The solution structure and fluxional behaviour of cyclic and acyclic (diene)Fe(CO)2L complexes (L = phosphine, phosphite, isonitrile)

Article abstract of DOI:10.1016/0022-328X(91)86199-Z

Variable temperature 13C and 31P NMR spectroscopy has been used to establish solution structures and conformer populations for a variety of cyclic and acyclic (diene)Fe(CO)2L complexes (L = phosphine, phosphite, isonitrile).The solid state structures of (2,3-dimethylbutadiene)Fe(CO)2PPh3 and (trans,trans-2,4-hexadiene)Fe(CO)2PPh3 have been determined and used as a basis for molecular modelling of steric effects in these complexes.

Full text of DOI:10.1016/0022-328X(91)86199-Z

91
Journal of Organometallic Chemistry, 401 (1991) 91-123
Elsevier Sequoia S.A., Lausanne
JaM 21208
The solution structure and fluxional behaviour of cyclic
and acyclic (diene)Fe(CO)2L complexes
(L = phosphine, phosphite, isonitrile)
James A.S. Howell *, Gary Walton, Marie-Claire Tirvengadum, Adrian D. Squibb,
Michael G. Palin
Chemistry Department, University of Keele, Keele, Staffordshire ST5 5BG (U.K.)
Patrick McArdle, Desmond Cunningham,
Chemistry Department, University College, Galway (Ireland)
Zeev Goldschmidt, Hugo Gottlieb and Gila Strul,
Chemistry Department, Bar llan University, Ramat Gan (Israel)
(Received June 18th, 1990)
Abstract
Variable temperature 13C and 31p N M R spectroscopy has been used to establish solution structures
and conformer populations for a variety of cyclic and acyclic (diene)Fe(CO)2 L complexes (L = phosphine,
phosphite, isonitrile). The solid state structures of (2,3-dimethylbutadiene)Fe(CO)z PPh 3 and (trans,trans-
2,4-hexadiene)Fe(CO)2PPh3 have been determined and used as a basis for molecular modelling of steric
effects in these complexes.
Diene and dienyl complexes of tricarbonyliron continue to attract attention as
synthetic intermediates, particularly for enantioselective synthesis. Though the
chemistry of cyclic six- and seven-membered ring complexes has already been
developed extensively [la-c], there has been a recent renewed interest in the use of
acyclic diene complexes targetted towards several molecules of pharmacological
interest [2]. The ease of preparation of related (diene)Fe(CO)2L derivatives (L =
phosphine, phosphite) [3a-c] provides alternative complexes which exhibit increased
reactivity towards electrophiles and cycloaddition, [4a-c] altered and increased
regiospecificity in nucleophilic attack [4b,5a,b], and the potential for resolution or
asymmetric induction via the use of chiral auxiliary ligands [4b,6a-d].
Particularly in the case of regiospecificity, such changes may be the result of site
preference of the auxiliary ligand for non-equivalent sites within the square pyra-
midal molecular structure. We wish to report here NMR and single crystal studies
which establish the solid-state and solution structures of a variety of simple
diene-substituted cyclic and acyclic (diene)Fe(CO)2L complexes (L = phosphine,
phosphite, isonitrile), together with molecular modelling studies which at least for
flfCT~-~gRY/ql /~0"~ 50
© 1991 - Elsevier Seauoia S.A.

92
the most sterically demanding PPh3 complexes, provide some discrimination be-
tween steric and electronic effects [7].
Though a number of experimental studies have been reported on fluxional
processes in (diene)M(CO)3 complexes (M = Fe, Ru, Os) [8a-g], systematic studies
of (diene)M(CO)2L complexes are less numerous and confined to cases where
L = PF3 [9a,b], isonitrile [10a-d] and P(OCH2)3CEt [8f] and to phosphine and
phosphite substituted (~4-enone)Fe(CO)2L derivatives [lla,b].
Results
(i) Preparation and fluxionality
The phosphine, arsine, stibine and isonitrile complexes 1-7, 9-17 and 2 0 - 2 7
R 3
R2
1
4
Fe(CO)2L
Fe(CO)2L
R 1
R 2
R 3
L
R 1
R2
L
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
H
H
H
H
PPh 3
PPh 3
(9)
(10)
(11)
(12)
(13)
(14)
(15)
(16)
(17)
(18)
Me
H
H
H
H
H
H
H
H
H
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
PPh 3
H
Me
PPh 3
H
H
Me
H
PPh 3
PPh2Me
PPhMe2
AsPh 3
SbPh3
CNMe
Me
H
Me
PPh 3
Ph
H
PPh3
Me
Me
Me
C02Me
CHO
CHO
H
PPh 3
H
PPh 3
H
P(OMe) 3
CNEt
CNBu t
P(OMe) 3
Fe(CO)2L
Fe(CO)[P(OMe)3]2
(19)
n
L
(20)
2
3
2
3
2
3
PPh3
PPh3
CNMe
CNMe
CNEt
CNEt
(21)
(22)
(23)
(24)
(25)
(26)
(27)
2
3
CNBu t
CNBu t
(28)
3
P(OMe) 3
(29)
3
P ( O P h ) 3

93
were prepared by substitution of the tricarbonyl in the presence of Me3NO [3a]. The
phosphite complexes 8, 18 and 19 were prepared photolytically, while 28 and 29
were prepared by direct thermal substitution of the tricarbonyl [5a].
Satisfactory analytical and spectroscopic data were obtained for all complexes.
(Tables 1 and 2) Though a small number of the complexes have been previously
prepared, variable temperature NMR studies have been reported only on 24, a
complex where our conclusions differ from those previously published [10b].
Proton spectra of these complexes (relative to the tricarbonyl) show consistent
shielding of the terminal diene hydrogens. There is no clearly consistent trend in the
shielding of the averaged room temperature chemical shifts of the diene carbons,
nor in the magnitude of the long range P-C and P-H coupling constants. P-C
coupling is confined exclusively to the terminal diene carbons; P-H coupling is
observed to both terminal and internal hydrogen, but for acyclic complexes is
strongest to Ha protons. Small long range P-C coupling to the methyl substituent is
also observed in the isoprene complexes 10-12 and 18. All complexes show a
deshielding of both axial and basal CO resonances [12"] which increases in the
order CNR ~ P(OR)3 < PR3; for phosphine complexes, P-CO coupling constants
provide a reliable aid to structure elucidation in the low temperature limiting
spectra and have the values J(Paxial-CObasal) ~ 5 Hz, J(Pbasal-COaxial) ~- 4 Hz
and J(Pbasal-CObasal)~ 25 Hz. Couplings for phosphite complexes are in the
same order, but slightly larger.
The fluxional process in these Fe(CO)2L complexes can be represented in the
following way:
81
L
(A)
(B')
Mechanistically(,B) exchange can occur by either sequential Berry pseudorotation or
simple rotation of the diene relative to the Fe(CO)EL fragment (turnstile mecha-
nism); the two processes are indistinguishable by NMR [13a,b]. Because of strain
involved in the axial/equatorial site occupancy of the diene in the trigonal bipyra-
midal intermediate of the Berry process, the turnstile mechanism is preferred, and is
the one assumed here in the modelling studies.
For a complex possessing no mirror plane (e.g. X = Me), A, B and B' represent
chemically distinct conformational isomers in which two non-interconverting sets of
CO ligands (bl/b3/a2 and bE/al/b4) undergo exchange. The appearance of the
averaged tSC spectrum will depend on the identities and relative populations of the
conformers, but exchange does not completely average the CO resonances and two
averaged resonances are observed for all compounds of this type, with the exception
of the isonitrile complex 15. For complexes possessing a mirror plane (e.g. X = H),
B and B' form an enantiomeric pair, and complete carbonyl scrambling occurs,
*
Reference n u m b e r with asterisk indicates a note in the list of references.

94
Table 1
~3C and 31p NMR spectral data a
Complex Averaged spectrum
(20 ° C)
Limiting low temperature spectrum Temper-
Isomer ratio
A
B
B'
ature ( ° C)
i
1,4
2,3
CO
PPh3
P
40.6 [40.7] b
84.3 [85.5]
217.8 (13.9)
c
a
33.9, 43.4
80.9, 84.6
215.1 (26.2)
221.1 (3.6)
72.3
- 8 5
A : B / B ' •l:6
71.9
76.5
e
2
1
2
3
4
Me
CO
42.7 (6.9) [39.4]
82.2 [80.91
86.6 I89.01
52.8 (7.9) [58.51
19.0
215.8 (23.4)
221.2 (4.9)
71.6
- 9 0
A : B = I : 1 8
P
73.2
72.2
- 8 5
B : B ' f f i 2 . 5 : I
35.8
89.6
87.3
48.1 (br)
13.6
215.3 (26.5)
222.9 (4.9)
71.1
3
1
2
3
4
Me
CO
40.8 I40.91
90.1 [90.41
86.9 [88.21
49.7 (4.9) [53.9]
13.8
217,8 (16.6)
219.0 (10.7)
68.9
43.5
91.6
85.4
47.3 (8.8)
14.4
215.8 (25.5)
222.4 (br)
68.4
P
-
85
A : B / B ' ffi 52 : 1
4
5
1,4
2,3
Me
CO
P
58.1 (3.6) [56.9]
85.3 [84.8]
17.7
217.1 (4.9)
68.0
~
69.1
e
66.3
,
1
2,3
4
43.3 (6.8) I38.8]
81.6, 82.6 [81.2, 82.11
56.2 (5.9) [59.4]
142.1 (a)
- 8 5
A : B f f i l : 1 3
Ph
128.0 (fl)
125.9 (y)
124.9 (8)
CO
P
214.8 (23.4)
220.4 (5.9)
71.6
68.2
72.5
- 85
axial only
6
1
2
3
4
61.8 [59.6]
88.6 [88.31
83.6 [82.81
48.5 (5.9) [45.61
18.3
e,/
Me
CO2Me 50.6, 174.1
CO
211.8 (5.9)
214.6 (8.8)
64.9
P

95
Table 1 (continued)
Complex Averaged spectrum
(20 o C)
Limiting low temperature spectrum Temper-
Isomer ratio
A
B
B'
ature ( ° C)
7
1,4
61.7 (1.9) [60.31
59.3 (4.9) [54.5]
90.4 [89.5]
82.3 [81.2]
18.1
~
- 6 0
A : B / B ' = 32:1
2
3
Me
CHO
CO
197.5
211.7 (4.9)
214.8 (8.8)
63.8
P
64.0
•
58.8
•
8
1,4
54.9
- 6 0
A : B : B ' • 1 : 4 : 2 2
57.6
2
88.5
3
80.3
Me
CHO
CO
17.8
197.2
210.5 (br)
212.8 (16.8)
182.6
P
183.3
181.9
182.9
9
1,4
2,3
Me
CO
P
47.5
98.3
19.8
214.1 (4.9)
74.8
e,/
- 85
axial only
10
1
41.6 (5.3) [37.4] 35.9
43.1
-100
A : B ~ I
2
86.1 [84.1]
100.4 [102.4]
42.7 [43.1]
22.0
214.9 (16.1)
218.2 (4.3)
74.4
84.2
97.9
40.0
21.1
213.7 (3.2)
213.9 (5.3)
73.3
86.1
101.6
48.0
22.3
214.6 (25.6)
221.7 (6.4)
77.1
3
4
Me
CO
P
!1
1
2
40.4 (5.9)
84.9
37.3
84.9
40.8 (br)
- 8 5
A : B f f i 2 : I
3
100.5
102.2
98.2
4
Me
CO
43.5
22.8
214.7 (13.7)
217.0 (5.9)
41.8
23.7
214.3 (4.9)
214.4 (br)
47.6 (br)
22.9
215.0 (31.8)
222.5 (4.9)
PPh2Me 19.7 (28.3)
128.2 (9.3) (~,)
129.5 (3.0) (~)
131.5 (10.9) (fi)
138.7 (38.0) (a)
138.8 (38.0) (a')
53.3
P
53.3
56.7
12
I
2
38.9 (5.9)
84.3
39.9 (5.9)
84.0
37.2 (5.6)
85.4
- 8 5
A : B f f i 5 : I
3
100.6
101.7
98.7
4
Me
CO
43.1 (3.0)
23.0
214.3 (12.6)
215.8 (6.9)
45.8 (7.8)
22.8
213.9 (6.8)
214.4 (7.8)
38.7 (br)
23.1
215.1 (29.3)
222.2 (5.8)
(continued)

96
Table 1 (continued)
Limiting low temperature spectrum Temper-
Isomer ratio
Complex
Averaged spectrum
(20 ° C)
A
B
B'
ature ( ° C)
12
PPhMe2 19.5 (24.0)
19.6 (24.0)
129.0 (9.0) (7)
128.2 (2.0) (8)
127.8 (8.8) (fl)
141.0 (29.1) (a)
35.6
P
36.6
37.7
13
1
2
3
4
Me
CO
39.2
84.8
100.5
43.1
22.9
214.3
215.9
40.6
84.8
101.8
45.9
23.6
36.7
85.0
98.0
39.5
22.8
215.6
222.1
-
85
A : B = 6:1
213.9
214.0
AsPh 3
137.4 (a)
132.6(#)
129.2 (8)
128.4 ('y)
14
1
35.1
e
- 80
axial only
2
82.5
3
99.4
4
40.8
Me
CO
22.6
213.8
214.4
SbPh s
127.9 (8)
128.7 ('},)
135.1 (fl)
135.9 (a)
37.2
83.7
- 7 5
A : B : B ' = 1 : 5 : 5
15
1
2
3
4
Me
CO
35.1
36.1
82.2
99.9
a
36.8
81.6
99.6
42.4 (br)
22.1
219.5
213.4
29.6
83.5
100.8
41.3
22.0
216.2
100.5
g
22.4
a
212.1
212.4
d
213.9
31.6
MeNC
29.7 (br)
158.5
163.4 (br)
162.4
157.1
h
16
1
2
35.0
83.4
h
- 7 5
A : B : B ' = I : 5 : 8
83.7
82.2
81.5
3
4
100.8
41.4
99.8
h
100.3
h
99.5
h
22.4
Me
CO
22.9
a
22.1
219.5
213.8
14.9
*
216.1
216.5
15.6
212.2
212.4
d
h
213.4
14.6
h
EtNC
39.4 (br)
156.8
163.4 (br)
161.8
158.2

Table 1 (continued)
97
Complex Averaged spectrum
(20 ° C)
Limiting low temperature spectrum Temper-
Isomer ratio
B : B ' = I : 3
A
B
B'
ature ( o C)
36.7 (br)
80.9
99.6
g
17
1
39.2
- 8 0
2
82.6
83.2
99.3
g
3
4
100.1
41.2
42.1
Me
CO
22.4
21.6
215.5
216.6
30.1
219.1
213.4
29.1
g
212.9
31.2
155.5
B u t N C
161.1 (br)
156.5
37.2 (10.7)
a
18
1
2
35.8 (6.7)
82.2
36.2
82.4
-105
A : B : B ' = 8 : 1 7 : 1
3
4
Me
CO
99.2
40.5
21.4
213.9 (20.2)
216.4 (10.8)
100.3
42.1
22.1
98.7
38.5 (14.8)
g
212.1 (br)
212.1 (34.9)
220.1 (12.1)
P(OMe)3 51.1 (2.9)
P
189.9
197.5
189.4
187.5
e.f
20
1,4
2,3
CH 2
CO
P
61.5 (3.0) [61.11
84.6 [84.3]
24.7
218.7 (13.6)
70.2
basal only
basal only
21
1,4
2,3
57.3 [59.5]
87.8 [87.9]
60.4, j
-
85
85.5, 89.7
28.5 (br)
24.4
C H 2 ( a ) 28.5
C H 2 ( f l ) 24.5
CO
218.9 (13.8)
216.2 (24.4)
222.7 (br)
/
P
70.3
22
1,4
2,3
CH 2
CO
58.6
84.7
24.1
58.6
82.4
23.1 (br)
213.3
60.2
- 8 5
A : B / B ' = I : 7
82.9, 85.2
g
216.5
214.7
218.9
29.6
MeNC
29,7 (br)
*
30.5
163.4
157.5
23
1,4
2,3
55.7
87,2
a
a
58.7, j
86.4, 88.5
27.9, 28.2
24.3
-
80
A : B / B ' = 1 : 10
C H 2 ( a ) 28.1
CH2(fl) 24.4
a
a
CO
216.4
213.2
214.7
220.8
MeNC
29.8 (br)
30.7
30.6
164.2 (br)
167.9
159.1
1,4
1,4
2,3
CH 2
CO
58.6
84.7
24.1
58.4
82.3
22.5
60.3
83.0, 85.4
23.2
214.7
219.1
14.5
- 8 0
A : I I / B ' = 1 : 3
216.6
213.4
EtNC
15.6
14.5
39.3 (br)
*
38.6
162.9
38.6
157.3
(continued)

98
Table 1 (continued)
Com- Averaged spectrum
Limiting low temperature spectrum
Temper-
ature
(°C)
Isomer
ratio
plex
(20 o C)
B'
A
B
25
1,4
2,3
C H 2 ( a)
CH2(fl)
CO
55.8
87.3
28.2
24.4
216.4
a
a
57.5, j
85.5, 87.7
27.0, 27.2
23.3
- 8 0
A : B / B ' =1:8
26.6
a
212.2
212.2
213.7
14.7
EtNC
15.6
15.3
39.3 (br)
164.3 (br)
38.7
166.5
38.6
157.8
26
1,4
2,3
CH 2
CO
58.7
84.7
24.2
216.5
58.3
82.2
23.1
213.5
60.4
83.2, 85.6
- 8 5
A : B / B '
= 1 : 2
214.8
219.1
g
But NC
30.7
*
29.4
161.3
156.5
--85
A : B / B '
= 1 : 5
27
1,4
2,3
C H 2 ( a)
CH2(fl)
CO
55.7
87.2
28.2
24.5
55.3
85.1
26.4
23.2
55.6, 57.6
85.5, 87.7
26.8, 27.1
23.2
213.4
216.3
212.4
220.0
ButNC
30.7
k
30.1
164.7
29.4
156.9
57.9, y
85.5, 87.8
27.8
--85
basal
only
28
1,4
2,3
C H 2 ( a)
CH2(B)
CO
57.0
86.4
28.1
24.3
24.2
217.3 (20.4)
213.6 (34.2)
221.8 (7.8)
50.7
P(OMe)3
P
50.9
184.0
/
60.0 (7.8), j
85.2, 87.2
27.8
-
80
basal
only
29
1,4
2,3
C H 2 ( a)
58.4 (4.9)
86.3
27.9
23.9
212.8 (31.3)
219.9 (3.9)
CHE(fl)
CO
24.1
216.0 (18.5)
P(OPh)3
151.5 (5.9) (a)
129.6(8)
124.6 (8)
121.3 (5.0) (V)
169.0
/
P
35.5
(t, 9.8, 9.8) t
19
1
d,m
2
3
4
80.1
95.9
38.6
(dd, 9.3, 16.1)

99
Table 1 (continued)
Limiting low temperature spectrum Temper- Isomer
ratio
Corn- Averagedspectrum
plex (20 ° C)
A
B
B'
ature
(°C)
19
Me
CO
22.2
213.6
(dd, 8.8, 40)
49.5
P(OMe)3
49.9 (3.0)
204.0
(J(P-P)
= 23.4)203.3
188.5
A B : AB'
38: I ( J ( P - P)
=
=16.1)
185.6
a 13C chemical shifts in ppm relative to TMS; 31p chemical shifts in ppm relative to 85% H3PO4; spectra
in CD2CI2/CH2CI2 solvent; J(P-C) values in parentheses, b Values for (diene)Fe(CO)3 complex given
in square brackets, c136.7(37.6) (a), 133.1(10.8) (fl), 129.4(2.7) (~), 128.0(9.4) ('I'), other PPh3 complexes
are similar, d l3c resonances for minor isomer not seen. e13C spectrum essentially invariant with
temperature. / 31p spectrum essentially invariant with temperature, g Coincident resonance, h Overlap
of C(1,4) and CH2 resonances, j Under solvent resonance at 53.8. k CN resonance not seen at + 20 ° C.
t Major AB isomer. " Minor AB' isomer.
though the position and coupling constant of the averaged resonance will also
depend on the relative populations of axial and basal conformers.
(ii) Solution structures of the phosphine complexes
Solution structures were established using variable temperature 31p and 13C
spectroscopy; the results are summarized in T a b l e 1. The sorbaldehyde complexes 7
and 8 exhibit additional low temperature restricted rotation a b o u t the C - C H O
b ond , and are discussed separately, together with the methylsorbate complex 6.
The asymmetric complexes 2, 3, 5 and 1 0 - 1 2 all exhibit two 31p resonances of
varying relative intensity in the limiting low temperature spectra which are averaged
to a single resonance at 20 ° C. Spectra of (isoprene)Fe(CO)2PPh3 (10) are shown in
Fig. 1. The two 3~p resonances of equal intensity may be assigned to axial and basal
conformers on the basis of the 13C spectrum which shows clearly axial/basal and
b a s a l / b a s a l pairs of resonances. The single basal conformer is assigned to B rather
than B' on the basis of modelling studies (vide infra) and on the basis of the low
temperature 1H spectrum ( - 98 o C, CDECI 2):
4s
4a
- 1.26
2
Is
l a
- 1.66
Me
A
B
1.75
5.28
1.55
2.23
J ( 4 a - P ) --- 11,
J ( l s - P ) = 3,
J ( l a - P ) - 8.5
1.62
0.14
4.09 0.63
- 0.05
2.11
J ( l s - P ) = 6
Of particular note are the large upfield shifts for la, 4a in A and Is,
2 in B
attributable to shielding by PPh3 and the strong P - H coupling to la, 4a, ls of A
and Is of B as a result of small P - F e - C - H dihedral angles. Conformer interconver-

100
sion results in b2/a 2 a n d bl/b 4 C O e x c h a n g e to y i e l d two a v e r a g e d ~3C r e s o n a n c e s
w h o s e c h e m i c a l s h i f t s a n d c o u p l i n g c o n s t a n t s are in a g r e e m e n t with p r e d i c t i o n .
The 13C s p e c t r u m of (trans-pentadiene)Fe(CO)EPPh 3 (2) is essentially t e m p e r a -
ture i n v a r i a n t a n d c o n s i s t e n t with only b a s a l c o n f o r m e r b e i n g p r e s e n t ( a s s i g n e d to B
Table 2
1H NMR a infrared b and analytical data
Complex 1H NMR
Infrared Analysis
(cm- i) CHN (found) CHN (calc) M.p. ( * C)
!
1,4a
1,4s
2,3
-0.11 (m)[-0.031 c
J(a-s) = 2.1, J ( a - P ) = 4.3
1.35 (m) [1.46]
J ( s - P ) = 2.1
4.83 (m) I4.891
1983
1927
67.1
5.13
67.3
4.91
146-147
J(a-2,3) = 7.9, J(s-2,3) ~ 5.3
J(P-2,3) = 1.8
PPh 3
a
67.8
5.09
67.9
5.13
99-101
la
ls
-0.01 (t)
0.97 (m)
1976
1918
4a
2
0.72 (m)
4.35 (m)
3
4.98 (m)
Me
1.42 (dd, 6.3, 1.4)
68.5
5.36
67.9
5.13
106-107
la
ls
1.15 (m)
1.62 (m)
1914
1904
4s
2
2.42 (m)
4.92 (m)
3
4.70 (m)
Me
0.82 (d)
J(4s-Me) ~ 7.2
68.7
5.82
68.4
5.48
136-138
132-133
1,4a
- 0 . 1 2 (m)
J(a-2,3) = 7.4, J ( a - P ) = 8.6
4.94 (m)
1.25 (d)
J ( a - M e ) = 6.2
1974
1914
2,3
Me
71.3
4.90
71.4
4.96
0.21 (m)
1.21 (m)
1.66 (d)
4.53 (m)
1982
1928
$
6
7
la
ls
4a
2
5.62 (m)
3
Ph
e
64.9
5.07
64.8
5.00
158-160
143-144
-
0.28 (m)
1986
1924
la
4a
2
3
Me
CO2Me
0.09 (m)
4.59 (m)
5.87 (m)
1.03 (d, 6.1)
3.19 (s)
66.2
4.96
66.4
4.89
0.16 (m) [0.75]
4.62 (m) [4.291
5.31 (m) 15.10]
1.04 (d, 6.2)
1986
1930
1,4a
2
3
Me
CHO
9.10 (d, 5.6)

101
Table 2 (continued)
Complex 1H N M R
Infrared Analysis
( c m - 1 )
CHN (found) CHN (calc) M.p. (*C)
8
la
4a
1.04 (m)
0.89 (m)
1999
1935
39.7
5.48
39.8
5.12
55-57
2
4.57 (m)
3
5.16 (m)
Me
CHO
P(OMe) 3
1.26 (dd, 6.1, 1.3)
9.20 (d, 6.1)
3.28 (d, 11.4)
9
1,4a
-0.92 (dd)
1980
1920
68.7
5.57
68.4
5.48
116-118
122-124
J ( p - l , 4 ) = 11.o,
J(a-s) =1.1
1.59 (d)
1,4s
Me
2.06 (d, 2.2)
10
la
- 0 . 3 5 (td) [0.0]
J(la-ls) = 2.1,
J(la-P) = 9.1
- o . 1 0 (m) [0.301
J(4a-4s) = 2.1,
J ( 4 a - P ) = 3.9
1.40 (m) [1.63]
J(ls-P) = 4.6
1.81 (t) 11.8Ol
J(4s-P) = 1.7
4.81 (t) [5.26]
J ( l a - 2 ) = 9.1,
J(ls-2) = 7.0
2.17 (d, 0.6)
1980
1922
67.6
5.20
67.9
5.13
4a
ls
4s
2
Me
II
la
4a
-
-
0.59 (m)
0.28 (m)
1978
1922
63.7
5.76
63.2
5.53
100 o C / 0 . 0 4 mmHg
ls
1.28 (m)
4s
1.70 (m)
2
4.95 (t)
Me
PPh2Me
2.16 (d, 1.2)
1.80 (d, 7.4)
7.2-7.7 (m)
12
la
4a
-
-
0.81 (m)
0.50 (m)
1974
1918
57.3
6.02
56.6
5.97
85" C / 0 . 0 4 mmHg
ls
1.16 (m)
4s
1.47 (m)
2
4.81 (t)
Me
PPhMeq
2.02 (d, 1.5)
1.36 (d, 8.1)
7.1-7.6 (m)
13
la
4a
-
-
0.65 (dd)
0.39 (m)
1970
1908
61.4
4.97
61.7
4.73
84-86
1.46 (dd)
1.69 (m)
ls
4s
2
4.96 (t)
Me
AsPh 3
2.04 (s)
7.01-7.61 (m)
(continued)

102
Table 2 (continued)
Infrared
Analysis
Corn-
plex
1H NMR
( e r a - 1 )
CHN
CHN
(caic)
M.p. (°C)
85-87
(found)
14
la
4a
ls
4s
-
0.71 (dd)
1976
1912
56.7
4.20
56.3
4.32
- 0 . 4 5 (m)
1.51 (dd)
1.67 (m)
2
5.12 (t)
Me
SbPh 3
2.12 (s)
7.07-7.71 (m)
2138 f
1990
1946
40 o C/0.1 mmHg
49.0
5.18
5.92
48.9
4.98
6.34
15
la
- 0.20 (dd)
J ( l a - 2 ) = 8.4,
J(la-ls) = 2.2
0.07 (dd)
4a
J(4a-4s) = 2.0,
J(4a-2) = 0.9
1.42 (dd)
ls
J(ls-2) = 6.7
1.56 (t)
4s
J(4s-2) = 1.8
4.98 (t)
2
Me
1.96 (s)
1.90 (s)
MeNC
51.1
5.73
6.06
51.1
5.54
5.96
40 ° C/0.1 mmHg
-0. 17 (dd)
0.11 (m)
1.43 (dd)
1.57 (t)
2128 /
1992
1944
16
la
4a
ls
4s
5.03 (t)
2
1.96 (s)
0.43 (t)
Me
EtNC
2.35 (q, J = 7.4)
54.8
6.66
5.16
54.7
6.47
5.33
40 ° C/0.1 mmHg
2103 !
1992
1944
17
la
4a
Is
- 0.19 (dd)
0.11 (t)
1.42 (dd)
1.56 (t)
4.99 (t)
1.98 (s)
0.77 (s)
4s
2
Me
ButNC
55 ° C/0.1 mmHg
1990
1932
39.2
5.40
39.5
5.59
-
0.14 (td)
18
la
J(la-ls) = 2.3,
J(la-P) = 9.3
0.19 (m)
4a
J(4a-4s) = 1.7,
J ( 4 a - P ) = 5.8
1.59 (m)
J(ls-P) = 4.1
1.74 (m)
Is
4s
2
J(4s- P) = 1.6
5.00 (t)
J ( l a - 2 ) = 9.2,
J(ls-2) = 6.7
2.06 (d, 1.9)
3.36 (d, 11.5)
Me
P(OMe)3

103
Table 2 (continued)
Complex IH NMR
Infrared Analysis
(cm- i) CHN (found) CHN (calc) M.p. ( * C)
19
la
4a
ls, 4s
2
- 0.48 (m)
- 0.06 (m)
1.73 (m)
1910
36.2
6.60
36.0
6.50
40 * C/0.01 mmHg
5.00 (m)
Me
P(OMe)3
2.22 (d, 2.8)
3.45 (dd, 11.2, 7.7)
20 ~
1,4
2,3
2.62 (m) [2.69]
4.80 (dd) [4.57]
J(P-2,3) = 3.9,
1975
1922
69.1
4.81
68.7
5.07
118-120
•/1-2 = 3.0
1.40 (d)
1.86 (d)
C H 2 ( e x o )
CH 2( endo )
J( exo - endo ) = 10.2
21
22
23
1,4
2.57 (m)
4.62 (m)
1.6-2.2 (m)
1,2-1.4 (m)
1972
1918
68.7
5.25
69.2
5.34
132-134
56-57
2,3
C H 2 ( a )
CH2(#)
1,4
2,3
CH2
MeNC
2.85 (m)
4.97 (m)
1.3-1.8 (m)
1.90 (s)
2136 /
1986
1940
51.9
4.91
5,99
51.5
4.72
6.00
1,4
2,3
C H 2 ( a )
CH2(/3)
MeNC
2.70 (m)
4.95 (m)
1.6-2,1 (m)
1.2-1.4 (m)
1.91 is)
2136 /
1986
1940
53.4
5.17
5.60
53.4
5.26
5.67
45-46
24
25
1,4
2,3
CH2
EtNC
2.84 (m)
4.98 (m)
1.3-1.8 (m)
0.43 (t)
2.37 (q, s = 7.4)
2126 /
1984
1938
53.1
5.08
5.18
53.4
5.26
5.67
50 o C/0,1 mmHg
50° C/0.1 mmHg
1,4
2,3
C H 2 ( a )
Cl-12(fl)
EtNC
2.71 (m)
4.98 (m)
1.6-2.0 (m)
1.2-1.4 (m)
0.46 (t)
2124 I
1982
1938
55.7
5.67
5.45
55.2
5.75
5.36
2.40 (q, J = 7.4)
26
1,4
2,3
CH2
ButNC
2.84 (m)
4.96 (m)
1.4-1.8 (m)
0.77 (s)
2122 /
1986
1938
56.8
6.15
5.01
56.7
6.18
5.09
49-50
88-89
27
1,4
2,3
C H 2 ( a )
CH2(~)
But NC
2.68 (m)
4.94 (m)
1.6-2.0 (m)
1.2-1.4 (m)
0.79 is)
2144 !
1982
1926
58.1
6.18
4.90
58.1
6.57
4.84
(continued)

104
Table 2 (continued)
Infrared
(cm-1)
Analysis
Complex
1H NMR
CHN (found)
CHN (calc)
M.p. (°C)
28
1,4
2,3
2.90 (m)
4.86 (m)
1991
1934
43.9
5.83
43.7
5.76
60°C/0.1 mmHg
CH2(a)
CH2(fl)
P(OMc)3
1.6-2.0 (m)
1.2-1.4 (m)
3.20 (d, 11.5)
29
1,4
2,3
2.86 (m)
4.82 (m)
1994
1942
63.2
4.91
62.8
4.84
87-88
CH2(a)
CH2(fl)
P(OPh)3
1.6-2.0 (m)
1.2-1.4 (m)
7.1-7.5 (m)
a C6D6 solution; Ha = inner terminal diene proton; Hs = outer terminal diene proton; values for
tricarbonyi in square brackets. ~' Hexane solution, c from reference 8g. d 6.9-7.7 (m); other complexes
are similar, d Under PPh3. 1 CN vibration, g 270 MHz; other spectra at 100 MHz.
on the basis of modelling studies), though 31p spectra show the presence of a second
conformer which may be A or B', but is arbitrarily assigned to A in Table 1. The 13C
spectra of the cis-pentadiene complex 3 (Fig. 1) show clearly two axial/basal pairs,
consistent with population of only B and B'; the spectra do not distinguish between
them, though B is arbitrarily assigned as the major conformer in Table 1. Con-
former interconversion results in a2/b3 and al/b4 exchange to give the two
expected averaged resonances.
~3C spectra of the symmetric complexes 4 and 9 show a single, temperature
invariant resonance consistent with exclusive population of the axial conformer,
though the 3Zp spectrum of the hexadiene complex shows that the B/B' enantiomer
pair has a small population. The spectra of (cycloheptadiene)Fe(CO)2PPh3 (21)
(Fig. 1) are as expected for a symmetrical complex populated exclusively by the
basal conformer, showing a single, temperature invariant 31p resonance and in the
~/P~hL3=
b~
PPh3
a2
b2
PPh3/
b4
(10A)
(2A)
(3B)
a2
a2
at
Ph3P
b
PhaP~//~b4
b ? / ' ~ P P h s
(10B)
(2B)
(3B')

105
31p
13 C
+20°C
1
- 1 S°C
( a )
"
,
~2
,m ' 2 / b '
- 100°(2 / .
,
b41t/
-90°C
80
70
220
210
O°C
(b)
- 8 5 ° C
2
I-t
. 8 5 ° C
B'
__J
al•a
74
69 ppm
220
215
'3
~
70.2
O°C
+20°C
(c)
~ axial
" 8 0 ° C
" 8 0 ° C
[l-
~ua~al
'
15 ppm
2
220
Fig. 1. High and low temperature 13C and 31p spectra of (a) (isoprene)Fe(CO)zPph3, (b) (cis-
pentadiene)Fe(CO)2PPh 3, and (c) (cycloheptadiene)Fe(CO)2pph3.

106
13C spectrum, one axial/basal pair at low temperature which is averaged to a single
resonance at high temperature. The cyclohexadiene complex 20 exhibits a tempera-
ture invariant ~3C spectrum down to - 9 0 °C due to a decreased rotational barrier,
but the chemical shift and coupling constant of the averaged resonance are con-
sistent with population of the basal conformer only, the conformer which is also
observed in the solid state [14]. The changes in the ~3C spectrum of the butadiene
complex 1 resemble those of the cycloheptadiene complex 21, though the 31p spectra
show a significant population of the B/B' pair. Changes consistent with these
exchange equilibria are also observed in the diene/alkyl region of all of the 13C
spectra.
The effects of alkyl substitution may be summarized as follows, taking the
unsubstituted (butadiene)Fe(CO)2PPh 3 as reference:
(a) cis-terminal substitution (as in 3, 20 and 21) depopulates completely the axial
conformer.
(b) terminal or internal disubstitution (as in 4 and 9) depopulates the basal
conformer.
(c) trans-terminal monosubstitution depopulates the axial conformer while inter-
nal monosubstitution depopulates the basal isomer.
(d) in the isoprene series, methyl substitution of the phosphine increases the
population of the axial conformer in the order PPh3 < PPh2Me < PPhMe2. In all
cases, the basal conformer is assigned to B rather than B' structure.
(iii) Single crystal and modelling studies of phosphine complexes
In order to confirm the solution NMR studies, and to provide a basis for the
modelling studies, the structures of the hexadiene and 2,3-dimethylbutadiene com-
plexes 4 and 9 have been determined by single crystal X-ray diffraction. The results
(Table 3, Fig. 2) confirm the axial orientation of the PPh3 ligand. Relative to the
structure of (butadiene)Fe(CO)3 [16], the main structural change is manifested in an
increase in the axial ligand-Fe-diene angle ( P / C a - F e - Z ) coupled with a decrease
in C O - F e - C O basal angle. As with (cyclohexadiene)Fe(CO)2PPh3 [14], the steric
effect of the phosphine appears to affect more its attitude relative to the diene
rather than the CO ligands. In the hexadiene complex 4, the two terminal methyls
are twisted out of the diene plane towards the iron by an average of 3.5 °, though
this has been set to zero in the modelling studies. The Fe-P bond lengths and
internal bond lengths and angles in the PPh3 ligand are common to other
(diene)Fe(CO)2PPh 3 structures, all of which contain PPh3 in a basal position
[4c,14,16a-d].
Though full molecular mechanics calculations on transition metal complexes
remain rare due to a lack of metal parameterization, recent results [17a-d] show
that calculation of simple interligand Van der Waals interactions can provide an
accurate estimate of relative conformer stabilities. Using CHEM-X [18], we have
applied this approach to evaluate steric contributions to the relative stabilities of
axial/basal conformers of the (diene)Fe(CO)2L series.
Though some of the gross features of the rotational profiles in Fig. 3 can be
generated by rotation of the diene relative to a rigid Fe(CO)2PPh3 fragment using
structures generated from the crystallographic coordinates, the more subtle features
emerge only if

108
(9)
(4)
(a) hydrogens omitted
13
2
4
~
2
2
13
1
2 s U , . . _
. W . , ~ ,
P!
IS
CI ~" 0 2
Feet
J,;'
C7
Fe'l
C01~=~02
C';( )'~C4
esk..)
(b) space filling model, hydrogens and Fe omitted
Fig. 2. Molecular structures of 4 and 9.
(a) the molecular geometry is idealized [19] through equalization of symmetry
related bond lengths and angles with the diene, within the PPh3 ligand and within
the square pyramidal polyhedron in symmetric complexes such as 9. To isolate the
effect of changes in the diene or metal substituent, all acyclic complexes were
generated by modification of the idealized structure of 9, thus retaining a constant
PPh 3 and constant bond lengths and angles within the diene and the square
pyramid. For the idealized cyclohexa- and cyclopentadiene complexes 20 and 30,
bond lengths and angles for the (diene)Fe moiety were taken from the basal
literature structures [14,16c] but for consistency, superposed on an axial
Fe(CO)2PPh3 fragment for the purposes of modelling.
(b) energy minimization about conformationally mobile F e - P , P-C and C-C
bonds is allowed. In particular, for symmetric complexes such as 4, 9 or 20, P-C
minimization is necessary to generate isoenergetic profiles for the enantiomerically
related clockwise/anticlockwise rotation of the diene relative to Fe(CO)2L. The
implication is that, as in the case of PPh3 complexes chiral at iron [20], the screw
sense of the PPh3 propeller is linked to the molecular chirality. For asymmetric,

109
--- 2 0
- - 3 0
:/".,
2 0
- - -
1
40 !
-
-
(
b
u
t
a
d
i
e
n
e
)
Fe(CO)3 ~
,
30:
~
//
i,i
w
10.
0
2oo
100
20o
o
IO0
Rotation
Rotation
- - - 10
---9
4 0 ¸
30
p 20
w
p- ,~
/ s ~
1 0
"
0
100
2 0 0
3 0 0
460
tO0
2 0 0
0
R o t a t i o n
R o t a t i o n
F i g . 3. Rotational energy profiles.
chiral complexes such as 3 and 10, both enantiomers were generated and provided
isoenergetic rotational profiles.
It may also be noted that axial/basal exchange requires not only turnstile
rotation but also small changes in bond angle. As shown for (butadiene)Fe(CO)3 in
Fig. 3, neglect of bond angle change during rotation results in slightly unequal
energy maxima at 60 and 180 °, and a small overestimate of the energy of the
equivalent ground state structure generated on 120 ° rotation. These inequalities
disappear if all Z - F e - C O and C O - F e - C O angles are equalized. In Fig. 3, relative
energies are plotted as a function of rotation, where 0 ° represents the axial isomer
and rotation of the Fe(CO)2L fragment relative to the diene is clockwise. For
symmetric complexes, only half of the symmetry related 360 ° rotation is shown.
Since our interest is primarily in relative ground state energies, we have neglected
bond angle changes during the turnstile rotation, though because of this small
uncertainty (shown also by other tricarbonyls), we have deemed isoenergetic any
conformer pair which differs in energy by less than two kcal.
Rotational profiles for the symmetric 2,3-dimethylbutadiene and hexadiene com-
plexes 9 and 4 show clearly an enhanced stability of the axial PPh3 conformer by
about 5 kcal, in agreement with experiment. Profiles for the cyclic complexes 20 and
30 show in contrast an enhanced stability of the basal conformer by approximately
15 and 7 kcal respectively, again in agreement with experiment. Though we have yet
to model the cis-pentadiene and cycloheptadiene complexes 3 and 21 [21"], a
similar destabilization of the axial position is anticipated. It is interesting to note
that in contrast to 9, the o-xylylene complex 31 adopts the basal conformation in
the solid state [16b]. Modelling of 31 reveals that the planarity of the o-xylylene unit
is now sufficient to render the axial and basal positions isoenergetic in terms of
steric interactions.

110
~ P P i 1 3
(30)
~
PPh3
(31)
More subtle is the variation in conformer population between 1, 2 and 10.
Modelling of the butadiene complex 1 shows no energy difference between axial and
basal PPh~ site occupancy; modelling of both the trans-pentadiene complex 2 and
the isoprene complex 10 shows a destabilization of approximately 5 kcal for the
basal site cis to the methyl, but no energy difference between the axial and the
remaining unhindered basal site tran_s to methyl. Observance of only two populated
conformers for 2 and 10 is thus consistent with these predictions; variation of the
axial/basal ratio, which shifts towards basal on terminal substitution but towards
axial on internal substitution, may be the result of electronic fine tuning. Both 13C
NMR [23a-c] and photoelectron spectra [24] show that the electronic effects of
terminal versus internal methyl substitution are different. Most relevant are infrared
force constant studies [9a], which show that though all methyl substituted dienes are
net electron donors compared to butadiene, the increased back donation to axial
and basal sites is not equal, and depends on the position of diene substitution. For
(isoprene)Fe(CO)3, the greatest relative change occurs in the basal force constant,
implying
a relatively greater back donation to the basal sites compared to
(butadiene)Fe(CO) 3. Thus, in the more electron rich (isoprene)Fe(CO)2 PPh 3, a shift
towards the axial PPh 3 conformer increases the basal site occupancy by the stronger
~r-accepting CO ligands. Conversely, terminal substitution (cis or trans) results in a
greater and more pronounced relative change in the axial force constant. A shift
towards the basal PPh3 conformer thus maximizes axial site occupancy by CO. This
appears also to be reflected in the ordering of the 31p chemical shifts, which is
Pbasal
> Paxial for internal substitution, but Paxial > Pbas,I for terminal substitution
and for (butadiene)Fe(CO)2PPh 3.
The observed shift towards axial conformer in the (isoprene)Fe(CO)2PPh 3_ xMe~
series (x = 0-2) 10-12 as phosphine basicity increases [25] is also consistent with
maximization of basal ~r-acceptance by CO. Though the effect may be partially
steric, modelling studies of complexes 10-12 show no difference in energy between
the A and B conformers. Though B' increases in relative stability, reflecting the
decreasing cone angle [PPh3(145) > PPh2Me(136) > PPhMe2(122 °)], even for the
PPhMe2 complex, B' remains approximately 3 kcal higher in energy than the A/B
pair.
The phenylbutadiene complex 5 provides a link between ~4-diene complexes and
~4-enone complexes of structure 32 which we have reported in a previous study
[lla].
L
Ph---"~
OR
(A)
(B)
(32a) R= H, L= PPh3
13)R= H, L= PPh2Me
c) R= H, L= PPhMe2
d) R= Me, L= PPh3

111
Though both 5 and 32a show the presence of only conformers A and B in
solution, there is a significant shift towards the axial conformer (for 5, A : B = 1 : 13,
for 32a, A : B = 0.6 : 1). Modelling studies for both 5 and 32a predict an isoenergetic
A/B pair, with the B' conformer being > 10 kcal higher in energy for both
complexes. The shift in conformer population thus seems a consequence of the
stronger ~r-accepting character of the enone ligand. As in the isoprene series, there is
an increase in axial conformer population for 3 2 a - e in the order PPh 3 < PPh2Me <
P P h M e2. Introduction of a methyl group in 32d also results in increased axial
population (A:B = 59:1); and modelling studies of 32d reveal an increase in the
energy of B relative to A by approximately 4 kcal, with B' remaining > 10 kcal
higher in energy.
We have not investigated in detail the dependence of total intramolecular energy
on the orientation of the phenyl groups of the PPh3 ligand relative to the diene
along the Fe-P axis. One feature which seems to emerge is that in the higher energy
areas of the energy profiles in Fig. 3, this orientation has a large influence on
determining the energy minima, whereas at energies close to relative zero, the energy
profiles with repect to Fe-P bond rotation are much flatter. The axial conformer of
(cyclohexadiene)Fe(CO)2PPh 3 (20) is an example of the former, with well defined
minima at structures 33 and 34 (the latter slightly lower in energy) in which an
approximate mirror plane bisects the C H 2 - C H2 bond. The energy maximum occurs
at 35 in which CH 2 is directly eclipsed by a phenyl ring.
(33)
(35,)
OP
@o
viewed down C4-C5;
Fe omitted
(35)

112
The results raise the possibility of a concerted, cogwheel rotation in some of these
complexes, a possibility we are investigating with sterically more demanding phos-
phines.
(iv) Solution structure and modelling of phosphite, isonitrile, arsine and stibine com-
plexes
For the (isoprene)Fe(CO)2L series 15-18 containing sterically less demanding
trimethylphosphite (cone angle 107") and isonitrile ligands, modelling studies
indicate no energy difference between the A / B / B ' conformer trio, and solution
NMR studies show that three conformers are populated in all cases. The phosphite
complex 18 shows three 31p resonances at low temperature (Fig. 4), the two major
conformers being axial and basal. Though the spectra do not differentiate between
B and B', the major conformer is assigned the B structure on the basis of results on
the PPh3 complex. The bis(phosphite) complex 19 shows two unequal sets of
doublets, the major set being assilg~3able to the axial/basal conformer AB on the
basis of couplings observed in the i C spectrum. The minor resonances are assigned
to the alternative axial/basal conformer AB'. The basal/basal conformer is not
populated; the reason may be steric, with the larger axial-basal angle minimizing
interligand repulsions. The ratio of AB : AB' is similar to the B : B' ratio observed
for 18. Turnstile rotation in 19 does not completely scramble the phosphite ligands,
and two broadened 31p resonances are observed at 20°C, still below the high
temperature limiting spectrum.
P(OMe)3
P(OMe)3
Z-'--XP(OMe~3
(19AB)
(19AB')
The cycloheptadiene complexes (28) and (29) show temperature invariant 3]p
spectra which, in combination with axial/basal pair of resonances observed in the
limiting low temperature ]3C spectra, indicate no detectable population of the axial
phosphite conformer. Similar 31p results have been reported for analogous iron
complexes containing the cage phosphite P(OCH2)3CEt, though the ruthenium
analogues do show some population of the axial conformer [8f].
]3C spectra of the isonitrile complexes, of which those of the CNEt derivatives 16
and 25 are typical (Fig. 4) show population of basal and axial conformers for both
acyclic and cyclic complexes. All the cyclohexadiene and cycloheptadiene complexes
exhibit a single CO resonance at 20 °C which is resolved at low temperature into an
axial/basal pair due to the major basal CNR conformer and a single basal
resonance due to the two symmetry related CO ligands of the axial CNR conformer.
Two resonances which reflect the same conformer distribution are seen in the CN
region at low temperature, though at room temperature the averaged CN resonance
is broadened considerably by quadrupolar relaxation [26]. Our results on (cyclohe-
xadiene)Fe(CO)ECNEt differ from those previously published [10c] where no axial
CNEt conformer was detected, though axial/basal conformer mixtures have also
been observed for (~4-cyclooctatetraene)Fe(CO)2CNR derivatives [10d]. Spectra of
the isoprene complexes are similar, though four rather than two basal CO reso-

113
nances and three rather than two CN resonances are observed due to the inequiv-
alence of the B/B' conformer pair.
Compared to (isoprene)Fe(CO)2PPh3 (10), there is a shift of conformer popula-
tion towards occupancy by phosphite or isonitrile of the better ~r-accepting basal
site in the square pyramid [27], consistent with the accepted decreasing o-donor/~r-
acceptor ratio of the ligand in the order PPh3 > P(OMe)3 > CO [25" ]. The position
of isonitrile is more problematical; photoelectron results on Fe(CO)aCNR show
that the isonitrile functions as a net electron acceptor, with ~r-acceptor and particu-
larly o-donor capacity reduced relative to CO; ~r-donor interactions may also play a
role in the destabilization of the metal-centred orbitals [28]. This study also shows
little variation in the electronic character of CNR as a function of R. The lower
population of the axial isomer in the cycloheptadiene complexes 23, 25 and 27,
compared to their cyclohexadiene analogues 22, 24 and 26 and the increasing
asymmetry in the B/B' population of the isoprene complexes 15-17 may have a
steric origin. Observations whose explanations remain unclear are the increased
population of basal conformer in the isoprene series in the order CNMe < CNEt <
CNBu t and the increased population of what is expected to be the sterically more
hindered axial isomer in the order CNMe < CNEt < CNBut for both the cyclohe-
xadiene and cycloheptadiene complexes.
The origin of the shift in conformer population towards axial in the isoprene
complexes 10, 13 and 14 in the order PPh3 < AsPh3 < SbPh3 is also unclear.
(v) Rotational barriers in (diene)Fe(CO): L complexes
Experimental rotational barriers for tricarbonyl complexes are in the range 9-13
kcal mo1-1, with a calculated electronic barrier for (butadiene)Fe(CO)3 of 14.5 kcal
mol -] [13a]. We make no claim for the accuracy of the steric contribution to
rotational barriers represented in Fig. 3, since they take no account of torsional or
angular changes in the diene or polyhedron or movement of the iron relative to the
diene at the transition state. Nevertheless, several features may be noted:
(a) in all cases, steric factors favour the staggered structure 36s rather than the
eclipsed 36e as the ground state, thus reinforcing the electronic preference for the
staggered geometry [29].
(36s)
(36e)
(b) in all cases except the pentadiene and hexadiene complexes 3 and 4, energies
of the eclipsed transition state for the Fe(CO)2PPh3 and Fe(CO)3 complexes are
close (as shown in Fig. 3 for (butadiene)Fe(CO)2L), thus indicating that the large
PPh 3 should not sterically influence rotational barriers. The pentadiene and
hexadiene complexes show substantially larger barriers for passage of PPh3 past the
trans-ternfinal methyl group, reaching a maximum of ca. 80 kcal in each case at 70 °
rotation. There is no evidence for such a massively increased barrier for the
axial/basal exchange in the variable temperature 31p spectra of the hexadiene
complex 4. The possible reduction in energy of the 70 ° transition state was
modelled via elongation along the Fe-Z axis or twisting about C 5 - C 6 to move the
terminal methyl further from the PPh 3. In both cases, the energy fails dramatically.

115
Elongation along the Fe-Z axis such as to increase the Fe-C distances by ca. 0.3 ,~
or twisting of methyl away from the iron by ca. 35 ° both reduce the transition state
to an energy comparable with other Fe(CO)2PPh3 and Fe(CO)3 complexes. The
energy cost of such distortions in terms of orbital overlap is unknown, though it
may be noted that twisting angles larger than 35 ° are found in the ground state
structures of C6 and C7 cyclic (diene)Fe(CO) 3 complexes. Energy profiles for the
smaller isonitrile ligands are essentially superposable on those of the tricarbonyls, as
shown for (cyclohexadiene)Fe(CO)2CNMe in Fig. 3.
(c) Through line shape analysis of variable temperature 31p and 13C spectra
[30* ], we have determined rotational barriers for (isoprene)Fe(CO)2L of A H * =
6.8 -+ 0.1 kcal mo1-1, AS * =
- 13.9 4- 0.5 cal K -1 mo1-1 for L = PPh3 and A H *
=9.8_+0.5 kcal tool -1, AS * = --2_+2 cal K -1 tool -1 for L = C O . In common
with other PPh3 and PF3 complexes [9a,b,lla], phosphine substitution thus appears
to decrease the rotational barrier. The origins of this remain obscure, since calcula-
tions [13a] predict an increase in rotational barrier with increasing o-donor capacity
due to tilting changes in the 2es and le~ orbitals of the Fe(CO)3 fragment which
reduce the asymmetry in the important 2es-~r3 (butadiene) and les-~r3 (butadiene)
orbital overlaps in the staggered and eclipsed configurations. Within the precision of
such measurements, rotational barriers for Fe(CO)ECNR [10a] and Fe(CO)3 [Sa]
complexes are identical.
(vi) Restricted C - C H O rotation in (sorbaldehyde)Fe(CO)2L (L= CO, PPh 3,
P(OMe)z)
Introduction of an electron withdrawing substituent to a terminal diene carbon is
known to substantially increase the barrier to rotation. In complexes such as 6-8
and their Fe(CO)a analogues the conjugative effect is maximized through coplanar-
ity of the CHO or CO2Me substituent, and for (sorbaldehyde)Fe(CO)3, indirect
evidence from protonation studies indicates an equally populated mixture of the
cis/trans isomers 37 and 38 [31]. We here present direct NMR evidence for this
rotational process.
Diene rotation in (sorbaldehyde)Fe(CO)a is slow on the NMR time scale at
- 6 0 ° C , as evidenced by the well resolved axial/basal CO resonances (Fig. 5).
Below - 7 0 o C, all 13C resonances (except the methyl) show additional broadening.
Though a limiting low temperature spectrum cannot be obtained, that at - 1 2 5 ° C
shows a clear splitting into unequally populated resonances of one of the basal CO
resonances and the C-3 inner diene resonance. From modelling studies, the cis/trans
forms of (sorbaldehyde)Fe(CO)3 appear equal in energy; the experimentally ob-
served asymmetry in the cis/trans distribution evident in the low temperature
NMR spectrum is at most in the range 2-3 : 1.
R
O
°
Fe(CO)2L
Fe(CO)2L
(37) cis
(38~ trans
L = CO, PPh3, P(OMe)3
R = H , OMe

116
C-3
M- CO
-65°C
-125
.
.
.
.
.
.
i
•
l
.
.
.
.
"
"
.
.
.
.
.
.
.
010 ppm
2
t
0
ppm
Fig. 5. 13C spectra of sorbaldehyde Fe(CO)3 (Et20-dl°).
alp spectra of the sorbaldehyde complexes 7 and 8 both show features at low
temperature (Fig. 6) which seem best explained by slowing of C - C H O bond
rotation.
The 31p spectrum of the PPh3 complex 7 exhibits a singlet at 20°C which is
resolved into two resonances at - 6 0 ° C ; on the basis of the 1ac NMR at this
temperature which shows clearly a basal pair of resonances [32*], the major
conformer contains axial PPh a. Identification of the minor basal conformer as B or
B' is not possible; though modelling studies show clearly a greater stability of the
axial conformer by ca. 8 kcal, they do not distinguish between the two basal
conformers. Below - 6 0 °C, the resonance due to the axial conformer broadens
again; the spectrum at - 105 ° C, though not quite low temperature limiting, exhibits
two resonances of slightly unequal population. Modelling studies of the axial PPh 3
conformer indicate that the cis conformer 37 is more stable than 38 by ca. 3 kcal.
The resonance due to the minor basal conformer remains sharp down to -105 ° C,
indicating that only 37 or 38 is populated. Modelling studies on both possible basal
conformers indicate that 37 is more stable by > 10 kcal.
Both the 13C and 3~p spectra of the methylsorbate complex 6 are essentially
temperature invariant, indicating population of only the axial PPh3 conformer and,
most probably, only the trans conformer 38 based on the solid-state structure of
(methylsorbate)Fe(CO) 2PPh2(neomenthyl) [6a].
13C and 31p spectra of the P(OMe)3 complex 8 (Fig. 5) resemble those of the PPh3
complex 7 except that P(OMe)3 exhibits a greater preference for basal occupation,

117
(a)
(b)
+20°C
+20°C
CD2CI2
CD2CI2/EtzO
1
A[ xlO
-50°C
xlO
~A
-60°C
8
xlO
-80°C~
"I0°C e~
-8°°c
•/•
,ol/A
-100°C
/
x lo
-115°CA
J_
__A
,h
' - pp~
'70
60
ppm
Fig. 6. Variable temperature 3tp spectra of (a) (sorbaldehyde)Fe(CO)2PPh3 and (b) (sorbaldehyde)-
Fe(CO)2P(OMe)3.
and both basal conformers are populated, though unequally. On cooling to - 115 ° C,
resonances due to the axial and minor basal isomers exhibit changes attributable to
slowing of C - C H O rotation, though due to overlap, only one of the resonances due
to the minor basal isomer is apparent. Though a distinction between the two basal
isomers is not possible on the basis of NMR or modelling data, the resonance due to
the major basal isomer remains sharp down to -115 ° C, indicating, as in the PPh 3
case, probable population of only the cis conformer 37.
The ability of auxiliary ligands to control the cis/trans ratio may provide an
important influence on the diastereoselectivity of reactions at the aldehydic carbon,
which are important in the synthetic utility of these complexes [2].

118
Experimental
All reactions were performed under nitrogen using distilled and degassed solvents.
NMR spectra were recorded on JEOL FX-100 or JEOL GSX-270 spectrometers;
infrared spectra were recorded on a Pye Unicam SP2000 spectrometer, trans-1-
Phenyl-l,3-butadiene [34], and the isonitrile ligands [35] were prepared by literature
methods. With the exception of (butadiene)Fe(CO)3 which was purchased, all
(diene)Fe(CO)3 complexes were prepared by ultrasonic reaction of the diene with
Fe2(CO)9 [36].
(a) Preparation of (isoprene)Fe(CO)2PPh3 (10)
Me3NO-2H20 (7.7 g, 68.8 mmol) was added to a stirred solution of
(isoprene)Fe(CO)3 (8 g, 38.4 mmol) and PPh3 (16 g, 61.1 mmol) in acetone (150 ml).
The mixture was vigorously stirred at reflux until infrared sampling indicated
disappearance of starting material. Diethyl ether (200 ml) was added to the cooled
mixture which was filtered and evaporated. The residue was dissolved in diethyl
ether (75 ml) and stirred with excess MeI for one hour to remove unreacted PPh3.
After filtration and removal of solvent the residue was extracted with 10% ethyl
acetate/petroleum ether (30-40) and chromatographed on alumina using 5% ethyl
acetate/petroleum ether (30-40). Evaporation of solvent from the yellow band
collected gave the product 10, as a yellow solid (10.6 g, 55%). An analytical sample
was obtained by crystallization from petroleum ether (30-40).
Other PPh3 complexes and the PPh2Me, PPhMe2, AsPh3, SbPh3 and isonitrile
complexes were prepared in the same way; analytical samples were obtained either
by crystallization from petroleum ether (30-40) or in the case of oils, microdistilla-
tion at the temperatures and pressures shown in Table 2.
(b) Preparation of (isoprene)Fe(CO)x [P(OMe)3] 3- ~ (x = 1, 2)
(i) (isoprene)Fe(CO)3 (3 g, 14.4 mmol) and P(OMe)3 (2 g, 16.1 mmol) were
stirred in toluene (600 ml) and irradiated using a 90W medium pressure Hg lamp
until infrared sampling indicated almost complete disappearance of starting material.
After removal of solvent, the crude product was purified by preparative tic on silica
using 3% ethyl acetate/petroleum ether (30-40). The product was collected as the
main yellow band which separated from traces of the faster moving tricarbonyl and
the slower moving disubstituted complex. Extraction and microdistillation provided
18 (1.5 g, 35%).
(ii) (isoprene)Fe(CO)3 (2 g, 9.6 mmol) and P(OMe)3 (4 g, 32.2 mmol) were
stirred and irradiated in toluene (600 ml) with further additions of phosphite until
infrared sampling indicated complete disappearance of the monosubstituted com-
plex 18. After removal of solvent, the residue was purified by preparative tlc on
silica using 3% ethyl acetate/petroleum ether (30-40). Collection of the faster
moving yellow band, followed by extraction and rnicrodistillation yielded 19 (650
mg, 17%). The slower moving yellow band was identified by NMR as the fully
substituted (isoprene)Fe[P(OMe)3] 3 complex [37a,b].
Complexes 28 and 29 were prepared by reaction of (cycloheptadiene)Fe(CO)3
with phosphite in di-n-butylether [5a].
(c) Crystallographic data for complexes 4 and 9
Data were collected on a Hilger Watts Y290 diffractometer using Mo-K~ radia-

119
Table
4
Details of data collection for compounds 4 and
4
9
9
C26H2502PF e
triclinic
"
C26H2502PFe
monoclinic
space group
Z
P1
2
P 2 1 / c
4
(two molecules per
lattice point).
a
(A)
(,k)
10.492(3)
10.263(2)
10.297(3)
28.782(6)
b
c (,g,)
( o )
11.623(2)
77.21(2)
15.708(5)
90
a
fl ( ° )
y ( o )
84.00(2)
67.68(2)
98.99(2)
90
U
(~3)
1128.78
18.55
476
4598.1
18.22
1908
/~ (cm -1)
/7(000)
range
reflections I > 3 o ( 1 )
2 < 0 < 28
3115
2 < 0 < 24
3124
variable parameters
151
272
maximum shift/esd
R
Rw
< 0.001
5.55%
6.67
< 0.001
6,88
7,17
maximum excursion
minimum excursion
0.29 e / h ,3
- 0.21
0.18 e / A3
- 0.19
tion (0.71069 A). Structures were solved by a combination of Patterson search and
direct methods (SHELX86) [38] and refined by full matrix least squares (SHELX76) [39].
Data were corrected for Lorentz and polarization effects, but not for absorption.
Hydrogen atoms were included in calculated positions. For 4, the iron, phosphorus
and carbonyl groups were refined anisotropically; for 9, only iron and phosphorus
were refined anisotropically. Atomic scattering factors for hydrogen and non-hydro-
gen atoms and the anomalous dispersion correction factors for non-hydrogen atoms
were taken from the literature [40-42]. Calculations were performed on a VAX
1 1 / 7 8 5 computer. Tables of calculated and observed structure factors and aniso-
tropic and isotropic thermal parameters are available from the authors as supple-
mentary material. Atomic coordinates are listed in Table 5.
For complex 9, there are two independent molecules per lattice point, but bond
length differences between the two are close to experimental error and are not
considered chemically significant.
(d) Molecular modelling studies
The idealized structure of (2,3-dimethylbutadiene)Fe(CO)2 PPh3 (9) (Table 3) was
generated by modification of the experimentally determined structure, and all
acyclic (diene)Fe(CO)2L and (T/a-enone)Fe(CO)2 L complexes were generated from
this by changing the substitution of the diene or alteration of L. Idealized diene
geometries for 20 and 30 were generated from literature structures of 20 [14] and
(5-exo-benzyl-l,3-cyclopentadiene)Fe(CO)2PPh 3 [16c], but the Fe(CO)2PPh3 moiety
was altered such that PPh3 occupied an axial position. The Fe(CO)3 complexes were
generated from Fe(CO)EPPh3 by replacement of PPh3 with a linear CO having

120
Table 5
Fractional atomic coordinates for complexes 4 and 9
Atom
x
y
z
Complex 4
Fel
P1
O1
0 2
C1
C2
C7
C3
C4
C5
C6
C8
0.06858(5)
0.20248(10)
0.1627(4)
0.1863(4)
0.14392(5)
0.13168(10)
0.0716(4)
0.15149(5)
0.29412(8)
0.2656(4)
0.1028(4)
0.2254(4)
0.1227(4)
-
- 0.1447(4)
0.0998(5)
- 0.0307(5)
0.1254(6)
0.2223(5)
0.2420(5)
0.3188(5)
0.3711(4)
0.4418(6)
- 0.0487(4)
- 0.1376(5)
- 0.2758(5)
- 0.3275(6)
- 0.2426(6)
- 0.1027(5)
0.1944(4)
0.1971(5)
0.2444(5)
0.2880(5)
0.2831(6)
0.2364(5)
0.2388(4)
0.1800(5)
0.2673(6)
0.4121(5)
0.4719(6)
0.3854(5)
- 0.0690(5)
0.1420(5)
0.3090(5)
0.1860(4)
0.0524(5)
- 0.0619(5)
- 0.0378(4)
- 0.1595(5)
0.3267(4)
0.4025(4)
0.4945(5)
0.5130(5)
0.4385(5)
0.3460(5)
0.1217(4)
- 0.0088(5)
- 0.0694(5)
- 0.0012(5)
0.1310(6)
0.1921(5)
0.3120(4)
0.4454(5)
0.5189(6)
0.4634(5)
0.3318(5)
0.2559(5)
- 0.0511(5)
0.0099(4)
- 0.0229(4)
0.0409(4)
0.1365(4)
0.2126(5)
0.3541(3)
0.2751(4)
0.3159(5)
0.4362(5)
0.5148(5)
0.4737(4)
0.4299(3)
0.4650(4)
0.5693(5)
0.6368(4)
0.6044(5)
0.5002(4)
0.2467(3)
0.2025(4)
0.1547(5)
0.1534(4)
0.1975(4)
0.2437(4)
C9
C10
C11
C12
C13
C14
C 15
C16
C17
C18
C19
C20
C21
C22
C23
C24
C25
C26
Complex 9
Fel
Fe2
P1
P2
O1
0 2
03
0 4
C1
C2
C3
C4
C5
C6
C7
C8
C9
0.58529(15)
0.79780(14)
0.7522(3)
0.7933(3)
0.3930(8)
0.6406(10)
1.0769(9)
0.7226(8)
0.4719(11)
0.6218(12)
0.6761 (11)
0.5402(11)
0.4498(11)
0.5014(11)
0.5007(13)
0.3017(12)
0.9045(9)
0.11231(5)
- 0.07170(5)
0.1089(1)
- 0.1412(1 )
0.1608(3)
0.1961(4)
- 0.0562(3)
- 0.0083(3)
0.1415(4)
0.1615(4)
0.0680(4)
0.0747(4)
0.0648(4)
0.0480(4)
0.0950(5)
0.0714(4)
0.1377(3)
0.1437(4)
0.1654(4)
0.1815(4)
0.93935(9)
1.44871(9)
0.8646(2)
1.5070(2)
0.8177(5)
1.0393(6)
1.4973(5)
1.5773(5)
0.8663(7)
0.9968(8)
1.0379(7)
1.0428(7)
0.9722(7)
0.8968(7)
1.1238(8)
0.9649(8)
0.9116(6)
1.0004(7)
1.0365(8)
0.9851(7)
CIO
C11
C12
0.9355(11)
1.0532(12)
1.1373(12)