Synthesis of strained macrocyclic biaryls for enthalpy-driven ring-opening polymerization

Article abstract of DOI:10.1021/ma0517796

Polymerizable macrocyclic biarylene-ether-ketones and biarylene-ether- sulfones are accessible from linear, bis(chloro)-terminated oligomers via nickel-catalyzed, intramolecular coupling under pseudo-high-dilution conditions. Single-crystal X-ray analyses

Full text of DOI:10.1021/ma0517796

Macromolecules 2005, 38, 10413-10420
10413
Synthesis of Strained Macrocyclic Biaryls for Enthalpy-Driven
Ring-Opening Polymerization
Howard M. Colquhoun,*,† Zhixue Zhu,† Christopher C. Dudman,‡
Caroline A. O’Mahoney,§ David J. Williams,§ and Michael G. B. Drew†
School of Chemistry, University of Reading, Whiteknights, Reading, U.K. RG6 6AD, Imperial
Chemical Industries plc, The Heath, Runcorn, Cheshire, U.K. WA7 4QE, and Chemical
Crystallography Laboratory, Department of Chemistry, Imperial College, South Kensington,
London, U.K. SW7 2AY
Received August 11, 2005; Revised Manuscript Received October 10, 2005
ABSTRACT: Polymerizable macrocyclic biarylene-ether-ketones and biarylene-ether-sulfones are
accessible from linear, bis(chloro)-terminated oligomers via nickel-catalyzed, intramolecular coupling under
pseudo-high-dilution conditions. Single-crystal X-ray analyses of the resulting cyclo-oligomers reveal
extremely distorted and highly strained geometries, with 4,4′-biphenylene units showing deviations of
up to 70° from linearity.
1. Introduction
Perhaps as a result of the reversible nature of
nucleophilic aromatic substitution, the macrocyclic oli-
gomers so far obtained by this route are almost exclu-
sively strain-free, so that there is no significant enthalpy
change associated with their ring-opening polymeriza-
tion.7-11,15,16 The latter reaction is then dependent
almost entirely on entropic factorssmainly the increase
in conformational entropy resulting from the opening
of a constrained cyclic structure. However, this is
opposed by a corresponding decrease in translational
entropy on polymerization, resulting in a delicately
balanced situation whereby quite modest dilutions can
shift the thermodynamic equilibrium strongly in favor
of cyclic oligomers.17 Indeed, even in the absence of
solvent, the “self-diluting” effect of polymer formation
means that significant levels of macrocycles are invari-
ably present at equilibrium.
To provide a higher thermodynamic driving force for
ring-opening polymerization, a rather different type of
aromatic macrocycle is clearly required, i.e., one having
a substantial degree of ring-strain analogous to that of
an oxirane or oxetane in aliphatic chemistry. Nucleo-
philic cyclo-etherification, for the reasons given above,
generally fails to yield macrocycles of this type although
trace quantities of the strained cyclic thioethers [-S-
1,4-C6H4-SO2-Ar-]2 and [-S-1,4-C6H4-]4 have been
isolated from reactions involving displacement of aryl
halide substituents by sulfur-based nucleophiles.18,19
Electrophilic aromatic acylation has also been been
briefly described as a route to polymerizable macrocyclic
ether-ketones,20,21 but no evidence for the formation of
strained macrocycles has so far been reported for these
reactions. In the present paper we describe the first
general route to highly strained, polymerizable aromatic
macrocycles via intramolecular, nickel-catalyzed forma-
tion of biaryl linkages.
Aromatic polyethers such as the poly(ether-ketone)s
and poly(ether-sulfone)s exhibit high thermooxidative
stabilities, excellent thermomechanical properties and
very good impact resistance, leading to their widespread
application as high-performance engineering thermo-
plastics.1 However, the high melt viscosities of these
polymers place limits on their ability to be fabricated
into complex shapes, especially at very short length
scales, and to be used in composite materials where high
levels of reinforcing particles or fibers are present. A
potential solution to this problem involves the use of
low viscosity macrocyclic precursors, which undergo
ring-opening polymerization in situ to give high molar
mass polymers without generating byproducts.2 There
is thus considerable current interest in the synthesis
of macrocyclic aromatic ethers and in the chemistry of
their interconversion with polymeric materials.3
Such syntheses (other than those described in pre-
liminary comunications of the present work)4,5 have
focused mainly on activated nucleophilic (SNAr) cyclo-
etherification chemistry, carried out under pseudo-high-
dilution conditions.6 For example, homologous series of
cyclic ether-ketones have been obtained (i) by self-
condensation of 4-fluoro-3′-hydroxybenzophenone7 or
4-fluoro-4′-hydroxybenzophenone,8 (ii) by condensation
of 1,3-bis(4-fluorobenzoyl)benzene with 4,4′-biphenol,9
(iii) from 1,2-bis(4-fluorobenzoyl)benzene,10 or its tet-
raphenyl derivative,11 by reaction with a range of
bisphenols, and (iv) by condensation of 4,4′-difluoroben-
zophenone with either resorcinol,12 bisphenol A,13 or
3,3′,5,5′-tetramethyl-4,4′-biphenol.14 Pure, monodisperse
macrocycles have been isolated from a number of these
“one-pot” reactions by chromatographic fractionation,
but an alternative approach involving stepwise con-
struction of long-chain intermediates has also enabled
good yields of specific macrocyclic ether-ketones15 and
ether-sulfones16 to be obtained.
2. Experimental Section
2.1 Materials. Reagents and solvents were obtained from
Aldrich, Acros, or Lancaster and, other than N,N-dimethylac-
etamide, were used as received. The latter was distilled from
calcium hydride under dry nitrogen immediately prior to use.
The intermediates 4,4′-diphenoxydiphenyl sulfone,22 4-chloro-
4′-fluorobenzophenone,23 2,2-bis(4-phenoxyphenyl)hexafluoro-
* Corresponding author. E-mail: h.m.colquhoun@rdg.ac.uk.
† University of Reading.
‡ Imperial Chemical Industries plc.
§ Imperial College.
10.1021/ma0517796 CCC: $30.25 © 2005 American Chemical Society
Published on Web 11/15/2005

10414 Colquhoun et al.
Macromolecules, Vol. 38, No. 25, 2005
propane,24 2,8-diphenoxydibenzofuran,22 3,6-dihydroxyxantho-
ne,22 4-chloro-4′-hydroxydiphenyl sulfone,25 and 4,4′-bis(4′′-
chlorobenzenesulfonyl) diphenyl ether26 were synthesized
according to literature methods. Syntheses of the linear
precursor compounds 1-4 and 9-16 and characterization data
for these precursors and for macrocycles 17-26 are provided
as Supporting Information.
Scheme 1. Nickel-Catalyzed Dehalogenative Coupling
of Chloroarenes To Yield Biaryls
9.0 Hz, 4H), 7.32 (d, J ) 8.7 Hz, 4H), 6.85 (d, J ) 9.0 Hz, 4H).
13C NMR (CH3SO3H:CD2Cl2 1:5, 75 MHz), δ (ppm): 204.2,
166.8, 158.4, 144.4, 139.1, 137.5, 134.5, 131.2, 130.6, 129.4,
128.0, 123.7, 116.9. IR (Nujol): 1691, 1677 cm-1 (νCO). MS
(FAB): calcd for C38H24O6S, m/z 608.13; found, 609 [M + H]+.
Anal. Calcd for C38H24O6S: C, 74.98; H, 3.97; S, 5.26. Found:
C, 74.77; H, 3.72; S, 5.46. Crystal data for 6: C38H24O6S‚C6H5-
CH3, M ) 700.8, monoclinic, space group P21/n, a ) 11.403(2)
Å, b ) 18.815(3) Å, c ) 17.273(3) Å, â ) 106.63(1)°, V ) 3550.9-
(10) Å3, T ) 293 K, Z ) 4, Dc ) 1.311 g cm-3, µ(Cu KR) ) 1.22
2.2. Instrumental Analyses. 1H and 13C NMR spectra were
recorded on Bruker DPX-250, A-300, and AMX-400 spectrom-
eters. Spectra were obtained, unless otherwise stated, using
a 10:1 v/v mixture of deuteriochloroform and trifluoroacetic
acid as solvent. Chemical shifts (δ) are quoted as parts per
million (ppm) downfield from TMS, with multiplicities given
as singlet (s), doublet (d), double-doublet (dd), triplet (t),
multiplet (m), and broad (b). Infrared spectra were obtained
from Nujol mulls on a Perkin-Elmer PE1700 FT-IR spectro-
photometer. Mass spectra (EI/CI) were recorded on a VG
Autospec spectrometer, and FAB-MS spectra were obtained
on a Kratos Concept-IS instrument. Matrix-assisted laser
desorption/ionization time-of-flight (MALDI-TOF) mass spec-
tra were obtained using Micromass Tofspec 2E and SAI LT3
LaserTof spectrometers, with 1,8,9-trihydroxyanthracene as
matrix and sodium trifluoroacetate as cationizing agent.
Melting points were measured under nitrogen using a Mettler
DSC-20 system, at a heating rate of 10 °C/min, and are quoted
as peak values. Elemental analyses were carried out by Medac
(UK) Ltd. Thin-layer chromatography (TLC) was performed
on Polygram Sil G/UV 254 plates with visualization by UV
light at 254 nm, and column chromatography on Merck silica
gel 60 (particle size 0.040-0.063 nm). X-ray data were collected
on a Siemens P4/RA diffractometer with graphite-monochro-
mated Cu KR radiation using ω-scans, and on a Mar Research
image-plate system with Mo KR radiation. Structures were
solved by direct methods and the non-hydrogen atoms were
refined anisotropically.
mm-1, F(000) ) 1464; 3648 independent reflections, R1
)
0.043, wR2 ) 0.106 for 2915 independent observed reflections
[2θ e 100°, I > 2σ(I)].
1
Macrocycle 7. Yield: 53.7%. Mp: 353 °C. H NMR (CH3-
SO3H:CD2Cl2 1:5, 300 MHz), δ (ppm): 7.88 (d, J ) 8.4 Hz, 4H),
7.72 (d, J ) 2.4 Hz, 2H), 7.68 (d, J ) 8.9 Hz, 2H), 7.62 (d, J )
8.4 Hz, 4H), 7.52 (d, J ) 9.0 Hz, 4H), 7.35 (dd, J ) 8.9 and 2.4
Hz, 2H), 7.14 (d, J ) 9.0 Hz, 4H). 13C NMR (CH3SO3H/CD2Cl2
1:5, 75 MHz), δ (ppm): 205.3, 165.6, 154.8, 148.9, 135.4, 134.2,
131.7, 129.8, 129.4, 128.1, 124.7, 122.1, 116.1, 114.3, 114.1.
IR (Nujol): 1674 (νCO), 1232, 1167 cm-1 (ν C-O-C). MS
(FAB): calcd for C38H22O5, m/z 558.15; found, 559 [M + H]+.
Anal. Calcd for C38H22O5: C, 81.71; H 3.97. Found: C, 81.63;
H, 3.93. Crystal data for 7: C38H22O5.0.5C6H5CH3, M ) 604.7,
triclinic, space group P1h, a ) 9.962(1) Å, b ) 12.777(2) Å, c )
13.369(2) Å, R ) 69.12(1)°, â ) 73.45(1)°, γ ) 89.13(1)°, V )
1517 Å3, T ) 293 K, Z ) 2, Dc ) 1.324 g cm-3, µ(Cu KR) )
0.70 mm-1, F(000) ) 630; 4075 independent reflections, R )
0.048, Rw ) 0.057 for 3555 independent observed reflections
[2θ e 116°, I > 1.5σ(I)].
2.3. General Procedure for Ni(0)-Catalyzed Cycliza-
tion (Yielding Macrocycles 5, 6, 7, 8, 17, 18, 19, 20, 21, 22,
23, 24, 25, and 26). The synthesis of macrocycle 6 is given as
an example. A mixture of triphenylphosphine (38.0 g, 145
mmol), tetra-N-butylammonium iodide (27.0 g, 73 mmol) and
the nickel(II) chloride-pyridine complex Ni(Py)4Cl2 (8.25 g, 19
mmol) in dry DMAc (800 mL) was stirred under nitrogen for
10 min to give a green solution. Finely powdered zinc (13.2 g,
200 mmol) was then added, and the mixture was heated to 60
°C and stirred for a further 0.5 h to give a red-brown
suspension of Ni(PPh3)4. A solution of the linear oligomer 2
(14.0 g, 20 mmol) in dry DMAc (1200 mL) was added dropwise
with stirring over 5 h. After the addition was completed, the
mixture was stirred at 60 °C for 0.5 h and then cooled and
filtered to remove unreacted zinc. The filtrate was poured into
5 M hydrochloric acid (3000 mL), and the precipitate was
collected by filtration, washed with aqueous methanol and
dried. The solid was then extracted with hot ethanol, the
extract filtered hot, and the solid dried. Macrocycle 6 was
obtained in 45% yield as a white crystalline solid by column
chromatography (4% ethyl acetate in DCM).
Macrocycle 8. Yield: 24.9%. Mp: 409 °C. 1H NMR (CDCl3/
TFA, 250 MHz), δ (ppm): 8.36 (d, J ) 8.9 Hz 2H), 7.68 (d, J
) 8.4 Hz, 4H), 7.52 (d, J ) 8.4 Hz, 4H), 7.49 (d, J ) 8.8 Hz,
4H), 7.32 (dd, J ) 8.9. 2.4 Hz, 2H), 7.23 (d, J ) 8.8 Hz, 4H),
6.85 (d, J ) 2.4 Hz, 2H). 13C NMR (CH3SO3H/CD2Cl2 1:5, 75
MHz), δ (ppm): 204.4, 174.4, 167.6, 160.4, 155.8, 144.3, 138.6,
138.1, 130.5, 129.5, 128.9, 128.7, 121.0, 120.7, 110.9, 102.9.
MS (FAB): calcd for C39H22O6, m/z 586.14; found, 587 [M +
H]+. Anal. Calcd for C39H22O6: C, 79.86; H, 3.78. Found: C,
79.90; H, 3.82. Crystal data for 8: C39H22O6, M ) 586.6,
triclinic, space group P1h, a ) 10.567(3) Å, b ) 15.223(4) Å, c
) 20.307(7) Å, R ) 99.62(3)°, â ) 103.81(2)°, γ ) 107.69(2)°, V
) 2919.0(2) Å3, T ) 293 K, Z ) 4, Dc ) 1.335 g cm-3, µ(Cu KR)
) 0.732 mm-1, F(000) ) 1216; 5856 independent reflections,
R1 ) 0.096, wR2 ) 0.254 for 3969 independent observed
reflections [2θ e 100°, I > 2σ(I)].
3. Results and Discussion
Homogeneous, nickel-catalyzed biaryl formation in-
volving zinc-promoted dehalogenative coupling of chlo-
roarenes (Scheme 1)27 is an extremely versatile reaction,
and it seemed to us that this type of chemistry might
provide a successful approach to highly strained aro-
matic macrocycles. In particular, (i) the biaryl-forming
reacton is generally irreversible, so avoiding the poten-
tial problem of equilibration to larger, unstrained mac-
rocycles, (ii) the kinetics of coupling are rapidsespecially
when electron-withdrawing aromatic substituents are
presentssuggesting that the pseudo-high-dilution con-
ditions which favor intramolecular over intermolecular
reaction should be readily achieved, and (iii) molecular
modeling studies showed that the diarylnickel(III)
intermediate, from which the final biaryl product is
formed,27a would itself be virtually unstrained, yet
capable of yielding a highly strained macrocyclic product
on reductive elimination (Scheme 2).
Macrocycle 5. Yield: 27.8%. Mp: 387 °C. 1H NMR (CDCl3/
TFA; 250 MHz), δ (ppm): 7.74 (d, J ) 8.3 Hz, 4H), 7.53 (d, J
) 8.3 Hz, 4H), 7.42 (d, J ) 8.9 Hz, 4H), 7.38 (d, J ) 9.0 Hz,
4H), 7.12 (d, J ) 9.0 Hz, 4H), 7.04 (d, J ) 8.9 Hz, 4H). 13C
NMR (CH3SO3H:CD2Cl2 1:5, 62.5 MHz), δ (ppm): 205.9, 167.0,
155.0, 146.0, 138.4, 134.7, 132.6, 131.6, 131.4, 129.6, 128.9,
121.1, 118.4; IR (Nujol) 1677 cm-1 (νCO). MS (FAB): calcd
for C41H24F6O4, m/z 694.16; found, 695 [M + H]+. Anal. Calcd
for C41H24F6O4: C, 70.89; H, 3.48. Found: C, 70.67; H, 3.47.
Crystal data for 5: 2(C41H24O4F6)‚Et2O‚MeOH, M ) 1495.37,
monoclinic, space group P21/c, a ) 26.873(1) Å, b ) 12.652(1)
Å, c ) 20.611(1) Å, â ) 98.84(1)°, V ) 6924.7(6) Å3, T ) 173
K, Z ) 4, Dc ) 1.434 g cm-3, µ (Cu-KR) ) 0.975 mm-1, F (000)
) 3088; 10905 independent reflections, R1 ) 0.053, wR2
)
0.129 for 8259 independent observed reflections [2θ e 124°, I
> 2σ(I)].
1
Macrocycle 6. Yield: 45.2%. Mp: 354 °C. H NMR (CH3-
SO3H:CD2Cl2 1:5, 300 MHz), δ (ppm): 8.10 (d, J ) 8.7 Hz, 4H),
7.86 (d, J ) 8.4 Hz 4H), 7.65 (d, J ) 8.4 Hz, 4H), 7.46 (d, J )

Macromolecules, Vol. 38, No. 25, 2005
Highly Strained Macrocyclic Biaryls 10415
Scheme 2. Reductive Elimination of a Highly
Strained Macrocyclic Ether-Ketone from an
Unstrained Diaryl-Nickel(III) Intermediate
Chart 1. Linear Precursors to Macrocyclic Aromatic
Ether-Ketones
Figure 1. X-ray structure of the linear precursor oligomer 1,
displaying a preferred conformation that is preorganized for
macrocyclization. The internal valence angles at the bridge-
bond atoms C(17), O(27), and C(37) are 122, 116, and 113°
respectively.
Chart 2. Strained Macrocyclic Aromatic
Ether-Ketones Obtained by Nickel-Catalyzed
Coupling Chemistry
To test this idea, a linear, bis-4-chloro-terminated
oligomer (1) was synthesized by reaction of the potas-
sium salt of 4,4′-hexafluoroisopropylidenediphenol with
4-chloro-4′-fluorobenzophenone, a reaction in which
nucleophilic attack on the fluoro-substituted (rather
than chloro-substituted) aromatic ring occurred with a
very high degree of selectivity. Precursor 1 was also
synthesized in good yield by Friedel-Crafts acylation
of hexafluoroisopropylidenebis(1,4-phenoxyphenylene)
with 4-chlorobenzoyl chloride. A series of analogous
reactions led to linear oligomers 2-4 (Chart 1).
The X-ray structure of linear oligomer 1 is shown in
Figure 1. The molecule has crystallographic C2 sym-
metry about an axis passing through the bridging
isopropylidene carbon, whose adjoining aromatic rings
are symmetrically skewed and subtend an angle of ca.
111° at this carbon. The diaryl ether unit has an
essentially orthogonal conformation, whereas the diaryl
ketone fragment is symmetrically skewed. Although the
molecule will have substantial conformational freedom
in solution, its solid-state structure shows that a
conformation favoring intramolecular coupling is clearly
favored.
The macrocyclic structure of compound 5 (Figure 2)
was confirmed by single crystal X-ray diffraction, which
showed the presence of two independent molecules with
very similar conformations in the unit cell. The hexaflu-
oroisopropylidenediphenylene unit has a conformation
virtually unchanged from that observed in the linear
precursor 1, with the angle at the bridging carbon atom
being ca. 111° in 5. The conformations of the pairs of
diaryl ether units are asymmetrically skewed so as to
accommodate the newly formed 4,4′-biphenylene unit.
The latter is twisted by about 30° about the biaryl
linkage and is severely distorted, with the terminal
arene-to-carbonyl bonds [C(29)-C(30) and C(1)-C(39)]
subtending an angle of ca. 126°sa deviation of 54° from
the conventional linear geometry of a biaryl unit.
The methodology developed for synthesis of macro-
cyclic ether-ketone 5 was extended to other strained
macrocyclic ether-ketones (Chart 2) by replacing the
hexafluoroisopropylidene linkage with other linking
units such as sulfone (macrocycle 6). The cyclic structure
of 6 was also confirmed by single-crystal X-ray analysis
(Figure 3) which showed the diaryl sulfone unit to adopt
a conventional open-book conformation (C-S-C ) ca.
107°), but with the conformation of the remainder of the
macrocycle little changed from that observed in 5. The
increased C-S bond lengths, relative to the correspond-
ing C-C linkages in 5, brings about a slight relaxation
of the overall macrocyclic strain, though the biphenyl
unit remains highly distorted, with the biarylarene-to-
carbonyl bonds now subtending an angle of ca. 129°.
Slow addition of a solution of 1 in N,N-dimethylac-
etamide (DMAc) to a solution of Ni(PPh3)4, generated
in situ from Ni(Py)4Cl2, PPh3, and Bu4NI by reaction
with excess powdered zinc in the same solvent, did
indeed result in intramolecular biaryl formation. Mac-
rocycle 5 (Chart 2) was isolated by column chromatog-
raphy in 28% yield and was characterized by IR, MS,
1
and H and 13C NMR spectroscopy and by elemental
analysis.

10416 Colquhoun et al.
Macromolecules, Vol. 38, No. 25, 2005
Figure 4. Molecular structure of macrocycle 7. Internal
valence angles at O(5), C(14), C(27), and O(31) are 120, 116,
115, and 119°, respectively.
Figure 2. Molecular structure of macrocycle 5. Internal
valence angles at C(1), O(8), C(15), O(22), and C(29) are 115
(114), 119 (120), 111 (111), 120 (119) and 115 (116)°, respec-
tively. Values in parentheses correspond to the second inde-
pendent molecule in the unit cell.
noxydibenzofuran with 4-chlorobenzoyl chloride and,
after purification, this was successfully cyclized to give
7 (mp 353 °C) in 54% yield. The X-ray structure of 7
(Figure 4) showed that replacement of the hexafluor-
oisopropylidenediphenylene and diphenylenesulfone
bridging units in 5 and 6 respectively by the more rigid
and more compact dibenzofuran fused-ring system
results in an appreciable increase in the level of mac-
rocyclic ring-strain, as reflected in the bowing of the
biphenylene unit. Here the two phenylene rings are
twisted by ca. 31° and their arene-to-carbonyl bonds
subtend an angle of ca. 115°, representing a 65° devia-
tion from linearity. It is surprising to note that in
structures of this type, the strain-distortion appears to
be almost entirely concentrated in the biphenylene-
dicarbonyl unit (the internal angle at the two carbonyl-
carbon atoms being also significantly contracted, from
120 to ca. 115°) whereas the remainder of the macro-
cycle is little distorted from ideal geometry.
A second example of a strained macrocycle containing
a fused-ring system was obtained from the linear
precursor 4, synthesized in 81% yield by reaction of 3,6-
dihydroxyxanthone with 4-chloro-4′-fluorobenzophe-
none. Macrocycle 8 (mp 410 °C) was produced in 25%
yield by cyclization of 4 under pseudo-high-dilution
conditions and crystallizes with two independent mol-
ecules in the unit cell. X-ray analysis (Figure 5) shows
that replacement of the dibenzofuran unit by a xanthone
residue brings about a significant relaxation of the
macrocyclic ring-strain, reflected in an increase in the
angle subtended by the two biphenylene-to-carbonyl
bonds to ca. 138° (from 115° in macrocycle 7). The
twisting of the biphenylene unit is essentially un-
changed at ca. 30°.
Figure 3. Molecular structure of macrocycle 6. Internal
valence angles at S(1), O(5), C(14), C(27), and O(31) are 107,
117, 116, 114, and 116°, respectively.
The unique ability of nickel(0)-catalyzed coupling to
generate strained macrocycles of this type was demon-
strated by the failure of efforts, in this work, to
synthesize macrocycle 6 by electrophilic (Friedel-
Crafts) acylation of 4,4′-diphenoxydiphenyl sulfone with
4,4′-biphenyldicarbonyl dichloride under pseudo-high-
dilution conditions.
Macrocycles 5 and 6 contain only isolated phenylene
rings, so that the structures are still relatively flexible,
but the introduction of fused aromatic rings such as
dibenzofuran and xanthone into the macrocyclic struc-
ture should increase the rigidity of the macrocycle, and
thus also enhance the glass transition temperature of
any polymer resulting from ring-opening polymeriza-
tion. The first such molecule obtained was the diben-
zofuran-based macrocycle 7. The linear precursor 3 was
obtained in high yield from the reaction of 2,8-diphe-
Synthesis of Strained Macrocyclic Ether)Sul-
fones. The Ni(0)-catalyzed coupling reaction is known
to be favored by electron-withdrawing substitutents,27
and so should also be applicable to the synthesis of
macrocyclic ether-sulfones by replacing the ketone unit
with the even more strongly electron-withdrawing sul-
fone group (Scheme 3). Molecular simulation studies
(Cerius-2) indicated that the cyclic dimer 17 would be
the smallest cyclic oligomer to be expected from Ni(0)-
catalyzed cyclization of precursor 9. Formation of the

Macromolecules, Vol. 38, No. 25, 2005
Highly Strained Macrocyclic Biaryls 10417
strained cyclic monomers of this type it seemed logical
either: (a) to move to longer chain (e.g., six-ring) linear
precursors, as used successfully in the synthesis of
strained macrocyclic polyketones, or (b) to reduce the
macrocyclic ring strain by replacing the para-substitut-
ed dioxybenzene unit in 17 with a meta- or ortho-
substituted aromatic ring. To test the first approach, a
six-ring linear precursor 11 was first synthesized by
electrophilic sulfonylation of 4,4′-diphenoxydiphenyl
sulfone with 4-chlorobenzenesulfonyl chloride in the
presence of iron(III) chloride. Although a pure oligomer
was obtained in moderate yield, the presence of mono-
sulfonylation products gave rise to difficulty in workup
and purification, and so alternative routes to linear
ether-sulfone precursors were investigated.
A general method for synthesis of poly(aryl ether-
sulfone)s involves nucleophilic aromatic substitution
reactions between bis(phenols) and sulfone-activated
aromatic dihalides in the presence of potassium carbon-
ate, and this approach has been used in the synthesis
of an oligomeric bis(phenol) by nucleophilic substitution
of 4,4′-dichlorodiphenyl sulfone with a 20-fold excess of
4,4′-isopropylidenediphenol.16 Such reactions can in
principle yield linear aromatic ether-sulfones with
either a single main product (by using a very large
excess of bis(phenol)), or else a series of linear oligomers
can be formed by working with a smaller excess. In the
present work, bis(chlorophenylene)-terminated precur-
sors were required for intramolecular coupling reactions,
so that here the activated dihalide (A) was always used
in excess over the bis(phenol) (B). If reactant A is excess
over reactant B, then several bis(chlorophenylene)-
terminated precursors (2A + B, 3A + 2B, 4A + 3B, etc.)
for macrocycle synthesis can in principle be obtained
from a single reaction. Precursor 11 was thus also
obtained in 15% yield by nucleophilic aromatic substitu-
tion of excess 4,4′-dichlorodiphenyl sulfone with 4,4′-
dihydroxydiphenyl sulfone in a 2.2:1 mole ratio. The
MALDI-TOF mass spectrum of the crude product
confirmed that a series of higher linear oligomers were
also formed, but attempted fractionation by column
chromatography failed to resolve these.
Nickel-catalyzed coupling of the linear ether-sulfone
precursor 11 again yielded a series of linear oligomers,
up to the heptamer, but chromatographic fractionation
also afforded the target cyclic monomer 19 (mp 400 °C)
though in only 8.0% yield. Single-crystal X-ray analysis
of 19 (Figure 6) showed that the molecule still has a
strained structure, although the degree of distortion of
the biphenyl residue is much smaller here than in the
corresponding diketone macrocycle 6. This reduction in
ring-strain results from (a) the narrower bond angle at
sulfone (C-S-C ) ca. 103° in 19, compared to ketone
C-C-C at ca. 115° in 6), and (b) the increase in bond
length on replacing the carbonyl bridges in 6 (C-C )
1.50 Å) by sulfone units (C-S ) 1.76 Å). The overall
effect of these changes is to open up the total angle
subtended by the terminal S-aryl bonds of the biphe-
nylene unit to 146°, from a value of 129° in the diketone
analogue 6. A comparatively rare diaryl sulfone confor-
mation is observed in 19, where two of these units
display a near-orthogonal relationship between their
adjoining aromatic rings (cf. the open-book or slightly
skewed conformations normally observed).
Figure 5. Molecular structure of macrocycle 8. Internal
valence angles at O(9), C(16), C(29) and O(36) are 122, 111,
115, and 118°, respectively.
corresponding cyclic monomer would not be expected
due to the extreme strain energy associated with such
a molecule. Cyclization of 9 was conducted under the
same conditions as those used for the synthesis of
macrocyclic ether-ketones, but characterization of the
crude product by MALDI-TOF mass spectrometry
showed that the product in fact comprised mainly a
series of linear, H-terminated oligomers up to the linear
pentamer. Fractionation by column chromatography did
afford the pure macrocyclic dimer 17, but in only 3.9%
yield. The formation of linear H-ended oligomers is
consistent with previous reports that Ni(0) can promote
hydrogenolysis of chloroarenes as a competing reaction
with biaryl formation,27a and the very low yield of
macrocyclic product obtained in this work suggests that
syntheses of cyclic biarylene-sulfones are especially
susceptible to this side-reaction. Computational model-
ing of the cyclic diarylnickel(III) intermediates (cf.
Scheme 2) for cyclization of analogous ketone- and
sulfone-based precursors indicates that the narrower
bond angle at sulfone often leads to a more strained
intermediate (see Supporting Information), and this
may explain why hydrogenolysis competes more ef-
fectively with cyclization in sulfone-based systems.
Cyclization of the four-ring ether-sulfone precursor
9 afforded cyclic dimer 17, but molecular simulation
studies suggested that the corresponding reaction of a
five-ring precursor 10 might possibly yield a cyclic
monomer. With the aim of obtaining such a macrocycle,
precursor 10 was synthesized by electrophilic sulfony-
lation of 1,4-diphenoxybenzene with 4-chlorobenzene-
sulfonyl chloride. Cyclization of precursor 10 was carried
out under conditions similar to those used in the
synthesis of macrocycle 17, and the crude product was
analyzed by MALDI-TOF mass spectrometry, which
again showed that the product mainly comprised linear
oligomers, now up to linear hexamer. Chromatographic
fractionation did afford a pure macrocyclic dimer 18 in
6.2% yield, but no trace of cyclic monomer was obtained,
indicating that the proposed monomeric five-ring ether-
sulfone macrocycle is in fact too highly strained to be
formed.
The cyclization of sulfone-based linear precursors
containing four or five p-substituted aromatic rings thus
afforded only cyclic dimers (17 and 18). To synthesize
A cyclic dimer (20) was also isolated from this
reaction, in 3.1% yield. The 1H NMR spectrum of
macrocycle 20 shows that the protons are all deshielded

10418 Colquhoun et al.
Macromolecules, Vol. 38, No. 25, 2005
Scheme 3. Synthesis of Macrocyclic Aryl
Ether-Sulfones by Ni(0)-Catalyzed Coupling
Figure 6. Molecular structure of macrocycle 19. Internal
valence angles at S(1), O(27), S(2), S(3), and O(57) are 106,
120, 102, 104, and 118°, respectively.
large strain-induced distortions occur within the 4,4′-
biphenylene unit, with the terminal S-C bonds sub-
tending an angle of ca. 110°, i.e., a deviation from
collinearity of 70°! This extreme level of strain is
reflected both in boatlike deformations of the rings
comprising the biaryl unit and in pyramidalization at
carbon atoms C(3) and C(6), which lie 0.09 and 0.11 Å
respectively out of the planes of their substituent atoms.
The internal angles at sulfone are compressed from a
conventional value of 105° to only 100°. Despite the
small ring size of this macrocyclescomprising only five
aromatic ringssthere is still a free pathway through its
center (defined by van der Waals radii) some 3.6 Å in
diameter.
The Ni(0)-catalyzed cyclization reactions so far de-
scribed all involved relatively short chain (five- or six-
ring) precursors, but cyclization of the more extended
(8-ring) precursor 13, isolated as a byproduct from the
synthesis of precursor 12, afforded a macrocyclic biaryl
(22) in 14.5% yield (Scheme 3). Further investigation
of nucleophilic oligomerization as a route to bis(chloro-
terminated) macrocycle precursors yielded oligomer 14
in 20.6% yield from the 1:1 reaction of 4,4′-bis(4-
chlorobenzenesulfonyl)-biphenyl with 4-chlorophenyl-4′-
hydroxyphenyl sulfone. The unsymmetrical precursor 14
(Scheme 3) then afforded, on cyclization, a chemically
symmetrical cyclic monomer (23) in 10% yield. The
structure of 23 was confirmed by single-crystal X-ray
analysis (Figure 8) which showed that in this molecule
the ring-strain is, uniquely, distributed between two
4,4′-biphenylene units (the two pairs of terminal S-aryl
bonds subtend angles of 139 and 146°), in contrast to
the other macrocycles described here where the ring
strain is concentrated at only a single biphenylene unit.
The disulfone analogue of the diketone-macrocycle 8
should still be significantly strained even though, as
noted above, the internal angle at sulfone is significantly
smaller than that at the carbonyl group. Precursor 15
was therefore synthesized by reaction of 3,6-dihydrox-
yxanthone with excess 4,4′-dichlorodiphenyl sulfone,
and cyclization of this precursor under pseudo-high-
dilution conditions afforded cyclic monomer 24 in 14%
isolated yield, together with a series of linear oligomers.
The cyclic structure of 24 was confirmed both spectro-
scopically and by single-crystal X-ray analysis (Figure
9). It is clear that macrocycle 24 is indeed less strained
than its ketone analogue 8, as evidenced by the reduced
Figure 7. Molecular structure of macrocycle 21. Internal
valence angles at S(7) and O(14) are 100 and 123°, respectively.
1
relative to their positions in the H NMR spectrum of
the smaller cyclic monomer 19, an effect which probably
results from a reduction in transannular ring-current
shielding in the larger macrocycle.
Reduction of macrocyclic ring strain by replacing the
para-substituted diaryloxybenzene in 10 with a meta-
substituted central ring was investigated using precur-
sors 12 and 13, which were both obtained from the
reaction of a slight excess of 4,4′-dichlorodiphenyl sul-
fone (A) with resorcinol (B). The MALDI-TOF mass
spectrum of the crude product showed a series of linear
oligomers [(2A + B), (3A + 2B), (4A + 3B) etc.], and
chromatographic fractionation of the crude product
afforded precursors 12 and 13 in 31.4% and 13.1% yield,
respectively. Cyclization of 12 yielded the five-ring cyclic
monomer 21 (mp 410 °C) in 26.7% yield, in contrast to
the result for precursor 10 where only cyclic dimer and
linear oligomers could be isolated. The X-ray structure
of 21 (Figure 7) showed the molecule to have crystal-
lographic C2 symmetry and to be very highly strained,
more so than any other biarylene-containing macrocycle
obtained in this (or indeed in any other) work. Very

Macromolecules, Vol. 38, No. 25, 2005
Highly Strained Macrocyclic Biaryls 10419
or phenoxide initiators and that, in keeping with the
high levels of ring-strain implicit in the structures
reported here, these polymerizations are highly exo-
thermic.4 The following paper analyzes the strain ener-
gies associated with macrocycles of this type and
describes a more detailed investigation of their ring-
opening polymerization chemistry.
4. Conclusions
Extremely strained macrocyclic biaryls containing
both ether-ketone and ether-sulfone units can be
obtained by homogeneous, nickel-catalyzed, intramo-
lecular coupling of short-chain bis(4-chlorophenyl)-
terminated oligomers under pseudo-high-dilution con-
ditions. Unstrained macrocycles can be obtained by
cyclization of longer-chain precursors. Crystallographic
analyses show that the disortions necessary to accom-
modate the high levels of ring-strain are generally
concentrated in the 4,4′-biphenylene unit(s), implying
an unsuspected degree of flexibility for this type of
molecular fragment.
Figure 8. Molecular structure of macrocycle 23. Internal
valence angles at S(1), O(8), S(15), and S(30) are 103, 117, 102,
and 100°, respectively.
Acknowledgment. We thank the University of
Reading for grants in support of this work and Univer-
sities U.K. for the award of an Overseas Research
Studentship to Z.Z. Experimental assistance in the early
stages of this work was provided by Mr. M. Thomas.
Supporting Information Available: Text giving syn-
thetic procedures and characterization data for the precursor
oligomers 1, 2, 3, 4, 9, 10, 11, 12, 13, 14, 15, and 16, and for
macrocycles 17, 18 19, 21, 22, 23, 24, 25, and 26, Figure S1,
showing energy-minimized models (re-optimised Dreiding-II
forcefield) for cyclic organo-nickel intermediates (A and B) in
cyclization of the ketone- and sulfone-based precursors 6 and
11, respectively, and figures showing thermal ellipsoid plots
and CIF files giving full crystallographic data for precursor
oligomer 1 and macrocycles 6, 8, 19, 23, and 24. This material
is available free of charge via the Internet at http://pubs.ac-
s.org. Crystallographic data for compounds 5, 7, and 21 were
deposited at the Cambridge Crystallographic Data Centre at
the time of their publication in preliminary communications,
and they have the following CCDC refcodes: RUYYAL, KEG-
CAA, and RUYXIS respectively.
Figure 9. Molecular structure of macrocycle 24. Internal
valence angles at S(1), S(27), O(37), and O(66) are 103, 103,
120, and 120°, respectively.
References and Notes
(1) For a recent overview see: Hergenrother, P. M. High Perform.
Polym. 2003, 15, 3-45.
bowing of the biphenylene unit; the total angle sub-
tended by the terminal S-aryl bonds increases from a
value of 138° in macrocycle 8 to 148° in macrocycle 24.
(2) This approach was initially optimised for the synthesis of
bisphenol A polycarbonate, though macrocyclic oligomers
have long been known to form as coproducts of step-growth
polymerization. See: (a) Jacobson, H.; Stockmayer, W. H. J.
Chem. Phys. 1950, 18, 1600-1606. (b) Brunelle, D. J.; Boden,
E. P.; Shannon, T. G. J. Am. Chem. Soc. 1990, 112, 2399-
2402. (c) Brunelle, D. J.; Shannon, T. G. Macromolecules
1991, 24, 3035-3044.
(3) Reviews: (a) Jayakannan, M.; Ramakrishnan, S. Macromol.
Rapid Commun. 2001, 22, 1462. (b) Hodge, P.; Colquhoun,
H. M.; Williams, D. J. Chem. Ind. (London) 1998, 162-167.
(c) Wang, Y. F.; Chan, K. P.; Hay, A. S. React. Funct. Polym.
1996, 30, 205-227.
(4) Colquhoun, H. M.; Dudman, C. C.; Thomas, M.; Mahoney,
C. A.; Williams, D. J. J. Chem. Soc., Chem. Commun. 1990,
336-339.
(5) Colquhoun, H. M.; Zhu, Z. X.; Williams, D. J. Org. Lett. 2001,
3, 4031-4034.
(6) Cella, J. A.; Talley, J. J.; Fukuyama, J. M. Polym. Prepr. (Am.
Chem. Soc., Div. Polym. Chem.) 1989, 30 (2), 581.
(7) (a) Teasley, M. F.; Hsiao, B. S. Macromolecules 1996, 29,
6432-6441. (b) Teasley, M. F.; Wu, D. Q.; Harlow, R. L.
Macromolecules 1998 31, 2064-2074.
(8) Ben-Haida, A.; Colquhoun, H. M.; Hodge, P.; Williams, D. J.
J. Mater. Chem. 2000, 10, 2011-2016.
The methodology for synthesis of the strain-free
macrocycle 22, by cyclization of an eight-ring ether-
sulfone precursor oligomer (13) was also extended to a
ten-ring precursor. Nucleophilic substitution of 4,4′-
biphenol with excess 4,4′-dichlorodiphenyl sulfone gave
a series of linear oligomers, as demonstrated by MALDI-
TOF MS, and the 2:3 condensation product 16 was
isolated chromatographically in 13.1% yield. Cyclization
of oligomer 16 was achieved under the same conditions
as those used for the eight-ring precursor 13, yielding
the ten-ring cyclic monomer 25 in 12.5% yield and the
20-ring cyclic dimer 26 in 6.6% yield.
Since all the macrocyclic compounds reported here
contain aromatic ether linkages activated toward nu-
cleophilic cleavage by electron-withdrawing carbonyl or
sulfone substituents, potential clearly exists for nucleo-
philically initiated ring-opening polymerization of these
macrocycles. Preliminary results4 have shown that such
polymerizations can indeed be achieved using fluoride

10420 Colquhoun et al.
Macromolecules, Vol. 38, No. 25, 2005
(9) Baxter, I.; Colquhoun, H. M.; Hodge, P.; Kohnke, F. H.; Lewis,
D. F.; Williams, D. J. J. Mater. Chem. 2000, 10, 309-314.
(10) (a) Chan, K. P.; Wang, Y. F.; Hay, A. S. Macromolecules 1995,
28, 653-655. (b) Chan, K. P.; Wang, Y. F.; Hay, A. S.
Macromolecules 1995, 28, 6705-6717. (c) Gao, C. P.; Hay,
A. S. Polymer 1995, 36, 4141-4146. (d) Wang, Y. F.; Paventi,
M.; Chan, K. P.; Hay, A. S. J. Polym. Sci., Part A: Polym.
Chem. 1996, 34, 375-385.
(18) Baxter, I.; Ben-Haida, A.; Colquhoun, H. M.; Hodge, P.;
Kohnke, F. H.; Williams, D. J. Chem.sEur. J. 2000, 6, 4285-
4296.
(19) Sergeev, V. A.; Nedel’kin, V. I.; Astankov, A. V.; Nikiforov,
A. V.; Alov, E. M.; Moskvichev, Yu. A. Izv. Akad. Nauk SSSR
Ser. Khim. 1990, 4, 854-858.
(20) Chen, M. F.; Guzei, I.; Rheingold, A. L.; Gibson, H. W.
Macromolecules 1997, 30, 2516-2518.
(21) Zolotukhin, M. G.; Colquhoun, H. M.; Sestiaa, L. G.; Rueda,
D. L.; Flot, D. Macromolecules 2003, 36, 4766-4771.
(22) Colquhoun, H. M.; Dudman, C. C.; Thomas, M. Eur. Patent
317,226, 1989 (to ICI).
(23) Ridd, J. H.; Yousaf, T. I.; Rose, J. B. J. Chem. Soc., Perkin
Trans. 2 1988, 1729.
(24) Bridger, R. F.; Kinney, R. E.; Williams, A. L. US Patent
3 450 772, 1969 (to Mobil).
(25) Attwood, T. E.; Barr, D. A.; Feasey, G. G.; Leslie, V. J.;
Newton, A. B.; Rose, J. B. Polymer 1977, 18, 354.
(26) Darsow, G. German Patent 1 932 068, 1971 (to Bayer).
(27) (a) Colon, I.; Kelsey, D. R. J. Org. Chem. 1986, 51, 2627-
2637. (b) Kwiatkowski, G. T.; Colon, I.; El-Hibri, M. J.;
Matzner, M. Makromol. Chem. Macromol. Symp. 1992, 54/
55, 199-224. (c) Kwiatkowski, G. T.; Matzner, M.; Colon, I.
J. Macromol. Sci. Pure Appl. Chem. 1997, A34, 1945-1975.
(d) Ben-Haida, A.; Colquhoun, H. M.; Hodge, P.; Williams,
D. J. J. Mater. Chem. 2000, 10, 309.
(11) Ding, Y.; Hay, A. S. Macromolecules 1996, 29, 3090-3095.
(12) Colquhoun, H. M.; Sestiaa, L. G.; Zolotukhin, M. G.; Williams,
D. J. Macromolecules 2000, 33, 8907-8909.
(13) Shibata, M.; Yosomiya, R.; Chen, C. H.; Wang, J. Z.; Wu, Z.
W. Angew. Makromol. Chem. 1998, 263, 15-20.
(14) Shibata, M.; Yosomiya, R.; Chen, C.; Zhou, H.; Wang, J. Z.;
Wu, Z. Eur. Polym. J. 1999, 35, 1967-1974.
(15) (a) Chen, M. F.; Gibson, H. W. Macromolecules 1996, 29,
5502-5504. (b) Chen, M. F.; Fronczek, F.; Gibson, H. W.
Macromol. Chem. Phys. 1996, 197, 4069-4078. (c) Xie, D. H.;
Ji, Q.; Gibson, H. W. Macromolecules 1997, 30, 4814-4827.
(16) (a) Ganguly, S.; Gibson, H. W. Macromolecules 1993, 26,
2408-2412. Xie, D. H.; Gibson, H. W. (b) Xie, D. H.; Gibson,
H. W. Macromol. Chem. Phys. 1996, 197, 2133-2148.
(17) (a) Ben-Haida, A.; Baxter, I.; Colquhoun, H. M.; Hodge, P.;
Kohnke, F. H.; Williams, D. J. Chem. Commun. 1997, 1533-
1534. (b) Baxter, I.; Ben-Haida, A.; Colquhoun, H. M.; Hodge,
P.; Kohnke, F. H.; Williams, D. J. Chem. Commun. 1998,
2213-2214. (c) Colquhoun, H. M.; Lewis, D. F.; Hodge, P.;
Ben-Haida, A.; Williams, D. J.; Baxter, I. Macromolecules
2002, 35, 6875-6882.
MA0517796