Addition of aryl substituants to cyclohexadienyliron electrophiles in the development of routes to C12 central building blocks for alkaloid synthesis

Article abstract of DOI:10.1016/s0020-1693(99)00390-4

Novel iron-aroyl complexes or synthetically prized arylcyclohexadienyl complexes can be formed from tricarbonyl(η⁵-1,4-dimethoxycyclohexadienyl)iron(1 +) hexafluorophosphate(1-) by correct control of reaction sequences that exploit the addition of aryllithium reagents to introduce the aromatic group. Bimetallic products are obtained when a tricarbonylchromium-bound aryllithium reagent or the corresponding cuprate are employed.

Full text of DOI:10.1016/s0020-1693(99)00390-4

www.elsevier.nl/locate/ica
Inorganica Chimica Acta 296 (1999) 139–149
Addition of aryl substituents to cyclohexadienyliron electrophiles
in the development of routes to C12 central building blocks for
alkaloid synthesis
Andrej V. Malkov a, Audrey Auffrant b, Christophe Renard b, Eric Rose b,
Franc¸oise Rose-Munch b, David A. Owen a, Elizabeth J. Sandoe a,
G. Richard Stephenson a,*
a School of Chemical Sciences, Uni6ersity of East Anglia, Norwich NR4 7TJ, UK
b Uni6ersite´ Pierre et Marie Curie, Laboratoire de Synthe`se Organique et Organome´tallique, UMR CNRS 7611, Tour 44, 4 Place Jussieu,
75252 Paris, Cedex 05, France
Received 1 June 1999; accepted 7 September 1999
Abstract
Novel iron–aroyl complexes or synthetically prized arylcyclohexadienyl complexes can be formed from tricarbonyl(h5-1,4-
dimethoxycyclohexadienyl)iron(1+) hexafluorophosphate(1−) by correct control of reaction sequences that exploit the addition
of aryllithium reagents to introduce the aromatic group. Bimetallic products are obtained when a tricarbonylchromium-bound
aryllithium reagent or the corresponding cuprate are employed. © 1999 Elsevier Science S.A. All rights reserved.
Keywords: Acyliron complexes; Aryllithium nucleophiles; (Aryllithium)chromium complexes; Cyclohexadienyliron electrophiles; Alkaloid synthesis
1. Introduction
trophilic component around which syntheses can be
based. Examples of this approach (Fig. 1) can be seen
in a demonstration synthesis [3] of O-methyljouber-
tiamine (1), a formal total synthesis [4] of lycoramine
(2), and the synthesis [5] of the ABC ring section of the
hippeastrine (3) carbon skeleton. An important chal-
lenge in the optimisation of procedures of this type
concerns the proper definition and control of regioselec-
tivity properties in reactions of organometallic elec-
trophiles with nucleophiles, and over recent years, work
in Norwich [6] and in Paris [7] with h5 and h6 elec-
trophiles has explored these reactions. The choice of
types of nucleophiles, solvents, and reaction conditions
can all play an important role, and on occasions, small
changes in structure can result in wide differences in
reactivity properties. The development of C12 elec-
trophilic intermediates provides an important illustra-
tion that has in part been the subject of a preliminary
communication [8], and is now reported in full.
Versatile building blocks for use in organic synthesis
are especially valuable when they correspond to the
central portion of a common feature of a class of
important target structures. When organometallic com-
plexes are used for this purpose, additional advantages
such as potent stereocontrol effects are available, and
since the metal can be brought into use repeatedly
during a synthesis, the organometallic reactions can
play a central role in the development of the synthetic
strategy [1]. For example, many alkaloid structures
contain in the central part of the molecule, aromatic
and partially saturated six-membered rings joined at a
stereogenic centre [2]. Because of their wide ranging
biological activity, these compounds are extensively
addressed as target structures, and recent work has
identified the possibility that aryl-substituted cyclohexa-
dienyliron complexes can provide a central C12 elec-
The lycoramine class of targets [9] contain an ortho
carbon substituent (R1=CH2–X in 4) on the aromatic
ring and although often this is introduced late in a
* Corresponding author. Tel.: +44-1603-592 019; fax: +44-1603-
259 396.
E-mail address: g.r.stephenson@uea.ac.uk (G. Richard Stephenson)
0020-1693/99/$ - see front matter © 1999 Elsevier Science S.A. All rights reserved.
PII: S0020-1693(99)00390-4

140
A.V. Malko6 et al. / Inorganica Chimica Acta 296 (1999) 139–149
Fig. 1. Examples of C12 organoiron building block (indicated by bold lines) for alkaloid synthesis (for 1: 4, R=OMe, R1=R3=R4=H,
R2=OMe; for 2: 4, R=OMe, R1=CH2X, R2=H, R3=OMe, R4=O(prot.); for 3: 5, R=CH2CH2X, R1=CO2H, R2–R3=O–CH2–O,
R4=H).
synthesis [10], more efficient routes would put this
substituent in place at the beginning. Because of the
easy access [11] to (arylcyclohexadienyl)iron complexes
by elaboration of more simple cyclohexadienyl com-
plexes equipped with suitably placed leaving groups, in
the case of organometallic building blocks based on
these structures it is possible to address the inclusion of
this additional carbon atom on the initial C12 compo-
nent (i.e. a ‘C12–CH2–X’ organometallic electrophile).
Early attempts to prepare such components, however,
quickly established that these reactions were intrinsi-
cally more complicated and less efficient than those
which introduce aromatic groups without ortho sub-
stituents. Besides products from aryl addition to the
dienyl intermediate, attack at carbonyl ligands was also
observed, and in some cases, the resulting acyl products
could be isolated and fully characterised for first time.
Such complexes have often been proposed to explain
inefficient examples of the normally reliable nucleophile
addition procedure, but until these investigations, the
identity of the acyl products had not been established in
the case of h5 organoiron electrophiles (corresponding
structures from h6 organomanganese complexes are
known [12]).
these conditions, the acyl structure is sometimes the
only nucleophile addition product observed although
yields are then generally very low. Furthermore, if the
property of the nucleophile is sufficiently modified, the
access to acyl structures is no longer restricted [8] to
aryl nucleophile addition to 1-alkoxy-substituted cyclo-
hexadienyl complexes (e.g. 6), and the simple parent
(C6H7)Fe(CO)+3 cyclohexadienyl electrophile 12 has
now also been converted into an (acyl)(cyclohexa-
dienyl)Fe(CO)2 product. In this work, a selection of
novel acyliron complexes (11) (Scheme 1), including a
bimetallic example (12) (Scheme 2), have been pre-
pared, and reliable methods have been defined to switch
between this reaction pathway and the access route to
C12 building blocks which are now generally accessible
for evaluation in alkaloid synthesis.
Working at 0° in THF, the aryllithium reagent pre-
pared from o-MeOCH2C6H4Br and n-butyllithium was
added to the 1,4-dimethoxy-substituted cyclohexadienyl
complex 6 [3] to afford (Table 1, entry 1) the
organoiron acyl 11 (R1=CH2OMe, R2=R3=R4=H)
as the sole product of nucleophile addition, together
with traces (TLC) of the cyclohexadienone complex 10
[14] arising from demethoxylation of the C1 OMe
group. Under these conditions, the acyl product was
formed in only 10% yield, but it was found to be
possible to improve the yield by switching (entry 2) to
the more strongly coordinating diether solvent DME.
At 0°C, the yield of the acyl product was improved to
30% and sufficient cyclohexadienone was now formed
(5%) for this by-product to be isolated. The reaction is
sensitive to temperature effects. In THF at −78°C
(entry 3), no nucleophile addition products were formed
at all, while in DME at −78°C (entry 4), the
demethoxylation to the cyclohexadienone was suffi-
2. Results and discussion
We have found that the choice of solvent has a large
effect on the product distribution in the organoiron
series, with the original Birch dichloromethane (DCM)
procedure [13] providing efficient access to the (‘C12–
CH2–X’)Fe(CO)+3 building blocks needed to address
the routes to lycoramine, while ether solvents such as
THF and DME allow substantial quantities of the
unstable acyl (aroyl) products to be isolated. Under

A.V. Malko6 et al. / Inorganica Chimica Acta 296 (1999) 139–149
141
Scheme 1.
Scheme 2.
ciently favoured to allow the production both of larger
amounts of this complex (10%) and the product of its
reaction with the aryllithium reagent. This aryl-substi-
tuted hydroxydiene complex 9 was now the main
product (35%) of the reaction. Other ortho-substituted
aryllithium nucleophiles also afforded the acyl prod-
ucts, though in much lower yields (Table 1, entries
5–7), reflecting the much reduced stability of these
products. In contrast to these results, phenyllithium
itself afforded none of the acyl product when used in
DME (entry 8), and substantial amounts of the aryl-
substituted hydroxydiene complexes were formed in-
stead. The reaction of the unsubstituted cyclohexa-
dienyliron complex has been examined with an aryl-
lithium reagent 13 generated from a tricarbonyl-
chromium complex of 1,3-dimethoxybenzene. Even in
THF (Table 1, entry 18), a metal acyl complex 15 could
be isolated (13%), and with the complications arising
from demethoxylation now unable to intervene, the
diene complex 14 formed by nucleophile addition at the
cyclohexadienyl ring was also obtained (34%). With 1-
alkoxy-substituted cyclohexadienyl complexes, organo-
cuprate reagents have been shown to be less effective
than their lithium counterparts, but when this C1 sub-
stituent is absent, improved results are often obtained
with organocopper reagents. Thus in the case of the
electrophile 12, it was decided to compare results from
the lithium reagent (entry 18) with those obtained
(entry 19) with the diarylcuprate formed from 2 equiv.
of the metallated chromium complex and copper(I)
iodide. As anticipated, the overall efficiency of the
reaction improved from 48 to 75%. Selectivity for the
acyl structure was also increased, and this procedure
gave the most efficient (48%) access to this new struct-
ural class. Because of their considerable instability, the
acyl structures were characterised mainly by IR and
NMR spectroscopy. In addition to metal carbonyl vi-
brational stretching bands at frequencies typical for
neutral products, a characteristic acyl CO band was
observed between 1611 and 1597 cm−1 (Fig. 21). The
NMR spectra corresponded to the normal patterns and
multiplicities of peaks expected for dienyl complexes,
but shifted to higher field as should be the case for a
neutral product. In some cases, characteristic molecular
ions could be seen in the mass spectra of the acyl
products, together with fragmentations involving losses
of Fe(CO)2 and H2, presumably leaving a diarylketone-
based fragment ion. Attempts to emulate this fragmen-
tation by deliberate decomplexation, however, were
unsuccessful, nor could acyl migration be achieved
chemically by heating the acyl complex with a phos-
1 When reported in the preliminary communication (Ref. [8]) the
wavenumber scale in this figure was miss-numbered in the region
2000–1600 cm−1; the illustration is reproduced here with the cor-
rected scale, placing the characteristic acyl vibrational band at 1607
cm−1 for 11a.

142
A.V. Malko6 et al. / Inorganica Chimica Acta 296 (1999) 139–149
Table 1
Entry
a
Electrophile
Nucleophile
Solvent b
Proc. c
Products
Yields (%)
Solvent d
Temperature
(°C)
Subst
R1
R4
H
1
2
6
6
7
7
CH2OMe
CH2OMe
THF/hex
DME/hex
A
A
11a
10
10
trace
30
THF
0
0
H
11a
10
DME
5
–
e
e
3
4
6
6
7
7
CH2OMe
CH2OMe
H
H
THF/hex
DME/hex
A
A
–
THF
DME
−78
−78
9a
11a
10
35
10
10
6
5
6
7
OMe
H
DME/hex
A
10
DME
−60
11b
11b
10
11c
9d
1
1
10
5
37
10
6
7
6
6
7
7
OMe
CH2NMe2
H
H
DME/hex
DME/hex
A
A
DME
DME
0
0
8
6
7
H
H
DME/hex
A
DME
−78
10
e
e
9
10
6
6
7
7
H
H
H
H
DME/hex
DME/hex/TMEDA
A
B
–
–
DME
DME
0
9d
10
8d
9d
10
8d
8d
8d
8d
8d
60
10
72
13
9
−78
11
6
7
H
H
Et2O/cy
C
DCM
−78
12
13
14
15
16
17
18
6
6
6
6
6
6
12
7
7
7
7
7
7
13
H
H
H
H
H
H
OMe
H
H
H
H
H
H
THF/hex
Et2O/hex
Et2O/hex
Et2O
THF/hex/1/2CuI
THF/hex/1/2CuI
A
C
C
B
D
D
E
3
THF
Et2O
DCM
DCM
THF
DCM
THF
−78
−78
−78
−78
−78
−78
−78
21
58
70
72
e
e
–
–
OMe THF/hex
14
15
15
14
34
14
48
27
–
73
5
50
68
43
24
17
58
15
10
50
42
14
4
19
12
13
OMe
OMe THF/hex/1/2CuI
F
THF
−78
e
e
20
21
12
6
13
7
OMe
OMe
OMe THF/hex
OMe Et2O/hex
E
G
–
DCM
DCM
−78
−100
8e
10
8b
10
8b
9b
10
8a
9a
10
10
10
8c
9c
10
22
23
24
6
6
6
7
7
7
OMe
OTBDMS
OMe
H
H
H
Et2O/hex
Et2O/hex
Et2O/hex
G
H
A
DCM
DCM
DCM
−100
−100
−78
25
6
7
CH2O
MeH Et2O/hex
A
DCM
−78
26
27
28
6
6
6
7
7
7
CH2O
CH2NMe2
CH2NMe2
MeH DME/hex
H
H
A
A
A
DCM
DCM
DCM
−78
−78
−110
Et2O/hex
Et2O/hex
5
a Nucleophile (substituents R2 and R3 in Scheme 1=H).
b Solvents/reagents used in the generation of the organolithium or cuprate reagents.
c See Section 3.
d Solvent for the electrophile, to which the solution of the nucleophilic reagent was added.
e No products isolated.
(MeOCH2)C6H4Li with the 1,4-dimethoxycyclohexadi-
enyliron complex 6) was selected for full charac-
terisation. Even in this case, the product is heat sensi-
phine ligand. In the mass spectrum, however, formation
of M+–Fe(CO)2–H2 was seen to be facile. The most
stable acyl product (arising from reaction of o-

A.V. Malko6 et al. / Inorganica Chimica Acta 296 (1999) 139–149
143
small amount of DME present was sufficient to com-
pletely disfavour the addition of the aryl group to the
cyclohexadienyl ligand (Table 1, entry 26), but under
these conditions and at −78°C no acyl products were
observed, and the cyclohexadienone complex from
demethoxylation was isolated in 50% yield.
We have observed before that including nitrogen-
containing substituents in aryl nucleophiles can lead to
inefficient nucleophile addition reactions [15]. With sim-
ple cyclohexadienyliron complexes, organocuprate re-
agents can be used to overcome this limitation [15,16],
but since these are not normally suitable for use with
1-alkoxy-substituted cyclohexadienyl complexes, the use
of Me2NCH2C6H4Li at very low temperature was ex-
amined instead, to obtain diene complexes comparable
to the FeCOC6H4CH2NMe2 acyl product 11 that was
obtained in low yield in the earlier stages of these
investigations. Performing the reaction in DCM at
−110°C, 18% yield of a 3:1 mixture of methoxy- and
hydroxydiene complexes was possible (Table 1, entry
28) while the yield of the cyclohexadienone complex
dropped (from 42% at −78°C) to just 5%.
The formation of mixtures of methoxy- and hydroxy-
substituted cyclohexadieneiron complexes complicates
product purification at the intermediate stage, but does
not make inefficient the route to the ‘C12–CH2–X’
organometallic electrophiles because both products can
be converted into the same aryl-substituted cyclohexa-
dienyl complex by reaction in the normal way with
acid. Either HBF4·Et2O or trifluoroacetic acid followed
by addition of ammonium hexafluorophosphate can be
used for this purpose, with comparable results, and
yields are good (Scheme 3). Thus if the intermediate
diene complexes prepared in DCM are taken on by this
method without separation, high overall yields of cyclo-
hexadienyliron complexes with ortho-substituted aryl
substituents at C1 are possible. From the point of view
of access to alkaloid structures, the benzyl ether case is
the best prospect, and this can be made in over 55%
yield ‘salt-to-salt’ from the simple 1,4-dimethoxycyclo-
hexadienyl complex 6, which is itself available on a
large scale by Birch reduction of 1,4-dimethoxybenzene,
followed by complexation and hydride abstraction [3].
The addition of the aryllithium reagents in this proce-
dure, however, is very sensitive to the precise fashion
the reaction is performed. For example, to gain large-
scale access to the ‘C12–CH2–X’ electrophiles it has
been possible to establish protocols (Table 1, entries 21,
22) that avoid demethoxylation and the formation of
alcohols, allowing the preparation of a single arylcyclo-
hexadiene complex as the intermediate. In this version
of the reaction, the metallation to form the organo-
lithium reagent is performed in ether in the usual way,
but the solution is then cooled to −100°C and added
to a suspension of the salt in DCM at −100°C. This
Fig. 2. Typical IR spectrum for novel metal acyl complexes of
dicarbonyl(h5-cyclohexadienyl)iron(II).
tive and unstable even under vacuum, so characterisa-
tion was performed immediately after the final chro-
matographic purification of the product. IR and NMR
data match the general pattern of the whole series, and
the mass spectrometric properties were fully defined. By
FAB ionisation, M++Na was confirmed by high reso-
lution mass measurements, and by CI, accurate mass
measurements confirmed the identity of an MH+–
Fe(CO)2–H2 ion at 287, corresponding to the M+–
Fe(CO)2–H2 fragment at 286 in the EI mass spectrum.
Under EI, M+–2CO was observed at 344, but no
parent molecular ion could be observed.
As with Birch’s original investigations of the use [13]
of organolithium reagents with cyclohexadienyliron
electrophiles, the use of DCM as solvent proved to
favour direct addition at the dienyl ligand, at least with
simple aryllithium nucleophiles (Table 1, entries 11, 14,
15). The tricarbonylchromium-complexed lithium re-
agent, on the other hand, gave neither acyl nor diene
complexes when used in DCM (entry 20), and in this
case the best access to the diene complexes was in
THF/hexane (entry 18). Typically, though, in DCM
using a procedure comparable to that which affords
acyl structures in DME, considerable quantities of the
required aryl-substituted methoxy- and hydroxycyclo-
hexadiene complexes (8 and 9, respectively) can be
obtained (65–87%) together with the cyclohexadienone
product 10 (17–9%). In these reactions, the aryllithium
was used as purchased in the case of phenyllithium (in
cyclohexane/ether), or generated from the aryl bromide
in hexane/ether by addition of n-butyllithium in hex-
anes, except in one case where the lithium/halogen
exchange was performed in DME, allowing sufficient of
the coordinating solvent to be present to modify the
properties of the nucleophile. This experiment was per-
formed to establish whether the formation of the acyl
products was possible as a competing reaction in DCM
through the influence of lithium-bound DME. The

144
A.V. Malko6 et al. / Inorganica Chimica Acta 296 (1999) 139–149
usually has little effect on overall yields ‘salt-to-salt’ but
is experimentally more convenient.
3.1. A: General procedure for the reaction of the
1,4-dimethoxy-substituted salt with aryllithium reagents
n-Butyllithium (2 ml of a 1.27 M solution in hexanes,
2.54 mmol) was added to a solution of the aryl bromide
(2.6 mmol) in dry ether, THF or DME (3–5 ml) at
−78°C, and the mixture was stirred at that tempera-
ture for 30 min and then transferred into a solution of
the cyclohexadienyl salt 6 (360 mg, 1 mmol) in dry
THF, DME or DCM (25-30 ml) at 0, −78 or
−110°C. After addition of water and extraction with
ether, chromatography on silica gel or alumina (in the
case of products from the nitrogen-containing nucle-
ophile) with 5% ethyl acetate in hexane as eluant gave
in order of elution methoxy- and hydroxy-substituted
cyclohexadiene complexes 8 and 9. Using mixture of
hexane/ethyl acetate (2:1) the cyclohexadienone com-
plex 10 [14] was eluted.
3. Experimental
All reactions were performed under an atmosphere of
dry, oxygen-free nitrogen and in the cases that yield
acyl products, glassware was oven-dried and twice evac-
uated and filled with the nitrogen. Ether refers to
diethyl ether; water refers to distilled water. All solvents
were of reagent grade and used as supplied commer-
cially unless specified as dry, in which case they were
dried and distilled immediately before use as follows:
ether and tetrahydrofuran (THF) from sodium/ben-
zophenone; dimethoxyethane (DME) from sodium hy-
dride; dichloromethane (DCM) from calcium hydride.
Reaction temperatures of −110, −78 and 0°C refer to
absolute ethanol/liquid-nitrogen, acetone/dry-ice and
water/ice bath cooling, respectively. Filtration refers to
filtration under water-pump suction. Analytical TLC
was performed on silica or alumina plates, and visu-
alised by ultraviolet irradiation (254 nm). Column chro-
matography was performed using Merck silica gel or
BDH neutral alumina. Low resolution EI mass spec-
trometry (Kratos MS25 mass spectrometer) and ele-
mental analyses were performed at the University of
East Anglia by Mr A.W.R. Saunders; other mass spec-
tra were measured at the EPSRC National Mass Spec-
trometry Service Centre at University of Wales,
Swansea. IR spectra were recorded as a thin film using
a Perkin–Elmer 1420 or 1720X FT-IR spectrometer.
NMR spectra were recorded on Jeol EX90 (1H, 90
MHz), Jeol EX270, Bruker AC250 or Bruker ARX (1H,
400 MHz) spectrometers. Chiral metal complexes were
prepared in racemic form and illustrations in Schemes
1–3 indicate relative stereochemistry only.
3.2. B: Modified procedures for the reaction of the
1,4-dimethoxy-substituted salt with phenyllithium —
TMEDA
The procedure was the same as described in A, using
DME and hexanes, except that TMEDA (1 equiv.) was
added to the solution of the organolithium reagent
before reaction with the organoiron salt (Table 1, entry
10).
3.3. C: Modified procedures for the reaction of the
1,4-dimethoxy-substituted salt with phenyllithium
Phenyllithium (1.5 ml of a 1.70 M solution in cyclo-
hexane/ether (7:3), 2.55 mmol) was added to a solution
of the salt 6 (360 mg, 1 mmol) in dry THF, DME or
DCM (25–30 ml). Work-up and chromatography was
performed as described in A. Similarly, phenyllithium
Scheme 3.

A.V. Malko6 et al. / Inorganica Chimica Acta 296 (1999) 139–149
145
(7.80 ml of a 0.78 M solution in ether, 6.08 mmol) was
added to a solution of the salt 6 (2.00 g, 4.72 mmol) in
dry DCM (30 ml) at −78°C. After the work-up, flash
chromatography (30% DCM in light petroleum) af-
forded complex 8d (1.18 g, 70%). In other cases (Table
1, entries 13, 14), phenyllithium was prepared by stir-
ring n-butyllithium (0.63 ml of a 1.45 M solution in
hexanes, 0.91 mmol) with bromobenzene (14.4 mg, 0.92
mmol) in dry ether (2 ml) at room temperature (r.t.) for
20 min. This reagent solution was reacted, as described
above, with a solution of the salt 6 (300 mg, 0.71 mmol)
in dry DCM (10 ml) at −78°C. The work-up and flash
column chromatography gave the product 8d (146 mg,
58%). The same procedure was used with suspensions
of the salt in ether or THF.
ether) to separate recovered tricarbonyl(h6-1,3-dime-
thoxybenzene)chromium(0) (0.11 g, 0.4 mmol, 40%),
and then 45% ether in petroleum ether to elute the
bimetallic h6,h4 product 14 (0.167 g, 34%) and with
60% ether in petroleum ether to afford the bimetallic
h6,h5 acyl complex 15 (0.064 g, 13%).
3.6. F: Modified procedure for the reaction of the
unsubstituted salt with a chromium-complexed
diarylcuprate reagent
Procedure E was employed except that the flask was
fitted with a solids addition side-arm charged with
copper(I) iodide (0.5 mmol) and after generation of the
chromium-complexed aryllithium reagent, the copper
salt was added under nitrogen at −78°C and the
mixture was stirred for a further 20 min before the
reagent was transferred to the flask containing the
cyclohexadienyliron complex.
3.4. D: Modified procedure for the reaction of the
1,4-dimethoxy-substituted salt with lithium
diphenylcuprate
n-Butyllithium (4.20 ml of a 1.52 M solution in
hexanes, 6.37 mmol) was added to a solution of bromo-
benzene (1.00 g, 6.37 mmol) in dry THF or DCM at
−78°C. The mixture was warmed to r.t. and added by
syringe to a stirred suspension of copper(I) iodide (591
mg, 3.10 mmol) in dry THF (5 ml) at 0°C. After 5 min,
the organoiron salt (600 mg, 1.42 mmol) was added
against nitrogen back-pressure, and stirring was contin-
ued 5 min at 0°C. The mixture was then poured into a
separating funnel charged with saturated aqueous am-
monium chloride (50 ml) and extracted with ether
(3×30 ml). The combined extracts were washed with
water (30 ml), dried (MgSO4), and filtered, and the
solvent was removed under reduced pressure. Flash
chromatography with 30% DCM in light petroleum as
the eluant afforded the product 8d (72%).
3.7. G: Modified procedure to prepare 8e a6oiding the
formation of alcohol and dienone byproducts (9 and 10)
Following the method of Snieckus [17], 1,3-
dimethoxybenzene (2 equiv., 1.66 g, 12 mmol) was
dissolved in dry ether (10 ml) under nitrogen. n-Butyl-
lithium (1.6 M in hexanes, 12.0 mmol, 7.5 ml) was
added and the mixture was heated at reflux for 2 h after
which time 1-lithio-2,6-dimethoxybenzene was formed
as a yellow suspension. The cyclohexadienyliron salt
(2.196 g, 6.0 mmol) was dissolved in dry DCM (20 ml)
and cooled to −100°C. The solution of the nucleophile
at −100°C was added through a cannula at −100°C.
The mixture turned black and was stirred for 2 h. The
reaction was quenched with water (25 ml) and ether (25
ml) at −100°C and warmed to r.t. The mixture was
extracted into ether (3×25 ml) and water (3×25 ml).
The combined organic extracts were washed with water
(3×25 ml), dried (MgSO4) and filtered. The solvent
was removed under reduced pressure to afford a brown
oil which was purified by column chromatography on
silica eluted with ether/cyclohexane (1:1).
3.5. E: Modified procedure for the reaction of the
unsubstituted salt with a chromium-complexed
aryllithium reagent
n-Butyllithium (1.27 M solution in hexanes, 1.1
mmol) was added to a solution of tricarbonyl(h6-1,3-
dimethoxybenzene)chromium(0) (0.274 g, 1 mmol) in
THF (15 ml) at −78°C. After stirring at this tempera-
ture for 1 h, the resulting solution of the aryllithium
reagent was transferred by syringe to a solution of
tricarbonyl(h5-cyclohexadienyl)iron(1+) hexafluoro-
phosphate(1−) (0.44 g, 1.2 mmol) in THF (10 ml) at
−78°C. The mixture was allowed to warm and was
stirred at r.t. for 5 min. Water (20 ml) was added and
the mixture was extracted with ether (2×30 ml). The
extracts were washed with brine and dried (MgSO4),
filtered through Celite, and evaporated without heating
with a stream of nitrogen gas. The residue was purified
by chromatography (silica/30% ether in petroleum
3.8. H: Procedure for the attempted reaction of the
1,4-dimethoxy-substituted salt with 1-lithio-2-t-
butyldimethylsilyloxybenzene
A solution of 2-bromophenol (5 g, 29 mmol) in dry
THF (20 ml) was added over 1 h to NaH (60% suspen-
sion in mineral oil) (1.2 g, 29 mmol) in dry THF (20 ml)
at 0°C. The reaction mixture was stirred at 0°C for 1 h
to give a pale brown solution. t-Butyldimethylsilyl chlo-
ride (4.36 g, 29 mmol) was added at 0°C. The reaction
mixture was stirred at r.t. for 18 h during which time a
white precipitate formed. The mixture was quenched
with water (50 ml) and ether (50 ml) and extracted with

146
A.V. Malko6 et al. / Inorganica Chimica Acta 296 (1999) 139–149
ether (3×50 ml). The combined organic extracts were
washed with water (3×25 ml), dried (MgSO4) and
filtered. The solvent was removed under reduced pres-
sure. Column chromatography (10% ether/89%
petroleum ether/1% Et3N) afforded 1-bromo-2-t-butyl-
dimethylsilyloxybenzene [18] as a clear liquid (6.96 g, 24
mmol, 85%). HRMS (EI) (Found: 286.0389.
C12H19OBrSi (M+) requires 286.0389). lH (CDCl3) 0.24
(6H, s, SiMe2), 1.04 (9H, s, Me3CSi), 6.79 (1H, ddd,
J=8.9, 7.3, 1.7, Ar), 6.86 (1H, d, J=8.3, 1.7, Ar), 7.14
(1H, ddd, J=8.9, 7.3, 1.7, Ar), 7.50 (1H, dd, J=8.3,
1.7, Ar). This product (0.89 g, 3.1 mmol) was dissolved
in dry ether (5 ml) and cooled to −30°C under nitro-
gen. n-Butyllithium (1.6 M in hexanes, 2.0 ml, 3.1
mmol) was added after stirring for 1 h at −30°C,
1-lithio-2-t-butyldimethylsilyloxybenzene was formed as
a white suspension and was cooled to −100°C. The
organoiron salt (1.1 g, 3 mmol) was dissolved in dry
DCM (10 ml) and cooled to −100°C. The solution of
the nucleophile was added through a cannula and the
mixture was stirred for 30 min. The mixture was
quenched with water (25 ml) and ether (25 ml), and
extracted with ether (3×25 ml). The combined extracts
were washed with water (3×25 ml), dried (MgSO4) and
filtered. The solvent was removed under reduced pres-
sure to afford a brown oil. Column chromatography on
silica (ether/petroleum ether gradient) afforded the cy-
clohexadienone complex 10 [14] (0.54 g, 68%).
3.8.3. Dicarbonyl[1%-carboxy(2%-methoxymethylphenyl)]-
[(1,2,3,4,5-p)-2,5-dimethoxy-2,4-cyclohexadienyl]-
iron(II) (11a) by procedure A
Yellow solid, heat sensitive, unstable in vacuum. IR:
wmax (cm−1) 1999, 1942 (CO); 1607 (ArCOFe). lH
(CDC13) 2.59 (1H, d, J=14.5 Hz, 6exo-H), 3.05 (1H,
ddd, J=14.9, 5.7, 1.5 Hz, 6endo-H), 3.33 (3H, s, 2%-
OMe), 3.37 (3H, s, 1-OMe), 3.60 (3H, s, 4-OMe), 3.63
(2H, m, 2,5-H), 4.32 (2H, m, 2%-CH2), 5.64 (1H, dd,
J=5.7, 2.6 Hz, 3-H), 7.14–7.34 (4H, m, 3%-6%-H). lH
(benzene-d6) 2.06 (1H, d, J=15.4 Hz, 6exo-H), 2.59
(3H, s, OMe), 2.72 (1H, dd, J=15.4, 5.1 Hz, 6endoH),
2.90 (2H, m, 2-H and 5-H), 3.18 (3H, s, 1-OMe), 3.22
(3H, s, 4-OMe), 4.47 (1H, d, J=10.0 Hz, 2%a-H), 4.66
(1H, d, J=10.0 Hz, 2%b-H), 5.13 (1H, m, 3-H), 7.00–
7.48 (4H, m, 3%-6%-H). m/z (EI) 344 (M+–2CO, 1.05%),
286 (M+–Fe(CO)2–H2, 2%). Found m/z (FAB)
M++Na, 423.0514. C19H20O6Fe requires M++Na,
423.0507. Found m/z (CI) MH+–Fe(CO)2–H2,
287.1283. C19H20O6Fe requires MH+–Fe(CO)2–H2,
287.1283.
3.8.4. Tricarbonyl[(1,2,3,4-p)-2,5i-dimethoxy-5h-
(2%-methoxyphenyl)-1,3-cyclohexadiene]iron(0) (8b) by
procedures A and G
Grey–white solid. IR: wmax (cm−1) 2050, 1985, 1964
(CO). lH (CDC13) 2.14 (1H, dd, J=15.1, 3.5 Hz,
6endo-H), 2.39 (1H, dd, J=15.0, 2.6 Hz; 6exo-H), 2.90
(1H, d, J=7.5 Hz, 4-H), 2.95 (3H, s, 5-OMe), 3.27
(1H, m, 1-H), 3.58 (3H, s, 2-OMe), 3.74 (3H, s, 2%-
OMe), 5.01 (1H, dd, J=6.8, 2.4 Hz, 3-H), 6.76–7.48
(4H, m, 3%-6%-H). m/z (EI) 358 (M+–CO, 1%), 330
(M+–2CO, 2.5%). Found: C, 56.38; H, 4.65.
C18H18O6Fe requires: C, 55.98; H, 4.70%. Found m/z
(CI) MH+–MeOH, 355.0269. C18H18O6Fe requires:
MH+–MeOH, 355.0269.
Using these procedures, the following compounds
were prepared (for yields, see Table 1):
3.8.1. Tricarbonyl[(1,2,3,4-p)-2,5i-
dimethoxy-5h-(2%-methoxymethylphenyl)-1,3-
cyclohexadiene]iron(0) (8a) by procedure A
Viscous pale yellow oil. IR: wmax (cm−1) 2051, 1982,
1964 (CO). lH (CDC13) 2.12 (1H, dd, J=14.7, 2.6 Hz,
6exo-H), 2.34 (1H, dd, J=14.7, 3.5 Hz, 6endo-H), 2.91
(3H, s, 5-OMe), 3.01 (1H, d, J=6.8 Hz, 4-H), 3.28
(1H, m, 1-H), 3.40 (3H, s, 2%-OMe), 3.63 (3H, s, 2-
OMe), 4.60 (2H, s, 2%-CH2), 5.32 (1H, dd, J=6.8, 2.4
Hz, 3-H), 7.10–7.59 (4H, m, 3%-6%-H). m/z (EI) 372
(M+–CO, 0.1%), 346 (M+–2CO, 0.2%). Found m/z
(CI) MH+–MeOH, 369.0425. C19H20O6Fe requires
MH+–MeOH, 369.0425.
3.8.5. Tricarbonyl[(1,2,3,4-p)-5i-hydroxy-2-methoxy-
5h-(2%-methoxyphenyl)-1,3-cyclohexadiene]iron(0) (9b)
by procedure A
Viscous pale yellow oil. IR: wmax (cm−1) 3534 (OH),
2046, 1975 (CO). lH (CDC13) 2.25 (2H, m, 6-CH2),
2.86 (1H, d, J=7.0 Hz, 4-H), 3.27 (1H, m, 1-H), 3.60
(3H, s, 2-OMe), 3.83 (3H, s, 2%-OMe), 5.17 (1H, dd,
J=7.0, 2.8 Hz, 3-H), 6.70–7.45 (4H, m, 3%-6%-H). m/z
(EI) 344 (M+–CO, 0.3%), 316 (M+–2CO, 0.3%).
Found m/z (CI) M+–CO, 344.0347 and MH+–H2O,
355.0269. C17H16O6Fe requires M+–CO, 344.0347 and
MH+–H2O, 355.0269.
3.8.2. Tricarbonyl[(1,2,3,4-p)-5i-hydroxy-2-methoxy-
5h-(2%-methoxymethylphenyl)-1,3-cyclohexadiene]iron(0)
(9a) by procedure A
Viscous pale yellow oil. IR: wmax (cm−1) 3402 (OH),
2048, 1978 (CO). lH (CDC13) 2.34 (2H, m, 6-CH2),
2.84 (1H, d, J=6.6 Hz, 4-H), 3.33 (1H, m, 1-H), 3.38
(3H, s, 2%-OMe), 3.66 (3H, s, 2-OMe), 4.42 (1H, d,
J=11.6 Hz, 2%a-H), 4.82 (d, 1H, J=11.6 Hz, 2%b-H),
5.20 (1H, dd, J=6.7, 2.5 Hz, 3-H), 7.21–7.52 (4H, m,
3%-6%-H). m/z (EI) 330 (M+–2CO, 2%).
3.8.6. Dicarbonyl[1%-carboxy(2%-methoxyphenyl)]-
[(1,2,3,4,5-p)-2,5-dimethoxy-2,4-cyclohexadienyl]-
iron(II) (11b) by procedure A
Yellow solid, heat sensitive, unstable in vacuum. IR:
wmax (cm−1) 1996, 1937 (CO); 1597 (ArCOFe). lH

A.V. Malko6 et al. / Inorganica Chimica Acta 296 (1999) 139–149
147
(benzene-d6) 2.07 (1H, d, J=15.4 Hz, 6exo-H), 2.64
3.8.11. Tricarbonyl[(1,2,3,4-p)-5i-hydroxy-2-
(1H, m, 6endo-H), 2.96 (3H, s, 1-OMe), 3.14 (1H, m,
5-H), 3.34 (6H, s, 4- and 2%-OMe), 5.12 (1H, m, 3-H),
6.60–7.20 (4H, m, 3%-6%-H). m/z (EI) 272 (M+
−Fe(CO)2-H2, 4%), 242 (M+−Fe(CO)2 −MeOH,
1.5%).
methoxy-5h-phenyl-1,3-cyclohexadiene]iron(0) (9d) by
procedures B and C
Viscous pale yellow oil. IR: wmax (cm−1) 3448 (OH),
2047 and 1975 (CO). lH (CDC13) 2.27 (2H, m, 6-CH2),
2.70 (1H, d, J=7.0 Hz, 4-H), 3.49 (1H, m, 1-H), 3.74
(3H, s, 2-OMe), 5.09 (1H, dd, J=7.0 and 3.0 Hz, 3-H),
7.15–7.50 (5H, m, Ph). m/z (EI) 314 (M+–CO, 0.4%),
258 (M+−3CO, 2%).
3.8.7. Tricarbonyl[(1,2,3,4-p)-2,5i-dimethoxy-5h-
(2%-dimethylaminomethylphenyl)-1,3-cyclohexadiene]-
iron(0) (8c) by procedure A
Viscous pale yellow oil. IR: wmax (cm−1) 2045, 1976,
1963 (CO). lH (CDC13) 2.15 (2H, m, 6-CH2), 2.19 (6H,
s, NMe2), 2.85 (3H, s, 5-OMe), 3.03 (1H, d, J=7.2 Hz,
4-H), 3.24 (1H, m, 1-H), 3.54 (2H, m, 2%-CH2), 3.60
(3H, s, 2-OMe), 5.29 (1H, dd, J=6.8, 2.5 Hz, 3-H),
7.10–7.72 (4H, m, 3% and 6%-H). m/z (EI) 385 (M+–CO,
0.6%), 329 (M+–3CO, 2%). Found m/z (CI) MH+,
414.1004. C20H23NO5Fe requires MH+, 414.1004.
3.8.12. Tricarbonyl[(1,2,3,4-p)-2,5i-dimethoxy-5h-
(2%,6%-dimethoxyphenyl)-1,3-cyclohexadiene]iron(0) (8e)
by procedure G
IR: wmax (cm−1) 2040, 1966, (CO), 1589, 1488, 1251,
1110, 622, 583. lH (CDC13) 7.17 (1H, t, J=8.5, ArH-
4), 6.54 (2H, d, J=8.5, ArH-2, ArH-5), 5.02 (1H, dd,
J=6.9, 3.0, H-3), 3.80 (6H, s, Ar–OMe), 3.66 (1H, d,
J=6.9, H-4), 3.56 (3H, s, C2-OMe), 3.27 (1H, m, H-1),
3.00 (3H, s, C5-OMe), 2.47 (1H, dd, J=15.5, 2.5,
H-6b), 2.36 (1H, dd, J=15.5, 3.8, H-6a). m/z (EI)
(M+–(CO)2 388. Found m/z (EI) 388.0609 (M+–CO).
C18H20FeO6 (M+–CO) requires 388.0609.
3.8.8. Tricarbonyl[(1,2,3,4-p)-5i-hydroxy-2-methoxy-
5h-(2%-dimethylaminomethylphenyl)-1,3-cyclohexadiene]-
iron(0) (9c) by procedure A
Viscous pale yellow oil. IR: wmax (cm−1) 3430 (OH),
2048, 1980, 1969 (CO). lH (CDC13) 2.16 (6H, s, NMe2),
2.28 (2H, m, 6-CH2), 2.88 (1H, d, J=6.6 Hz, 4-H), 3.02
(1H, d, J=12.3 Hz, 2%a-H), 3.61 (3H, s, 2-OMe), 3.92
(1H, d, J=12.5 Hz, 2%b-H), 5.22 (1H, dd, J=6.6, 2.6
Hz, 3-H), 6.99–7.62 (4H, m, 3%-6%-H). m/z (EI) 343
(M+–2CO, 2%). Found m/z (CI) MH+, 400.0847.
C19H21NO5Fe requires MH+, 400.0847.
3.8.13. Nonacarbonyl{(1,2,3,4-p)-2,5i-
dimethoxy-5h-[(1%,2%,3%,4%,5%,6%-p)-2%,6%-
dimethoxyphenyl]chromium(0)]-1,3-cyclohexadiene}-
iron(0) (14) by procedures E and F
Pale yellow solid. IR: wmax (cm−1) 2030, 1960, 1865
(CO). lH (400 MHz, CDC13) 1.74 (1H, dm, J=14.2
Hz, 6a-H), 2.06 (1H, ddd, J=14.2, 11.7, 4.1 Hz, 6b-H),
2.87 and 3.10 (2H, m, 1-H and 4-H), 3.69 (1H, m, 5b-H
and 6H, s, 2MeO), 4.61 and 4.66 (2H, d, J=7 Hz, 3%-H
and 5%-H), 5.36 (2H, m, 2-H and 3-H), 5.47 (1H, t, J=7
Hz, 4%-H). Found: C, 48.67; H, 3.26. C20H16O8Fe re-
quires: C, 48.81; H, 3.28%.
3.8.9. Dicarbonyl[1%-carboxy(2%-dimethylamino-
methylphenyl)][(1,2,3,4,5-p)-2,5-dimethoxy-2,4-
cyclohexadienyl]iron(II) (11c) by procedure A
Yellow solid, heat sensitive, unstable in vacuum. IR:
wmax (cm−1) 1998, 1940 (CO), 1611 (ArCOFe). lH
(CDC13) 2.17 (6H, s, NMe2), 2.59 (1H, d, J=15.0 Hz,
6exo-H), 3.05 (1H, ddd, J=15.0, 5.7, 1.5 Hz; 6endo-H),
3.35 (3H, s, 1-OMe), 3.36 (1H, d, J 5.7 Hz, 2-H), 3.60
(3H, s, 4-OMe), 3.60 (3H, m, 5-H, 2%-CH2), 5.64 (1H,
dd, J=5.7, 2.4 Hz, 3-H), 7.00–7.26 (4H, m, 3%-6%-H).
m/z (EI) 299 (M+–Fe(CO)2–H2, 4%). Found m/z (CI)
MH+ Fe(CO)2–H2, 300.1600. C20H23NO5 requires
MH+–Fe(CO)2–H2, 300.1600.
3.8.14. Octacarbonyl{1%-carboxy[(1%,2%,3%,4%,5%,6%-p)-
2%,6%-dimethoxyphenyl]chromium(0)}[(1,2,3,4,5-p)-
2,5-dimethoxy-2,4-cyclohexadienyl]iron(II) (15) by pro-
cedures E and F
Yellow solid. IR: wmax (cm−1) 2040, 2000, 1965, 1865
(CO), 1605 (ArCOFe). lH (400 MHz, CDC13) 1.83 (1H,
d, J=14.2 Hz, 6a-H), 2.62 (1H, m, 6b-H), 3.50 (2H, m,
1-H and 5-H), 3.73 (6H, s, 2MeO), 4.64 (2H, d, J=6.3
Hz, 3%-H and 5%-H), 4.72 (2H, m, 2-H and 4-H), 5.43
(1H, t, J=6.3 Hz, 4%-H), 6.49 (1H, t, J=5.3 Hz, 3-H).
3.8.10. Tricarbonyl[(1,2,3,4-p)-2,5i-dimethoxy-5h-
phenyl-1,3-cyclohexadiene]iron(0) (8d) by procedures A,
C, D and G
Viscous pale yellow oil which solidified on refrigera-
tion. M.p. 84–86°C. IR: wmax (cm−1) 2046 and 1966
(CO). lH (CDC13) 2.21 (2H, m, 6-CH2), 2.80 (1H, d,
J=7.0 Hz, 4-H), 3.05 (3H, s, 5-OMe), 3.41 (1H, m,
1-H), 3.64 (3H, s, 2-OMe), 5.05 (1H, dd, J=7.0, 2.8
Hz, 3-H), 7.20–7.60 (5H, m, Ph). Found m/z (CI)
MH+, 357.0391. C17H16O5Fe requires MH+, 357.0425.
3.8.15. Preparation of cyclohexadienyliron complexes
with TFA or HBF4·Et2O
The mixture of complexes 8 and 9 (1 mmol) was
stirred with trifluoroacetic acid (TFA) (10 mmol) or
HBF4 · Et2O (10 mmol) at 0°C for 30 min. Addition of
a saturated solution of ammonium hexafluorophos-
phate afforded corresponding salt as a yellow powder
which was reprecipitated from acetonitrile/ether.

148
A.V. Malko6 et al. / Inorganica Chimica Acta 296 (1999) 139–149
3.8.16. Preparation of cyclohexadienyliron complexes
with aqueous HPF6
3.8.20. Tricarbonyl[(1,2,3,4,5-p)-1-phenyl-4-methoxy-
2,4-cyclohexadienyl]iron(1+) hexafluorophosphate(1−)
(4d, R=OMe)
Hexafluorophosphoric acid (75% in water, 0.5 ml,
4.68 mmol) was added to a cooled solution (0°C) of the
cyclohexadiene complex (0.145 g, 0.38 mmol) in acetic
anhydride (20 ml), and was stirred for 30 min. The
mixture was warmed to r.t. and cold (0°C) dry ether (25
ml) was added dropwise. A brown oil formed. Satu-
rated aqueous ammonium hexafluorophosphate solu-
tion was added to produce an orange precipitate which
was collected by filtration and washed with cold dry
ether.
72% overall yield via 8 and 9 by procedure A (85%
yield) followed by the TFA/NH4PF6 method (85%
yield), (see Scheme 3), or 70% overall yield via purified
8 by procedure C (72% yield) followed by the TFA/
NH4PF6 method (98% yield). IR: wmax (cm−1) 2100,
2045 (CO). lH (acetone-d6) 2.75 (1H, d, J=14.9 Hz,
6a-H), 3.97 (1H, dd, J=14.9, 6.6 Hz, 6b-H), 3.99 (3H,
s, OMe), 4.43 (1H, dd, J=6.6, 2.7 Hz, 5-H), 6.63 (1H,
d, J=6.2 Hz, 2-H), 7.29 (1H, dd, J=6.2, 2.7 Hz, 3-H),
7.5 (5H, m, Ph). lC (62.5 MHz, acetone-d6) 30.4 (C-6),
43.4 (C-5), 58.0 (OMe), 73.5 (C-3), 86.5 (C-1), 93.4
(C-2), 127.3 (C-2%, C-6%), 130.5 (C-3%, C-5%), 131.6 (C4,),
135.1 (C-1%), 151.1 (C-4), and 206.3 (CO). Found: C,
40.8; H, 2.65. C16H13F6FeO4P requires: C, 40.9; H,
2.8%.
3.8.17. Preparation of cyclohexadienyliron complexes
with Ph3CPF6
The cyclohexadiene complex 8e (1.81 g, 4.36 mmol)
was dissolved in DCM and added to a solution of
triphenylcarbenium hexafluorophosphate (1.69 g, 4.36
mmol) in DCM (5 ml) at 0°C. The mixture darkened
and was stirred at 0°C for 2 h. The reaction mixture
was added dropwise into dry ether (200 ml) at 0°C to
produce the product as a yellow precipitate.
Using these procedures, the following compounds
were prepared.
3.8.21. Tricarbonyl[(1,2,3,4,5-p)-1-(2%,6%-
dimethoxyphenyl)-4-methoxy-2,4-cyclohexadienyl]-
iron(1+) hexafluorophosphate(1−) (4e, R=OMe)
54% overall yield via purified 8 by procedure G (73%
yield) followed by the Ph3CPF6 method (74% yield),
(see Scheme 3). IR: wmax (cm−1) 2103, 2052 (CO), 1595,
1499, 1250, 837, 558. lH (acetone-d6, 250 MHz) 2.73
(1H, d, J=16.5, 6a-H), 3.29 (1H, dd, J=16.5, 6.0,
6b-H), 3.82 (3H, s, 4-OMe), 3.85 (6H, s, Ar–OMe),
3.96 (1H, m, 5-H), 5.70 (1H, d, J=6.3, 2-H), 6.71 (2H,
d, J=8.5, Ar), 6.98 (1H, dd, J=6.3, 2.5, 3-H), 7.42
(1H, t, J=8.5, Ar). m/z (FAB) 385 (M+−PF6).
Found: 385.0375 C18H17FeO6 (M+−PF6) requires:
385.0375).
3.8.18. Tricarbonyl[(1,2,3,4,5-p)-1-(2%-methoxy-
methylphenyl)-4-methoxy-2,4-cyclohexadienyl]iron(1+)
hexafluorophosphate(1−) (4a, R=OMe)
56% overall yield via 8 and 9 by procedure A (72%
yield) followed by the TFA/NH4PF6 method (98%
yield), (see Scheme 3). IR: wmax (cm−1) 2108, 2053
(CO). lH (acetone-d6) 2.92 (1H, d, J=15.2 Hz, 6a-H),
3.22 (3H, s, 2%-OMe), 3.49 (1H, dd, J=15.9, 5.9 Hz,
6b-H), 3.99 (3H, s, 4-OMe), 4.36 (1H, m, 5-H), 4.37
(2H, s, 2%-CH2), 6.22 (1H, d, J=5.3 Hz, 2-H), 7.28–
7.42 (5H, m, 3-H, 3%-6%-H). Found: C, 42.25; H, 3.24.
C18H17O5FePF6 requires: C, 42.05; H, 3.33%.
Acknowledgements
We thank the Royal Society, the Underwood Fund,
the EPSRC, Glaxo Wellcome and the CNRS for finan-
cial support, and the EPSRC National Mass Spec-
trometry Service Centre (University of Wales, Swansea,
UK) for mass spectrometric measurements.
3.8.19. Tricarbonyl[(1,2,3,4,5-p)-1-(2%-methoxyphenyl)-
4-methoxy-2,4-cyclohexadienyl]iron(1+)
hexafluorophosphate(1−) (4b, R=OMe)
54% overall yield via 8 and 9 by procedure A (67%
yield) followed by the TFA/NH4PF6 method (80%
yield), (see Scheme 3), or 29% overall yield via purified
8 by procedure G (72% yield) followed by the HPF6
method (58% yield). IR: wmax (cm−1) 2100, 2052, 2044
(CO). lH (acetone-d6) 2.98 (1H, d, J=15.5 Hz, 6a-H),
3.76 (1H, dd, J=15.5, 6.3 Hz, 6b-H), 3.93 (3H, s,
OMe), 4.05 (3H, s, OMe), 4.36 (1H, m, 5-H), 6.49 (1H,
d, J=6 Hz, 2-H), 7.11 (1H, t, J=7.6 Hz, Ar) 7.20 (1H,
d, J 8.3, Ar), 7.29 (1H, dd, J=6.3, 2.6, 3-H), 7.42 (1H,
t, J=7.6, Ar), 7.53 (1H, t, J=8.3, Ar). Found: C,
40.84; H, 2.97. C17H15O5FePF6 requires: C, 40.83; H,
3.02%.
References
[1] G.R. Stephenson, Advanced Asymmetric Synthesis, Blackie,
London, 1996, p. 313.
[2] For example, see Amaryllidaceae alkaloid structures surveyed by
S.F. Martin, in: A. Brossi (Ed.), The Alkaloids, Academic Press,
30 (1987) 251; O. Hoshino, in: G.A. Cordell (Ed.), The Alka-
loids, 51 (1998) 323.
[3] G.R. Stephenson, H. Finch, D.A. Owen, S. Swanson, Tetrahe-
dron 49 (1993) 5649.
[4] E.J. Sandoe, G.R. Stephenson, S. Swanson, Tetrahedron Lett. 37
(1996) 6283.

A.V. Malko6 et al. / Inorganica Chimica Acta 296 (1999) 139–149
149
[5] S.T. Astley, G.R. Stephenson, Synlett, (1992) 507. S.T. Astley,
M. Meyer, G.R. Stephenson, Tetrahedron Lett. 34 (1993)
2035.
[6] G.R. Stephenson, S.T. Astley, I.M. Palotai, P.W. Howard, D.A.
Owen, S. Williams, in: K.H. Dotz and R.W. Hoffmann (Eds.),
Organic Synthesis via Organometallics, Vieweg, Braunschweig,
Vol. 3, 1991, p. 169.
[9] H. Kondo, S. Ishiwata, Chem. Ber. 70B (1937) 2427. W.D.
William, D. Rogers, Proc. Chem. Soc. (1964) 357. M. Kihara, K.
Konishi, L. Xu, S. Kobayashi, Chem. Pharm. Bull. (1991) 1849.
[10] M. Ishizaki, K. Ozaki, A. Kanematsu, T. Isoda, O. Hoshino,
Chem. Pharm. Bull. 58 (1993) 3877. S.F. Martin, P.J. Garrison,
J. Org. Chem. 47 (1982) 1513. D.J. Ackland, J.T. Pinhey, J.
Chem. Soc., Perkin Trans. 1 (1987) 2695.
[7] L. Besson, M. le Bail, D. Aitken, H.P. Husson, F. Rose-Munch,
E. Rose, Tetrahedron Lett. 37 (1996) 3307. F. Rose-Munch, C.
Le Corre-Susanne, F. Balssa, E. Rose, J. Vaisserman, E.
Licandro, A. Papagni, S. Maiorana, W.D. Meng, G.R.
Stephenson, J. Organomet. Chem. 545–546 (1997) 9. C. Renard,
R. Valenic, F. Rose-Munch, E. Rose, J. Vaisserman,
Organometallics 17 (1998) 1587. F. Rose-Munch, V. Gagliardini,
C. Renard, E. Rose, Coord. Chem. Rev. 178–180 (1998) 249. E.
Rose, C. Le Corre-Susanne, F. Rose-Munch, C. Renard, V.
Gagliardini, F. Teldji, J. Vaisserman, Eur. J. Inorg. Chem.
(1999) 421.
[11] D.A. Owen, G.R. Stephenson, H. Finch, S. Swanson,
Tetrahedron Lett. 31 (1990) 3401.
[12] M. Brookhard, A.R. Pinhas, A. Lukacs, Organometallics 1
(1982) 1730.
[13] B.M.R. Bandara, A.J. Birch, T.-C. Khor, Tetrahedron Lett.
(1980) 3625.
[14] A.J. Birch, P.E. Cross, J. Lewis, D.A. White, S.B. Wild, J.
Chem. Soc. A (1968) 332.
[15] I.M. Palotai, PhD Thesis, University of East Anglia, 1989.
[16] M. Tinkl, A. McKillop, G.R. Stephenson, unpublished results.
[17] V. Snieckus, Chem. Rev. 90 (1990) 879.
[8] A.V. Malkov, G.R. Stephenson, J. Organomet. Chem. 489
(1995) C44.
[18] P.G. McDougal, J.G. Rico, Y.-I. Oh, B.D. Condon, J. Org.
Chem. 51 (1981) 3388.
.