ( C 0 )and
~ , C5H5Mo(CO);[39*109’
it is not possible to isolate a binuclear complex having a
CH, bridge without a metal-metal bond.
The products obtained by starting from the mixed
bis(phosphane) complexes 123, 124 and the phosphane(phosphite) compounds 129- 131 are summarized
in Scheme 17[27,102J.Two striking points are that in the
rearrangement it is always the better donor ligand which
migrates and that in the case of 132-134 the isomerization takes place even without adding NEt3 or another base.
Kinetic measurement^"^^ reveal that the rate of the reaction of 132 to 135 is-after an induction period-practically independent of the concentration of the educt and increases with increasing polarity of the solvent. Since addition of iodide ions has no effect on the rate, the mechanism might be compared to that of some anionotropic rearrangements of organic compounds in which a concerted
pathway is very probable on the basis of labeling experiment~“~~
’ . observation that the rate of isomerization
of 132 to 135 is significantly increased by addition
NEt3 and
fact that
does not rearrange are
consistent with a non-stepwise 1,2-shift of the iodide and
phosphane. Such a mechanism could also apply to the
reactions of other phosphanemetal compounds, e. g .
Co(PMe& Pt(PPh3)4, or Pt(PEQ3 with dihalomethanes,
which lead to methylenephosphorane complexes[’04-’061.
Similarly to 19 (Scheme 16), 123 and 124, the chelate
complexes 141 and 142 also react with CH212by oxidative
addition to yield the iodomethyl compounds 143 and
Although 144 is inert and does not isomerize even
in the presence of NaOH, the reaction of 143 with NEt3,
involving insertion of CH, into one of the Rh-P bonds,
leads to the quantitative formation of the metallaheterocycle 145 (Scheme 18). The Rh-I bond in the latter is fairly
labile and thus at room temperature in the presence of
Angew. Chem. Int. Ed. Engl. 22 (1983) 927-949
123, 124
125, 126
127, 128
123, 125, 127, 130, 133, 136, 139: P R 3 = P M e z P h
124, 126, 128, 131, 134, 137, 140: P R 3 = P M e P h ,
129, 132, 135, 138: P R 3 = P M e ,
138 -140
CSH5RhCOCH3(PMe3)1, cannot be observed['"!
course of the reaction of 83 with CH212 depends on the
solvent used. Whereas in ether only 147 is formed, in methanol mainly the compound C,H,Rh12(PMe3) is produced
uiu elimination of CHz11071.
The reactivity of the C-I bond towards nucleophiles
renders the complex 147 (rather like 117) a very suitable
starting material for the synthesis of numerous novel organorhodium compounds (see Scheme 19). In particular, formation of the complexes 152, 154, and 155 should be
mentioned in which the ylides Ph3AsCHz, MezSCH2 and
Et3NCH2, which are unstable in the free state, are fixed at
the metal"071.
Scheme 17.
PMe3 (and subsequent addition of NH4PF6) the complex
146 is formed. The relatively broad 'H-NMR signals indicate that in solution a rapid interconversion of the two
possible chair conformations probably takes place[271.An
alternative method of preparing 146 from 147 and
C2H4(PMeZ),proved not to be successful.
I Rh I\
142, 144: R
/CHzNE t,
Scheme 19. [Rhj=CsH5RhPMe3.
The reactions of the cobalt compounds analogous to 19,
129, and 141 with CH2Cl2. CHZCII, and CHz12 proceed
very rapidly even at -78"C, although they give no products containing a Co-CH,Cl or Co-CHZI bond. In addition to paramagnetic species, all that can be isolated from
the reaction mixture are the halometal complexes
[C5H5CoXL2]X(X = C1, I)[loZ1.
Investigations of the reactivity of CSHsCo(CO)PMe3with dihalomethanes have proven
that this result is not due to the instability of methylenephosphoranecobalt(Ir1) compounds.
The metal bases C5HSRh(CO)PMe3 63 and
CSH,Rh(C2H4)PMe3 83, which are weaker nucleophiles
than the bis(phosphane) compounds C5H5Rh(PMe3)PR3
19, 123, and 124, also react under very mild conditions
with CHz12[1071.In this case the neutral complex 147 is
formed. The unique course of the reaction of the carbony1 compound 63 is remarkable insofar as an
acyl complex, C5H,RhCOCH21(PMe3)I, analogous to
63: L = C O
83: L = C,H,
+ L
Angew. Chem. Int. Ed. Engl. 22 (1983) 927-949
The isomer 156, obtained together with 155 in the reaction of 147 with NEt3, is formed quantitatively when 147
reacts with NaOH in a two-phase benzene/water system in
the presence of triethylbenzylammonium chloride acting
as phase transfer catalyst (FTC)"091. The expected
C,H,RhCH20H(PMe3)I - by analogy to the synthesis of
153 from 147 and NaOMe - does not occur, although the
former could be a suitable starting material for the preparation of Rh=CH, complexes.
In contrast to OH-, the weaker bases SH-, SeH- and
TeH- effect a substitution of the C-I bond in 147 and
thereby open up a simple route to thio-, seleno- and telluroformaldehyderhodium compoundsI1lol.The structure of
158 has been confirmed by X-ray structure
The synthetic pathway adopted may involve proton abstraction from the initially formed intermediate 0 by EHe
(which is present in excess) and a subsequent substitution
of iodide by the more nucleophilic chalcogen atom in P to
form the three-membered ring of the product.
Independently, but at the same time as us, Roper et
obtained the chalcogenoformaldehydeosmium complexes Os(q2-CHzE)(C0)2(PPh3)2(E = S , Te) from osmiumhalomethyl or osmiumformaldehyde precursors. Moreover, Herberhold et ~ l . [ " and
~ ] Herrrnann et ~ 1 . ~ recently
succeeded in preparing binuclear manganese compounds
157: E = S; 158: E
159: E = Te
[ R h I,
,C HX2
+ EHo
[ R h ] = C5H5RhPMe3
168: X = Rr; 169:
= I
Scheme 21. IRhl=C5HsRhPMe3.
with CH2S, CH2Se, or CH2Te as the bridging ligand by allowing [CSRsMn(C0)2]2E,(E=S, Se; R=H, Me; n = l or
2) or [C5H5Mn(CO)J3Teto react with CHZN2.
Several results pertaining to the reactivity of complexes
157 and 158 toward Lewis acid substrates are summarized
in Scheme 20. Accordingly, it is clear that electrophilic addition occurs directly at the chalcogen atom in the reactions with CF3S03Me and C5HSMn(CO)2THF.However,
in the reaction with CF3C02H/NH4PF6it is possible that
the metal is initially protonated and that this is followed by
a hydrogen shift to the CH2E carbon atom. 162 and 163
are the first examples of binuclear compounds in which a
thioformaldehyde or selenoformaldehyde ligand links two
different metal centers'"''.
157, 158
160, 161
162, 163
c H3
= S
160, 162, 164: E
161, 163, 165: E
= Se;
CSH5RhBrz(PMe3)or CSHSRh12(PMe3).The complexes
CsMe5RhCHX2(CO)X,which correspond to 168 and 169
insofar as their properties are concerned, are formed from
CSMeSRh(CO)z and CHX3 (X=Br, I)[1161. 169 and its
CSMes analogue represent the first isolated diiodomethyl
transition-metal compounds whose reactivity is of interest
on the basis of the results presented in Scheme 19 and
reaction (24).
The carbonyl(phosphane)cobalt complex 56 is considerably more reactive toward dihalomethanes than its rhodium analogue 63. At -60°C with CH2CII the compound
170, which has been characterized by NMR spectroscopy,
is obtained; even at -50°C in acetone, this eliminates
CH2 to form 171. Addition of an equimolar amount of
PMe3 or P(OMe), to 170 results in formation of the cationic chloromethylcobalt complexes 172 and 173, which
react at room temperature with trimethylphosphane to give
174 and 175, respectively. These compounds can also be
prepared from 56 or 176 and CH2X2/PMe, simply by
mixing the reactands (molar ratio 1 : 1); C5H5Co(CO)2is
also formed. In this case it can be assumed that an intermediate analogous to 170 is formed in which the cobalthalogen bonds are evidently so labile that a very rapid subsequent reaction with the phosphane takes place. Astonishingly, the formation of 174 and 175 is independent of
the polarity of the solvent; in toluene, ether, or acetone no
difference is observed either in the reaction rate or in the
composition of the products (Scheme 22)" ''1.
Scheme 20. [Rh]=CsHSRhPMe3.
-6 0 T
The ethyIene(phosphane) complex 83 reacts not only
with CH212[see reaction (23)] but also with other di- and
trihalomethanes["61. The reaction with CH2CII proceeds at
almost the same rate as with CH212, which is plausible
since in both cases the cleavage of a C-I bond occurs. In
the case of CH2Br2 only a very slow reaction takes place,
which (in ethedpentane) affords 167 after two days at
room temperature or after 15 h at 45°C (Scheme 21).
CH2C1, does not react at all under similar conditions. The
observed difference in the reaction rate reflects very clearly
the C-X bonding energy differences; in going from C-I
to C-Cl there is an increase of cu. 105 kJ/mol.
The compounds 168 and 169 obtained using CHBr3 and
CHI3, respectively, can readily be handled in the solid
state, although they are labile in solution. In nitromethane,
for example, they form, among other products,
172, 174: L = P M e ,
173, 175: L = P ( O M e ) 3
176: C5HSCo(CO)P(OMe),
Scheme 22. [Co]= CSHsCoPMe3
Angew. Chem. I n t . Ed. Engl. 22 (1983) 927-949
Preparation of the thioformaldehyde- and selenoformaldehyde cobalt complexes 177 and 178, respectively, is possible via the conjectured intermediate Q formed from 56
and CH2XZ.At least as important as this result is the isolation of the first dihapto-bonded thioacetaldehydemetal
compound, which is formed-in addition to small amounts
of CSH,CoBr2(PMe3)-according to reaction (26)[1171.
the chalcogenoformaldehydes, CHzE, thioacetaldehyde is
also unstable in the monomeric form and has previously
not been fixed in a complex, despite ligand characteristics
which are comparable to those of CH2S. In this context it
should be mentioned that Herrmann et uZ.[”~’ were able to
generate not only CH,Se and CHzTe but also the dimethylseleno ketone Me,CSe-which is also unstable in the
free state-as bridging ligand between two Mn atoms.
177: E
= S;
178: E = Se
‘ S
[CoI = C’,fI,Co€’hle3
The reactivity of the complexes 177 and 178 corresponds to that of the rhodium compounds 157 and 158.
Methylation can be achieved even with methyl iodide, a
fact which confirms the great nucleophilicity of the sulfur
and selenium atoms in CoCE three-membered rings[”71.
180: E = S; 181: E = Se
[CO] = (‘,iISCoPMe,
3. Metal Bases with Six-Membered Ring Ligands
3.1. ds-Systems
From the structural and bonding point of view, the aromatic metal compounds C6&ML2 and C6&MLL’ (M = Fe,
Ru, 0 s ) are comparable to the cyclopentadienylmetal complexes, C,R,ML, and C5RSMLL’(M = Co, Rh, Ir). The metal atom now is in the oxidation state zero and, like Co’,
Rh’ and Ir’, thus represents a ds system. In contrast to the
cyclopentadienyl complexes C5R5ML2,until recently principally only compounds of the type C6&MLZ were known
with L, = diolefin (e. 9.C4H6, 1,3-C6Hs, 1,3-C7Hlo7and 1,5CSHI,). Most of these were prepared by the “Gngnard
method”[”*]. Before 1978 (Le. when we began our studie~[”~I),
only a few iron complexes with manodentute ligands L existed, namely with L=PF3[’”] and CO[’21’.Because of the pronounced n-acceptor properties of trifluorophosphane or carbon monoxide, these complexes are
probably only weak metal bases.
Angew. Chem. Inl. Ed. Engl. 22 (1983) 927-949
3.1.1. Synthetic Methods
The synthesis of the bis(trifluorophosphane) complexes
(C6HsMe)Fe(PF3)2and C6H6Fe(PF3)2has been carried out
using metal atom vaporization techniques”201. By this
means, the corresponding phosphite compounds
C6%Fe[P(OMe)3]2 (c6%= C6H6, C6HsMe, 1,4-C6H,Mez7
1,3,5-C&Me3), and (C6HsMe)FeIP(oEt),lZ, as well as the
phosphane complex (C6H5Me)Fe(PMe3), are also accessiblei“21. In the cocondensation of iron atoms with arenes it
seems likely that molecules such as Fe(C6&) or Fe(C6%)2
are formed initially and that these then react with
the phosphite or phosphane. The preparation of
C6Me6Fe(C0)2 proceeds Via the Complex Fe(C6Me6)2,
which can be isolated as a solid. In the reaction, an q4bonded six-membered ring is probably substituted by the
CO groups[”ll.
For the synthesis of the ruthenium and osmium compounds C6%ML, (L=PR,, P(oR),) and C&MLL’
(L= PR,; L‘= P(OR),, CO, CNR, CzH3R), a reductive procedure employing NaCloHs as the reducing agent has
The starting materials for the proproved to be
cedure are the salts of the cations [C6&MXLZ]+ or
[C6%MXLL’] *, which can be obtained from the halidebridged compounds [C6%MX2I2by one of the routes indicated in Scheme 231’24~’31,’34!
It depends primarily upon
the ligands L and L’ which is the best preliminary step for
the preparation of the salts. The acetone complexes can be
isolated only for M = Ru (R= H[1241,Me[’2s1; X = CI;
L=PMe3, PMeZPh, PMePh,, PPh3); for M = O s only the
binuclear dications [ ( C ~ H ~ ~ S L ) ~ ( ~are
L - Xstable.
[C&RUCl(PR3)2]PF6 (PR3= PMezPh, PMePh,, PPh3) has
also been carried OUt by allowing [ ( C ~ H ~ R U ) Z ( ~ - C ~to) ~ ] P F ~
react with the appropriate phosphane[’261.
Apart from NaClOH,, experiments were also performed
using Na, Na/Hg, NaH, NaNH,, Li,CH,, Li2CsH8, Mg,
Co(C5H&, and diphenylketyl Na as reducing agents although no satisfactory result could be achieved. The successful synthesis of the metal(0) complexes C6&ML, and
C6&MLL’ using NaCloHs has the disadvantage that it is
not at all easy to completely separate the naphthalene
formed during the reduction from the highly volatile and
very soluble products. In the case of thermally relatively stable compounds (e. g. M = Os, L = PMe,, L‘= C2H4, C3H6),
this is possible uia the hydridometal cations and their deprotonation with NaH[lz7].
Unexpectedly, the complex C6H60s(PMe3), could not
be obtained by reducing [C6H60sI(PMe3)z]PF6 with
NaCloHs. Even the attempt to exchange the olefink ligand
in 183 and 184 by trimethylphosphane did not succeed.
Rather the hydrido(phenyl)osmium(Ii) compound 186
forms by oxidative addition of an (sp2)C-H bond to the
electron-rich metal center. When R = H it is possible to isolate the intermediate 185. The conversion of the C&OS
into the OsH(C6Hs) moiety occurs strictly intramolecularly, as may be seen from the quantitative formation of 186
It may be that in the initial step
in C6D6 as
after attack of the trimethylphosphane at the metal a reorientation of the benzene from an $-to an q4-mode of
coordination occurs. This has been postulated by Muetter94 I
PF 6
2 NaCloHs
ties et al., for example, in the catalytic hydrogenation process based on areneruthenium
Kinetic measurements are consistent with the conjectured reaction sequence in Scheme 24[1301.
- C,H3R
3.1.2. Reactions with Electrophiles
As might be expected, the complexes C6&ML2 and
C6&MLL‘ react with Bransted acids and with methyl iodide or [OMe,]BF4 to form salts of the cations
[C6&MELz] + and [c6&MELL’] +,respectively. These are
considerably more stable (thermally and towards oxidizing
agents) than the metal(o) compounds. Here, too, the metal
basicity depends quite significantly on the donor strengths
of the ligands L and L’. Whereas, for example, the
(c&= C6H6, 1,4-C6H4(Me)iPr, C6Me6) react even at
-78°C with NH4PF6 in methanol to form the compounds
[C6&RuH(PMe,),]PF6, the formation of PMePh, and PPh3
analogues occurs only at room temperaturei1231.Moreover, for the synthesis Of [C6Me6RUH(PMe3)CO]PF6,
[C6H6RuH(PMeS)CzH41BF4, and
[(C6H5Me)FeH(P(OMe),),]BF, it is necessary to use acids stronger than
NHZ, such as CF,CO,H or HBF4[1Z2a~1z3~1271.
It is generally found that 1) the presence of a good n-acceptor ligand such as CO or C,H4 weakens the metal basicity and 2) the osmium(0) compounds are more reactive
than the homologous iron and ruthenium complexes.
Methylation proceeds just as smoothly as protonation.
Here, the reaction rate here also depends characteristically
on the metal and on the nature of the ligands. Thus, for example, C6H60s[P(OMe),]2 reacts with methyl iodide to
form [C6H60SCH3(P(OMe)3)2]Ii’231,
whereas [OMe3]BF4is
necessary for the formation of the corresponding cationic
iron complex[122a1.
In the case of bis(ph0sphane)ruthenium
compounds of the type c6&Ru(PR3)2,there is a decrease
in the rate of methylation along the sequence PMe, >
PMe2Ph > PMePhz > PPh3 (for C6&=C6H6) and C6H6
> 1,4-C6H4(Me)iPr > C6Me6 (for PR, = PMe3)11231.
indicates that, in addition to electronic effects, steric effects also play a role in determining the ease of attack of
the electrophile at the metal atom in C6&MLZ complexes.
Angew. Chem. Int. Ed. Engl. 22 (1983) 927-949
Only C6H6Ru(PMe3)2 191 reacts smoothly with ethyl
iodide to give the corresponding ethyl compound
(which is isolated as the PF6 salt
When 191 is allowed to react with n-propyl and
n-butyl bromides or iodides, in addition to the expected
products [C,H,RUR(PM~~)~]X,
the corresponding metal
halide complexes [C6H6RUX(Pkk3)2]Xare also formed. As
for the analogous reactions of the cyclopentadienyl compounds, C5R5M(PMe3)2(see Section 2.2.3), this result suggests the formation of radical intermediates. Of the Lewis
acidic trimethylelement halides Me3ECI (E = Si, Ge, Sn),
only the tin compound affords a product having the composition [C6H6R~EMe3(PMe3)2]Xupon reaction with
Under the same conditions, Me3SiC1
and Me3GeC1form mixtures of products in which only the
hydridoruthenium cation [C6H&H(PMe3)2]+ can be unequivocally detected.
Recently, two complexes containing an osmium-gold
bond have been prepared by oxidative addition of
PPh3AuCI to C6H6OS(PMe3)CNR (187, 188)['311.
According to the IR data of 189 and 190, the possibility of the
isocyanide forming a bridge between the two metal atoms
can be excluded. The PPh3Au+ cation thus behaves completely analogously to a proton towards the compounds
187 and 188, which is certainly to be expected from their
isolobal relati~nship"~~].
M e , P /OS\CNR
187, 188
189, 190
187, 189: R = M e
188, 190: R = p-Tolyl
Overall, the metal basicity of the arenemetal complexes
C6&MLZ would even appear to exceed that of the cyclopentadienylmetal compounds C5R5ML2. While the PF6
salts of the cations [C5HSMH(PMe3)2]+16 and 17 readily
react with NaH to form the neutral complexes
C5H5M(PMe3)218 and 19, respectively, deprotonation of
the analogous benzeneruthenium and benzeneosmium
compounds [C6H6RuH(PMe3)2]+and [C.&OSH(PPh3)2]
does not occur under the same conditions['231.These cations react very slowly even with MeLi or tBuLi to form
191 and C6H60~(PPh3)2r
respectively. Measurements (MS
and PES) currently being made to determine the ionization
potentials should provide information on how these values
differ for typical derivatives of the two complex series
C5RSML2and C6&ML2, and how this may be correlated
with the difference in metal basicity.
been possible to eliminate a hydride ion from the methyl
compound [C6H6RUCH3(PMe3)2]C.Reaction of the hydrido complex [C6H6RuH(PMe3)2]PF6with [CPh3]PF6 in
acetone leads to the PF6 salt of the dication
[C6H6RU(PMe3)2(0CMe2)]'+. This salt contains an unusually inert Ru-OCMe2 bond, and is therefore not a
suitable starting material for the preparation of octahedral ruthenium(I1) complexes of general type
(L= PR3, P(OR),, co, e t ~ . ) " ~ ~ ] .
195: R
197: R
= M e ; 196: R = i P r
= Ph; 198: R = O M e
Scheme 25. [Rh]=C6H6Ru(PMe3)2.
Nucleophilic addition of a phosphane to coordinated
ethylene has also been observed in the reactions of
[C6H6MCH3(PMe3)C2H4]PF6(M = Ru, 0 s ) and 201 with
PMe3[IZ7'.In the latter case, according to reaction (30) a
mixture of 202 and 203 is obtained which contains the two
isomers in a ratio of 3 :2. The ruthenium complex analogous to 201 yields exclusively the ethyl compound
192 with trimethylphosphane. The variable mobility of
the ligands in the [C6H6RuH(PMe3)C2H4J+ and
[C6H60sH(PMe3)C2H41+cations is, in general, reflected in
the NMR spectra, of which only that for the ruthenium
complex shows a temperature dependence consistent with
an equilibrium between the ethylene(hydrid0) and ethyl
% [Osl,
[ O s ]= C 6 H 6 0 s P M e 3
PF6 ( 3 0 )
,H, P hI e ,
32.3- Subsequent Rea&*ons of tlie Primary Products
The ethyl compound [ C & R U C ~ H S ( P M ~ ~ ) ~
from 191 reacts with [CPh3]PF6 to give the dication
[ C ~ H ~ R U ( P M ~ ~ ) ~ ~which
Z H ~behaves
] ' + , as an electrophile
and reacts with phosphanes or phosphites even at room
temperature. The process, which involves the conversion
of the Ru-C2H4 n-bond into an Ru-C2H4PR3 o-bond,
yields the complexes 195-199 (Scheme 25)1133J.
It has not
Angew. Chem. Int. Ed. Engl. 22 (1983) 927-949
In addition to the [C6H6MR(PMe3)C2H4]+cations obtained from C6H&U(PMe3)C& and 200, the structurally
related [C.&RUCI(PR3)PR;]
complexes are also capable
of adding a phosphane. According to reaction (31), in the
reaction of 204 and PMe3, for example, the (q5-6-phosphoniocyclohexadienyI)ruthenium(u) compound 205 is
obtained. In this process, addition of phosphane to
the ring occurs initially and is followed by substitution
of the
At low temperatures, rotation about
with R=Ph or NMe2 is so
strongly hindered that rigid rotamers can be observed. The
C-PMe3 bond in 205 can easily be cleaved by trifluoroacetic acid and thus opens up a route to half-sandwich
complexes Of the type [c6H6RU(PR3)(PR;)L]2+ (e.9. 206)
which are not accessible by any other method"351.
Surprisingly, the carbonyl(methy1) complexes 209 and
210, prepared according to reaction (32), are highly inert
the hydrido(isocyanide) compounds 212-214 react analog o ~ s l y [ ' ~ lThe
~ ] . stability of the 0s-CH, bond in 221-223
and of the 0s-H bond in 218-220 evidently prevents migration of a methyl or hydride ligand to the isocyanide
group. AccordingIy, even in the presence of an excess of
PMe, (or CNR), the formation of corresponding cationic
imidoylosmium complexes cannot be observed.
3.2. d6-Systems
and react neither with PMe, nor with iodide ions by CO
insertion (or by methyl group migration) to form the acetyl
compounds [C6&MCOCH3(PMe3)Z]PF6and
C6&MCOCH3(PMe3)I, r e s p e ~ t i v e l y " ~ ~For
~ ' ~adjacent
CO and CH3 groups such immobility is very unusual
and is paralleled only in the behavior of
O S C H ~ ( C O ) ~ ( P P ~ ~The
) ~ I acetylruthenium
cation of
composition [C6H6RuCOCH3(PMe3)2] also cannot be obtained by reacting 191 with CH,COCI; in this instance
only decomposition (even at - 78°C) is
207, 208
209, 210
207, 209: [MI
= C6Me6RuPMe3;
208, 210: M = C6H6OSPMe,
The related isocyanide(methy1) compounds 215-217
are similarly as inert as 209 and 210. In contrast to the cobalt complexes [C,H5CoCH,(PMe3)CNR]X (see Scheme
9), the cations of 215-217 do not rearrange either in acetone or in acetonitrile-not even in the presence of a large
excess of iodide ions-to form acetimidoyl compounds;
nor do they react to form metallaheterocycles"31a1.
trimethylphosphane in acetone, the complexes 221-223
having the CH, and CNR ligands in the trans position are
formed by displacing the six-membered ring (Scheme 26);
M e 3 F0&s C H 3
3 PMe3
187, 212, 215, 218, 221: R = M e
188, 213, 216, 219, 222: R = P-Tolyl
211, 214, 217, 220,223: R = rBu
Scheme 26.
In contrast to the cyclopentadienylmetal compounds
C5H5Co(CO)zand C5H5Rh(CO), which-as mentioned in
the introduction-are rapidly decomposed in
BF, .H20, the arenechromium tricarbonyls C6&Cr(CO)3
are transiently stable in strongly acidic media (e.g. HS03F
or CF3C02H/BF3-HzO= l/l['39J)and form hydridometal cations [C6&CrH(CO),] + which can be detected by IR and NMR spectroscopy. Substitution of a CO
group in C&Cr(C0)3 by a phosphane such as PPh3 increases the electron density at the metal, and thus protonation, even with CF3C02H in toluene (ratio 1 : lo), becomes
possible['401. If the compounds C6&M(C0)2PMe2Ph
(M = Cr, Mo, W) are used as starting materials, the formation of two isomers of [C6&MH(CO)2PMezPh]f (with H
and PMeZPh in the cis and trans positions respectively) as
well as their rearrangement have been detected by NMR
spectrosc~py"~''. To isolate stable salts of the type
[C6&MHL3]X, it is necessary to start out from the
appropriate tris(phosphane)
complexes, such as
C6H6Mo(PMePh2), or C6H6MO(PMe2Ph),, which react
with CF3C02Hor HCl in ethanol to yield the corresponding hydridomolybdenum
The manganese compounds C5H5Mn(PR3),(R= Me, OMe, OEt)['431,which are
structurally related to C6H6Mo(PR3),, rapidly decompose
when dissolved in acidic media. Accordingly, it is not yet
possible to make any statement about the existence of the
hydrido complexes [C5H5MnH(PR3)3]f.
Of the presently known areneruthenium and areneosmium compounds having a d6-electron configuration, the
dihydrido complexes C6&RU&[PiPr3], C~MC?~RUH~(PR~)
(R= Me, Ph), and C6H60sH2[PiPr3]are metal bases['441.
They react even at -78°C with CF3C02H/NH4PF6 or
HBF, in ether to form [C6&MH3(PR3)]+ cations which
possess a fluxional structure in solution at room temperature. This behavior corresponds to that of the
[MH3(PMe3)J+ cations in which the metal is similarly in
the oxidation state + 4 and which are formed from the
compounds MH2(PMe3)4(M=Ru, 0 s ) and NH4PF$14s1.It
is especially noteworthy that the complex 226, which is
closely related to 224, yields a binuclear cation upon protonation['*]. The low tendency of rhodium to form compounds having the oxidation state + 5 (as would be the
case in a cation of the type [C,H5RhH3(PR3)]+)obviously
prevents formation of a mononuclear product.
As well as the dihydrido complexes, the analogous dimethylruthenium(I1) complexes also react with Br~nsted
acids. The reactions carried out using 229 are summarized
in Scheme 271144b21471.
Since the ligands H - and CH 3 are
comparable in their ability to act as o-donors, it may also
Angew. Chem. Int. Ed. Engl. 22 (1983) 927-949
[ Rh] = C&;KhP(iPr)3
be assumed that in the reactions of 229 an initial protonation at the metal takes place followed by a reductive elimination of methane. The dimethylcobalt compounds
(CSH4R’)C~(CH3)2PMe3(R’= H, iPr, tBu) behave
The bis(trifluoroacetato) complexes
(C5H4R)Co(OCOCF3)2PMe3,formed from the dimethylcobalt complexes in the presence of excess trifluoroacetic
acid, yield with P(OMe), in a sequence of reaction steps
the phosphonate compounds mentioned in Scheme 3 (see
Section 2.2.2). The circle is thereby completed to the starting point of our own studies on electron-rich half sandwich complexes.
Cooper et al. have discovered a completely analogous
conversion of an M(CH3)2 moiety into an MH(C,H,)+
group for M = (CsH5)2W[1S01.
In this case it was possible to
show that in the primary step an electron transfer takes
place from the dimethyl compound to the trityl cation and
that the trityl radical reacts with the paramagnetic moiety
[(CsHs)2W(CH,)2]+. There are no indications of similar
radical processes in the formation of 230 (nor in the formation of the homologous osmium complex[’44b1).
The CC
bond formed during the conversion of T into U [see reaction (35)] represents, in addition, a model reaction for an
elementary step in the heterogeneous Fischer-Tropsch synthetic process. The most generally accepted mechanism for
this is that CH2 and CH3 groups are formed on the surface
of the catalyst and that the chain is built up by successive
insertion of methylene units into metal-alkyl bonds[’”].
After @-Helimination (analogous to that in the conversion
of U into 230 according to reaction (35)), a-olefins are
formed which are, of course, known to be primary products of the Fischer-Tropsch synthesis.
Finally, a d6-metal system should be mentioned which
can be protonated and was discovered during our attempts
to synthesize the metal base C6H60~(PMe3)2.
The compound [C6H60sI(PMe3)2]PF6231, which-as indicated in
Scheme 24-cannot be reduced by NaCloHs to form
C6H60s(PMe3),, reacts with lithiumorganyls, LiR, by exoaddition of the carbanion to the arene ring to produce the
cyclohexadienyl complexes 232-2371’521.The latter react
with [CPh3]PF6to yield the corresponding ring-substituted
areneosmium(I1) compounds 238-243 together with
CHPh3. This result is surprising insofar as there are only a
few examples currently known where cyclohexadienylmetal complexes (qs-c6H6R)MLnin which the group R is in
the exo-position react with trityl salts by hydride eliminati~n[”~I.
Scheme 27. [Ru]=C6Me,RuPPh3.
The compound 229 reacts very readily not only with
HBF, or CF3C02H but also with Lewis acids such as
[CPhJ +. Thereby, the ethylene(hydrid0) complex 230,
whose structure has been determined by X-ray analysis, is
obtained in high yield”491.As established by labeling experiments the solvent does not participate in this reaction.
The mechanism of the CC bond formation presumably involves primary attack of the trityl cation on a C-H bond
of one of the Ru-CH3 groups to eliminate a hydride ion,
and subsequent formation of the coordinatively unsaturated ethyl compound U from the intermediate T by intramolecular rearrangement. The compound U stabilizes itself by a P-H migration to form the ethylene(hydrid0) complex.
Angew. Chem. Int. Ed. Engl. 22 (1983) 927-949
( 3 5)
232, 238: R
233,239: R
234, 240: R
= Me
= Et
= nPr
235, 241: R = n B u
236, 242: R = fBu
237, 243: R = P h
Mechanistic investigations on the conversion of 232231 into 238-243 indicate that the hydride abstraction
does not take place directly but occurs only in the presence
of protons (formed, for example, from [CPhs]+ in acetone)[’541.In the primary step a hydridoosmium(1v) cation
is formed which for R = tBu can be isolated as the PF6 salt.
However, with the other R groups it rearranges intramolecularly to afford an isomer in which R is found in the olefinic part of the ring. In the intermediate V either a cyclo945
hexadiene ligand (formed by hydride migration from the
metal to the C' carbon atom) or an OsCH three-center
bond could be present-as in similar manganese comp l e x e ~ ~ ' Deprotonation
of the hydrido cation formed
from V (which for R = Me and nBu was isolated as the PF,
salt) yields a cyclohexadienyl compound with a CH, group
in the ring which then reacts with [CPh,]+ by hydride elimination.
Fig. 2. Electron density distribution in an occupied MO of CSRsML2 or
C 6 % M b 11571.
Scheme 28.
The reaction sequence shown in Scheme 28 demonstrates that even the complexes 232-237 function as metal
bases. Because of the coordination of two trimethylphosphane ligands, the metal clearly has a relatively high electron density and accordingly is able to react by proton addition. The cyclopentadienylruthenium and cyclopentadienylosmium(I1) compounds, C5H,MX(PMe3)2 (M = Ru,
X=CI; M=Os, X=Br) recently synthesized by Bruce et
al., also form stable hydridometal cations with HPF, and
behave completely analogously to 232--237['561.
This, together with results obtained using [C6H60~I(PMe3)2],
increases the number of examples known which show that
not only half-sandwich complexes having a d8 configuration at the metal but also those with a d6 metal center are
capable of oxidative additions.
4. Conclusions
Following in the footsteps of the square-planar Vaskatype compounds (including Wilkinson's catalyst), the halfsandwich complexes C5R,ML, and C6%ML2 have established themselves in coordination chemistry over the past
decade-primarily due to the contributions discussed in
this review-as an additional group of metal bases having
a d8-metal system. Their nucleophilicity is determined primarily-as has been documented in many instances-by
the donor properties of the ligands C,Rn and L. As the donor strength increases electron density increases at the metal and hence facilitates the attack of any electrophile. As
far as is known there are only minor differences between
the cyclopentadienyl- and arenemetal complexes. This can
be understood qualitatively from MO calculations carried
out by Hoffmann and Albright"571which show that halfsandwich complexes C,R,ML2 (n=5 or 6) possess an energetically relatively high-lying occupied molecular orbital
which has non-bonding character with respect to the metal-ligand bonds present. Its electron density distribution
can be described as shown in Figure 2.
Such an MO (whose "shape" very closely resembles the
dzz orbital of the metal in a planar ds system) represents,
according to Roald Hoffmannr'*81,"the spearhead of nucleophilic activity of the compounds C,R,ML,".
"spearhead" clearly exceeds even the nucleophilicity of
the phosphorus atom in a phosphane, as is evidenced
by the reaction of C5Me5Rh(CO)PMezPMe2 with methyl
[C5Me5RhCH3(CO)PMezPMe2]+ to the compound
Kinetic measurements
by Pearson and F i g d ~ r e ~support
' ~ ~ ' the conclusion that the
rate of methylation of the C,H5M(CO)PPh3 (M = Co, Rh,
Ir) complexes at room temperature is 10-100 times that
for uncoordinated PPh3.
In addition to the electronic factors determining the metal-base character of the half-sandwich compounds
C,RSMLz and C6&ML2, stenc factors cannot be entirely
neglected when considering the reactivity. Results obtained to date suggest that a substantial number of electrophiles are capable of addition to a metal to form a metalelement bond. Such results also indicate that when bulky
(space-filling) Lewis acids are used radical reactions frequently occur which are in most cases associated with decomposition. This behavior can be understood by recalling
that the metal in the five- and six-membered ring ML,
complexes is relatively well shielded by the ligands, as
should be apparent from Figure 2. A fairly large electrophile thus has difficulty in interacting with the nucleophilic center. In several cases the reactivity is directly paralleled by the space-filling capacity of L (e.g . expressed in
terms of the Tolman cone angle"601).Thus, for instance, the
rate of the reaction of C5H5CoL, with methyl iodide decreases with increasing bulkiness of the ligand L (according to the sequence: PMe3 < PMe2Ph < PEt3 < PMePh,
< PPh3)[','].
Interestingly enough, the metal basicity of cyclopentadienylcobalt and rhodium compounds is not confined to
species having only one metal center. The binuclear complexes 244['621
and 2451'631,
in which the two metal atoms
are coordinated in a formalZy similar way to those in the
compounds C5R5M(PMe3),, react for example with
Brernsted acids and SO, by electrophilic addition to the
. The hydrido bridge formed upon
protonation can be opened by Lewis bases thus allowing
formation of binuclear cobalt(rI1) and rhodium(m) hydrido complexes. These may be viewed as potential reducing agents and are capable, for example, of intramolecularly converting an isocyanide- into a formimidoyl-ligand
and an alkyne- into a ~ i n y l - g r o u p ~ 'From
~ ~ ] . these observaAngew. Chern. Int. Ed. Engl. 22 (1983) 927-949
tions it seems that bridge building could be a means of increasing the reactivity of metal clusters (and perhaps also
of metal surfaces).
244, 246,248:
245, 247:
= CajCo
Scheme 29. P= PMe2.
What are the trends for the future? An extremely relevant topic-which is connected at least indirectly with the
chemistry of metal bases C,R,ML,-is
the “activation”
(or, more precisely, the facile cleavage) of C-H bonds by
Although there is still no concrete
evidence that compounds of the type C5R5MLz or
C6&ML, react directly with aliphatics or arenes, the corresponding moieties [C5R5ML]and [C6%ML] are definitely
able to do so. Thus, Janowicz and Bergman”671and, independently, Hoyano and Graham[’681have recently found
(see Scheme 30) that UV irradiation of solutions of both
249 and 250 in benzene, cyclohexane or neopentane leads
to oxidative addition with formation of the corresponding
aryl(hydrid0) or alkyl(hydrid0) iridium complexes
C5Me51rH(R)L (R= Ph, C6Hllr CH,CMe,). The authors
assume that in these reactions a 16-electron moiety
[C5Me51rL](W, X) is initially formed which then reacts
with a C-H bond in the hydrocarbon.
Results obtained over the last few months provide evidence that not only iridium but also analogous rhodium
and ruthenium compounds can effect such a C-H
bond cleavage. Low temperature photolysis of
C5Me,RhHz(PMe3)in pure propane yields the spectroscopically characterized complex C5Me5RhH(nPr)PMe3.This
complex reacts at - 15°C by reductive elimination of
C3H8,but upon addition of CHBr3 it can be converted into
the stable product C5Me5Rh(nPr)(PMe3)Br[’691. The
hydrido(methy1) compound
C5Me5RhCH3(PMe3)C1is also extremely labile and very
rapidly forms the complex 258 in C6D2’691.
Angew. Chern. Inr. Ed. Engl. 22 (1983) 927-949
25 7
With other phosphane ligands it is possible to stabilize
an intermediate such as W or Y [see Scheme 30 and reaction (37)] even by means of intramolecular C-H addition.
Thus for example, photolysis of C6H6RuH,[P(iPr),] 224 at
room temperature in C6H12 or C6DI2and leads to isolation
of 259, which can be characterized by elemental analysis
and NMR spectroscopy. By dissolving 259 in benzene or
toluene and allowing the reaction to take place in the dark,
the corresponding aryl(hydrid0) complex 260 or 261 is obtained. It seems probable that the moiety actually reacting
is the intermediate 2. This C-H bond cleavage without
any simultaneous irradiation may offer a possibility to
make other, more sensitive hydrocarbons react with transition metals and thereby bring a new stimulus to the problem of “C-H activation”.
H; 261:R = M e
Scheme 31. [Ru]=C6H6Ru.
In an article which appeared in 1970 and made the concept of metal basicity popular, Shriverf51wrote that “It can
be anticipated that metal basicity will become an increasingly important concept because current vigorous activity
in the synthesis of new low oxidation-state complexes will
lead to new metal bases and because available spectroscopic tools allow convenient recognition of metal-base interaction with Lewis acids”. This expectation has been
more than fulfilled, not least in investigations of the behavior of electron-rich half sandwich complexes. It may well
be conceivable that d8 systems of the type CnRnML2or
C,R,MLL’, where M is Fe’, Ruo and Oso, as well as Co’,
Rh’ and Ir’, will soon find counterparts in other systems
having Ni”, Pd“, and Ptrl as the central atom. This again
would represent an extension of the palette of nucleophilic
transition-metal compounds, which nowadays are relatively common and are no longer curiosities.
I should likefirst and foremost to thank my co-workers in
this field whose names appear in the literature citations.
Without their involvement and enthusiasm it would not have
been possible to reach our goals. Thanks are also due to colleagues who by carving out numerous X-ray analyses have
contributed to the success of these investigations: Dr. C.
Burschka, Dr. G. Evrard, Pro$ C. Kriiger, Pro$ U. Schubert, Pro$ M . L. Ziegler and their co-workers. Finally, our
success is due in large measure to support by the Deutsche
Forschungsgemeinschaft, the Fonds der Chemischen Zndustrie and the companies BASF, Bayer and Degussa, all of
which Z heartily thank. Last but not least, Z extend cordial
thanks to Professor Sir Jack Lewis, Cambridge (U.K.),
whose kind invitation made it possible for me to prepare this
article in the stimulating atmosphere of the Chemical Laboratory in Lensfield Road and in Robinson College.
Received: July 7, 1983 [A 478 IE]
German version: Angew. Chem. 95 (1983) 932
E. 0. Fischer, W. Hafner, Z . Naturforsch. B 9 (1954) 503.
E. 0. Fischer, H. P. Fritz, Adu. Inorg. Chem. Radiochem. I (1959) 56.
G. Wilkinson, F. A. Cotton, h o g . Inorg. Chem. I (1959) 1.
G. E. Coates, M. L. H. Green, K. Wade: Organometallic Compounds,
Yol. ZI, Chapman and Hall, London 1968.
[5] D. F. Shriver, Aec. Chem. Res. 3 (1970) 231.
161 a) L. Vaska, Acc. Chem. Res. I (1968) 335; b) J. P. Collman, W. R. Roper, Adu. Organomet. Chem. 7(1968) 53; c) J. Halpern, Acc. Chem. Res. 3
(1970) 386.
[71 A. Davison, W. McFarlane, L. Pratt, G. Wilkinson, J. Chem. Sac. 1962,
[8] D. J. Cook, J. L. Dawes, R. D. W. Kemmitt, J . Chem. SOC.A 1967,
191 1. N. Nowell, D. R. Russell, Chem. Commun. 1967, 817.
[lo] J. L. Dawes, R. D. W. Kemmitt, J. Chem. SOC.A 1968, 1072.
Ill] a) A. J. Oliver, W. A. G. Graham, Znorg. Chem. 9 (1970) 243; b) A. J.
Oliver, W. A. G. Graham, ibid. 9 (1970) 2653; c) A. J. Hart-Davis, W. A.
G. Graham, ibid. 9 (1970) 2658; d) A. J. Oliver, W. A. G. Graham, ibid.
I0 (1971) 1165; e) A. J. Hart-Davis, W. A. G. Graham, ibid. 10 (1971)
[12] T. Kruck, W. Hieber, W. Lang, Angew. Chem. 78 (1966) 208; Angew.
Chem. Int. Ed. Engl. 5 (1966) 247.
[13] V. Harder, J. Miiller, H. Werner, Helu. Chim. Acta 54 (1971) 1.
1141 B. Juthani, Dissertation, Universitat Wiirzburg 1980.
1151 H. Yamazaki, N. Hagihara, Bull. Chem SOC.Jpn. 44 (1971) 2260.
1161 B. Klingert, H. Werner, Chem. Ber. 116 (1983) 1450.
[I71 H. Werner, W. Hofmann, Chem. Ber. 110 (1977) 3481.
[18] H. Yamazaki, Y. Wakatsuki, J. Organomet. Chem. 139 (1977) 157.
[19] H. Werner, W. Hofmann, Chem. Ber. 114 (1981) 2681.
[ZO] H. Werner, B. Heiser, B. Klingert, R. Dolfel, J. Organornet. Chem. 240
(1982) 179.
[21] P. V. Rinze, J. Lorberth, H. Noth, B. Stutte, J. Organomet. Chem. 19
(1969) 399.
[22] Y . Wakatsuki, H. Yamazaki, J. Organomet. Chem. 64 (1974) 393.
[23] F. Faraone, G. Tresoldi, G. A. Loprete, J. Chem. SOC.Dalton Trans.
1979, 933.
[24] H. Werner, J. Wolf, U. Schubert, K. Ackermann, J. Organomet. Chem.
243 (1983) C63.
I251 H. Werner, A. Hohn, unpublished results; see A. Hohn, Diplomarbeit,
Universitat Wiirzburg 1983.
[26] H. Neukomm, H. Werner, Helu. Chim. Acta 57 (1974) 1067.
[27] H. Werner, L. Hofmann, W. Paul, J. Organomet. Chem. 236 (1982)
[28] H. Werner, R Feser, J. Organomet. Chem. 232 (1982) 351.
[29] H. Werner, R. Feser, W. Buchner, Chem. Ber. 112 (1979) 834.
[30] R. A. Jones, F. M. Rea1.G. Wilkinson, A. M. R. Galas, M. B. Hursthouse, J. Chem. Sac. Dalton Trans. 1981, 126.
[31] H. Werner, R Feser, 2. Naturforsch. 8 3 5 (1980) 689.
[32] R A. Jones, F. M. Real, G. Wilkinson, A. M. R. Galas, M. B. Hursthouse, K. M. Abdul Malik, J . Chem. Sac. Dalton Trans. 1980. 511.
[33] R Feser, Dissertation, Universitat Wiirzburg 1981; R. Feser, L. Hofmann, H. Werner, unpublished results.
[34] H. G. Schuster-Woldan, F. Basolo, J . Am. Chem. SOC.88 (1966) 1657.
I351 A. Spencer, H. Werner, J . Organomet. Chem. I71 (1979) 219.
[36] T. Aviles, M. L. H. Green, .
Chem. Sac. Dalton Trans. 1979, 1116.
[37] R Feser, H. Werner, J. Organornet. Chem. 233 (1982) 193.
[38] H. Werner, B. Klingert, J . Organomer. Chem. 218 (1981) 395.
(391 L. Hofmann, H. Werner, unpublished results.
[40] R Cramer, L. P. Seiwell, J . Organomet. Chem. 92 (1975) 245.
[41] W. Hofmann, Dissertation, Universitat Wiirzburg 1980.
[42] H. Werner, S. Lotz, B. Heiser, 1. Organomet. Chem. 209 (1981) 197.
[43] P. M. Maitlis, Chem. Soc. Reu. I0 (1981) 1.
I441 J. W. Kang, P. M. Maitlis, J . Organomet. Chem. 26 (1971) 393.
(451 a) F. Calderazzo, Angew. Chem. 89 (1977) 305; Angew. Chem. Inr. Ed.
Engl. 16 (1977) 299; b) J. Plank, D. Riedel, W. A. Herrmann, ibid. 92
(1980) 961; 19(1980) 937.
[46] R. D. Cramer, J. Am. Chem. Soc. 87 (1965) 4717.
[47] L. P. Seiwell, Inorg. Chem. IS (1976) 2560.
1481 H. Werner, Angew. Chem. 89 (1977) 1; Angew. Chem. Znt. Ed. Engl. 16
(1977) 1.
[49] V. Harder, E. Dubler, H. Werner, 1. Organomet. Chem. 71 (1974) 427.
[SO] W. Klaui, K. Dehnicke, Chem. Ber. 111 (1978) 451.
[Sl] D. K. Towle, S . J. Landon, T. B. Brill, T. H. Tulip, Organometallics I
(1982) 295.
[52] H. Neukomm, Dissertation, Universitat Zurich 1976.
I531 For the recently described synthesis of trinuclear complexes of composee: W.
sition [C~MeSRh(p-P(OMe)ZO)3M(p-OP(OMe)2)3RhCsMeS]
Klaui, H. Otto, W. Eberspach, E. Buchholz, Chem. Ber. 115 (1982)
[541 H. Neukomm, H. Werner, J. Organomet. Chem. 108 (1976) C26.
I551 H. Werner, H. Neukomm, W. Kllui, Helu. Chim. Acta 60 (1977) 326.
[56] H. Werner, R. Feser, Z . Anorg. Allg. Chem. 458 (1979) 301.
[571 H. Werner, Tri Ngo-Khac, R. Feser, unpublished results; see also: H.
Werner, Tri Ngo-Khac, D. Friebel, P. Kohler, D. Reinen, 2. Naturforsch. 8 3 6 (1981) 322.
[58] H. Werner, W. Hofmann, Chem. Ber. I15 (1982) 127.
[591 U.Schubert, R. Werner, L. Zinner, H. Werner, J. Organomet. Chem.
253 (1983) 363.
[601 K. Dey, H. Werner, J. Organomet. Chem. 137 (1977) C 2 8 ; Chem. Ber.
I12 (1979) 823.
[61] K. Dey, unpublished results.
[621 W. Hofmann, W. Buchner, H. Werner, Angew. Chem. 89 (1977) 836;
Angew. Chem. Znt. Ed. Engl. 16 (1977) 795.
1631 T. A. Albright, R. Hoffmann, Chem. Ber. I l l (1978) 1578.
1641 R. K. Pomeroy, D. J. Harrison, J. Chem. SOC.Chem. Commun. 1980,
[651 D. L. Morse, M. S. Wrighton, J . Organomet. Chem. 125 (1977) 71.
[661 a) R. J. McKinney, J. Chem. Sac. Chem. Commun. 1980, 603; b) Znorg.
Chem. 21 (1982) 2051.
[67] R. Eisenberg, D. E. Hendriksen, Adu. Catal. 28 (1979) 119.
1681 M. C. Baird, G. Wilkinson, Chem. Commun. 1966, 514; J. Chem. SOC.
A 1967, 865.
[691 M. C. Baird, G. Wilkinson, Chem. Commun. 1966, 267.
(701 H. Werner, K. Leonhard, C. Burschka, J. Organomet. Chem. 160, 291
(711 H. Werner, 0. Kolb, Angew. Chem. 91 (1979) 930; Angew. Chem. Int.
Ed. Engl. 18 (1979) 865.
1721 H. Werner, 0.Kolb, U. Schubert, K. Ackermann, Angew. Chem. 93
(1981) 583; Angew. Chem. Int. Ed. Engl. 20 (1981) 593.
(731 0. Kolb, Dissertation, Universitat Wiirzburg 1983.
[74] H. Werner, M. Ebner, W. Bertleff, U. Schubert, Organometallics 2
(1983) 891.
1751 H. Werner, 0. Kolb, R. Feser, U. Schubert, J. Organomet. Chem. 191
(1980) 283.
[761 G . R. Clark, T. J. Collins, K. Marsden, W. R. Roper, J. Organomet.
Chem. 157 (1978) C23.
[77] B. D. Dombeck, R. J. Angelici, Inorg. Chem. I5 (1976) 2397.
[78] H. Werner, 0. Kolb, U.Schubert, K. Ackermann, J. Organomet. Chem.
240 (1982) 421.
[79] a) W. Strohmeier, Angew. Chem. 76 (1964) 873; Angew. Chem. Znt. Ed.
Engl. 3 (1964) 730; b) E. 0. Fischer, M. Herberhold: Essays in Coordination Chemistzy, Exper. Suppl. IX,Birkhauser, Base1 1964, p. 259.
[SO] H. Werner, B. Juthani, J. Organomet. Chem. 209 (1981) 211.
1811 0. Kolb, H. Werner, Angew. Chem. 94 (1982) 207; Angew. Chem. Znt.
Ed. Engl. 21 (1982) 202.
[82] G. Evrard, unpublished results.
1831 H. Werner, 0. Kolb, U. Schubert, unpublished results.
1841 H. Werner, B. Heiser, A. Kiihn, Angew. Chem. 93 (1981) 305; Angew.
Chem. Int. Ed. Engl. 20 (1981) 301.
[SS] H. Werner, B. Heiser, U.Schubert, unpublished results; see: B. Heiser,
Dissertation, Universitat Wiirzburg 1983.
[86] H. Werner, B. Heiser, C. Burschka, Chem. Ber. I15 (1982) 3069.
I871 B. Heiser, H. Werner, M. L. Ziegler, unpublished results.
1881 a) J. Fortune, A. R. Manning, J . Organomet. Chem. I90 (1980) C95; b)
A. R. Manning, personal communication.
[89] a) K. Stanley, M. C. Baird, J. Am. Chem. Sac. 97 (1975) 6598; b) H.
Brunner, Adu. Organornet. Chem. 18 (1980) 186.
(901 For further work in this connection, see a) J. W. Byrne, H. V. Blaser, J.
A. Osborn, J. Am. Chem. Sac. 97 (1975) 3871; b) J. W. Byrne, J. R.
Kress, J. A. Osborn, L. Picard, R. E. Weiss, J . Chem. SOC.Chem. Commun. 1977, 662; c) H. F. Klein, R. Hammer, J. GroO, U. Schubert, Angew. Chem. 92 (1980) 835; Angew. Chem. Int. Ed. Engl. I9 (1980) 809;
d) R. B. A. Pardy, M. J. Taylor, E. C. Constable, J. D. Mersh, J. M. K.
Sanders, J . Organornet. Chem. 231 (1982) C25.
I911 S . Otsuka, A. Nakamura, Adu. Organomet. Chem. 14 (1976) 245.
[92] U. Schubert, P. Erk, unpublished results.
I931 K. Leonhard, H. Werner, Angew. Chem. 89 (1977) 656; Angew. Chem.
Int. Ed. Engl. 16 (1977) 649.
[941 H. Werner, J. Wolf, R. Zolk, U. Schubert, Angew. Chem. 95 (1983)
1022; Angew. Chem. Int. Ed. Engl. 22 (1983) 981.
Angew. Chem. Int. Ed. Engl. 22 (1983) 927-949
[95] J. Wolf, H. Werner, 0. Serhadli, M. L. Ziegler, Angew. Chem. 95 (1983)
428; Angew. Chem. Int. Ed. Engl 22 (1983) 414.
[961 a) A. B. Antonova, N. E. Kolobova, P. V. Petrovsky, B. V. Lokshin, N.
S. Obezyuk, J . Organomef. Chem. 137(1977) 55; b) M. I. Bruce, R. C.
Wallis, Ausf. J . Chem. 32 (1979) 1471; c) A. Davison, J. P. Selegue, J .
Am. Chem. Sac. 102 (1980) 2455.
1971 J. Wolf, unpublished results.
1981 a) U. Behrens, F. Edelmann, J . Organomet. Chem. 118 (1976) C41; b)
D. Wormsbacher, F. Edelmann, U. Behrens, Chem. Ber. 114 (1981)
153; I15 (1982) 1332.
[99] M. Green, R. B. L. Osborn, F. G. A. Stone, J . Chem. SOC.A 1970,
[loo] H. Schmidbaur, Nachr. Chem. Tech. Lob. 27 (1979) 620.
[I011 R. Feser, H. Werner, Angew. Chem. 92 (1980) 960; Angew. Chem. Int.
Ed. Engl. 19 (1980) 940.
[lo21 L. Hofmann, Diplomarbeit, Universitat Wiirzburg 1981.
[I031 H. R. Christen: Grundlagen der organischen Chemie, Verlag Sauerlander-Diestenveg-Salle. Aarau-FrankfurtIMain 1970, chapt. 18.
[lo41 H. F. Klein, R. Hammer, Angew. Chem. 88 (1976) 61; Angew. Chem.
Int. Ed. Engl. I5 (1976) 42.
[lO5] J. R. Moss, J. C. Spiers, J . Organomet. Chem. 182 (1979) C20.
I1061 N. J . Kermode, M. F. Lappert, B. W. Skelton, A. H. White, J. Holton, J .
Chem. Sac. Chem. Commun. 1981, 698.
[I071 H. Werner, R. Feser, W. Paul, L. Hofmann, J . Organornet. Chem. 219
(1981) C29.
[lo81 B . Heiser, Diplomarbeit, Universitat Wiirzburg 1980.
[lo91 W. Paul, Diplomarbeit, Universitat Wiirzburg 1981.
Ill01 W. Paul, H. Werner, Angew. Chem. 95 (1983) 333; Angew. Chem. Int.
Ed. Engl. 22 (1983) 316; Angew. Chem. Suppl. 1983, 396.
111 I] G. Miiller, C. Kriiger, unpublished results.
[I121 a) T. J. Collins, W. R. Roper, 1. Chem. Soc. Chem. Commun. 1977, 901;
b) C. E. L. Headford, W. R. Roper, J. Organomef.Chem. 244 (1983)
c 53.
11 131 M. Herberhold, W. Ehrenreich, W. Biihlmeyer, Angew. Chem. 95 (1983)
332; Angew. Chem. Int. Ed. Engl. 22 (1983) 315.
[ I 141 W. A. Herrmann, J. Weichmann, R Serrano, K. Blechschmitt, H. pfsterer, M. L. Ziegler, Angew. Chem. 95 (1983) 331; Angew. Chem. int.
Ed. Engl. 22 (1983) 314; Angew. Chem. Suppl. 1983. 363.
[1151 H. Werner, W. Paul, G. Miiller, C. Kriiger, unpublished results.
[I161 H. Werner, W. Paul, J. Organornet. Chem. 236 (1982) C71.
[117] L. Hofmann, H. Werner, unpublished results.
[118] a) E. 0. Fischer, H. Werner: Metal n-Complexes, Vol. I (Complexes
with Di- and Oligo-olefinic Ligands), Elsevier, Amsterdam 1966; b) J.
Miiller, C. G. Kreiter, B. Mertschenk, S. Schmitt, Chem. Ber. 108 (1975)
273; c) P. Pertici, G. VituIli, M. Paci, L. Porri, J. Chem. Soc. Dalton
Trans. 1980, 1961, and literature cited therein.
[119] H. Werner, R Werner, Angew. Chem. 90(1978) 721; Angew. Chem. Int.
Ed. Engl. I7 (1978) 683.
[I201 a) D. L. William-Smith, L. R Wolf, P. S. Skell, J. Am. Chem. Sac. 94
(1972) 4042; b) R. Middleton, J. R Hull, S. R. Simpson, C. H. Tomlinson, P. L. Timms, J. Chem. SOC.Dalton Trans. 1973, 120.
I1211 S. R. Weber, H. H. Brintzinger, J. Organomef. Chem. 127 (1977) 45.
[122] a) S. D. Ittel, C. A. Tolman, J. Organomet. Chem. 172 (1979) C 4 7 ; b) S.
D. Ittel, F. A. Van-Catledge, J. P. Jesson, J. Am. Chem. SOC.101 (1979)
3874; c) S. D. Ittel, C. A. Tolman, Organometallics I (1982) 1432.
[123] R. Werner, H. Werner, Chem. Ber. 115 (1982) 3781.
[124] H. Werner, R. Werner, Chem. Ber. 115 (1982) 3766.
[125] H. Kletzin, Diplomarbeit, Universitat Wiirzburg 1981.
[126J D. R. Robertson, T. A. Stephenson, T. Arthur, J. Organomet. Chem. 162
(1978) 121.
[I271 R. Werner, H. Werner, Chem. Ber. 116 (1983) 2074.
[I281 R. Werner, H. Werner, Angew. Chem. 93 (1981) 826; Angew. Chem. h t .
Ed. Engl. 20 (1981) 793.
[129] E. L. Muetterties, J. R. Bleeke, A. C. Sievert, J. Organomet. Chem. 178
(1979) 197.
11301 K. Zenkert, H. Werner, unpublished results.
11311 a) R. Weinand, Diplomarbeit, Universitat Wiirzburg 1982; b) H. Werner, R. Weinand, Z . Nafurforsch. 8 3 8 (1983), in press.
Angew. Chem. Int. Ed. Engl. 22 (1983) 927-949
11321 R. Hoffmann, Angew. Chem. 94 (1982) 725; Angew. Chem. Int. Ed.
Engl. 21 (1982) 711.
11331 H. Werner, R. Feser, R. Werner, J. Organomet. Chem. 181 (1979) C 7 .
11341 R. Werner, Dissertation, Universitat Wiirzburg 1982.
11351 a) R Werner, H. Werner, J. Organomet. Chem. 210 (1981) C 11; b) H.
Werner, R. Werner, Chem. Ber. 116 (1983), in press.
[136] H. Werner, R. Werner, J. Organomet. Chem. 174 (1979) C67.
11371 K. R. Grundy, W. R. Roper, J . Organomet. Chem. 216 (1981) 255.
[1381 C. P. Lillya, R. A. Sahatjian, Inorg. Chem. 11 (1972) 889.
[139] B. V. Lokshin, V. I. Zdanovich, N. K. Baranetskaya, V. N. Setkina, D.
N. Kursanov, J. Organomet. Chem. 37 (1972) 331,339.
11401 L. A. Fedorov, P. V. Petrovskii, E. I. Fedin, N. K. Baranetskaya, V. I.
Zdanovich, V. N. Setkina, D. N. Kursanov, J. Organomet. Chem. 99
(1975) 291.
[141] T. C. Flood, E. Rosenberg, A. Sarhangi, J. Am. Chem. SOC.99 (1977)
(1421 M. L. H. Green, L. C. Mitchard, W. E. Silverthorn, J . Chem. Sac. A
1971, 2929.
11431 H. Werner, B. Juthani, J. Organornet. Chem. 129 (1977) C39.
[144] a) H. Werner, H. KIetzin, J . Organomet. Chem. 243 (1983) C59; b) H.
Kletzin, unpublished results; see H. Kletzin, Dissertation, Universitilt
Wiirzburg, in preparation.
11451 a) H. Werner, J. Gotzig, Organometallics 2 (1983) 547; b) J. Gotzig, unpublished results.
11461 H. Werner, J. Wolf, Angew. Chem. 94 (1982) 309; Angew. Chem. Inf.
Ed. Engl. 21 (1982) 296.
[I471 H. Werner, H. Kletzin, J. Organomet. Chem. 228 (1982) 289.
[148] W. Hofmann, H. Werner, Chem. Ber. 115 (1982) 119.
11491 H. Kletzin, H. Werner, 0. Serhadli, M. L. Ziegler, Angew. Chem. 95
(1983) 49; Angew. Chem. Int. Ed. Engl. 21 (1983) 46.
11501 J. C. Hayes, N. J. Cooper, J . Am. Chem. SOC.104 (1982) 5570.
I1511 W. A. Herrmann, Angew. Chem. 94 (1982) 118; Angew. Chem. Int. Ed.
Engl. 21 (1982) 117.
11521 H. Werner, R. Werner, C. Burschka, Chem. Ber. 116 (1983). in press.
11531 a) I. U. Khand, P. L. Pauson, W. E. Watts, J. Chem. Sac. C 1969, 2024;
b) A. N. Nesmeyanov, N. A. Volkenau, L. S. Shilovtseva, V. A. Petrakova, J. Organomet. Chem. 85 (1975) 365.
[154] R. Werner, H. Werner, Chem. Ber. 116 (1983), in press.
[I551 M. Brookhart, W. Lamanna, W. R. Pinhas, Organometallics 2 (1983)
[156] M. I. Bruce, I. B. Tomkins, F. S. Wong, B. W. Skelton, A. H. White, J .
Chem. SOC.Dalton Trans. 1982, 687.
[157] T. A. Albright, R. Hoffmann, unpublished results; see also: P. Hofmann, Angew. Chem. 89 (1977) 551; Angew. Chem. In{. Ed. Engl. 16
(1977) 536.
I1581 R. Hoffmann, personal communication (7. June 1976).
El591 R. G. Pearson, P. E. Figdore, J. Am. Chem. Soc. 102 (1980) 1541.
11601 C. A. Tolman, Chem. Reu. 77 (1977) 313.
11611 W. Hofmann, Dissertation, Universitat Wiirzburg 1980.
11621 a) H. Werner, W. Hofmann, Angew. Chem. 91 (1979) 172; Angew.
Chem. Int. Ed. Engl. 18 (1979) 158; b) H. Werner, W. Hofmann, R.
Zolk, L. F. Dahl, J. Kocal, A. Kiihn, Organometallics. in press.
[1631 B. Klingert, H. Werner, J . Organomet. Chem. 252 (1983) C47.
[1641 W. Hofmann, H. Werner, Angew. Chem. 93 (1981) 1088; Angew. Chem.
Int. Ed. Engl. 20 (1981) 1014.
[1651 a) R. Zolk, H. Werner, J. Organomef. Chem. 252 (1983) C53; b) H.
Werner, B. Klingert, R. Zolk, unpublished results.
[1661 a) G. W. Parshall: Homogeneous Cafalysls, Wiley-Interscience. New
York 1980; b) F. A. Cotton, G. Wilkinson; Aduanced Inorganic Chemisfry,4th edit., Wiley, New York 1980.
[1671 A. H. Janowicz, R. G. Bergman, J . Am. Chem. SOC.104 (1982) 352.
I1681 J. K. Hoyano, W. A. G. Graham, J . Am. Chem. Sac. 104 (1982) 3723.
[1691 W. D. Jones, F. J. Feher, Organometallics 2 (1983) 562.
[I701 a) H. Werner, Plenary lecture XII. Sheffield-Leeds International Symposium on Organometallic, Inorganic and Catalytic Chemistry, Sheffield 28th March 1983; b) H. Kletzin, H. Werner, Angew. Chem. 95
(1983) 916; Angew. Chem. Int. Ed. Engl. 22 (1983) 873.
Без категории
Размер файла
2 274 Кб
base, complexesчmetal, par, excellent, electro, sandwich, half, rich
Пожаловаться на содержимое документа