Electroorganic reactions. Part 55. Quinodimethane chemistry. Part 3. Transition metal complexes as inter- and intra-molecular redox catalysts for the electrosynthesis of poly(p-xylylene) (PPX) polymers and oligomers

Article abstract of DOI:10.1039/b103274g

The role of metal complexes as redox mediators in the electrosynthesis of poly(p-xylylenes) (PPXs) has been explored, with a view to designing metal-containing precursors that can act both as mediators and starting materials for metal-containing polymers. A number of transition metal complexes [Cr(III), Ni(II) and Co(II)] are efficient redox catalysts for production of quinodimethanes, and hence PPXs. Following encouraging results from experiments using mediators based on anthranilic acid and salicylaldehyde ligands a macrocyclic compound was designed, and successfully prepared by a convergent route that incorporated both a 1,4-bis(chloromethylarene) function as a precursor to a quinodimethane and a Ni(II) salen unit as an intramolecular redox catalyst. The macrocycle was successfully reduced cathodically to yield a PPX polymer with bound Ni(II). Evidence is presented for the operation of intramolecular redox catalysis (homomediation).

Full text of DOI:10.1039/b103274g

View Article Online / Journal Homepage / Table of Contents for this issue
Electroorganic reactions. Part 55.† Quinodimethane chemistry.
Part 3. Transition metal complexes as inter- and intra-molecular
redox catalysts for the electrosynthesis of poly(p-xylylene) (PPX)
polymers and oligomers‡
2
Robert G. Janssen, James H. P. Utley,* Emmanuelle Carré, Evelyne Simon and
Heike Schirmer
Department of Chemistry, Queen Mary (University of London), Mile End Road, London,
UK E1 4NS
Received (in Cambridge, UK) 11th April 2001, Accepted 12th June 2001
First published as an Advance Article on the web 23rd July 2001
The role of metal complexes as redox mediators in the electrosynthesis of poly(p-xylylenes) (PPXs) has been
explored, with a view to designing metal-containing precursors that can act both as mediators and starting
materials for metal-containing polymers. A number of transition metal complexes [Cr(), Ni() and Co()] are
efficient redox catalysts for production of quinodimethanes, and hence PPXs. Following encouraging results from
experiments using mediators based on anthranilic acid and salicylaldehyde ligands a macrocyclic compound was
designed, and successfully prepared by a convergent route that incorporated both a 1,4-bis(chloromethylarene)
function as a precursor to a quinodimethane and a Ni() salen unit as an intramolecular redox catalyst. The
macrocycle was successfully reduced cathodically to yield a PPX polymer with bound Ni(). Evidence is presented
for the operation of intramolecular redox catalysis (homomediation).
the presence of palladium13,14 and cobalt11 complexes. An
Introduction
important distinction must be made between those reactions
that proceed by an inner sphere mechanism (e.g. oxidative addi-
tion to Ni0) and those in which redox catalysis involves homo-
geneous outer sphere electron transfer (e.g. probably from NiI).
Benzyl halides undergo15,16 oxidative addition to Ni0 complexes
and bibenzyl results from reduction of benzyl chloride in the
presence of NiBr2bipy (bipy = 2,2Ј-bipyridine)17 or NiCl2-
(PPh3)2.18 In the context of the present paper we have shown16
that 1,4-bis(chloromethyl)benzene is electrochemically reduced
to poly(p-xylylene) (PPX) in the presence of catalytic amounts
of NiIICl2L2 (L2 = dppe or dppp) [dppe = 1,2-bis(diphenylphos-
phino)ethane, dppp = 1,2-bis(diphenylphosphino)propane]. In
this case the balance of evidence was in support of an inner
sphere mechanism involving oxidative addition of the dihalide
to Ni0L2.
The electrochemical reduction of 1,4-bis(bromomethyl)-
benzene gives1,2 poly(p-xylylene) (PPX), presumed to arise via
2 F 1,6-elimination of two bromide anions with formation of
p-xylylene, the parent quinodimethane (QDM). Because the
conditions (room temperature, atmospheric pressure) are less
drastic than for pyrolytic generation the electrochemical route
has been extensively investigated for the formation3,4 of PPXs
and poly(p-phenylvinylenes) (PPVs). Furthermore, o-quino-
dimethanes, important intermediates for cycloaddition reac-
tions, may be generated in the presence of dienophiles, and thus
effect useful synthetic conversions.5,6 A limitation of this
electrochemical reduction route was that with few exceptions
bis(bromomethyl)arenes must be used, as the cheaper and less
lachrymatory chloro derivatives are more difficult to reduce and
in many cases give cathodic hydrogenolysis.
We present here the results of experiments in which we
explore the formation of QDMs, with subsequent polymerisa-
tion or oligomerisation, in electrolyses of 1,4-bis(halomethyl)-
arenes in the presence of relatively simple transition metal
complexes. We have also constructed a QDM precursor in
which the 1,4-bis(halomethyl)arene function and the complexed
transition metal (NiII) are combined in the same molecule.
The purpose is to demonstrate for practical purposes “intra-
molecular redox catalysis”, in our case to produce a metal-
containing polymer or oligomer. Intramolecular electron trans-
fer between donor and acceptor groups has a long history19
and recently a systematic study has explored20 conditions and
kinetics for intramolecular dissociative electron transfer in
rigid 1,4-disubstituted cyclohexanes acting as donor–spacer–
acceptor systems.
We have shown7 that quinodimethanes may be generated
electrochemically and characterised by cyclic voltammetry and,
in some cases, by NMR spectroscopy. The major follow-up
reactions are polymerisation to give PPXs, oligomerisation or,
for co-electrolysis, co-polymerisation. We also found7 that
generation of the QDMs, with subsequent oligomerisation
or polymerisation, could be effected by redox catalysis using
organic mediators.
Transition metal complexes have also been used as redox
mediators in organic electrosynthesis. In particular, nickel
complexes have been used for the electroreductive coupling of
aryl halides8,9 and the electropolymerisation10 of 1,4-dihalo-
benzenes to poly(p-phenylene). Coupling between benzal
dichloride (PhCHCl2) and stilbene has been catalysed11,12 by
the reduction of NiII(salen) [salen = bis(salicylidene)ethylene-
diamine]. Furthermore, carbon–carbon bond formation results
from electroreductions of a variety of alkyl and aryl halides in
Results and discussion
Strategy and substrates: cyclic voltammetry of some transition
metal complexes in the presence of 1,4-bis(halomethyl)arenes
† For Part 54 see: J. H. P. Utley, S. Ramesh, X. Salvatella, S. Szunerits,
M. Motevalli and M. F. Nielsen, J. Chem. Soc., Perkin Trans. 2, 2001,
153.
The first step was to establish whether simple transition metal
complexes were effective as external catalysts for the mediated
‡ Dedicated to the late Professor Lennart Eberson.
DOI: 10.1039/b103274g
J. Chem. Soc., Perkin Trans. 2, 2001, 1573–1584
This journal is © The Royal Society of Chemistry 2001
1573

View Article Online
Table 1 Cyclic voltammetry of metal complex mediators and QDM precursors
Complex
Working electrode
ϪEp(red)a/V vs. SCE
ϪEp(ox)a3/V vs. SCE
ϪE 0/V vs. SCE
Precursor
ϪEp/V vs. SCE
Co(salen)
Ni(salen)
Cr(acac)3
CrCl3ؒ6H2O
Hg bead
Hg bead
Hg bead and Au
Hg bead
Au
1.30
1.67
1.21
1.60
—
0.69
0.4
1.25
1.63
—
1a
1b
1c
1d
1.20e
1.34e
1.52e
1.96
1.84b/2.2c
1.10b/1.6c
0.78b/1.65c
1.88
0.89b
0.59b
salen
Au
1e
1f
2. 03
1.70
a ν = 200 mV sϪ1, DMF–Et4NBr or Bu4NBF4. b First reduction peak. c Second reduction peak. d ν = 200 mV sϪ1, Hg bead, DMF–Et4NBr. e First
reduction peak.
electrolysis of 1,4-bis(halomethyl)arenes with consequent con-
version into polymers or oligomers. The 1,4-bis(halomethyl)-
arenes involved in this initial phase are given as structures
1a–f. The reduction potentials for these substrates, together
with those of some common transition metal complexes, are
displayed in Table 1.
Scheme 1 Probable reduction sequence for C (acac)3.
is most likely that R1 corresponds to the one-electron reduc-
tion of CrIII(acac)3 with subsequent dissociation and R2 to
the reduction of CrII(acac)2. In the presence of increasing
aliquots of 1d, or of 1e, the peak current at R1 increases and
the peak for R2 disappears; under these conditions ip/v1/2 is
not independent of the scan rate v. Consequently it seems
possible that reduction of the precursors 1d, and 1e, is medi-
ated by reduction of CrIII(acac)3 and that homogeneous
electron transfer outruns dissociation of CrII(acac)3.
The effectiveness of NiII(salen) and of CoII(salen) as medi-
ators for the reduction of organic halides has been well demon-
strated.11,12,23 In agreement with this we find that, in DMF
solution, both complexes give reversible one-electron reduction
peaks (to NiI and CoI respectively) that become irreversible
with a concomitant increase in current in the presence of 1d
It is evident that the complexes are generally less easily
and 1e [for NiII(salen)] and 1c and 1f [for CoII(salen)]. In each
reduced than the 1,4-bis(bromomethyl)arenes (1a–f) and con-
case the normal independence of ip/v1/2 on scan rate v is lost. It
sequently cannot be expected to act as redox catalysts. However,
is noteworthy that we found no cyclic voltammetric, or prepar-
the corresponding dichlorides are reduced directly at potentials
ative, evidence for interaction between the complexes in their
more negative than those for the transition metal complexes,
reduced state and 1,3-bis(bromomethyl)benzene, the isomer
and CrCl3 is reduced slightly more easily than the dibromides
that cannot form a quinodimethane by reductive elimination.
1a and 1b. Reductive cleavage to QDMs is in these cases
possible in principle.
Preparative-scale electrolysis: “external” mediation by transition
For the cyclic voltammetry of those complexes listed in Table
metal complexes
1, CrCl3 gives21 a quasi-reversible one-electron first reduction
peak at Ep(red) = Ϫ1.10 V, corresponding to the CrIII/CrII
couple, followed by an irreversible two-electron peak corres-
ponding to the CrII/Cr0 system. The addition of compound 1d
(Ep = Ϫ1.96 V) has no effect on the first redox couple for CrCl3
(∆E = 0.86 V). The separation in reduction potentials is well
outside the range normally associated with redox catalysis
(∆E = 0.5 V). In other combinations (CrCl3 in the presence of
1a and 1b) the reduction peaks of the components are too close
properly to observe the effect on reversibility of the first wave
for CrCl3. However, preparative-scale reduction was carried
out, using CrCl3 as mediator, at the foot of the CrCl3 reduction
wave (see below). For the tetrabromide 1c ∆E = 0.42 V and in
this case cyclic voltammetry of CrCl3 in the presence of 1c
shows an irreversible first wave, of increased peak height,
consistent with redox catalysis.
The cyclic voltammetric experiments suggested that in several
cases quinodimethane formation, and hence complexes of CrIII,
NiII, and CoII, might usefully mediate polymerisation. The
results of preparative experiments based on the cyclic voltam-
metric observations are summarised in Table 2. In order to
minimise the possibility of direct rather than mediated reduc-
tion the co-electrolysis of metal complex and 1,4-bis(halo-
methyl)arene was carried out at a potential corresponding to
the foot of the first reduction wave of the (more easily reduced)
metal complex.
The co-electrolyses (under nitrogen) were preceded by pre-
electrolysis of the electrolyte and then prior reduction of the
catalyst, in each case to 1 F, and the cell current dropped to the
background value. Electrolysis was continued after addition of
the 1,4-bis(halomethyl)arene until passage of the charge neces-
sary to convert the whole amount of precursor into polymers, 2
F for PPXs and 4 F for PPVs. The concentration of the catalyst
was either ca. 0.017 M or 0.03 M and the mediator : substrate
ratio was between 1 and 0.1. For details see the Experimental
section and footnotes to Table 2.
It is reported22 that cyclic voltammetry of CrIII(acac)3
(acac = acetylacetonato) gives three cathodic waves with the
first one reversible only in the presence of a 40-fold excess of
Me4N(acac). It was proposed that dissociation is coupled to
electron transfer (Scheme 1).
We find that cyclic voltammetry of CrIII(acac)3, at 0.2 V sϪ1
gave two reduction peaks, at Ϫ1.84 M and Ϫ2.2 V (vs. SCE).
The first peak (R1) remained irreversible at increased scan
rates, but the second peak (R2) was reversible at 0.5 V sϪ1. It
For the “batch procedure” experiments with a high ratio of
mediator to substrate (>0.5) the solution of the reduced com-
plex changed colour when the substrate was introduced. This is
strong evidence that homogeneous electron transfer takes place.
1574
J. Chem. Soc., Perkin Trans. 2, 2001, 1573–1584

View Article Online
Table 2 Polymer formation by indirect electrolysisa using transition metal complex mediators
1,4-Bis(halomethyl)arene
∆Ep(red)/V
Mediator
CrCl3
Mediator : substrate
ϪE(red)b/V vs. Ag wire
Yieldc (% PPX)
1a
1b
1c
0.10
0.24
0.42
0.22
0.86
0.12
0.29
0.19
0.36
0.1
0.25
1.0
0.5
0.5
0.5
0.5
0.5
0.5
0.8
0.9
1.0
1.1
1.2
1.35
1.4
1.35
1.4
100
82
100(PPV)
Not determined
0
30
50
42
83c
Co(salen)
CrCl3
Cr(acac)3
Ni(salen)
Cr(acac)3
Ni(salen)
1d
1e
a Hg pool cathode, divided cell, DMF–Bu4NBF4 (0.1 M). b For the second part of the electrolysis; see Experimental section for the relationship
between Ag wire and SCE reference electrodes. c Not optimised.
Table 3 Mediation via an anthranilic acid complex
Components
ϪEp/V vs. SCEa
Commentb
Anthranilic acid (2-aminobenzoic acid)
CrCl3ؒ6H2O
Anthranilic acid (3 equiv.) ϩ CrCl3ؒ6H2O
Anthranilic acid (3 equiv.) ϩ CrCl3ؒ6H2O ϩ 1f (2.3 equiv.)
1.74
1.1
1.33
Ep(ox) = ϩ0.58 V
PPV formed in 76% yield
1.30; 1.1c
a Hg/Pt cathode, DMF–Bu4NBF4 (0.1 M). 0.2 V sϪ1 b See text for further explanation. c Two-stage electrolysis according to procedure described in
.
text.
In one case that was attempted, a good preparative result was
obtained with the mediator in ca. 10 mol%.
precedent,26,27 although we failed to isolate it. However, the
complex formed in solution nicely mediates quinodimethane
and hence polymer formation, as shown by the co-electrolysis
with 1f to give PPV (Table 3).
By analogy salicylaldehyde derivatives form complexes that
should mediate reductive cleavage in our systems; indeed bis-
(salicylaldehydato)nickel() dihydrate and the corresponding
Co() complexes have been much studied.28,29 The route in
Scheme 3 describes the attempted preparation of a possible
Electrogeneration of PPX polymers via QDMs formed by
intramolecular redox catalysis (homomediation)—proof of
concept
The incorporation in the same molecule of the quinodimethane
precursor functionality (the 1,4-bis(halomethyl)arene) and a
complexed transition metal offers the possibility for formation
of polymers/oligomers with included metals. A candidate was
the anthranilic acid derivative 2, and preparation was unsuc-
cessfully attempted by the route24,25 shown in Scheme 2, in the
Scheme 2 Attempted route to anthranilic acid complex as potential
mediator.
expectation that the nickel complex 3a would ultimately be
formed. A Cr() complex of anthranilic acid has been
reported.26,27 The likelihood that an anthranilic acid complex
would work, in this case involving Cr()/Cr() was indicated by
the experiments summarised in Table 3. Anthranilic acid is less
easily reduced than CrCl3ؒ6H2O in DMF; reduction of the
Cr() salt in the presence of anthranilic acid results in the dis-
appearance of the CrCl3ؒ6H2O peak and the appearance of two
new peaks, a reduction at Ϫ1.33 V and an oxidation at ϩ0.58 V.
We propose that the peak at Ϫ1.33 V is due to the in situ form-
ation of a Cr() complex analogous to 3b, based on literature
Scheme 3 Synthesis of the bis(dimethylsalicylaldehydato)nickel()
dihydrate mediator 4.
precursor of the type of complex required, but the high reac-
tivity to electrophilic substitution resulted in ring rather than
side-chain bromination in the final step. However, the nickel
complex 4, closely analogous to diaquabis(salicylalde-
hydato-O,OЈ)nickel,28 was prepared as orange–red crystals and
characterised by 1H NMR.
Cyclic voltammetric and preparative electrolysis results
J. Chem. Soc., Perkin Trans. 2, 2001, 1573–1584
1575

View Article Online
Table 4 Mediation via a salicylaldehyde complexa
Component
ϪEp or ϪEo/V vs.SCE
Comment
Salicylaldehyde
Bis(salicylaldehydato)cobalt() dihydrate
1.33
1.60; 1.91
1.76
Ered = Ϫ1.45
Ered = Ϫ1.45
Ered = Ϫ1.50
Ered = Ϫ1.50
Reversible
Irreversible
Reversible
61% corresponding PPX
75% corresponding PPX
70% corresponding PPX
75% corresponding PPX
4
4 (10 mol%)b ϩ 1d (coelectrolysis)
4 (20 mol%)b ϩ 1d (coelectrolysis)
4 (10 mol%)b ϩ 1e (coelectrolysis)
4 (20 mol%)b ϩ 1e (coelectrolysis)
a Hg/Pt or Hg pool cathode, DMF–Bu4NBF4 (0.1 M), CV at 0.2 V sϪ1 b The mediator was recovered in 70–98% yield.
.
Fig. 2 Cyclic voltammetric evidence for redox catalysis of reductive
cleavage of compound 26 (Scheme 7); Hg/Pt cathode, DMF–Et4NBr
(0.1 M), 0.1 V sϪ1; (a) Ni(salen), 0.9 mM, (b) as for (a) ϩ26, 1.1 mM,
(c) as for (a) ϩ26, 2.8 mM.
Fig. 1 Cyclic voltammetric evidence for redox catalysis in reduction of
catechol derivative 7 (Scheme 4); Au cathode, DMF–Et4NBr (0.1 M),
0.3 V sϪ1; (a) Ni(salen), 1.16 mM, (b) as for (a) ϩ7, 0.82 mM.
small amount (<50%) of a brown solid containing no C–Cl
groups (by FTIR) and giving broad signals in the H NMR
concerning salicylaldehyde complexes are summarised in Table
4. Essentially, the complexes are less easily reduced than salicyl-
aldehyde. Bis(salicylaldehydato)cobalt() dihydrate,29 in DMF,
is reduced irreversibly whereas the Ni() complex 4 reduces
reversibly at a potential suitable for mediation of reductive
cleavage of substrates 1d and 1e. This is confirmed by efficient
preparative conversions into the corresponding PPXs.
Although the analogue of 4, containing a 1,4-bis(bromomethyl)
function, could not be prepared, the results encouraged further
attempts at demonstrating “homomediation”.
1
spectrum. Examination by gel permeation chromatography
gave a bimodal distribution with high and low molecular weight
bands in the peak area ratio of 32 : 1. The broad high molecular
weight band gave Mn = 1.3 × 106 with a polydispersivity of 17.
Hence we can conclude that catechol-based polymers can be
formed by mediated electrolysis but that improvement of the
route to the precursors is necessary before more progress can be
made, including the incorporation of metals into the precursors
or into the polymers.
An alternative, but less straightforward approach is separ-
ately to design the ligand and the QDM precursor function and
to link them by an electrochemically inert bridge. Consequently
compound 23 (see Scheme 6) was chosen as a target, because
the 1,4-bis(chloromethyl)arene and the salen-like ligand could
be linked using well-tried macrocycle synthetic methodology.
The successful route is outlined in Schemes 5 and 6 and full
details are given in the Experimental section; during the final
stages of the sequence part of the product had complexed with
sodium, which was detected by electrospray mass spectrometry.
At an early stage it was necessary to establish that the chosen
metal complex (essentially Ni()salen) would indeed mediate
reductive cleavage of the 1,4-bis(chloromethyl)-3,4-dialkoxy-
arene unit. The cyclic voltammogram in Fig. 2 involving the
model compound 26 (Scheme 7) indicates clearly that the
expected redox catalysis is observed.
The electrochemical conversion of a catechol-based precur-
sor was also explored because these are known30 to complex
metal ions, especially Fe(). The compound 7 was prepared
according to Scheme 4 (see also Scheme 6) and examination of
Scheme 4 (see also Scheme 6). Route to model compound 7.
its electrochemistry was encouraging. Compound 7 shows
an irreversible reduction peak at Ϫ1.90 V (vs. SCE) [Pt/Hg
cathode, DMF–Et4NBr (0.1 M), 0.2 V sϪ1]. Co-electrolysis with
the more easily reduced Ni()salen shows evidence of redox
catalysis (Fig. 1) and two-stage preparative electrolysis (prior 1
F reduction of Ni()salen, addition of 7 and further 2 F reduc-
tion) gave, after work up and isolation by centrifugation, a
Realisation of intramolecular redox catalysis (homomediation)
The macrocyclic ligand 22 (and the corresponding sodium
complex) was readily converted by reaction with nickel acetate
in methanol into the corresponding nickel complex and then
into 23, which was well characterised by NMR and by electro-
spray mass spectrometry. Preparative electrolysis, at 0.013 M
1576
J. Chem. Soc., Perkin Trans. 2, 2001, 1573–1584

View Article Online
Fig. 3 Cyclic voltammograms during preparative electrolysis of
macrocycle 23 (Scheme 6); [0.013 M, Hg/Pt cathode, DMF–Et4NBr
(0.1 M), 0.5 V sϪ1], (a) before electrolysis, (b) after 2 F electrolysis.
Scheme 5 Assembly of the amine 16: a) TsCl, Et3N, DMAP, CH2Cl2,
rt, 18 h, 75% (9); b) K phthalimide, DMF, 100 ЊC, 8 h, 84% (10);
c) TsCl, Et3N, DMAP, CH2Cl2, rt, 18 h, 64% (11); d) NaH, allyl
bromide, DMSO, rt, 50% (13); e) (11), K2CO3, MeCN, reflux, 48 h, 83%
(14); f ) (CH2OH)2, TsOH, CH(OMe)3, toluene, reflux, 8 h (15);
g) N2H4, MeOH, reflux, 5 h, 79% (16).
substrate concentration, was monitored by cyclic voltammetry
and the outcome is summarised in Fig. 3. Initially a substantial
reduction current [Fig. 3, curve (a)] is observed with little
reverse oxidation, which is consistent with electrolysis involving
redox catalysis. At the conclusion of electrolysis (2 F) a residual
redox couple [Fig. 3, curve (b)] is observed that is due to the
nickel complex bound into the oligomeric/polymeric product.
[N.B., because of the high concentration used in the cyclic
voltammetric monitoring, it is not easy to compare peak cur-
rents, and hence diffusion coefficients, for Ni()salen (Fig. 2)
and the Ni() species in the macromolecule/oligomer.]
anthranilic group with Ni or Co prior to halogenation was
problematical. An alternative synthetic strategy was developed
to obtain a molecule, which has both the desired bis(chloro-
methyl) functionality as well as a chelated transition metal
atom, the latter acting as a mediator in the electroreduction.
Chloromethyl groups are incompatible with the nucleophilic
properties of most of the metal-chelating ligands of interest.
Direct chlorination (Cl2, NCS) is too harsh and likely to result
in competing ring-substitution, which immediately excludes
1,4-dimethyl groups as precursors to the chloromethyl func-
tions. The reduction of methoxycarbonyl groups was therefore
explored with, after the incorporation of the transition metal,
conversion of the resulting hydroxymethyl groups into chloro-
methyl groups using mild substitution methods. This sequence
was chosen to avoid complications from the reactions of prob-
ably labile chloromethyl transition metal complexes during
work up and purification. We have shown the salen metal-
chelating function to be a good mediator for bis(chloromethyl)-
arenes, while a host of different metals, e.g. Ni(), Co(), Zn(),
Cu(), are known to complex to give planar diamagnetic com-
plexes. The synthesis of macrocyclic salens has been described
in literature;31 if the salen- and the bis(chloromethylarene) unit
are incorporated in a common macrocycle, the resulting PPX
materials produced by electrolysis or coelectrolysis should be
capable of complexing a variety of metal cations.
The observed E0 value is the same as that for Ni()salen,
which is understandable as the Ni() environment in the macro-
cycle is very similar to that in Ni()salen. The product of
electrolysis, obtained in 85% yield, was characterised by 13C
NMR spectroscopy, FTIR, electrospray mass spectrometry and
atomic absorption spectroscopy. The last technique revealed
that the nickel content was 66% of the theoretical maximum
occupancy. The remaining sites could well be occupied by Naϩ,
which is prevalent in these systems and for which we did not
test.
We have chosen to depict the electrochemical conversion
of compound as intramolecular with homogeneous electron
transfer from Ni() being preferentially across the macrocycle
(homomediation). This is based on the proximity of the nickel
atom to the edge of the arene unit; MM2 energy minimisation
of the structure 23 gives a preferred conformation in which the
intramolecular separation of the nickel atom and arene unit is
ca. 12 Å. For comparison, for a recently studied20 case of
intramolecular dissociative electron transfer, the distance
between the arene (donor) groups (centre of the ring) and the
frangible acceptor group (C–Br) in the 4-aroyloxy-1-methyl-
cyclohexyl bromides is about 6 Å. In our system the size
and polarity of the macrocycle will result in a relatively low
diffusion coefficient that in principle should discriminate
against outer sphere electron transfer by bimolecular reaction.
Furthermore, intramolecular inner sphere electron transfer
is sterically inhibited and it is unlikely, also for steric reasons,
that two of these macrocycles could combine to effect
intermolecular inner sphere electron transfer.
Synthesis of macrocyclic salen 23
A convergent synthetic route was designed in which a 1,4-
disubstituted catechol would serve as a progenitor both to a
1,4-bis(chloromethylarene) intermediate and a symmetrical
linker for macrocycle construction. The overall route is set out
in Schemes 5 and 6. Catechol was bis-carboxylated to give
compound
5 using a Kolbe–Schmitt carboxylation pro-
cedure.32,33 The dimethyl ester 6 was obtained in 65% yield by
esterification of the crude acid in refluxing methanol using
sulfuric acid as a catalyst. Surprisingly, alkylation of both
phenolic hydroxy groups, using K2CO3 as a base and methyl,
ethyl and propyl bromides or iodides as electrophiles,
failed even after prolonged reflux. The lower reactivity of the
phenolate anion is probably due to steric hindrance by the
o-methoxycarbonyl groups. However, using more reactive
alkylating agents such as dimethyl sulfate and benzyl bromide
gave the corresponding dimethylated and dibenzylated
catechols 8a and 8b, respectively. Similarly both bis(cyano-
Synthesis of the macrocyclic complex 23 and its precursors
The multistage convergent synthesis of 23 requires separate
description and explanation. The key steps are given in outline,
and reagents and conditions summarised, in Schemes 5 and 6.
Direct halogenation of both methyl groups of either 3,6-
dimethylanthranilic acid or 3,6-dimethyl-2-hydroxybenzalde-
hyde into the corresponding chloro- or bromomethyl deriv-
atives failed (Schemes 2 and 3) and complexation of the
J. Chem. Soc., Perkin Trans. 2, 2001, 1573–1584
1577

View Article Online
Scheme 7 Preparation of model compound 26 and 28: a) CF3COOH,
rt, 3 h, 100% (17); b) 1) SOCl2, reflux, 1 h; 2) n-BuNH2, Et3N, DMAP,
CH2Cl2, 0 ЊC–rt, 84% (24); c) 1) LiAlH4, THF, 0 ЊC to rt, 2 h
(25); 2) Hϩ, 71%; d) SOCl2, 2,6-lutidine, rt, 1 h, quantitative (26);
e) 1) LiAlH4, THF, 0 ЊC to rt, 2 h; 2) Hϩ, quantitative (27); f ) SOCl2,
CH2Cl2, reflux, 1 h, quantitative (28).
bis(benzylated) diester 8b, gave confidence in the particular
synthetic route towards this new class of compounds.
The planned synthesis of macrocycle 23 (Scheme 6) involved
linking two formyl-substituted catechol units to the diacid 17
via tetraoxyethane spacers with subsequent reduction of the
methyl esters to hydroxymethyl groups. The closure to a macro-
cycle required the condensation of both o-hydroxyformyl
groups with an appropriate diamine to produce the desired
salen functionality. Using Ba2ϩ as a template ensured proximity
of the formyl groups. After metallation of the salen, the
hydroxymethyl groups could be converted into chloromethyl
groups, thus completing the synthesis.
The mono-tosylated glycol 9 was prepared by reaction
using a two-fold excess of tetraethylene glycol and DMAP as a
catalyst, and isolated in 75% yield by column chromatography.
The protection of the amine was achieved by substitution with
phthalimide in hot DMF to give 10 (84% yield). Subsequent
tosylation of the remaining hydroxy group, using a procedure
similar to that giving 9, gave 11.
2,3-Dihydroxybenzaldehyde (12) was the starting material for
synthesis of the intermediate amine 16 (Scheme 5). Selective
allylation of the more nucleophilic (and less acidic) o-hydroxy
group was achieved by double deprotonation, using 2.2 equiv.
of NaH in DMSO, and reaction with 1 equiv. of allyl brom-
ide.31 The unprotected 3-hydroxy group was combined with the
tosyl-phthalimide 11 in MeCN–K2CO3 (14 in 83% yield after
purification). The formyl group of 14 was protected as a dioxole
ether (15, quantitative yield) by acid-catalysed condensation
with glycol and efficient water removal (toluene–trimethyl
orthoformate). Deprotection by removal of the phthalimide
group with a methanolic solution of hydrazine was straight-
forward and gave the amine 16 (79%).
Synthesis of precursor 20 (Scheme 6) started from the tert-
butyl ester 8d following the methodology described earlier for
the preparation of 26; thus conversion of 16 into the bisamide
18 was achieved (74%). The methyl ester groups of 18 were
successfully reduced (LiAlH4–THF) and acidic work-up hydro-
lysed the dioxole groups to give 19 in almost quantitative yield.
Finally, palladium catalysed reductive deallylation (in situ
palladium tetrakis(triphenylphosphine)) gave 20 (62%). The
progress of this reaction was monitored carefully by TLC, as
both purity and yield are adversely affected by prolonged
reaction times.
Scheme 6 Assembly of the proteced catechol and convergence to the
macrocycle 23: a) 1) NaOH, MeOH, rt; 2) 110 ЊC, 3 days; 3) CO2, 70
bar, 200 ЊC, 2 days, 40% (5); b) MeOH, H2SO4, reflux, 8 h, 65% (6);
c) K2CO3, MeCN, RX, reflux, 18–48 h, 70–95% (8a–d); d) CF3COOH,
rt, 3 h, 100% (17); e) 1) SOCl2, reflux, 1 h; 2) (16), Et3N, DMAP,
CH2Cl2, 0 ЊC–rt, 74% (18); f ) 1) LiAlH4, THF, 0 ЊC to rt, 2 h; 2) Hϩ,
99% (19); g) Pd(OAc)2, P(C6H6)3, Et3N, HCOOH, EtOH–H2O, reflux,
1.5 h, 62% (20); h) ( )-trans-cyclohexane-1,2-diamine, Ba(OTf )2, THF,
reflux, 50% (21); i) Ni(OAc)2, MeOH, 1 min, 95% (22); j) MsCl, Et3N,
CHCl3, 55 ЊC, 6 h, 95% (23).
methyl) and bis(tert-butoxycarbonylmethyl) derivatives (8c and
8d) could be prepared in good yields (70–95%). Hydrogenation
of the cyano groups of 8c failed to give aminomethyl com-
pounds, probably due to partial amidation of the methyl esters.
However, quantitative selective hydrolysis of the tert-butyl
esters of 8d, in the presence of methyl esters, was achieved in
anhydrous TFA at room temperature, yielding the diacid 17,
which proved to be a useful intermediate. The proposed syn-
thetic procedure was partly tested by the prior synthesis
of model compound 26 (Scheme 7), which represents the bis-
(chloromethyl) part of the target molecule. Diacid 17 was
reacted with SOCl2, to give the corresponding di-acid chloride.
Subsequent amidation with n-butylamine gave 24 in 84%
yield, using chilled CH2Cl2 as a solvent and triethylamine–
DMAP as a base–catalyst system. Selective reduction of the
more reactive methyl esters in 24 in the presence of amide
groups was achieved with LiAlH4 at room temperature. After
work-up the bis(hydroxymethylated) diamide 25 was obtained
in 71% yield. The hydroxymethyl groups of 25 were converted
into chloromethyl groups by reaction with SOCl2–2,6-lutidine.
Alternative chlorination methods were tested with similar
results. These results and the similar preparation of bis-
(chloromethylated) compound 28 (Scheme 7), starting from
Ring closure of 20 was achieved by first adding Ba(OTf )2 to
a dilute, refluxing solution of 20 followed by addition of trans-
cyclohexane-1,2-diamine. Purification by column chromato-
graphy gave two major fractions in a total yield of 50%.
1578
J. Chem. Soc., Perkin Trans. 2, 2001, 1573–1584

View Article Online
1H NMR spectroscopy and mass spectrometry indicated that
one fraction was the free ligand (21) whereas the other was a
sodium complex (see below). The macrocycle 21 was reacted
with nickel() acetate in methanol to give 22 and the synthesis
completed by conversion of the hydroxymethyl groups of com-
pound 22 into chloromethyl groups to give 23 (Scheme 6).
Alternative procedures using SOCl2 failed due to partial
hydrolysis of the nickel salen unit. As anticipated, the macro-
cyclic salen complex 23 could not be purified by column
chromatography.
potential coulometry DMF (Aldrich, HPLC grade) was used
and stored over freshly-baked (150 ЊC) 4 Å molecular sieves, as
was acetonitrile. Supporting electrolytes (Aldrich) were used
without further purification. Dry CH2Cl2 was distilled from
P2O5. THF was distilled from K-metal with added benzo-
phenone as an indicator. All other solvents were analytically
pure and were used without further purification. Petroleum
ether refers to the fraction with bp 60–80 ЊC. All reactions were
carried out under a N2-atmosphere.
Melting points were determined on an Electrothermal capil-
lary tube melting point apparatus and were uncorrected. Infra
red spectra were recorded on KBr discs using a Perkin-Elmer
Characterisation of the macrocycles
1600 series FTIR spectrophotometer. 1H NMR spectra and 13
C
1
The free ligand, 21, exhibits a complex H NMR spectrum.
NMR spectra (63 MHz) were measured on a Bruker AM 250
(250 MHz) spectrometer. Solid-state 13C NMR spectra were
recorded at 75 MHz on a Bruker MSL-300 multinuclear NMR
spectrometer of the University of London Intercollegiate
Research Service (ULIRS). EI mass spectra and direct pyrolysis
mass spectra were measured using a Kratos MS50RF/Kratos-
DS90 data system; FAB mass spectra were obtained on the
same spectrometer using m-nitrobenzyl alcohol or glycerol as a
matrix.
Multiple signals are present for both the hydroxymethyl groups
and the methylene protons of the carbamoylmethoxy units. The
resonance at δ = 13.5 ppm (phenolic OH group of the salen
unit) indicates strong hydrogen bonding to the imine nitrogen.
Two main fractions were isolated from the chromatographic
purification of 21. Both fractions gave a clear electrospray mass
spectrum, each showing molecular ion signals at m/z 955.5 for
the protonated species (calculated 954.5) and at 977.4, for
the sodium complex. Reacting each fraction separately or
alternatively reacting the mixture of both fractions of 21 with
Preparative-scale column chromatography separations used
Merck silica gel 60H (230–400 mesh) and pre-coated silica gel
plates (Merck, 60 F254) were used for analytical TLC.
1
nickel() acetate gave a product with a greatly simplified H
NMR spectrum corresponding to that of a single compound
with high symmetry. Hence, the two fractions of 21 are the
empty macrocycle and its sodium complex. The complexity of
the 1H NMR spectrum of 21 is most likely governed by hydro-
gen bonding involving interaction between the hydroxymethyl
groups or the amide groups and the phenolic oxygens. This
hydrogen bonding is effectively inhibited upon complexation
with Ni(). The chloromethylated compound 23 was character-
ised by 13C NMR spectroscopy and electrospray mass spec-
troscopy. The latter showed the molecular ion pattern of the
empty macrocycle at m/z 1046 and its sodium complex at 1069.
The observed multiple signals are due to the isotope distribu-
tion of nickel, chlorine and sodium and excellently matched the
calculated distribution. The PPX type polymer obtained after
electrolysis of 23 gave broad signals in the 1H NMR. The pres-
ence of highly symmetrical macrocyclic units, however, could
be deduced. Similar results were obtained from the 13C NMR
spectrum, which clearly showed the presence of all major sig-
nals, with the exception of those for the chloromethyl groups.
The methylene signal at δ = 30.4 and the methyl signal at
δ = 16.6 clearly show the formation of oligomeric material.
From quantitative 13C NMR spectroscopy, using an inverse
gated pulse sequence with a delay time of 20 s, an average
oligomerisation degree of 2 to 5 was calculated. Electrospray
mass spectrometry confirmed the presence of monomer and
its sodium complex at m/z 979 and 1001, respectively. A weak
dimer signal was observed at 1980 (Na complex) and a complex
signal pattern between 480 and 580. Firm conclusions with
respect to the degree of oligomerisation cannot be obtained
from electrospray mass spectrometry as the polymer forms a
poly-sodium complex, whose signal position depends on the
m/z ratio. As the sodium charge of the oligomers is not well
defined, oligomers with, for example, a single Na-ion per
macrocyclic unit will show a signal at around m/z 1000, those
with two sodium ions per macrocyclic unit will exhibit a signal
at around 512, while a dimer with three Na-ions has a signal at
675. However, the presence of a complex signal pattern between
480 and 580 indicates the material to be of a poly-sodium
loaded oligomeric nature.
HPLC experiments were carried out using a Hewlett Packard
HP1100 Series Liquid Chromatograph with UV detection at
270 nm. The instrument was equipped with a Zorbax Sil
column (5 µm, 4.6 mm id × 250 mm) and guard, run isocratic-
ally with EtOAc–hexane (30 : 70%, Merck HPLC grade) eluent
at 1.0 ml minϪ1 and an injection volume of 10 µl.
Gel permeation chromatography: a Hewlett Packard HP1100
series liquid chromatograph, operated by HP ChemStation
software, was employed with a quaternary gradient pump and
UV (270 nm) detector. Polymer molecular weights were calcu-
lated using a Polymer Laboratories PL caliber GPC system,
version 4.01. Polymer Laboratories PL-gel [poly(styrene–
divinylbenzene) copolymer gel] mixed C (5 µm), two 300 × 7.5
mm and guard columns, enclosed in a constant temperature
oven, were used. A calibration curve was constructed using
narrow standard poly(styrene) (molecular weight range 500 to
3 × 106). The mobile phase was DMF (Aldrich, HPLC grade)
containing 0.1% LiBr at 73 ЊC and at the elution rate of 1.0 ml
minϪ1
Atomic absorption measurements: about 12 mg of the metal-
containing polymer sample were accurately weighed and dis-
solved in 10 ml of concentrated nitric acid (AR, 69%). The
extract was diluted to 100 ml using distilled water, to give a final
acid concentration of about 10% v/v. A blank was also pre-
pared using the same reagents. Immediately prior to analysis
the samples were filtered through a Whatman 541 filter paper to
remove particulate carbon. The metals were determined by
atomic absorption spectroscopy using a Unicam 939 AAS with
an air–acetylene flame. Standards were prepared by diluting
Spectrosol (BDH) 1000 ppm standard solutions of nickel with
subsequent dilution. Nickel was measured at 232.0 nm with
background correction. No nickel was found in the blank
samples.
Electrochemical experiments
Cyclic voltammetric experiments were performed using either
a Princeton Applied Research (PAR) VersaStat or Model263A
potentiostat with controlling PAR software (Model 270/250
Research Electrochemistry Software v 4.00). Glass cells for
cyclic voltammetry were undivided and equipped with an Au
disc (1 mm) or a mercury-coated platinum disk (0.1–0.6 mm)
working electrode (cathode), Ag-wire reference electrode and
platinum coil counter electrode (anode). The experiments were
typically carried out in DMF–Et4NBr (0.1 M) solution. The
Experimental
Materials and general procedures
For preparative electrolyses DMF (Aldrich) was used without
further purification. For cyclic voltammetry and controlled
J. Chem. Soc., Perkin Trans. 2, 2001, 1573–1584
1579

View Article Online
concentration of the substrate was 1–3 mM and the scan rate
varied from 0.1–200 V sϪ1. Because the potential of the Ag wire
reference electrode can vary over a relatively short timescale, the
E0 value for anthracene was normally measured at the begin-
ning and at the end of a series of measurements, against the Ag
wire electrode, so that reduction potentials could be consist-
ently referred to the SCE electrode (E0 for anthracene = Ϫ1.92
V vs. SCE in DMF34).
Scheme
8
Numbering of aromatic carbon atoms in catechol
derivatives.
Preparative scale electrochemical reductions were performed
using conventional glass cells, with the anode and cathode
compartments separated by a glass sinter. The cells were
equipped with a mercury pool working electrode (cathode), a
Ag-wire reference electrode and a carbon rod counter electrode
(anode). Reduction potentials for controlled potential elec-
trolyses were determined by cyclic voltammetry on the solu-
tions immediately prior to electrolysis, i.e. at greater than ideal
concentrations. Consequently the peak potentials so deter-
mined, against the relatively unstable Ag wire reference, serve to
enable reduction at the first reduction wave in an electrolysis
immediately following measurement. However, actual values
fluctuate from day-to-day and cannot be compared with peak
potentials measured at low concentration with reference to
SCE (see above). The reactions were kept an under an inert
atmosphere by the slow bubbling of dry nitrogen through the
electrolytes. The catholyte was mechanically stirred.
143 ЊC (petroleum ether); 1H NMR δ 7.8 (br s, 1H, NH), 7.35 (s,
2H, ArH), 2.65 (s, 6H, CH3); 13C NMR δ 169.1 (s, CO), 136.4,
135.7, 129.6, 17.4 (q, CH3).
2-Amino-3,6-dimethylbenzoic acid. A mixture of calcium
hypochlorite (800 mg), potassium carbonate (570 mg) and
KOH (160 mg) in approx. 15 ml water was shaken for 5 min,
filtered and additional KOH (1.8 g) was added. To the resulting
solution 3,6-dimethylphthalimide (1.0 g, 5.7 mmol) was added.
After 30 min stirring at 50 ЊC the solution was carefully acidi-
fied with acetic acid to ensure that the pH did not fall below 4.
The solution was extracted with diethyl ether (3 × 100 ml). The
combined organic layers were dried over Na2SO4 and evapor-
ated to give a yellow solid, which was recrystallised from EtOH–
H2O to obtain off-white crystals. Yield: 650 mg (69%); mp
112 ЊC (EtOH–H2O); 1H NMR δ (DMSO-d6) 7.02, 6.44 (d, 1H,
J = 6 Hz, ArH), 3.73 (br s, 2H, NH), 2.45, 2.15 (s, 3H, CH3).
Controlled potential electrolysis: in a divided cell at an Hg
pool cathode the electrolyte was pre-electrolysed at the poten-
tial to be used. The mediator was then electrolysed until con-
sumption of the required amount of charge (1 F) occurred and
the cell current dropped to the background value. The substrate
was introduced and a second electrolysis carried out as far as it
was necessary to convert the whole amount of precursor into
polymers (2 F for PPXs and 4 F for PPVs). The mediator
concentration was in the range 0.017 to 0.03 M and the concen-
tration of the starting materials was between 1 and 10 equiv-
alents in relation to the mediator concentration. The working
electrode, a mercury pool (5 or 10 cm2), was continuously
stirred not only to improve mass transport but also to preclude
the covering of the electrode by polymer. Insoluble polymer
product was removed by filtration, washed with water and
dried. Dilution of the catholyte with weak aqueous HCl caused
precipitation of DMF-soluble product.
3,6-Dimethylsalicylaldehyde. 2,5-Dimethylphenol (12.2 g, 0.1
mol) was dissolved, under a nitrogen atmosphere, in anhydrous
toluene (30 ml). SnCl4 (2.6 g, 0.01 mol) and tributylamine
(6.9 ml, 0.01 mol) were added. The resulting yellow solution
was stirred for 20 min and paraformaldehyde (6.6 g, 0.22 mol)
added. The yellow slurry was heated to 100 ЊC and kept at that
temperature for 8 h. The resulting brown solution was added to
water (500 ml) and acidified to pH = 2 with hydrochloric acid.
The layers were separated and the aqueous layer extracted with
ether. The combined organic layers were washed with brine,
dried and the solvent evaporated. The crude product was steam
distilled to give a yellow oil that was recrystallised from EtOH
to give light yellow crystals. Yield: 10.2 g (68%); mp 62 ЊC
1
(EtOH); H NMR δ 12.17 (s, 1H, OH), 10.32 (s, 1H, CHO),
7.25, 6.72 (d, 1H, J = 8 Hz, ArH), 2.56, 2.22 (s, 3H, CH3).
1
Polymers were characterised by FTIR or H NMR spectro-
Bis(3,6-dimethylsalicylaldehydato)nickel(II) mediator (4).
[Scheme 3] Nickel() chloride (795 mg, 3.3 mmol) was dissolved
in water (10 ml) and 3,6-dimethylsalicylaldehyde (1 g, 6.6
mmol) was added with vigorous stirring. Ethanol was gradually
added (maximum 3 ml), until the reactants had dissolved.
Aqueous ammonia (0.88 ml) was added until the solution
turned from light green to deep orange, then more ammonia
(2 ml) was added. The mixture was kept at 50 ЊC for 1 h. The
deep orange product (88% yield) was removed by filtration,
washed (water) and dried in a vacuum oven. An aqueous ethan-
olic solution of the product (ca. 10 ml water, 2 ml ethanol) was
heated on a steam bath. The solution was concentrated to about
half volume and the complex 4 precipitated as a dark orange
solid in near quantitative yield. 1H NMR δ 8.3 (s, 1H, OH), 8.08
(s, 1H, CHO), 6.9, 6.25 (d, 1H, J = 8 Hz, ArH), 2.50, 2.25
(s, 3H, CH3); MS (EI) m/z 359 (Mϩ, 359).
scopy (when soluble in organic solvents). Characteristic absorp-
tions are well documented:3,35 PPXs (ν/cmϪ1): 3044 (aromatic
C–H), 2920 (sat. C–H), 1510 (benzene ring) and 821 (two
adjacent H on benzene ring); PPVs (ν/cmϪ1): 3013 (aromatic
C–H), 1595 and 1510 (benzene ring), 961 (trans-H–C᎐C–H)
᎐
and 825 (two adjacent H on benzene ring).
Starting materials
Compounds 5,32,33 6,32,33 8a, 8b, 1,4-bis(hydroxymethyl)-2,3-
dihydroxybenzene and 9 were prepared according to slightly
modified literature procedures, while compound 13 was pre-
pared according to a well established literature procedure.31
The numbering of carbon atoms for the catechol derivatives
described below is given in Scheme 8.
Attempted preparation of 3,6-bis(bromomethyl)-2-aminobenzoic
acid (2) [Scheme 2]
2,3-Dihydroxyterephthalic acid (5). Catechol (100 g, 0.90 mol)
was added portionwise to a methanolic NaOH (73 g, 1.82 mol)
solution (500 cm3). The deep green solution was stirred at room
temperature for 18 h and evaporated to dryness. The remaining
solid was mortared and dried for three days in a vacuum oven
over P2O5 at a pressure of 10 mm Hg and a temperature of
110 ЊC. The tan powder was divided into two portions, each of
which was heated for two days in an autoclave under a CO2
pressure of 70 bar and at a temperature of 200 ЊC. The resulting
greyish powder was stirred in a hot HCl solution (6 M, 2 l) upon
which the crude acid precipitated. After cooling to room
3,6-Dimethylphthalimide. Maleimide (3.0 g, 31 mmol) in
CH2Cl2 (80 ml) was reacted with redistilled 2,5-dimethylfuran
(4 ml, 33 mmol). After stirring at room temperature for 30 min
BF3ؒEt2O (4.3 ml, 35 mmol) was added and the colour of the
solution turned dark orange. After 12 h stirring at room tem-
perature the red–brown slurry was neutralised and the organic
layer separated and dried over MgSO4. The solvent was reduced
to approx. 10 ml and the product precipitated by addition of
petroleum ether to give a yellow solid. Yield: 2.1 g (40%); mp
1580
J. Chem. Soc., Perkin Trans. 2, 2001, 1573–1584

View Article Online
temperature, the precipitate was filtered, washed with three
portions of water (200 ml) and triturated with a boiling HCl
solution (1 M, 2 l). After cooling the solution overnight, the
precipitated acid was filtered, washed with water (2 × 200 ml)
and dried in a freeze dryer. The brown coloured product was
used without further purification. Yield: 40%.
(s, ArCO), 130.2 (s, ArC), 126.0 (d, ArC), 81.8 (s, C(CH3)3),
71.1 (t, OCH2), 52.5 (q, OMe), 28.1 (q, C(CH3)3); MS (EI)
m/z 454.2 (Mϩ, 454.2); elemental analysis, calcd C, 58.14; H,
6.65; found C, 58.28; H, 6.65%.
Tetraethylene glycol monotosylate (9). To an ice-cooled solu-
tion of tetraethylene glycol (400 g, 354 ml, 2.06 mol), triethyl-
amine (46 g, 62 ml, 0.44 mol) and DMAP (1.0 g) in CH2Cl2 (1 l)
was added portionwise tosyl chloride (76 g, 0.40 mol). The
solution was stirred for 8 h at the same temperature and was
then allowed to warm-up to room temperature. After another
10 h stirring, the solution was washed with 1 M HCl solution
(2 × 200 ml), water (3 × 300 ml), brine (1 × 100 ml) and dried
over Na2SO4. Evaporation of the solvent gave a clear oil, which,
according to 1H NMR spectroscopy, consisted of approx-
imately 88% of the monotosylated derivative and 12% of the
bistosylated derivative. Crude yield: 132 g. Pure monotosylated
polyether 9 was obtained by column chromatography of 40 g of
crude material (silica gel, gradient elution, EtOAc– petroleum
Dimethyl 2,3-dihydroxyterephthalate (6). To a solution of
diacid 5 (15.0 g, 75.7 mmol) in MeOH (400 ml) was added 45 ml
of concentrated H2SO4 (CAUTION). The black solution was
refluxed for 8 h, cooled to room temperature and kept at
Ϫ30 ЊC for 12 h. The black crystalline precipitate was filtered
on a sintered glass filter, rinsed with MeOH and dried in vacuo.
The crude product was used without further purification. Yield:
64%.
General procedure A for the alkylation of dimethyl 2,3-dihydroxy-
terephthalate (6)
To a suspension of crude diester 6 and three mol equiv. of
K2CO3 in dry acetonitrile (50 ml gϪ1) was added the appropriate
alkyl halide (2–3 mol equiv.). The mixture was refluxed for
18–48 h. The solvent was evaporated and the residue was taken
up in EtOAc. The organic layer was washed with 1 M HCl
(1 × 100 ml), brine (1 × 50 ml) and dried over Na2SO4. The crude
products were purified either by column chromatography (silica
gel; mixtures of EtOAc–petroleum ether) or by recrystallisation.
1
ether 7 : 3 to neat EtOAc). Yield: 28 g (80% recovered). H
NMR δ 7.77, 7.31 (d, J = 8.3 Hz, 2H, TsH), 4.13 (m, 2H,
OCH2), 3.7–3.5 (m, 14H, OCH2), 2.42 (s, 3H, CH3); 13C NMR
δ 144.8, 132.9 (s, ArC), 129.8, 127.7 (d, ArC), 72.4, 70.4, 70.2,
70.1, 69.4, 68.4, 61.3 (t, CH2O), 21.4 (q, CH3).
N-(11-Hydroxy-3,6,9-trioxaundecyl)phthalimide (10). A sus-
pension of tetraethylene glycol monotosylate 9 (27 g, 77.5
mmol) and potassium phthalimide (14.8 g, 80 mmol) in dry
DMF (250 ml) was heated at 100 ЊC for 8 h. The clear solution
was evaporated to dryness and the residue was taken up in
EtOAc (500 ml). The organic layer was washed with 1 M HCl
solution (2 × 100 ml), brine (1 × 100 ml) and dried over
Na2SO4. The crude product was purified by column chromato-
graphy (silica gel, gradient elution, EtOAc–petroleum ether
7 : 3 to neat EtOAc) to obtain a yellow, clear oil. Yield: 21.2 g
Dimethyl 2,3-dimethoxyterephthalate (8a). 8a was prepared
from diester 6 (500 mg, 2.2 mmol), K2CO3 (900 mg, 6.6 mmol)
and Me2SO4 (694 mg, 5.5 mmol). After 48 h of reflux and work-
up, the crude product was purified by column chromatography
(silica gel; EtOAc–petroleum ether 2 : 3). A light-yellow clear
oil was obtained. Yield: 448 mg (80%). 1H NMR δ 7.50 (s, 2H,
ArH), 3.95, 3.93 (s, 2 × 6H, COOMe, OMe).
1
(84%). H NMR δ 7.9–7.7 (m, 4H, PhthH), 4.2–3.5 (m, 16H,
CH2O, CH2N); MS (EI) m/z 324.1 (M ϩ Hϩ, 324.1).
Dimethyl 2,3-bis(benzyloxy)terephthalate (8b). 8b was pre-
pared from diester 6 (2.0 g, 8.84 mmol), K2CO3 (3.6 g, 26.4
mmol) and benzyl bromide (4.5 g, 26.4 mmol). After 18 h of
reflux and work-up, the crude product was decanted with hot
petroleum ether to remove a black tar. Upon filtering and cool-
ing the clear solution, pure 8b crystallised as a white powder.
Yield: 2.24 g (62%). 1H NMR δ 7.54 (s, 2H, ArH), 7.4–7.3 (m,
10H, BnH), 5.12 (s, 4H, CH2), 3.86 (s, 6H, COOMe); 13C NMR
N-[11-(Methylbenzene-4-sulfonyloxy)-3,6,9,-trioxaundecyl]-
phthalimide (11). To a solution of phthalimide 10 (20.7 g, 64
mmol), triethylamine (10 ml, 67 mmol) and DMAP (0.5 g) in
CH2Cl2 (350 ml) was added portionwise tosyl chloride (12.8 g,
67 mol). The solution was stirred for 6 h at room temperature
(check by TLC) and then washed with 1 M HCl solution
(1 × 100 ml), brine (1 × 50 ml) and dried over Na2SO4. The
crude product was purified by column chromatography (silica
gel, EtOAc–petroleum ether 1 : 1) to obtain a yellow, clear oil.
δ 165.9 (s, C᎐O), 152.8 (s, ArCO), 136.8 (s, BnC), 130.4 (s,
᎐
ArC), 128.8, 128.5, 128.3 (d, BnCH), 125.7 (d, ArC), 76.6 (t,
CH2), 52.4 (q, OMe).
1
Yield: 19.6 g (64%). H NMR δ 7.9–7.7 (m, 6H, TsH, PhthH),
Dimethyl 2,3-bis(cyanomethoxy)terephthalate (8c). 8c was
prepared from diester 6 (500 mg, 2.2 mmol), K2CO3 (900 mg,
6.6 mmol), chloroacetonitrile (1.0 g, 13.3 mmol) and a spatula
point of NaI. After 24 h of reflux and work-up, the crude prod-
uct was purified by column chromatography (silica gel; EtOAc–
petroleum ether 2 : 3) followed by recrystallisation from MeOH
at Ϫ30 ЊC. A greyish crystalline product was obtained. Yield:
380 mg (57%); mp 152–154 ЊC (MeOH); 1H NMR δ 7.76 (s, 2H,
ArH), 4.89 (s, 4H, CH2CN), 3.98 (s, 6H, OMe); elemental
analysis, calcd C, 55.27; H, 3.98; N, 9.20; found C, 53.98; H,
3.73; N, 8.89%.
7.35–7.3 (m, 2H, TsH), 4.2–3.5 (m, 16H, CH2O, CH2N), 2.44 (s,
3H, CH3); MS (EI) m/z 477.1 (Mϩ, 477.2).
N-[11-(2-Allyloxy-3-formylphenoxy)-3,6,9-trioxaundecyl]-
phthalimide (14). A suspension of 2-allyloxy-3-hydroxybenz-
aldehyde 13 (6.6 g, 37 mmol), monotosylate 11 (17.7 g, 37
mmol) and anhydrous K2CO3 (8.3 g, 60 mmol) in dry
acetonitrile (250 ml) was refluxed for 48 h. The mixture was
evaporated to dryness and the residue was taken up in EtOAc
(250 ml). The organic layer was washed with 1 M HCl (1 × 100
ml), brine (1 × 50 ml) and dried over Na2SO4. The crude prod-
uct was purified by column chromatography (silica gel, gradient
elution, EtOAc–petroleum ether 1 : 1 to 7 : 3) to obtain a
Dimethyl
2,3-bis[(1,1-dimethylethyl)oxycarbonylmethoxy]-
1
terephthalate (8d). 8d was prepared from diester 6 (11.0 g, 48.6
mmol), K2CO3 (20.2 g, 146 mmol) and tert-butyl bromoacetate
(28.4 g, 146 mmol). After 24 h of reflux and work-up, the crude
product was decanted with hot petroleum ether to remove a
black tar. Upon filtering and cooling the clear solution to
Ϫ30 ЊC, pure 8d crystallised as pale yellow crystals. Yield: 15.5 g
(70%); mp 54–55 ЊC (petroleum ether); 1C NMR δ 7.55 (s, 2H,
ArH), 4.63 (s, 4H, CH2), 3.91 (s, 6H, COOMe), 1.48 (s, 18H,
C(CH3)3); 13C NMR δ 167.6 (s, CH C᎐O), 165.5 (s, C᎐O), 151.8
yellow, clear oil. Yield: 14.9 g (83%). H NMR δ 10.43 (s, 1H,
CHO), 7.9–7.7 (m, 4H, PhthH), 7.41 (dd, J1 = 7.6 Hz, J2 = 1.9
Hz, ArH), 7.17–7.08 (m, 2H, ArH), 6.2–6.0 (m, 1H, ᎐CH), 5.4–
᎐
5.2 (m, 2H, ᎐CH ), 4.7–4.65 (m, 2H, ArOCH ), 4.2–4.15, 3.9–
᎐
2
2
3.6 (m, 16H, CH2O, CH2N); MS (EI) m/z 483.2 (Mϩ, 483.2).
N-[11-(2-Allyloxy-3-(4,5-dihydro-1,3-dioxol-2-yl)phenoxy)-
3,6,9-trioxaundecyl]phthalimide (15). A solution of phthalimide
14 (14.5 g, 30 mmol), ethylene glycol (1.96 g, 31.5 mmol),
᎐
᎐
2
J. Chem. Soc., Perkin Trans. 2, 2001, 1573–1584
1581

View Article Online
trimethyl orthoformate (6.4 g, 60 mmol) and a catalytic amount
of toluene-p-sulfonic acid in dry toluene (150 ml) was refluxed
for 8 h (check by TLC). After cooling the solution to room
temperature, it was washed with saturated NaHCO3 solution
(1 × 50 ml), brine (1 × 50 ml) and dried over Na2SO4. After
evaporation of the solvent, a yellow oil was obtained. Yield:
15.88 g (quantitative). 1H NMR δ 7.9–7.7 (m, 4H, PhthH), 7.2–
(74%); 1H NMR δ 7.75–7.6 (s ϩ br t, 4H, ArH, CONH), 7.13,
6.89 (dd, J1 = 7.9 Hz, J2 = 1.7 Hz, 2H, ArH), 7.03 (t, J1 = 8.0 Hz,
2H, ArH), 6.2–6.0 (m, 2H, ᎐CH), 5.64 (s, 1H, CHOO), 5.4–5.2
᎐
(m, 2H, ᎐CH ), 4.55–4.5 (s ϩ m, 8H, ArOCH CH᎐), 4.13, 3.83
᎐
᎐
2
2
(t, J1 = 4.6 Hz, 4H, OCH2), 3.92 (s, 6H, OCH3), 3.75–3.5 (m,
24H, CH2O, CH2NH), 3.36 (s, 4H, OCH2CH2O); 13C NMR
δ 167.9 (s, CH C᎐O), 164.6 (s, C᎐O), 151.7 (s, 2 × ArCO), 146.4
᎐
᎐
2
6.9 (m, 3H, ArH), 5.65 (s, 1H, CHOO), 6.2–6.0 (m, 1H, ᎐CH),
(ArCO), 134.5 (d, ᎐CH), 132.3 (Ar C1), 123.7, 119.3, 117.3,
᎐
᎐
5.4–5.2 (m, 2H, ᎐CH ), 4.7–4.65 (m, 2H, ArOCH ), 4.15–3.55
114.3 (Ar C4, -C5, -C6; ᎐CH ), 128.5 (s, ArCC᎐O), 126.7 (d,
᎐ ᎐
2
᎐
2
2
(m, 16H, CH2O, CH2N), 3.36 (s, 4H, OCH2CH2O); MS (EI)
ArCH), 99.8 (s, CHOO), 74.2, 73.0, 70.8, 70.6, 70.4, 69.8, 69.7,
m/z 527.2 (Mϩ, 527.3).
68.3 (t, CH O, OCH C᎐O), 53.8 (t, OCH CH O), 52.8 (q,
᎐
2
2
2
2
OMe), 39.0 (t, CH2NH).
11-[2-Allyloxy-3-(4,5-dihydro-1,3-dioxol-2-yl)phenoxy]-3,6,9-
trioxaundecylamine (16). A solution of phthalimide 15 (15.7 g,
29.8 mmol) and N2H4ؒH2O (9 ml, 180 mmol) in MeOH (250 ml)
was refluxed for 5 h. After the reaction mixture was cooled to
room temperature, it was kept overnight in the freezer (Ϫ30 ЊC).
The precipitated phthalhydrazide was filtered off and the
mixture was evaporated to dryness. The residue was stirred in
CH2Cl2 (250 ml). A second portion of precipitated phthal-
hydrazide was filtered off and the clear solution was washed
with brine (1 × 25 ml) and dried over Na2SO4. After evapor-
ation of the solvent, a yellow oil was obtained. Yield: 9.3 g
Dimethyl 2,3-bis(butylcarbamoylmethoxy)terephthalate (24).
Compound 24 was prepared according to procedure B from
diacid 17 (2.0 g, 5.8 mmol), n-butylamine (950 mg, 13.0 mmol),
triethylamine (1.77 g, 2.4 ml, 17.5 mmol). The crude product
was purified by column chromatography (silica gel, EtOAc–
petroleum ether 4:1) to obtain a clear oil which solidified
rapidly upon standing. Yield: 2.19 g (84%). An analytical
sample was obtained by recrystallisation from petroleum ether;
1
mp 93–94 ЊC (petroleum ether); H NMR δ 7.7 (s, 2H, ArH),
7.6–7.5 (br t, 2H, CONH), 4.51 (s, 4H, OCH2), 3.93 (s, 6H,
OCH3), 3.37 (q, J = 7.1 Hz, 4H, NHCH2), 1.65–1.35 (m, 8H,
CH2), 0.97 (t, J = 7.3 Hz, 6H, CH3); 13C NMR δ 167.6 (s,
1
(79%). H NMR δ 7.15–6.9 (m, 3H, ArH), 6.2–6.0 (m, 1H,
᎐CH), 5.65 (s, 1H, CHOO), 5.4–5.2 (m, 2H, ᎐CH ), 4.7–4.65
᎐
᎐
2
CH C᎐O), 164.7 (s, C᎐O), 152.1 (s, ArCO), 128.3 (s, ArC),
᎐
᎐
2
(m, 2H, ArOCH2), 4.57–4.54, 4.17–4.13, 4.0–3.85 (m, 3 × 2H,
CH2O), 3.77–3.1 (m, 8H, CH2O), 3.44 (t, 2H, CH2O), 3.30 (s,
4H, OCH2CH2O), 2.9–2.8 (br t, 2H, CH2NH2), 1.6–1.4 (br s,
126.6 (d, ArC), 73.2 (t, OCH2), 52.7 (q, OMe), 39.0 (t,
NHCH2), 31.5, 20.1 (t, CH2), 13.8 (q, CH3); MS (EI) m/z 452.2
(Mϩ, 452.3); elemental analysis calcd C, 58.40; H, 7.13; N, 6.19;
found C, 58.35; H, 7.12; N, 6.22%.
2H, NH2); 13C NMR δ 151.8, 146.5 (Ar C-2, C-3), 134.6 (᎐CH),
᎐
132.3 (Ar C1), 123.7 (ArCH), 119.4, 117.3, 114.4 (ArCH,
᎐CH ), 99.8 (s, CHOO), 74.2, 73.5, 70.9, 70.7, 70.3, 69.8, 68.4
᎐
2
General procedure C for the reduction of dimethyl 2,3-bis-
substituted-terephthalates with LiAlH4
(t, CH2O), 53.8 (s, OCH2CH2O), 41.8 (t, CH2NH2); MS (EI)
m/z 399.2 (Mϩ, 397.2).
A suspension of LiAlH4 (3 mol equiv.) in dry THF (25 ml gϪ1
substrate) was refluxed for 0.5 h and then cooled in an ice-bath.
A solution of the appropriate dimethyl terephthalate (1 mol
equiv.) in dry THF (10 ml gϪ1 substrate) was added dropwise to
the pre-treated suspension of LiAlH4. After addition, the reac-
tion mixture was stirred for 2 h at 0 ЊC. The ice bath was
removed and the solution was stirred for another 2 h, while
slowly warming up to room temperature. The excess of LiAlH4
was quenched with 1 M HCl solution until a final acidity of
pH = 1 was reached (CAUTION, H2 evolution). When all
aluminium salts had dissolved, the THF layer was separated
and the water layer was extracted with EtOAc (2 × 150 ml). The
combined organic layers were washed with brine (1 × 100 ml)
and dried over Na2SO4.
Dimethyl 2,3-bis(hydroxycarbonylmethoxy)terephthalate (17).
A solution of tert-butyl ester 8d (14.54 g, 32 mmol) in neat
CF3COOH (50 ml) was stirred at room temperature for 3 h. The
mixture was evaporated to dryness and the solid residue was
dried in vacuo. Yield: 10.9 g (quantitative). An analytical sam-
ple was obtained by recrystallisation from EtOAc; mp 169–
170 ЊC (EtOAc); 1H NMR (DMSO-d6) δ 7.50 (s, 2H, ArH), 4.60
(s, 4H, OCH2), 3.89 (s, 6H, OMe), 3.7–3.0 (br, COOH, H2O);
13C NMR (DMSO-d ) δ 169.8 (s, CH C᎐O), 165.4 (s, C᎐O),
᎐
2
᎐
6
150.6 (s, ArCO), 130.1 (s, ArC), 125.6 (d, ArC), 70.3 (t, OCH2),
52.7 (q, OMe); MS (EI) m/z 342.1 (Mϩ, 342.1); elemental
analysis, calcd C, 49.13; H, 4.12 found C, 49.13; H, 3.91%.
General procedure B for the conversion of dimethyl 2,3-
bis(hydroxycarbonylmethoxy)terephthalate (17) to bisamides
1,4-Bis(hydroxymethyl)-2,3-dihydroxybenzene. [Scheme 4]
was prepared according to procedure C from LiAlH4 (0.25 g,
6.6 mmol) and dimethyl terephthalate 6 (0.5 g, 2.2 mmol).
Yield: 0.3 g (80%); 1H NMR δ 7.3, 6.75 (d, 1H, ArH), 6.2 (s, 2H,
OH), 4.7 (s, 2H, CH2), 3.85 (m, 4H, CH2, OH); 13C NMR
δ 144.4 (s, ArCOH), 128.1 (s, ArCCH2), 119.8 (d, ArCH), 61.8
(t, CH2OH); FTIR ν (cmϪ1) 3323 (O–H), 1462 (CH2–OH), 1250
(C–O, phenolic), 982, 937 (C–O, primary alcohol).
A solution of diacid 17 (1 mol equiv.) in neat SOCl2 (5 ml gϪ1
substrate) was refluxed for 1 h. The solution was evaporated to
dryness and the solid residue was dried in vacuo. To an ice-
cooled solution of the appropriate amine (2 mol equiv.), tri-
ethylamine (3 mol equiv.) and a spatula point of DMAP in dry
CH2Cl2 (100 ml gϪ1 substrate) was added dropwise a solution of
the freshly prepared diacid chloride in dry CH2Cl2 (100 ml).
After addition, the mixture was stirred for 18 h, while it was
allowed to warm up to room temperature. The solution was
washed with water (2 × 50 ml), brine (1 × 50 ml) and dried
over Na2SO4. The crude product was purified by column
chromatography or recrystallisation.
2,3-Bis[N-(2-allyloxy-3-formylphenoxyl)-3,6,9-trioxaundecyl-
carbamoyl methoxy]-1,4-bis(hydroxymethyl)benzene (19). 19
was prepared according to procedure C from LiAlH4 (0.95 g, 25
mmol) and dimethyl terephthalate 18 (9.1 g, 8.26 mmol). After
work-up and evaporation, a sticky oil was obtained in almost
quantitative yield, which was not purified further. Yield: 7.88 g;
1H NMR δ 10.4 (s, 2H, CHO), 8.0–7.9 (br t, 2H, CONH), 7.5–7
1,4-Dimethyl
2,3-bis-{N-[2-allyloxy-3-(4,5-dihydro-1,3-
dioxol-2-yl)phenoxy]-3,6,9-trioxaundecylcarbamoylmethoxy}
terephthalate (18). Compound 18 was prepared according to
procedure B from diacid 17 (3.87 g, 11.3 mmol), amine 16 (9.0
g, 22.6 mmol), triethylamine (3.5 g, 4.8 ml, 34 mmol). The crude
product was purified by column chromatography (silica gel,
EtOAc–MeOH 9:1) to obtain a yellow, clear oil. Yield: 9.16 g
(m, 8 H, ArH), 6.2–6.0 (m, 2H, ᎐CH), 5.4–5.2 (m, 2H, ᎐CH ),
᎐ ᎐
2
4.62 (d, 4H, CH2OH), 4.43 (s, 4H, OCH2CO), 4.7–4.65, 4.2–3.4
(m, 36H, ArOCH CH᎐, CH O, CH NH); 13C NMR δ 190.4 (s,
᎐
2
2
2
CHO), 169.0 (s, CH C᎐O), 152.2, 151.6, 149.2 (s, ArCO [C2,
᎐
2
C3, C7]), 134.5 (d, ᎐CH), 133.4 (s, ArC(8)CH ), 130.1 (s, Ar
᎐
2
C1), 125.1 (d, ArC(9)H), 124.1, 119.9, 119.6, (d, Ar C4, -C5,
1582
J. Chem. Soc., Perkin Trans. 2, 2001, 1573–1584

View Article Online
-C6), 118.9 (t, ᎐CH ), 75.1 (t, OCH C᎐), 72.1, 70.7 (2×), 70.6,
EtMe2N) to obtain macrocyclic salen 21 as a deep yellow
᎐
᎐
2
2
70.2, 69.8, 69.6, 68.7 (t, CH O, OCH C᎐O), 60.6 (t, CH OH),
amorphous solid. Yield: 902 mg (50%); 1H NMR δ 14–13.5 (br,
᎐
2
2
2
38.9 (t, CH2NH); MS (FAB) m/z 957 (M ϩ Hϩ, 957.5).
2H, OH), 8.22 (s, 2H, CH᎐N), 8.2–8.0 (br m, 2H, CONH),
᎐
7–6.5 (m, 8H, ArH), 4.60–4.3 (m, 8H, CH2OH, OCH2CO),
4.2–3.2 (m, 32H, CH2O, CH2NH), 2.2–1.3 (br m, 8H, CH2
[cyclohexyl]); 13C NMR δ 169.1 (s, CH C᎐O), 164.8 (d, CH᎐N),
151.9, 148.9, 147.2 (s, Ar C2, C3, C7), 133.6 (s, ArC(8)CH2),
124.5 (d, ArC(9)H), 123.7, 118.0, 116.5 (d, Ar C4, C5, C6),
118.8 (s, Ar C1), 77.0 (d, C[cyclohexyl]HN), 72.8, 72.0, 70.7,
1,4-Bis(hydroxymethyl)-2,3-bis(butylcarbamoylmethoxy)-
benzene (25). 25 was prepared according to procedure C from
LiAlH4 (0.5 g, 13.3 mmol) and dimethyl terephthalate 24 (2.0 g,
4.42 mmol). After work-up and evaporation a white solid was
obtained, which was further purified by recrystallisation from
EtOAc–MeOH. Yield: 1.25 g (71%); mp 157–158 ЊC (EtOAc–
MeOH); 1H NMR (CDCl3–DMSO-d6) δ 7.7–7.5 (br t, 2H
CONH) 7.7 (s, 2H, ArH), 4.61 (d, 4H, CH2OH), 4.95 (t, 2H,
OH), 4.45 (s, 4H, OCH2CO), 3.37 (q, J = 7.1 Hz, 4H, NHCH2),
1.65–1.35 (m, 8H, CH2), 0.97 (t, J = 7.3 Hz, 6H, CH3); 13C
᎐
᎐
2
70.6, 70.4, 70.0, 69.8, 68.6 (t, CH O, OCH C᎐O), 60.9 (t,
᎐
2
2
CH2OH), 39.0 (t, CH2NH), 33.3, 24.3 (t, CH2 [cyclohexyl]); MS
(FAB) m/z 954.0 (Mϩ, 954).
Metallation of macrocyclic salen 21. To a slightly warmed
40 ЊC) solution of macrocyclic salen 21 (600 mg, 0.63 mmol)
(
NMR (CDCl –DMSO-d ) δ 168.5 (s, CH C᎐O), 148.9 (s,
᎐
2
3
6
in MeOH (20 ml) was added a methanolic Ni(OAc)2ؒ4H2O
solution (9.4 ml, 0.1 M). The deep red solution was gently
warmed for 1 min (the temperature must not exceed 50 ЊC) and
evaporated to dryness. The residue was taken-up in CH2Cl2 and
washed with water (3 × 100 ml) [it is crucial that distilled water
be used at this stage]. The organic layer was filtered over Hiflo
to clear it from residual dispersed water. After evaporation pure
metallated 22 was obtained as a deep red solid in almost quanti-
ArCO), 135.1 (s, ArC), 125.0 (d, ArC), 72.2 (t, OCH2), 60.0 (t,
CH2OH), 38.7 (t, NHCH2), 31.4, 20.0 (t, CH2), 13.7 (q, CH3);
MS (EI) m/z 397.2 (M ϩ Hϩ, 397.3); elemental analysis, calcd
C, 60.59; H, 8.13; N, 7.06; found C, 60.45; H, 8.08; N, 6.96%.
1,4-Bis(hydroxymethyl)-2,3-bis(benzyloxy)benzene (27). 27
was prepared according to procedure C from LiAlH4 (0.57 g,
15.0 mmol) and dimethyl terephthalate 8b (2.0 g, 4.92 mmol).
After work-up and evaporation almost pure 27 in quantitative
yield was obtained as a white solid. Yield: 1.71 g. An analytical
sample was obtained by recrystallisation from petroleum
ether (70%); mp 104–105 ЊC (EtOAc–petroleum ether); 1H
NMR δ 7.4–7.3 (m, 10H, BnH), 7.06 (s, 2H, ArH), 5.2 (s, 4H,
OCH2C), 4.57 (s, 4H, CH2O), 2.4–2.1 (br s, 2H, OH); 13C NMR
δ 149.6 (s, ArC), 137.1 (s, BnC), 135.4 (s, ArC), 128.7, 128.6,
128.5 (d, BnCH), 124.2 (d, ArC), 75.5 (t, OCH2), 61.3 (t,
CH2OH); elemental analysis, calcd C, 75.41; H, 6.33; found C,
75.50; H, 6.22%.
1
tative yield. Yield: 637 mg; H NMR (CDCl3) δ 8.5–8.4 (br t,
2H, CH᎐N), 7.37 (s, 2H, CONH), 7.0 (s, 2H, ArH(9)), 6.82,
᎐
6.75 (d, J = 7.7 Hz, 2H, ArH), 6.47 (t, J = 7.7 Hz, 2H, ArH),
4.61 (s, 4H, CH2OH), 4.55 (s, 4H, OCH2CO), 4.2–4.1 (m, 4H,
CH2O), 3.8–3.5 (m, 28H, CH2O, CH2NH), 3.2–3.1, 2.5–2.3,
2.0–1.8 (br m, 2H, CH, CH2 [cyclohexyl]), 1.4–1.2 (br m, 4H,
CH, CH2 [cyclohexyl]), 2.7–2.6 (br, 2H, OH); 13C NMR
(CD Cl ) δ 169.6 (s, CH C᎐O), 158.1 (d, C᎐N), 157.0 (s,
᎐
2
᎐
2
2
ArC(2)O), 150.2, 149.9 (s, ArC(3)O), ArC(7)O), 135.9 (s,
ArC(8)CH2), 125.7 (d, ArC(9)H), 121.7 (s, Ar C1), 126.7,
118.7, 114.4 (d, Ar C4, C5, C6), 77.0 (s, CHN᎐), 73.0, 71.3,
᎐
Bis-2,3-(N-[11-(2-hydroxy-3-formylphenoxy)-3,6,9-trioxa-
undecyl]carbamoylmethoxy)-1,4-bis(hydroxymethyl)benzene
(20). To a solution of compound 19 (3.0 g, 3.13 mmol), in
ethanol–water (4:1) (60 ml) was added 38 ml of a solution of
freshly prepared solution of triethylammonium formate (1 M,
ethanol–water 4:1, pH = 8). A spatula-point of Pd(CH3COO)2
and a few flakes of triphenylphosphine were added. The mix-
ture was refluxed for 1.5 h and regularly checked by TLC
(EtOAc–MeOH 4:1). The mixture was evaporated to dryness.
The residue was taken up in EtOAc (500 ml) and washed with a
minimum of water, brine and dried over Na2SO4. The crude
product was purified by column chromatography (silica gel,
EtOAc–MeOH 4:1). A light yellow oil was obtained, which was
almost pure. Yield: 1.70 g (62%); 1H NMR δ 10.0 (s, 2H, CHO),
8.0–7.9 (br t, 2H, CONH), 7.23, 7.14 (dd, J1 = 7.9 Hz, J2 = 1.7
Hz, 2H, ArH), 6.91 (t, J1 = 8.0 Hz, 2H, ArH), 7.03 (s, 2H, ArH),
4.66 (d, 4H, CH2OH), 4.51 (s, 4H, OCH2CO), 4.2–3.5 (m, 32H,
CH2O, CH2NH); 13C NMR δ 195.8 (s, CHO), 169.0 (s,
70.9, 70.5, 70.3, 69.1 (t, CH O, OCH C᎐O), 61.2 (t, CH OH),
᎐
2
2
2
39.7 (t, CH2NH), 29.3, 25.2 (t, CH2 [cyclohexyl]); MS (FAB,
glycerol) m/z 1034 (M ϩ Naϩ, 1033.5), 1050 (M ϩ Kϩ, 1049.5).
General procedures for the conversion of 1,4-bis(hydroxymethyl)-
2,3-bis(substituted)-benzenes to 1,4-bis(chloromethyl)-2,3-
bis(substituted)-benzenes
Procedure D. A solution of the appropriate bis(hydroxy-
methylated) derivative (1 mol equiv.) and SOCl2 (5–10 mol
equiv.) in dry CH2Cl2 (25 ml gϪ1 substrate) was refluxed for 1 h.
The solution was diluted with additional CH2Cl2 and washed
with NaHCO3 solution (stad. 50 ml portions) until the water
layer was basic. The organic layer was dried over Na2SO4. The
crude products after work-up had enough purity to be used
without any further purification. Analytical samples could be
obtained by recrystallisation.
Procedure E. This is similar to procedure D, however, SOCl2
(3 mol equiv.) and additional 2,6-lutidine (3.3 mol equiv.)
were used. The reaction was performed at room temperature
(reaction time 1 h). The organic layer was washed with NH4Cl
solution (satd., 1 × 50 ml).
CH C᎐O), 152.1 (s, ArC(2)O), 147.4 (s, ArC(3)O), 149.3 (s,
᎐
2
ArC(7)O), 134.4 (s, ArC(8)CH2), 125.2 (d, ArC(9)H), 121.5
(s, Ar C1), 124.9, 120.7, 119.6, (d, Ar C4, -C5, -C6), 72.1, 70.7,
70.5, 70.2, 69.9, 69.6, 69.2 (t, CH O, OCH C᎐O), 60.8 (t,
᎐
2
2
CH2OH), 38.9 (t, CH2NH); MS (FAB) m/z 877 (M ϩ Hϩ,
877.4), 899 (M ϩ Naϩ, 899.4).
Procedure F. A solution of the appropriate bis(hydroxymeth-
ylated) derivative (1 mol equiv.), mesyl chloride (3–4 mol equiv.)
and triethylamine (3–4 mol equiv.) in dry, ethanol free CHCl3
(50 ml gϪ1 substrate) was heated at 55 ЊC for 2 h. Et3NؒHCl (10
mol equiv.) was added in one portion to the mixture and heat-
ing was continued for another 4 h. After cooling the mixture
was diluted with additional CH2Cl2. The solution was washed
with water, brine and dried over Na2SO4. The crude products
were purified by recrystallisation.
trans-3,5,20,28-Tetraaza-8,11,14,17,23,25,31,34,37,40-
decaoxa-1,7(1,3),24(1,2)-benzena-4(1,2)cyclohexana-243,246-
bis(hydroxymethyl)-12,72-dihydroxytetracontacyclophane-2,5-
diene-21,27-dione (21). A solution of compound 20 (1.70 g, 1.94
mmol) and Ba(OTf )2 (1.7 g, 3.9 mmol) in dry THF (2 L) was
refluxed for 1 h. ( )-trans-Cyclohexane-1,2-diamine (222 mg,
1.94 mmol) was added and refluxing was continued for another
1.5 h. The mixture was evaporated to dryness and the residue
was taken up in CH2Cl2. The solution was washed with water
(1 × 250 ml), dilute Na2SO4 solution (1 × 100 ml) and dried
over Na2SO4. The crude product was separated by column
1,4-Bis(chloromethyl)-2,3-dihydroxybenzene (7). 7 was pre-
pared according to procedure E, using 1,4-bis(hydroxymethyl)-
2,3-dihydroxybenzene (0.08 g, 0.5 mmol). Yield: 0.075 g (90%);
1H NMR δ 7.3, 6.85 (d, 1H, ArH), 4.65–3.9 (br, 4H, CH2Cl);
chromatography (silica gel, CH2Cl2–MeOH 9:1;
1 vol%
J. Chem. Soc., Perkin Trans. 2, 2001, 1573–1584
1583

View Article Online
13C NMR δ 142.9 (s, ArCOH), 128.5 (s, ArCCH2), 120.1 (d,
ArCH), 40.2 (t, CH2Cl); FTIR ν (cmϪ1) 3470 (O–H), 1441
(CH2–Cl), 1263 (C–O, phenolic), 691 (C–Cl).
OCH2CO, CH2O, CH2NH), 3.2–3.1, 2.5–1.3 (br m, 14H,
CH, CH2 [cyclohexyl], ArCH2, ArCH3); 13C NMR (CD2Cl2)
δ 169.1 (s, CH C᎐O), 158.2 (d, C᎐N), 150.4 (br s, ArC(3)O),
᎐
᎐
2
ArC(7)O), 132.6, 130.1 (s, ArC(8)CH2), 127–126 (br d,
ArC(9)H), 121.6 (s, Ar C1), 127.5, 118.1, 114.4 (d, Ar C4, C5,
Ni-salen complex of trans-3,5,20,28-tetraaza-8,11,14,17,23,
25,31,34,37,40-decaoxa-1,7(1,3),24(1,2)-benzena-4(1,2)cyclo-
hexana-243,246-bis(chloromethyl)-12,72-dihydroxy-tetraconta-
cyclophane-2,5-diene-21,27-dione (23). 23 was prepared accord-
ing to procedure F using macrocyclic salen 22 (400 mg, 0.395
mmol), mesyl chloride (181 mg, 1.58 mmol), triethylamine (160
mg, 1.58 mmol) and Et3NؒHCl (544 mg, 3.95 mmol). After
reaction, the mixture was washed with water (4 × 100 ml) [it is
crucial that distilled water be used at this stage] and filtered over
Hiflo to clear it from residual dispersed water. After evapor-
ation the crude chloromethylated macrocycle 23 was not further
purified, but stored in a freezer. Crude yield: 442 mg; 1H NMR
C6), 73.0–69.0 (t, CH O, OCH C᎐O), 39.6 (t, CH NH), 30.4 (t,
᎐
2
2
2
ArCH2–CH2Ar), 29.3, 25.2 (t, CH2 [cyclohexyl]), 16.6 (q,
ArCH3).
Acknowledgements
We are grateful to the EPSRC for a grant under the “Innovative
Polymer Synthesis” programme and we also thank the Uni-
versity of London Intercollegiate Research Services (ULIRS)
for provision of the high field NMR facilities (at QMW). Mr S.
J. Adams (QMW) is thanked for performing the AAS experi-
ments and Dr G. P. Moss (QMW) for advice on nomenclature.
Electrospray mass spectrometry experiments were performed
by the EPSRC National Mass Spectrometry Service, Swansea.
(CD Cl ) δ 7.5–6.3 (br m, 12H, CH᎐N, CONH, ArH(9), ArH),
᎐
2
2
4.7–4.4 (m, 8H, CH2Cl, OCH2CO), 4.2–3.3 (m, 28H, CH2O,
CH2NH), 3.2–3.1, 2.5–2.3, 2.0–1.8 (br m, 2H, CH, CH2 [cyclo-
hexyl]), 1.4–1.2 (br m, 4H, CH, CH2 [cyclohexyl]); 13C NMR
(CD Cl ) δ 168.8 (s, CH C᎐O), 158.9 (d, C᎐N), 155.3 (s,
᎐
2
᎐
2
2
References
ArC(2)O), 150.2, 149.8 (s, ArC(3)O), ArC(7)O), 133.8 (s,
ArC(8)CH2), 126.2 (d, ArC(9)H), 121.3 (s, Ar C1), 127.2,
117.3, 115.6 (d, Ar C4, C5, C6), 73.0–68.0 (t, CH2O,
1 H. G. Gilch, J. Polym. Sci., 1966, 4, 1351.
2 F. H. Covitz, J. Am. Chem. Soc., 1967, 11, 5403.
3 J. H. P. Utley, Y. Gao, J. Gruber and R. Lines, J. Mater. Chem., 1995,
5, 1297.
OCH C᎐O), 41.7 (t, CH Cl), 39.6 (t, CH NH), 29.4, 25.1 (t,
᎐
2
2
2
CH2 [cyclohexyl]); MS (ES) m/z 1047.5, 1049.5 (M ϩ Hϩ,
1047.3, 1049.3), 1069.3, 1071.3 (M ϩ Naϩ, 1069.3, 1071.3).
4 J. H. P. Utley and C. Z. Smith, J. Mater. Chem., 2000, 10, 2642.
5 J. H. P. Utley, E. Oguntoye (née Eru), C. Z. Smith and P. B. Wyatt,
Tetrahedron Lett., 2000, 40, 7249.
6 E. Oguntoye (née Eru), S. Szunerits, J. H. P. Utley and P. B. Wyatt,
Tetrahedron, 1996, 52, 7771.
7 S. Szunerits, J. H. P. Utley and M. F. Nielsen, J. Chem. Soc., Perkin
Trans. 2, 2000, 4, 669.
8 M. F. Semmelhack, P. M. Helquist and L. D. Jones, J. Am. Chem.
Soc., 1971, 93, 5808.
9 M. A. Fox, D. A. Chandler and C. Lee, J. Org. Chem., 1991, 56,
3246.
10 J. F. Fauvarque, M. A. Petit, F. Pfluger, A. Jutand and C. Chevrot,
Makromol. Chem., Rapid Commun., 1983, 4, 455.
11 A. J. Fry, U. N. Sirisoma and A. S. Lee, Tetrahedron Lett., 1993, 34,
809.
12 A. J. Fry and A. H. Singh, J. Org. Chem., 1994, 59, 8172.
13 Y. Becker and J. K. Stille, J. Am. Chem. Soc., 1978, 100, 838.
14 K. S. Y. Lau, P. K. Wong and J. K. Stille, J. Am. Chem. Soc., 1976,
98, 5832.
1,4-Bis(chloromethyl)-2,3-bis(butylcarbamoylmethoxy)benzene
(26). 26 was prepared according to procedures D–F. In a typical
preparation, applying procedure E, compound 25 (500 mg, 1.26
mmol), SOCl2 (452 mg, 3.8 mmol) and 2,6-lutidine (450 mg, 4.2
mmol) were used. After work-up almost pure 26 was obtained
as a white solid in quantitative yield. An analytical sample was
prepared by recrystallisation from EtOAc–petroleum ether.
1
Yield: 545 mg; mp 129–130 ЊC (EtOAc–petroleun ether); H
NMR (CDCl3) δ 7.17 (s, 2H, ArH), 6.8–6.7 (br t, 2H, CONH)
4.56, 4.51 (s, 4H, CH2Cl, OCH2CO), 3.37 (q, J = 7.1 Hz, 4H,
NHCH2), 1.65–1.35 (m, 8H, CH2), 0.97 (t, J = 7.3 Hz, 6H,
CH3); 13C NMR (CDCl ) δ 167.8 (s, CH C᎐O), 149.2 (s, ArCO),
᎐
3
2
133.0 (s, ArC), 127.0 (d, ArC), 72.3 (t, OCH2), 40.7 (t, CH2Cl),
39.2 (t, NHCH2), 31.5, 20.1 (t, CH2), 13.7 (q, CH3); MS (EI)
m/z 432.2 (Mϩ, 432.0); elemental analysis calcd C, 55.40; H,
6.98; N, 6.46; found C, 55.77; H, 6.85; N, 6.49%.
15 J. K. Stille and A. B. Cowell, J. Organomet. Chem., 1997, 124, 253.
16 C. Amatore, F. Gaubert, A. Jutand and J. H. P. Utley, J. Chem. Soc.,
Perkin Trans. 2, 1996, 2447.
17 Y. Rollin, M. Troupel, D. G. Tuck and J. Perichon, J. Organomet.
Chem., 1986, 303, 131.
18 G. Shiavon, G. Bontempelli and B. Corain, J. Chem. Soc., Dalton
Trans., 1981, 1074.
19 G. L. Closs and J. R. Miller, Science, 1988, 240, 440.
20 S. Antonello and F. Maran, J. Am. Chem. Soc., 1998, 120, 5713.
21 D. W. Sopher and J. H. P. Utley, J. Chem. Soc., Perkin Trans. 2, 1984,
1361.
22 A. Urbanczyk, E. Debek and M. K. Kalinowski, J. Electroanal.
Chem., 1995, 389, 141.
1,4-Bis(chloromethyl)-2,3-bis(benzyloxy)benzene (28). 28 was
prepared according to procedure D using compound 27 (500
mg, 1.43 mmol) and SOCl2 (1.7 g, 1 ml, 14 mmol). After work-
up almost pure 28 was obtained as a slightly yellow solid in
quantitative yield. An analytical sample was prepared by
recrystallisation from petroleum ether. Yield: 550 mg; mp 89–
1
90 ЊC (petroleum ether); H NMR δ 7.5–7.3 (m, 10H, BnH);
7.16 (s, 2H, ArH), 5.15 (s, 4H, OCH2), 4.57 (s, 4H, CH2Cl); 13
C
23 A. J. Fry and P. F. Fry, J. Org. Chem., 1993, 58, 3496.
24 S. Gronowitz and G. Hansen, Ark. Kemi, 1967, 27, 145.
25 M. S. Newman and J. A. Cella, J. Org. Chem., 1973, 38, 3482.
26 M. Czerwinski, B. Kowalczyk and P. Bragiel, Thermochim. Acta,
1994, 231, 151.
NMR δ 150.3 (s, ArC), 136.8 (s, BnC), 133.2 (s, ArC), 128.6,
128.5, 128.4 (d, BnCH), 125.9 (d, ArC), 75.3 (t, OCH2), 40.8 (t,
CH2Cl); MS (EI) m/z 386.1 (Mϩ, 386.2); elemental analysis,
calcd C, 68.21; H, 5.20; found C, 68.91; H, 5.15%.
27 K. Gielzak and W. Wojciechowski, Chem. Pap., 1995, 49, 54.
28 J. M. Stewart, E. C. Lingafelter and J. D. Breazeale, Acta
Crystallogr., 1961, 14, 888.
29 F. A Cotton and R. H. Holm, J. Am. Chem. Soc., 1960, 82, 2979.
30 C. G. Pierpont and R. M. Buchanan, Coord. Chem. Rev., 1981, 38,
45.
31 C. J. van Staveren, J. van Eerden, F. C. J. M. van Veggel, S. Harkema
and D. N. Reinhoudt, J. Am. Chem. Soc., 1988, 110, 4994.
32 J. Cason and G. O. Dyke, J. Am. Chem. Soc., 1950, 72, 621.
33 F. L. Weitl, K. N. Raymond and P. W. Durbin, J. Med. Chem., 1981,
24, 203.
34 A. J. Bard and L. R. Faulkner, Electrochemical Methods,
Fundamentals and Applications, Wiley, New York, 1980.
35 J. H. P. Utley, Y. Gao, J. Gruber, Y. Zhang and A. Munoz-Escalona,
J. Mater. Chem., 1995, 5, 1837.
Electrolysis of macrocyclic Ni-salen 23 to the corresponding
PPX oligomer. A solution of 23 (400 mg, 0.38 mmol) in DMF–
Et4NBr (0.1 M, 30 ml) was electrolysed at a mercury pool at
constant potential of Ϫ1.22 V (Ag/AgBr). The electrolysis was
followed by recording CV’s at regular intervals. When a total
charge of 90 C was passed, the red DMF solution was evapor-
ated to dryness. The residue was suspended in distilled water
and extracted with CH2Cl2 (2 × 100 ml). The organic layer was
washed with distilled water (5 × 50 ml), but not dried over
dehydrated salts. Evaporation gave a red solid, which was dried
1
in vacuo. Yield 315 mg (85%); H NMR (CD2Cl2) δ 8.0–6.3
(br m, 12H, CH᎐N, CONH, ArH(9), ArH), 4.7–3.3 (m, 36H,
᎐
1584
J. Chem. Soc., Perkin Trans. 2, 2001, 1573–1584