r- vol. 34, pp. 3M7-ml7

8PawmPmuwlvm. himdbolul~

TETRAHEDRON REPORT NUMBER 57

NUCLEOPHILIC ADDITION TO ORGANOTRANSITION METAL CATIONS CONTAINING UNSA’ItJRATED

HYDROCARBON LIGANDS

A SURVEY AND INTERPRETATION

&WHEN G. DAVIES, IS~ALCOLM L. H. GREEN and D. A~KHAEL P. MINGOS Inorgmic Chemistry Laboratory, South Parks Road, Oxford OX1 3QR, England

(Received for publkation IS Scptrmber 19n)

Abdmct-‘lluee simple rules arc proposed that enable the prediction of the most favourabk position of nuckopbilic attack oo 18 electron organotransition metal cations containing unsaturated hydrocarbon ligands.

Unsatura@ hydrocarbons, e.g. ethylene, butadiene and benzene, do npt normally undergo nucleophilic addition or substitution. reactions. However, when these mole- cules act as ligands to metal cations they are attacked by a wide range of nucleophiles such as B. R-, CN-, MeO- and ILN.’ ‘IIe greater reactivity of the coor- dinated u~turated hydrocarbon can be attributed broadly to metal&and bonding effects which result in a net withdrawal of electron density from the unsaturated hydrocarbon ligand to the positively charged metal centre. Coordination to the metal ion is therefore akin to the introduction of electron withdrawing substituents onto the hydrocarbon chain. The transfer of electron density associated with metal-ligand bonding is inlluenced by the number of C atoms coordinated to the metal, i.e. the lid &to number (q)‘, and the parity of the @and hapro number, i.e. whether an cuen or odd number of ligand C atoms are attached to the metal. Furthermore, for the purpose of this review it is neces- sary to distinguish between liis which are cyclically conjugated and those which are not. The former are described as closed and the latter as open. llms $- cyclopentadienyl (1) is an odd closed ligand and T$- pentadienyl and $-cyclohexadienyl (2 and 3) are odd open ligands. Some euen ligands are illustrated in 4 to 6.

closed-$ (4)

clorrd-i4 (5)

TaMe 1 lists examples of organotransition metal cations which undergo nuckophilic addition, and classi& them accord& to the hupto numbers of the un&ra&d hydrou&mligands.

open-,* We find that it is possible to put forward some simple (63 rules which if used sequentially enabIe a prediction of

In organometallic cations with carbon monoxide coordinated to the metal nucleophilic attack’ can occur either at the organic &and or the carbon monoxide. Therefore, the discussion will be limited initially to complexes which do not have ancillary carbon monoxide ligMdS.

1. Nuckophific attack on cations which do not contain carbon monoxide as a &and

X-Ray crystallographic and spectroscopic studies have shown that nucleophilic attack invariably occurs onto the #o-face of the @and, i.e. on the side of the @and away from the metal. Ibe following examples are typical?&

I -. Q :

:. /

cs . \ L I

‘.,_’

m 6 .’ .,

* , ‘__’

Fe

Tab

k 1.

Nuc

ko~W

ic

Sna

ck o

n ~

~~

met

al c

atio

ns

whi

ch d

o no

t twa

in

carb

on m

onox

ide a

s II

ligw

d

hapt

o **

N

UC

LEC

BP

HIL

E

nu

mb

ers

CA

TIO

N

No

. v

RU

LE

S

4 P

RO

DU

CT

A

FPU

ED

R

EF.

s

10

11

12

7 13

14

15

18

H-.

Me-

w-,

ph

-, H

, R

”

CkH

s-, P

b-, H

-, R

Jf-*

oM

a-, a-

K,a

J-

K

H-

----_

--

I

1.23

2.3

2,3

7 8

a. lo

, 11

ae,t

t

13,1

4

14.1

5

16

17

Nuckophilk addition to orpnoburition metal cations

n 0. a.

E

a a

I -, w . .’

a a

z

a a ,-. a a ‘. . gv * .

a+o a : ,,

0 a+o a= 0

SM n rl:

I !‘:i-; -_’

hap

to

nu

mb

ers

CA

TIQ

N

NO

. N

UC

LE

OP

HIL

E

Y

PR

OD

UC

T

RU

LE

S

AP

PLI

ED

R

EF

.

37

H-,

CN

-, O

k&-,

Sk@

@-

1,2,

3 37

1 3%

3%

P

= 1.

2 b

isr(

dip

hen

ylp

ho

sph

lno

)eth

ane

P

Nuclcophilic addition to organotrensition metal cations 3053

the most favourable position of attack of a nucleophile on unsaturated hydrocarbon ligands in 18clectron organotransition metal cations for reactions that are kinetically rather than thermodynamically controlled:

Rule 1.t Nucleophilic attack occurs preferentially at even coordinated polyenes which have no unpaired elec- trons in the homo’s.#

Rule 2. Nucleophilic addition to open coordinated polyenes is preferred to addition to closed polyenes.

Rule 3.t For wen open polyenes nucleophilic attack at the terminal carbon atom is always preferred, for odd open polyenyls attack at the terminal carbon atom oc- curs only if ML.’ is a strong electron withdrawing group.

The following examples illustrate applications of the rules.”

above. Our literature search has only uncovered three apparent exceptions to the rules, i.e. the reactions of 19’* and 2719 with hydride and 39” with hydride, cyanide and acctylacetonate (see page 3054 and Table I).

It seems reasonable to postulate that the products of hydride attack on 19 and 27 are thermodynamically rather than kinetically controlled. &o-attack on the hexamethyl benzene ring of 19 would give a sterically unfavourable product with a Me group in close proximity to the metal. It is significant that for the unsubstituted complex (8) and the trisubstituted complex (18) nucleo- philic attack does occur at the benzene ring in ac- cordance with Rule I?*’

Compound 27 gives only the expected product (41) with Ph-= but a mixture of products (42 and 43) with hydride ion.19 It follows from our rules that the initial

The regioselectivity of the above reactions follows from the application of Rule 1. In the latter example the xemvalent compound Mo(@d16~~-C$IS) (40) is not formed even though it is isoelectronic with the well known Mo+Cd%J2’o complex and would have been the intuitively predicted product if one had regarded the cycloheptatrienyl ligand as approximating to the free aromatic tropylium cation.

Rule 2 is illustrated by the reactions of compound 13 given below.‘2”3

product must be 42 which subsequently rearranges to the thermodynamically more stable product (43). Such a rearrangement is more favourable for the labile hydride ion than the phenyl ion. The above explanation could be verified by variable temperature studies.

A similar explanation can be used to explain the reac- tions of 39 with nucleophiles, since it is known that hydride migrations occur very easily in such systems.““’

Cation 37= (see page 3054) demonstrates the need to apply the rules sequentially. Rule 1 (even before odd)

n I

Rh

(13) The large number of additional examples given in

Table 1 indicate the wide scope of the rules described eliminates the possiiility of attack on the ally1 ligand. Rule 2 (open before dosed) suggests nuckophilic attack on the butadiene and attack at the terminal posit@ of this ligand

tThesc roks have heen briefly described6 in previous papers in follows from Rule 3. terms of the retention of formal valency state of the metal The rules provide a ready explanation for the different

SAl&ougl~ cycb-butadiene is an ~MR polyenc it IUS unp&d modes of attack on the dications (26” and 31)” which ckctrons in its home aad therefore according to Rule 1 nucko- result in substitution into both rings for 26 and the same pbilic addition to other cum polyems is preferred. It is however, rb@ twice for 31. attacked in preference to odd polyeoyl ligands. It follows from Rules 1 and 2 that the predicted

3054 SIWHEH 0. DAVES eral.

_C--. 32 C,_!_’ F f H-

R ,c ---.,

R D . : -. _- R

Fe+ Hb

6 ‘I *, .

_-’

(6)R=H tI6)R=Me

R

141) (27) l42)

+ Ru

(26)

N~~~~~~~

reactivity of coordmated unsaturated hydrocarbons is as shown below

3oss

This series is supported by the examples in Table 1 sod by the observation that the cycloheptauiene Jigand in J is more susceptiile to nucleophilic attack than the benzene ring in 36.=

(35)

2. Z?wordictai basis of the n&s

(36)

In the introduction it was noted that the greater reac- tivity of the coordinated polyenes could be attributed to the net withdrawal of electron density from the Iii to the positively charged metal centre. This charge transfer process may be put on a more quantitative basis usiag the first order ~~~ theory ~n~.”

The w-systems of the linear poIyenes C!nEL+2 are aftemad’ and therefore for the neutral ligand the w electron density is equally distributed over all the C atoms. Coordination of these neutral l&tuds to a posi- tively charged ML fragment which has 18-n vaience electrons will generate the 18 electron K&H,,+&]+ complexes which concern us in this paper. Previous StUdkPU have sb0wntbatthcconical ML.+fragments have orbitals of the correct symmetry and ener8ies to enter into strong bonding interactions with the polyene r-orb&& of the free polyene @and in such a way that they will reflect the forward and back donation ~rn~n~

~~0~~~~~~~~ bondingmodel. complex the dominant charge transfer process will arise from donation of electron density from the ligand orbi- t& to the metal. Back donation effects will be relatively less important because of the initial positive charge on the A&+ fragment. The high carbonyl stretchin frequencies observed for metal CO cations (i.e. - 2100 cm-‘)” supports this assertion. Etectron donation from the highest occupied molecular orbital (burro) of the @and to an orbital of appropriate symmetry on the metal will be particularly iniluential in de&mining the dfstrfbution of electron density on the coo&m&d polyene. Tbe nievant interaction d&rams for even and ~~~~~~~~~~~~~0~~

ncgkted the bond&j molecular orbital 6 resultin8 from this interaction will have the norma&d form?

hom and the metal acceptor orbital respectfvely. The mixing coefficient wfh be set by the relative energies of the metal and polyene orbitals and the extent of inter- action between them. Fii 1 ihustmtes three possible situations for odd and even polyenes; e-90” where the rne~~p~~~~atrn~~~s~ the @and home and consequentty Iii charge transfer occurs, 6 -49 where the metal and @and orbitals are well matched and finally 8 - 00 whea the metal acceptor orbital ties at much lower energies than the polyene komo and extensive figand to metal charge transfer resulta.

~f~~~~of~~~~~y ~~~~of~~~~e~~s~y occupied fcads to an important general dftference be- tween these two types of @and. In the former case all the electron density in the bondin molecular orbital o&rates from the polyene home, but in the latter electron &n&y is contriiuted by both the polyene and the metal. From eqn (1) above it is clear that eIectron density asso&ed with the metal in & is equal to 2~0.5~ 8. Tbe resultant positive char8e on the eveu polyene ligand is therefore +2 cos2 B. For an odd polyene metal complex, however, the charge is 2 cos’ B - 1, because the metal acaptor orbital is occupied by au electron before it interacts with the pofyene homo. ~~,~~~~~n~for~e~~~ potyene Iigands are comparabk then a ~~ cacll polyene will have a unit extra of positive charge compared to an odd polyene. For the situations covered by Fig. 1 this means that the charge on an even polyene will vary from 0 to +2.0 and that for an odd polyene from-l.Oto+l.O.

Figure 2 busses the relative eaergies and nodal chara&G!uic!3 of the home’s of the polyene? c&H& withn=2-6.Tbealtemantpmpe&softhepolyenes Suepest that within the Httckel approximation tbe Ironzo’s of the odd polyenes are of equal ener8ies and non- bonding. The IJoulo’s of the euen polyenes are bonding and consequently more stable than the correspondiqt ~Of~~~l~*~~~~~~ to somewhat far8er values of the mixin coefficient for euen polyenes (i.e. electron donation from a more stable orbital to the metal will be smaller), however this diEerence is unlfkely to be sulISently large to undermine the fuudamental difference between odd and even polyenes noted above.

~n~~~~~ofn~~c attack on the coo&rated polyene it is the charge ou a parti&u C atom rather than total char8e on the polyene which is important. An estimate of the charge on a pa&ular C atom can be made if the explicit forms of the PO&mm hnno”s are substituted into eun (1). Witbin

STEOHEN G. DAVIES ef al.

EVEN POLYENE

8- 9o"

CHARGE

ON POLYENE - +O.O

\ I

Y0 Vb’& M

8 - 45O 0 - 0”

- +1*0 - +2-o

ODD POLYENE

I’ : I

I

I I

I I

I I

Jo P t’-___p;b

8 - 9o” CHARGE

ON POLYENE --to 0.0 - +l.O

IQ I. Interaction diagrams for polyene homo (&‘I and metal acceptor orbital ($,,,q.

n= 2 3 4 5

Fii. 2. Energies and nodal chmcteristics of polyene homo’s.

6

polyeae are given by

For the polyene home’s p = n/2 if II is cuen and (n + 1)/2 if n is odd Tbe charge on C atom c of tbe polyene is

therefore 2 cos’ Oc’,,, if rr is rven and (2 cos’ B - I)& if II is odd.

The charges calculated for the terminal C atom (i.e. p G 1, or n) of a series of polyenes are shown in Fig. 3 for two values of 0 corresponding to ML.’ being a strong electron withdrawing group (0 = 30”) and a weak electron withdrawing group (e = 609.

Nuckophilk addition to organotransition metal cations 3057

q(neved=$ 2cos’ 8 sin’ &

+0.5 - q(n odd)= 32cos%)

C-e cakubfed for ferminol carbon ofom In plyane 840 chaldql

2 \ \ 5 ; 4 ‘,“,’ = ‘,? No. of carbon atoms In

: I’ V ‘&6cP llnear polyene chain

u (n)

-45 t

Fig. 3. Charges calculated for the terminal carbon atom in coordinated polyenes.

From Fii. 3 it is clear that as long as 0 does not vary greatly with I then the charge on the terminal C atom is dominated by the parity of the polyene, i.e. the terminal C atom of an cuen polyene is more positively charged. The calculated positive charges diminish with increasing chain length for a series of polyenes with the same parity reflecting the smaller atomic coefficients at the terminal positions in the /W&S.

The rates of nucleophilic addition reactions of 18 electron cationic polyene complexes are likely to be charge rather than orbitally controlled especially if the nucleophile is small and highly charged and therefore the regioselectivity of such reactions is probably dominated by the positive charges on particular C atoms?’ In such cases the perturbation theory argument given above leads directly to Rule 1, i.e. nucleophilic attack occurs preferentially at wen coordinated polyenes.

From the perturbation theory expressions given above it can be shown that an even polyene will bear a positive charge whenever 0 ~90” and that the charge on the terminal atoms is always larger than that on the internal C atoms. The calculated charges for the coordinated butadiene ligand at different 0 values given below serve to illustrate this point..

0.28 0.21

no.,, A,.,,

8(‘) 0 30

In contrast, the terminal C atom in an odd polyene only bears the largest positive charge when 9 ~45”. For example for the coordinated ally1 ligand

0 0

effects will apply except that the positive charge will be more evenly distributed amongst the C atoms making nucleophilic attack less favourable than at the cor- responding open polyene. (Rule 2).

Although the perturbation theory arguments described above account for the rules for nucleophilic attack which have been proposed above it would also be useful to have the conclusions confirmed by more sophisticated m.o. calculations. In this respect, we note the recent im- portant work in this field by C1ack?*J’

3. E#ects of substituenls on the polyene ring Further evidence supporting the view that the nucleo-

phile attacks the C atoms with the least electron density is provided by examples on page 3058 which have electron withdrawi~ and releasing groups on the polyene.

For compounds 44 and 45 nucleophilic addition occurs preferentially onto the ring with the electron withdraw- ing carboxylate group and in the /3-positions.”

For compound 46 when X is an electron withdrawing group (e.g. -CO&le) nucleophilic attack occurs at the 2-position but when X is an electron releasing substituent (e.g. -0Me) attack occurs at the 3-position?”

Similarly compound 47 is attacked preferentially in the

0.07

no.le n,“I

60 90

meto-position when X is strongly electron releasing (X=NMed?’

Finally in compounds 48 and 49 the OMe substituent

0 0 0.5 -0.5 0.25AO.25 -0.25A-0.25 -0.5A-0.5

em 0 30

It follows that nucleophilic attack on an odd polyene will only be preferred when the metal ion is a strong eleciron withdrawing group. These conclusions lead naturally to Rule 3 above.

Although the arguments described above have not explicitly considered cyclic polyenes, clearly similar

60 90

deactivates the S-e&ion and attack occurs preferen- tially at position-I.

4. Nucleophific attack on cations containing one or more cation monoxide ligands

In Table 2 organometallic cations containing only one

STEFWEN G. Davies eral.

g ‘:.I..:

Ma

CO+ c02Me

& a- * : I

(44)

Me

T I : C02Me

*. .’

CO+

(45)

H H

.

@

: l .____a’ OMe r-rn-

Cr 7

(CO)3 CO)3

(46)

Ratio of otthometa attack

R=MeX=CI A=MeX=CI Y- Me-

Ratio 6!L a:26

R __ -- Q Y

L!_ XH

Mn + (CO13

ortho-attack

R

/--- Y

Q ‘-__- I H

X Mn tco),

meto-attack

R=MeX=CI R=HX=Ohde R=MeX=NMep Ph-

01% A3 3%

cI_ . y3c Fe

(CO)3

-Oh WOMe FO Fe (CO)3 CIY

(CO13

149)

TaM

e 2.

N~l

~~

atta

ck on

org

aaan

siti

on m

etal

cati

ons c

onta

inin

g one

uns

atur

ated

hy

droc

arbo

n lid

and

at k

ast o

n@ ca

rbon

mO

nOX

idc

&an

d

CA

TIO

N

NO

. P

RQ

DU

CT

S

DE

RIV

ED

F

RO

M

pfK

mu

mS

oE

fuvu

[)F

Ro

M

AT

rAcK

oN

co

AT

TA

CK

ON

- R

EF

.

..*.

a-

. . .

Fel

C0)

2NO

+

Fd

CO

l;

FeK

Of&

S}+

FeK

O),

KN

Me?

51

Fe

(CO

leN

O

58

Y*P

Ph

a

52

co

/ F

e -c

o

\ Y

H*N

Ha,

NZ

HS

,

00

Y’

PR

s Y

H*N

H3,

MeO

H

y-.H

- W

-$8

o-

: -2

P0

._.’

Fe-

P&

\

YH

=N,H

,

NC

0 Y

;Ns’

,-*.

54

:.*

y cl--

F

elC

O)e

(NC

Sl

Y’*

N;

co

Fe&

O

‘0s

Y -9

OR

”

RO

/

.-

a-

;-: Fe(CO

)#N

) Y

’.NC

S’,N

CO

’ b

_.’ .*

I\ cl--

:-

I_’

Fed

iNM

e Y

H=N

H

R

55

‘00

L

RH

N/

67

57

Y’“

QH

e-

,I

YH

4htie

NN

He

Y

=OM

e-

76

71

co

klnt

NO

c=O

Y

/ YH

= R

N&

fi

EB

t;/

TH

F

Y

H

x3-

\ N

nC

Q

~NO

){~h

~~

Y

H

i5-

Mn

GO

&

Y-=

n!i

e- o

f p

h-

v K

Ol4

y-=k

#-

Y

H

72.7

3

72

74

75

76

CA

TK

m

NO

.

E

Me

Y=M

e-, P

h- a

nd

H-

7%

72

73

Me,

N

8540

4

14

78

R

ON

-.i*

,acw

r.

y Ii c3

- \

Y-a

de-,

acac

-,

t 8f.i

. M

OlC

O1,

io

r, 18

8 0

Mo(

CO

l*( F

+R

,)

Y-=

khl

-,l4

-

y H

c3-

\ WKO

)~

Y=

OM

J, a

cac*

0 f+

Jp--

(~*

Fe(

CO

f,

108.

187

184

WM

W+

3SH

O

110-

118

t!!

YH

CaK

Me

Me

Fet

CO

)3

6 8

Tab

le 2.

(C

on

of)

CA

TIO

N

NO

. P

RO

DU

CT

S

DE

RIV

ED

F

RO

M

AT

TA

CK

ON

CO

P

RO

DW

CT

S D

ER

WE

D

FR

OM

A

TT

AC

K O

N H

YD

RO

CM

BO

N

RE

F.

Fe C

oo

);

85

a8

$7

Fel

CO

&

(CO

),

Y-=

H-,

CN

H 6

Osc

o),

/

141

142-

14s

148.

147

196

128

Y*=

acac

-, k

llmad

ancj

. . Nuckophilic addii to y metal caGons 3061

9 f

f K

I’\ b *

unsaturated hydrocarboo ligand and at least one carbon monoxide @and are listed. ‘Ihe compounds in Table 2 may undergo nucleophilic attack either on the hydro- carbon ligand or on the C atom of a carbon monoxide lii. As shown in the Table products front both types of reaction are observed. We note however that for those nucleophiles where attack is likely to be irreversible (e.g. alkyl) (i.e. the product will be kinetically controlled) it is gen&ally the hydrocarbon which is attacked. When the nucleophile is such that the product of addition to the unsaturated hydrocarbon is labile then the isolated product is that derived from attack on carboo monoxide. 10 the case of methoxide attack on compound SO attack occurs bdtially at the carbonyl and subfquent rear- rangement leads to the 0b3crved e~~-pmduct.~

WWO’ *&* - bs d’ ‘cro

co ‘*Me

I

d ,,/“iko co

IO Table 3 organometallic cations containing 2 or more unsaturated hydrocarbon ligands and at least one carbon monoxide ligand are listed. llie cations are again &s&d accor$ing to the hpo-humbers of the hydro- carbooligands.

As shown in the Table the Rules described ahove may often be used to predict the k&&ally preferred products of nucIeophIlic @a&. There are exceptions however. For examp&attaJdncotion(l~by~sndCN-givesthe predi&d produMs while attack by i-PrS- does not.‘” Such exceptionQnay arise because the ngioselectivity is not always g&eFned by the charge distriition oo the polyene. For the highly polarisable i-P& ligand the regioselectivity may be orbitally controIled.s’

&Lsp:~ & Fe Fe+

(CO13 KO) 3

(104)

Tab

le 3.

Nuc

koph

i at

tack

on o

rgan

Wti

a m

etal

cati

ons c

~nta

hhg

two

or m

ore

un

satu

rate

d

hyd

roca

rbo

n

Iii

and

at ie

ast o

ne c

arb

on

mo

no

xid

e Iii

rl

Wn

ON

N

O.

NU

CL

EO

PH

ILE

P

RO

DU

CT

R

EF

. R

UL

E N

O.

Ah&

u 10

6

M (c

o);

M&

8106

+C

?s1

07

106

106

110

ON

-

Y-4

r,t4

crtc

-

NH

s, R

NH

n, O

N-

tMo

jor B

(m

inor

l

Ru

lCO

l,

166,

166

1

ls?-

161

1

H

!A

162

1

t

1

163

1

163

1

P 51

I t

-.

2/5

. J

Q-

-i

Fe+

-CO

_*

’ I co

164,

165

1 C

N-,

Na-

11

2

113

Fe-

U-l

,-C

HrY

1 ‘P

Ph

, C

O

166,

167

1 I

- F

e*-P

Ph

3

I co

CN

. P

hsP

CH

o-,

-po(

Ow

z

ii I-

- O-r

x 17

;: e\

co

co

114

168

Ii-

Y,,

/--,

o

- 10

, -1 * _e

’ F

e\

/ co

co

CH

(CO

sEt)

n-

169

115

116

CH

R

CH

RY

I-

*\

0-F

I

-’ A

-/’ A

0 co

170.

171

1 H

-, E

&N

H, P

Phs

t I-

. o--

:

\ + 1

_’

/ e\

M

e co

P

P%

172

H-

117

,-.

0 : $ -\

--\

9 : \:

_* +

315

oNq”

>J

co

173

2,3

H-,

O&

tie-

, AcO

-, C

p-

iVl0

ON

’)

--L

Y

co

118

.-

. .._.

3070 !STIIPWH G. DAVIIIS ef al.

8

Nuckophilii addition to -n&ion metal C~I~DM 3u71

5. influence of the nucleophilc Evidence supporting the view that certain nuckophiles The influence of the nucleophile on the product can give labile products from addition to unsaturated

formed is emphasised by the reactions of compound 112 hydrocarbon ligands has been mentioned briefly above shown below. This difference can be interpreted in terms and may be further illustrated by the reversible reactions of reversible addition of azide to the ethykne lid but and reamWements iUustraWl on pace 3072 as irreversible addition to a carbon monoxide Iii leading to formation of coordinated isocyanate and the 6. Applications of de 3 liberation of molecular nitrogen. When the hydrocarbon @and being attacked is open

The difference in regkselectivity in the reactions of then it is necessary to apply Rule 3, since unlike the 114 and 1M with nucleophiles may also be explained in simple cfosed systems the carbon atoms are inequivalent. terms of kinetic vs thermodynamic control. For the ally1 lid there are two possible positions of

The many reactions of the cations [Fe+ attack, i.e. at the terminal C atoms (C-l or C-3) or at the

Kineticcontrol

OMe Thermodynamic

+ pps Thermodynamic

/y control

CSHS)(CO)~olefin)]’ have recently been reviewed.lW The unexpected variety of products obtained on reac-

tion of compounds 119 and la with nucleophiles is probably either the result of ready hydride rearrange- ments of the initially formed products or of initial attack on the metal followed by r~ments.“~‘n It is interesting that the only other case where a pentadienyl ligand is attacked in the 3-position is the reaction of the M-electron cation (122) with Phg.‘= The initial attack is presumably onto the metal as has been reported for the cation (123)Jm

oy PPh3 6’ . co

co

centre C atom (C-2). As shown on page 3073 examples of both modes of addition have been observed.

It is possible to correlate the position of attack with the electron richness of the metal. When the metal is ekctranrichasinlOsad11tbeallyllieandbebavesmore like( ~)andthenuckophileattacksthecarbonatom with the least electron density namely C-2 However when the metal is electron poor, by virtue of electron withdraw- in9 figands such as CO or NO, as in 78 and 118 the ally1 ligand behaves as ( + ) and the nuckophile attacks C-l

(1221

STEPHEN Cl. DAVIES et 01.

CQI,

167)

--.

Q \1 -_’ I

ilo+ \ P9 5

+ A0

OMe

GP Ii+ p’! \:

(/ _l

(37)

Lln (CO),

kf. 180

Ref. 30

Ref. 37

Q Fe (CO):

u.on ) qoMe A ) (pte Ref. ‘24

Fe(o), Fe(co1, (85)

(62)

RNH* Ref. 74 z

fvln(NOIWPh,)(CONHR)

pm3 ~

FelCO),(NO)e+ - Ref. 181

(61) FetCO12 (NO)

Ref. 109

FetbO), FekO),

Nuckophilic addition ti ~sition metal cations 3ln3

ref. 8

ref. 8

Y') /=Ly ref. 108

+

&,

(78)

Q ‘.j) -.

’ :

Q l -’

Mo+-_co y- *

g jNO Jy,Y ref. 173

(lb) Fig 4. Nuclcophilic attack on organometallic cations containing the ally1 ligand.

or C-3 since these are now the C atoms with the least electron density.

A similar situation arises for the pentadienyl ligands. Electron poor metals will be expected to lead to attack at C-l, CL3 or C-5 whereas for electron rich metals attack at C-2 or C-4 will be preferred. The examples shown below illu!3trate this.‘yJ Jl*‘R

0 \ ; I

’ _’

I M ICO) e PPh;

Y-

Y

G \/

H

Y- l&CO,,

i93)M=Fc(lOO)M=Ru

Y

@ - . H

MtCO), FPh,

+ Y

@

l -_ H

MKO),

Nuckophilic attack on the relatively electron rich coppounds (98 and Its) preferentially occurs at C-2 wha#rittthelewClU%ollrichcompoundS(93andl~a mixtureofC-landG2tmu&ctsisobserved.

Supportfortheassumptioathatexchangcofac8rbon

monoxide li8and for triphenylphosphine increases the electron density on the metal is provided by CO IR frequency data. For compounds 98 v,.= 2050, 2007 cm-‘,“’ 93 vW = 2100, 2055 cm-‘,‘** 52 v, = 2120, 2070 cm-‘,‘” and 53 v., = 2955,201O cm-‘.‘” The shifts to lower wavenumbers can be related to an increase in the electron density at the metal.

(52) (53)

For euen open Qands electron density will always tend to be least at the terminal C atoms and as expected attack always occurs there.

7. Miscellaneous examples Table 4 lists some examples of or8anotransition metal

cations that lead either to additon of the nucleophile to the metal or to li8and substitution. Addition to the metal occurs when the cation has 16 or fewer electrons. L&and substitution occurs when the M-electron cation can be rnadily cooverted to a 16&ctron cation by dissociation of a 24Mroo @and. The oucleophile subsequently attacks the 16&ctron intermediate catioo 8ivi118 the new 18 ekct=produet.

3011 S~WHEN G. DAVIES cf a/.

Tabk 4. Nucleophiii addition to the metal and Iii rubstitution reactions of m&ion metal cMions

CATION NO NUCLEOPHILE PRODUCT REF.

n-.*

Q ’ I__!

FelCO)~ --\ : I Q -2’ Fe+KO)2 (OCMep)

, --\ P .e’ Rh(CH,CN)(PPh&COl+

128

52

127

128

28

L = HpO, @H&CO. C&id

H-

Y=Br-, k-, BCN-, NOa-

y-=NCS. N&-(LrPPhs, Et&I

CN

130 PPhr

131 CN

132 CN

133

75

55

kICO),H

Mnb,CN

-. Q I : l - .’

Ni

NC / ‘P up ,--.

Q '..I)

79

188

189

lio(CO)eCN) IAsMe,C3Hs)

190

y-=Q-, BCN-, YBOa-==l’-% 191 RhlCH2CN)IPPh3)(Y)

r

3076 STEPliEN 0. DAVIES d al.

eoP. L. Pauson and J. A. Segal. Ibid. Dalton 1677 (1975). “P. H. Bird and hi. R ChurchiU, char. &mm. 777 (196’1). nc. Winkhaus and H. Singer, E Naturfwwh. ISa, 418 (1963). ?U. G. Connelly and R L. Kelly, J. &em, Sot. Dalton 2334

(1974). %. E. Ball and N. G. Connelly. 1 Cbgoaometal. Chm. 55, C24

(1973). %. H&ett and 0. Jamen, Inotg. Chim. Acfa 12 L19 (1975). 9. E. Baikie. 0. !I. Miis. P. L. Pauson. G. H. Smth and J.

Valentine. &vu. Comm. 425 (1%5). wJ. D. Munm and P. L. Pausoo. J. Chem. Sot. 3475 (l%D. wJ. D. Munro and P. L. Pauson. Ibid. 3479 (l%l). 1. D. Munm and P. L. Pauson. Ibid. 34% (1961). ““‘J. D. Munro and P. L. Pauson. hoc. Chem. Sot. 267 (1959). “‘P. L. Pauson, 0. H. Smith and J. H. Valentine, J. Qem. Sot.

(Cl. 1057 (1967). ““P. L. Pauson. G. H. Smith and J. H. Valentine, Ibid. (C). 1061

(1%7). ? A. Mansljeld, K. M. Al-Kathumi and L. A. P. Kane-

Magurie. 1. ogOnometul. Chem. 71 Cl I (1974). ““‘K. M. AI-Kathumi and L. A. Kane-Maguii, Ibid. 1@2, C4

(1975). ‘-M. L. Ziealer, H. E. Sassc and B. Nuber. 2. Naturforsch. Jgb

22 (197S);IJbid 26. “?I. Deganello, T. Boschi and G. Albertin. Log. Chim Acto 10,

L3 (19741. lurE. E. Isa&s and W. A. G. Graham, 1. Orgonometal. Ckm. 90,

319 (1975). ‘9. H. Whitesides. R W. Arhart and R. W. Slaven. J. Am.

Chem. Sot. 95,5792 (1973). ‘wA. J. Biih and I. D. Jenkins, Tmnsifion Mefal @anometa/-

lies in 0punic Synthesis I, (Edited by H. Alper). p. 13. Academic Press, New York (1973).

“‘J. E. Mahler and R. Pettit. J. Am. Chem. Sot. 84,15I1(1%2). “‘J. E. M&r and R Pdtit. Ibid. 65.3955 (1%3). “*J. E. Mahlcr. D. H. Giin and R Pettit.Ibid.~l35.3959 (1963). IuP. McArdk and H. Sherlock, J. ogcUrometa/. Chem. 52, C29

(1973). ‘I’M. Anderson, A. D. H. Clague. L. P. Blaauw and P. A.

Coupenls. Ibid. 56,307 (1973)._ “‘B. F. 0. Johnson. J. Lewis. D. 0. Parker and S. R. Postlc. 1.

them. see. Lkdtoil794 (19n). I%. Ma@o. A. Musco. R Palumbo and A. Sii. Chem. Comm.

loo (1971). “‘G. Maglio. A. Musco and R Pabunbo. J. Ogonomefal. Chem.

32, 127 (1971). ‘IsT. G. Banner. K. A. Holder, P. Powell and E. Styles, Ibid. 131.

105 (19713. “‘A. J. Birch, P. E. Cross, J. Lewis, D. A. White and S. B. Wild,

J. Clta. !Ioc. (AI), 332 (1968). ‘q. Becker, A. Eisenstadt and Y. Shvo, Tetrahedron Letters

3183 (1972). ‘*‘J. Evans, D. V. Howe, B. F. G. Johnson and J. Lewis. J.

Otganometul. Cha. 61. C48 (1973). ‘*A. J. Birch, I. D. Jenkins and A. J. Liepa. Tefmhufrur’ Letters

1723 (1975). ‘*‘A. J. Biih and A. J. Pearson, Ibid. 2379 (1975). ‘%. E. Hine. B. F. G. Johnson and J. Lewis, Chem. Comm. 81

(1975). IBM. A. Heshmi. J. D. Munro. P. L Pauson and J. M. Wiikm-

son, J. them. sot. (A), 249 (1%7). ‘=A. J. Birch. K. B. Chamberlain. M. A. Haas. D. J. Thomtwon.

Ibid. Perkhl I. 1882 (1973). ‘% R. John, C. A. Mansfield and L. ‘A. P. KaneMapuire. Ibid.

Dalton 574 (1977). ‘wL. A. P. Kane-Maguire and C. A. Mansfield, C&m. Comm. 540

(1973). ‘3. A. P. Kane-Ma&re. J. Cha. See. (A), I662 (1971). ‘wA. Pelter, K. J. Gould and L. A. P. Kane-Maguii. Chem.

Comm. 1029 (1974). “‘A. J. Birch and G. S. R. SubbaBao. Tetmhedmn L.#ers 3791

Wm. ‘OR. E. Ireland, G. G. Brown, R H. Stanford, and T. C.

McKenzie, 3. olspnomctal. Qn. 39,51(1974).

“‘A J. Birch, K. B. Chamberlain and D. J. Thompson. J. Cha. sot. Perkin I 1900 (1973).

‘~. Aumann. I. ogCmometa/. Chem. 47, C29 (1973). ‘% Edwards, J. A. S. Howell, B. F. G. Johnson and J. Lewis, 1.

Chem. sec. DaBon 2105 (1974). ‘~.M&haudhan and P. L Pauson. J. Otganometal. Gem. 5.73

InA. Da&on, W. McFarlane. L. Pratt and G. Wilkinson, Chem & lad. 553 (I%]).

‘*A. Davison, W. McFarlane. L Pratt and 0. Wilkinson. 1. Cha. Sot. 4821 (1%2).

‘wJ. Evans, D. V. Howe, B. F. G. Johnson and J. Lewis, 1. GrpoaomcJol. Chem. 61, C46 (1973).

‘@A. Eisenstadt and S. Winstein, Tetmhedmn L.eftrr 613 (1971). “IN. Genco, D. Marten, S. Ita& and M. Kosenblum. 1. Am.

C’hem. SOE. %, 848 (1976). ;z. Aumann. Aupew. Chem. lnt. Edn. 12.574 (1973).

’ . Davison, W. McFarlane and G. Wdkinson, Chem. & Ind. #20 (1962).

l”A. Davison, W. Mcarknc, L Pratt and G. Wilkinson, J. C&m. Sot. 4921 (1962).

l”J. D. Holmes and R. P&it, J. Am. Chem. Sot. lJ3,2531 (I%3). ‘3. E G. Johnson, J. Lewis, P. McArdle and G. L. P. Bandah,

1. Chem. !Ioc. IUton 456 (1972) “‘B. F. G. Johnson, J. Lewis, P. McArdk and G. L. P. Bandah,

Ibid. Dalton 2076 (1972). ““B. A. Sosinsky, S. A. R. Knox and F. 0. A. Stone, Ibid. Dalton

1633 (1975). ‘@A. J. P. Dominpos, B. F. G. Johnson and J. Lewis, Ibid. I)plton

22S8 (1975). ‘~. Schiivon, C. Paradisi and C. Boanini, Inorg. Chim. ACM 14.

“‘FE: J. Miller. P. L. Pauson and J. B. Pol Tripathi, J. c&r. Sot. (0 743 (1971).

IuP. J. C. Walker and R J. Mawby. Inot& ain. Acta 7. 621 (1973).

Is3P. J. C. Walker and R J. Mawby. J. OtganometaI. Chem. 55. c39 (1973).

‘%J. C.-Burt; S. A. R Knox. R. J. McKinney and F. G. A. Stone, J. Chem. Sot. Dalton I (1977).

“‘A. Eisenstadt, J. OrgPnometal. Chem. 60,335 (1973). ‘%A. Eisenstadt. Ibid. 113,147 (1976). InA. D. Charles. P. Divers. B. F. G. Johnson, K. D. Karlii.

J. Lewis. A. V. Rivers and G. M. Shddrick. Ibid. 128, C31 (1977,.

IwF. A. Cotton, A. J. Deeming, P. L. Josty. S. S. Gllah. A. J. P. Domingos. B. F. G. Johnson, and J. Lewis, J. Am. Cha. Sot. 93.4624 (1971).

‘9. A. Cotton, M. D. LaPrade, B. F. G. Johnson and J. Lewis. Ibid. 91.4626 (1971).

‘9. A. Cotton and M. D. LaPrade. J. Orgonomefal. CSm. 39, 345 (1972).

16’A. J. Deeming, S. S. D&h. A. J. P. Domiquw. B. F. G. Johnson and J. Lewis, J. Chem. Sot. Dalton 2993 (1974).

‘62M. Cooke, P. T. Dtqgpett, M. Green. B. F. G. Johnson, J. Lewis and D. I. Yarrow. Chem. Corns. 621(1971).

‘w. H. Knoth, Jnorg. Chem. 14.1566 (1975). ‘@A. M. Rosan. M. Rosenblum and J. Tancrede, J. Am. Chem.

Sot. 93.3062 (1973). 16”A. Rosan and M. Koscnbkun. I. Organometal. Chem. 116. 103

119741. ‘“D. L..Keger. Inorg. Chem. 14.660 (1979. ‘“‘D. L. Beger and E. C. Culbertson. J. Grponomefol. C&m. l31.

29l Wm. ‘-M. L H. Green and P. L. I. Nagy. Ibid. 1.58 (1%3). ‘%. Roscnblum. Ace. Chem. Res. 7, 122 (1974). ‘9. W. Lichtenbcrg and A. Wojcicki. J. Am. Chem. !kx. 94.

8271 (1972). “‘D. W. Lichtenbcrg and A. Wojcicki. J. Ogcmometal. Chem. 94,

311 (1975). ‘wM. L. H. Green. L C. Mitchard and hf. G. Swanwick, J. Chem.

SC. (A) 794 (1971). “N. A. Bailey, W. G. Kita, J. A. McCkverty, A. J. Murray, B.

E. Mann and N. W. Walker. Cha. Comm. 592 (1974).

Nuckophilic addition to organotransition metal cations 3077

“‘B. F. G. Johnson, J. Lewis, T. W. Matheson, 1. E. Ryder and M. V. Turii, Ibid. 269 (1974).

“.‘J. AshkySmith. D. V. Howe, B. F. G. Johnson, J. Lewis and I. E. Ryder, J. Organome~al. Chem. 112.257 (1974).

‘9. Edwards, J. A. S. Howell, B. F. G. Johnson and J. Lewis, 1. Chem. Sot. Dalton 2105 (1974).

‘=B. F. G. Johnson. J. Lewis, I. E. Ryder and M. V. Twigs. Ibid. Dalton 421 (1976).

‘“E. 0. Fischer and F. J. Kohl. Chcm. Ber. 98,2134 (l%S). ImA. Salzer. and H. Werner, J. Organometal. Chcm 87, 101

(1975). IaM. R. Churchill and F. R. Schokr, Jnotx. Chem. 8, 1950

(1969). ‘*ID. A. Sweigart and C. N. Wilker, Chem. Comm. 304 (1977). ‘“H. J. Dauben and D. J. Bert& J. Am. Chem. Sot. 83.497

(1961).

‘%. J. Angelici, P. A. Christian. B. D. Dombek and G. A. Pfeffer. J. Chganometal. clrrm. 6X 287 (1974).

‘9. M. Treichel, R. L. Shubkin. K. W. Barnett and D. Reichard. Inorg. chm. 5. I I77 (KM).

‘“‘B. Faschinetti and C. Floriani. Chem. Comm. 578 (1975). ‘9. D. Dombek and R. J. Angelici. Jnorg. Chim. Acta 7, 345

(1973); W. E. Williams and F. J. Lalor. J. Cha. Sot. Dalton 1329 (1973).

InrE. C. Johnson, T. J. Meyer and N. Winterton. Inorg. Ckm. 19. 1673 (1971).

‘=F. W. S. Benfield, B. R. Francis and M. L. H. Green, J. Organometacrl. Chem. 44, Cl3 (1972).

‘9. Sate. T. Uemura and M. Sate, Ibid. 39. C25 (1972). lwK. P. W&right and S. B. Wild. C/tern. Comm. 571 (1972). ‘*IF. Faraone. F. Cusmano. P. Piraino and R. Pietropaolo. 1.

Organometal. Chem. 44.391 (1972).