The dynamic covalent chemistry of mono- and bifunctional boroxoaromatics

Article abstract of DOI:10.1016/j.tet.2006.12.034

10-Hydroxy-10,9-boroxophenanthrene reacts rapidly and reversibly with both benzylic and alkane diols in non-polar solvents. The formation of 2:1 adducts between the boroxoaromatic and the diols is favoured. The diol component of the adduct can be exchange

Full text of DOI:10.1016/j.tet.2006.12.034

Tetrahedron 63 (2007) 2391–2403
The dynamic covalent chemistry of mono- and
bifunctional boroxoaromatics
*
Lyndsey M. Greig, Alexandra M. Z. Slawin, Melanja H. Smith and Douglas Philp
EaStCHEM and Centre for Biomolecular Sciences, School of Chemistry, University of St Andrews, North Haugh,
St Andrews, Fife KY16 9ST, United Kingdom
Received 29 June 2006; revised 7 December 2006; accepted 14 December 2006
Available online 17 December 2006
Abstract—10-Hydroxy-10,9-boroxophenanthrene reacts rapidly and reversibly with both benzylic and alkane diols in non-polar solvents.
The formation of 2:1 adducts between the boroxoaromatic and the diols is favoured. The diol component of the adduct can be exchanged
readily and rapidly by treatment of the boroxoaromatic–diol adduct with an alternative diol in solution at room temperature. This reversible
covalent chemistry would appear to be ideal for the dynamic assembly of more complex superstructures. However, attempts to extend this
dynamic equilibrium to the assembly of macrocycles using the bifunctional boroxoaromatic 5,9-dihydroxy-5,9-dibora-4,10-dioxopyrene
failed as a result of changes in the reactivity of the boron centre in the bifunctional boron-containing compound.
Ó 2006 Elsevier Ltd. All rights reserved.
1. Introduction
of synthetic supramolecular self-assembled architectures in
which the assembly process is directed by selective interac-
tions, for example, hydrogen bonds6 or metal–ligand7 coor-
dination. There is, however, one major disadvantage to the
use of non-covalent interactions in synthetic self-assembling
systems. On formation of the superstructure, a large and un-
favourable change in entropy arises from the ordering of the
components. The formation of non-covalent bonds affords
only a small amount of enthalpic stabilization per interaction
and, therefore, a significant number of these interactions is
required before this stabilization offsets the entropic loss.
The construction1 of large, regular superstructures in biologi-
cal systems is usually accomplished through the application
of self-assembly processes. Examples include the folding2 of
proteins, the stepwise assembly3 of the tobacco mosaic virus
and the self-regulated construction4 of the virus T4 bacterio-
phage. Several key principles govern self-assembly pro-
cesses: self-assembly processes are highly convergent and,
thus economical in the amount of information required to
encode the target structures. In general, self-assembling
systems use a relatively small number of simple and often
identical units to create functionally and structurally com-
plex arrays. Therefore, self-assembling systems typically re-
quire only a small set of different interactions to create the
final molecular assembly. The entire assembly process is a
dynamic one in which all the assembly/disassembly steps
are reversible. This reversibility is most often accomplished
using weak non-covalent interactions. The result of this
reversibility is a degree of error checking. Incorporation of
an incorrect subunit results in an increase in free energy
and is not tolerated. These spurious additions to the structure
are reversed in a drive to minimize the total free energy of
the system. Given these advantages, it is desirable for chemi-
sts to mimic self-assembling systems of this kind. Indeed,
considerable effort has been directed towards the formation5
One answer to this problem lies in the use of covalent bonds
as a means of constructing large regular assemblies. Cova-
lent bonds are stronger than hydrogen bonds and, hence,
a smaller number is required to produce a stable assembly.
Traditionally, covalently linked structures have been synthe-
sized using sequential, kinetically controlled, and irrevers-
ible reactions. However, the linear formation of such large
structures requires a high number of synthetic steps and is
extremely time consuming. To date, the use of covalent in-
teractions in the field of self-assembly has been relatively
under-exploited. The use of covalent bond formation8 in
self-assembly processes requires that these bonds form un-
der thermodynamic control, i.e., bond formation is revers-
ible, as, for example, in the cases of disulfides9 or borate10
esters. An example of a system that employs covalent bonds
in a thermodynamically controlled assembly process has
been described by Sanders and co-workers.11 The cycliza-
tion of the cinchonidine monomer, methyl-10,11-dihydro-
quinine-11-carboxylate, might be expected12 to generate
Keywords: Self-assembly; Molecular recognition; Aromatic compounds;
Dynamic covalent chemistry; Reversible reactions.
* Corresponding author. Tel.: +44 1334 467264; fax: +44 1334 463808;
e-mail: d.philp@st-andrews.ac.uk
0040–4020/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2006.12.034

2392
L. M. Greig et al. / Tetrahedron 63 (2007) 2391–2403
a wide range of oligomers. However, in practice, the trimer is
found to be the most stable product. The preference to form
the trimer is not a result of preorganization within the mono-
mer, but rather a result13 of predisposition. Ipaktschi and co-
workers14 have also exploited reversible covalent bond
formation in a self-assembled synthetic strategy. The ulti-
mate expression of the power of dynamic covalent chemistry
for the assembly of complex superstructures is described
in the synthesis of Borromean rings by Stoddart and co-
workers.15 In this system, a complex set of interlocked
macrocycles is created in a single step by metal-directed dy-
namic imine chemistry. The field of dynamic covalent chemi-
stry is hampered, however, by the relatively small group16
of reactions that are truly reversible under conditions that
are compatible with a wide range of functional groups.
dimeric macrocycle 2 by exploiting reversible homoanhy-
dride formation. Although the reversible chemistry18 pro-
vided by borazaaromatics is attractive building blocks for
use in covalent self-assembly processes, the spontaneous
formation of homoanhydrides limits their application19 in
this context by making the incorporation of structurally di-
verse subunits difficult. Additionally, their low reactivity to-
wards other nucleophiles limits the complexity of structures
that can be assembled using these building blocks. Through
the model compound 3, we have demonstrated20 that the cor-
responding oxygen-containing boroxoaromatics are suitable
compounds to extend the use of boron-containing hetero-
aromatics in covalent self-assembly processes. In solution,
they do not form homoanhydrides, but do form mixed esters
with alcohols reversibly and rapidly, even at low (<ꢀ40 ꢁC)
temperatures.
We have described17 the assembly of a dimeric macrocycle
through the exploitation of a novel class of compounds—
borazaaromatics. When the monomer 1 is dissolved in ace-
tone solution, the compound reacts with itself to form a
Here, we describe the design and synthesis of a bifunctional
boroxoaromatic and its behaviour in both solution and solid
states and attempts to exploit its solution phase properties in
dynamic assembly processes.
Drawing from our previous work on bis-borazaaromatics we
identified the bis-boroxoaromatic 4 as a suitable synthetic
target. We envisaged that this compound could react (Fig. 1)
with a series of mono-, di-, and triols to give a range of self-
assembled structures. Based on a rigid pyrene skeleton,
compound 4 should react rapidly and reversibly with nucleo-
philes such as alcohols in a similar fashion to that observed
for 3. It should be possible to monitor the formation of linear
hetero-oligomers 5 in solution (Fig. 1). The reaction between
4 and a nucleophile in solution would be under thermody-
namic control and should give the most stable product that
may be macrocyclic or polymeric.
1
N
N
H
B
OH
O
B
H
OH
– H2O
+ H2O
N
N
N
H
H
B
O
O
B
H
H
O
O
B
B
N
2
2. Synthesis
Compound 4 can be accessed readily from the precursor
diphenol 5 (Scheme 1), which in turn can be constructed
by applying standard metal-catalyzed cross-coupling meth-
odology21 to an appropriately protected 2,6-dihydroxyhalo-
genobenzene.
O
B
3
OH
Addition22 (Scheme 1) of sodium hydrogen carbonate to an
equimolar mixture of 1,3-dihydroxybenzene 6 and iodine in
water resulted in the precipitation of 2,4,6-triiodoresorcinol,
which was removed by filtration. Work-up of the filtrate
afforded 2-iodoresorcinol 7. Protection of the phenolic
groups as their benzyl ethers followed by metal-catalyzed
cross-coupling using either Kharasch or Suzuki methodol-
ogy afforded the benzyl-protected 2,6-dihydroxybiphenyl
9. Deprotection of the benzyl ethers using standard method-
ology afforded the target phenol 5. An alternative and shorter
synthesis was also employed. Treatment23 of 1,3-dimethoxy-
benzene 10 with t-BuLi led to selective deprotonation
of the 2-position of the aromatic ring, and reaction with iodine
gave 2,6-dimethoxyiodobenzene 11 in moderate yield. The
yield of the subsequent Ni-catalyzed Kharasch cross-
coupling with PhMgBr was found to vary greatly with
the identity of the catalyst. With [Ni(dppe)Cl2], the desired
2,6-dimethoxybiphenyl 12 was obtained only in low yield
(18%), with the homocoupling product, biphenyl, accounting
– H2O
+ H2O
O
B
O
B
O
B
O
B
HO
OH
HO
O
OH
n
Linear Oligomers
– H2O
+ H2O
4
+
HO
OH
Diol
O
B
O
B
O
O
n
Cyclic Oligomers
Figure 1. Bis-boroxoaromatic 4 may react with a diol in solution to form
linear oligomers and/or cyclic oligomers reversibly.

L. M. Greig et al. / Tetrahedron 63 (2007) 2391–2403
2393
and the O–B bond is cleaved before the cyclization has taken
place.
O
B
O
B
HO
OH
RO
OR
HO
OH
A third method (Scheme 2) employed uncomplexed BHCl2
as the boron source. BHCl2 may be generated in situ by
exploiting the greater Lewis acidity of BCl3. Reaction of
BHCl2$Me2S with BCl3 in toluene results in the release of
BHCl2 from its complex as the Me2S associates preferen-
tially with the stronger Lewis acid and, under the conditions
employed, the BCl3$Me2S complex precipitates from solu-
tion. The solution of BHCl2 was then transferred by a syringe
and used to treat a solution of 5 in toluene at room temper-
ature. Formation of the intermediate adduct was monitored
by the formation of H2 gas. Once adduct formation was com-
plete, the addition of AlCl3 and heating effected cyclization
to give the target compound in good yield.
X
4
5
6
7
10
MeO
(f)
OMe
HO
(a)
OH
67%
63%
11
12
HO
OH
MeO
(g)
OMe
76%
I
I
(b)
89%
8
MeO
OMe
BnO
OBn
I
HO
OH
O
B
O
B
(a)
(c) or (d) 16% or 44%
87%
HO
OH
(h)
89%
5
4
Scheme 2. Reagents and conditions: (a) (i) BHCl2, PhMe and (ii) AlCl3,
PhMe, D.
HO
OH
BnO
OBn
(e)
9
5
80%
In solution, the O–H bond vectors in 4 can adopt a variety of
conformations with respect to the B–O bond, we label these
conformations anti–anti, syn–anti and syn–syn (Fig. 2). The
singlet resonance at d 8.45 arises from the exchangeable
hydroxyl proton He, since this signal disappears slowly when
a drop of D2O is added to the NMR sample. By comparison
with the resonances observed in the 1H NMR spectra of the
precursors to 4, we assign the doublet at d 8.39, just upfield
of the resonance arising from He to Hc. The other resonances
Scheme 1. Reagents and conditions: (a) I2, NaHCO3, H2O; (b) PhCH2Br,
K2CO3, 18-C-6, Me2CO; (c) PhB(OH)2, Pd(PPh3)4, Na2CO3, DME,
EtOH, H2O, 18 h; (d) PhMgBr, [Ni(dppe)Cl2], Et2O, 18 h; (e) Pd/C, H2,
MeOH, rt; (f) t-BuLi, THF, 0 ꢁC, I2, THF; (g) PhMgBr, [Ni(PPh3)2Cl2],
Et2O, D and (h) BBr3, CH2Cl2.
for the majority of the recovered material. The yield was
significantly improved (78%) on changing the catalyst to
[Ni(PPh3)2Cl2], and, furthermore, this method used only
2 mol % catalyst, rather than the 10 mol % required previ-
ously. The coupling was also attempted using Stille and
Suzuki methodology, but neither resulted in an improved
yield. Deprotection of 12 using BBr3 gave the diphenol 5
in a 43% overall yield starting from 10.
The cyclization of 5 to afford the target 5,9-dibora-4,10-
dioxapyrene 4 proved problematic. In all, three different ap-
proaches were employed. When the same conditions utilized
in the synthesis of boroxophenanthrene 3, i.e., reaction of 5
with BCl3, and subsequent cyclization with AlCl3, only
starting material was recovered. Reaction of 5 with BCl3
results in the evolution of HCl, and it is possible that the
initial O–BCl2 adduct is not stable under the increasingly
acidic reaction conditions. With this in mind, we employed
BHCl2$Me2S as the boron source, hoping that this would
solve this problem, however, once again, only starting mate-
rial was recovered from the mixture. It is possible to follow
the reaction of 5 with BHCl2 by measuring the volume of H2
gas produced during the process. In all cases, the volume of
gas was found to be consistent with the volume expected for
the formation of the intermediate oxoborondichloride ad-
duct. Calculations have shown that the boron–oxygen bond
is somewhat weaker than the boron–nitrogen bond in bor-
azaaromatics. It may be that the oxoborondichloride adduct
is simply too fragile to withstand the reaction conditions,
Figure 2. (a) The possible conformations of the O–H bond with respect to
the B–O bond in 4. (b) Partial 300 MHz 1H NMR spectrum of 4, recorded at
293 K in d6-acetone. The resonance arising from He is assigned on the basis
of its disappearance on addition of D2O. The remaining four resonances can
be tentatively assigned by COSY analysis.

2394
L. M. Greig et al. / Tetrahedron 63 (2007) 2391–2403
are then readily assigned (Fig. 2) by COSY experiments. 1H
NMR spectroscopic analysis in d6-acetone shows the bis-
hydroxy monomer 4 only, as expected by analogy with 3.
Unfortunately, compound 4 is not soluble in CDCl3, there-
fore we cannot determine with absolute certainty whether
4 is capable of reacting with itself in non-polar solvents.
Some evidence for the preference of the syn–syn conforma-
tion comes from the lack of any observable NOE between
protons He and Hc. Although He is, in principle, exchange-
able, we have, in the past, been able to detect NOEs to and
from this type of proton as their exchange rates are very slow.
From this observation it is evident that the hydroxyl group
adopts a syn conformation with respect to the boron–oxygen
bond, rather than an anti conformation (in which an NOE
effect would be observed, as, for example, in the case of 1).
3. Electronic structure calculations
In order to compare the structural and electronic properties of
4 with the mono-functional compound 3 we carried out elec-
tronic structure calculations at the HF/6-31G(d,p) level of
€
theory. Selected bond lengths and Lowdin bond orders in 4
and, for ease of comparison, 3 are shown in Table 1. It would
appear that the incorporation of the second boroxoaromatic
ring in 4 has little or no structural effect when the data are
compared with that for compound 3. The calculated structure
is planar, and both O–H bond vectors are syn with respect
to the endocyclic B–O bond. At this level of theory, syn–
syn-4 is around 30 kJ molꢀ1 more stable than anti–anti-4.
As in 3, the hybridization of the exocyclic oxygen is
confirmed to be closer to sp2 than sp3 by the C–B–Oexo and
Oendo–B–Oexo bond angles (123ꢁ and 118ꢁ, respectively),
and the calculated bond orders predict considerable p-char-
acter for both the endocyclic and exocyclic O–B bonds.
Figure 3. (a) Packing of compound 4 in the solid state. Hydrogen bonds are
shown as dashed lines and the outline of the unit cell is shown as a solid line.
(b) Molecular structure of compound 4 determined by single crystal X-ray
diffraction. Carbon atoms are coloured green, boron atoms are mauve,
oxygen atoms are red and hydrogen atoms are light grey.
€
Table 1. Selected bond lengths and Lowdin bond orders in the calculated
(HF/6-31G(d,p)) structures of 4 and 3
˚
Length/A
Compound
Bond
Bond order
direction of the crystal. This difference in propagation has
a dramatic effect on the hydrogen bonding distances in the
two arrays. In array A, the hydrogen bonds are significantly
3
Endocyclic O–B
Exocyclic O–B
C–endocyclic O
C–B
1.37
1.35
1.35
1.55
1.32
1.45
1.17
1.04
˚
shorter (O–H/O distance¼1.867 A; O–H/O angle¼
ꢁ
˚
˚
160.3 ; O/O distance¼2.809 A) than those present in array
syn,syn-4
Endocyclic O–B
Exocyclic O–B
C-endocyclic O
C–B
1.37
1.35
1.35
1.55
1.30
1.45
1.18
1.04
ꢁ
B (O–H/O distance¼1.898 A; O–H/O angle¼172.0 ;
˚
O/O distance¼2.869 A). This difference may reflect more
effective polarization of the hydrogen bond donor atoms in
the cyclic array as compared to the extended array. Each mol-
ecule of 4 is associated with one array of type A and one array
of type B and the A and B arrays are related by translation
in the a and b directions. Additionally, molecules of 4 are
engaged in offset aromatic stacks along the c direction.
4. Solid state structure
Single crystals of 4, suitable for analysis24 and structure
determination by X-ray diffraction, were grown by slow
evaporation of a solution of 4 in acetonitrile. Compound 4
crystallizes in the space group I41/a. The packing of 4
(Fig. 3a) is dominated by cyclic arrays (A and B) of four
O–H/O hydrogen bonds between hydroxyl groups attached
to the boron atoms of 4. Interestingly, these cyclic arrays
propagate in different ways through the structure. In array
A, the hydrogen bonds are between four molecules of 4
at the same level in the stack—the array is discrete. By con-
trast, array B is helical and extends continuously along the c
The molecular structure of 4 in the solid state determined by
X-ray diffraction (Fig. 3b) is in close agreement with that
calculated (Table 1) using electronic structure methods.
5. Solid state reactivity
The solid state structure of 4 revealed that the hydroxyl
groups in one molecule of 4 with a boron atom in the

L. M. Greig et al. / Tetrahedron 63 (2007) 2391–2403
2395
Figure 4. DSC trace recorded on heating a sample of 4 from 100 to 350 ꢁC,
at 50 ꢁC minꢀ1. The onset of the solid state dehydration of the sample is
clearly visible at 220 ꢁC.
molecule of 4 in the layer above are in relatively close prox-
˚
imity (<4.5 A). This observation suggested that reaction
among molecules of 4 may well occur in the solid state.
We therefore investigated the behaviour of 4 in the solid state
by thermal analysis and powder X-ray diffraction.
Samples of 4 do not melt below 400 ꢁC, which is suggestive
of polymerization25 occurring in the solid state. Initially,
the solid state reactivity of 4 was investigated (Fig. 4) using
differential scanning calorimetry (DSC). When heated to
350 ꢁC at a rate of 50 ꢁC minꢀ1, an endotherm is observed
with an onset temperature of 220 ꢁC. When the sample is
subsequently cooled at the same rate, no exotherm is ob-
served. Therefore, this endotherm is clearly not a result of
the sample melting, and therefore suggests a phase change,
perhaps as a result of a chemical reaction. The DSC analysis
also exhibits a second small endotherm with an onset tem-
perature of around 175 ꢁC. The origins of this event have,
so far, not been identified, but it may be a consequence of
reorganization within solid state structure prior to reaction.
Figure 5. (a) Powder X-ray diffraction pattern for a sample of 4, recorded at
room temperature. (b) Powder X-ray diffraction patterns recorded (i) at
30 ꢁC, for a sample of 4, and (ii) after heating a sample of 4 at 300 ꢁC for
180 min.
The powder X-ray diffraction pattern of 4 at room tempera-
ture is shown in Figure 5a and is in close agreement with the
pattern predicted on the basis of the single crystal X-ray
analysis. In order to probe the change in the solid state
structure of 4 that occurs on heating, we employed variable
temperature powder X-ray diffraction experiments. As the
sample was heated to 300 ꢁC, the reflections gradually
became less intense. However, even after heating for 3 h,
the sample does not appear to be completely amorphous
(Fig. 5b). Clearly a major structural change is occurring
within the sample.
the poly(anhydride) would result in a mass loss of 7.6% of
the total. Initially, a small sample (typically 3 mg) of 4
was heated from 30 to 300 ꢁC at a rate of 50 ꢁC minꢀ1. At
this rate, a slight loss of mass from the sample is visibly evi-
dent at temperatures around 75 ꢁC, and above (Fig. 6a). At
about 200 ꢁC, the rate of mass loss is noticeably increased,
until a total mass loss of around 8% has occurred. After
this point, a further slower mass loss is recorded, possibly
as a result of sublimation or decomposition of the sample
at the highest temperatures employed in the experiment. It
is important to note that the mass loss observed by TGA is
synchronous with the endotherm recorded by DSC. In order
to study the mass loss more accurately, a sample of 4 was
heated rapidly (100 ꢁC minꢀ1) to 200 ꢁC, in order to mini-
mize any mass loss prior to the subsequent isothermal stage
(Fig. 6b). The sample was then held at 200 ꢁC for 10 min
and a mass loss of around 8% was again observed, which
is consistent with the formation of the poly(anhydride) 13.
In order to probe this change further, we decided to probe for
a chemical change by employing thermogravimetric analy-
sis. If the sample of 4 was converted to the oligo(anhydride)
13 under heating, we would expect water to be evolved
and, at temperatures above 100 ꢁC, it would be lost from
the sample. We would therefore expect that the sample
mass would decrease. In fact, complete conversion of 4 to

2396
L. M. Greig et al. / Tetrahedron 63 (2007) 2391–2403
Figure 7. (a) MALDI-TOF mass spectrum for a sample of 4 recorded after
heating at 200 ꢁC for 10 min. The spectrum shows oligomers 4 up to the
hexamer (n¼4). (b) MALDI-TOF mass spectrum of a sample of 4, which
had not been heated above 30 ꢁC.
In order to confirm the formation of the poly(anhydride) 13,
the sample was analyzed by MALDI-TOF spectrometry. It
was necessary to dissolve the sample in acetone prior to
application to the MALDI sample plate. Therefore, in order
to minimize hydrolysis of any higher oligomers present in the
˚
sample, the acetone was pre-dried over 4 A molecular sieves.
MALDI-TOF mass spectral analysis of the sample resulting
from the isothermal heating confirms the formation of oligo-
mers up to the hexamer (Fig. 7, n¼4). Although the most
intense peak is due to the monomer 4, it should be noted
that MALDI spectrometry does not provide any quantitative
measurement of the proportions of oligomers present.
Figure 6. (a) Heating a sample of 4 from 30 to 300 ꢁC results in continuous
mass loss above 75 ꢁC. The results shown were recorded over a period of
around 10 min. One in every 30 data points are shown. (b) Thermogravimet-
ric analysis of a sample of 4. Region ꢀA1: held at 30 ꢁC for 5 min before heat-
ing to 200 ꢁC at a rate of 100 ꢁC min . Region B: isothermal at 200 ꢁC for
15 min resulting in a mass loss of around 8%. Region C: continued heating
For comparison purposes, the MALDI-TOF mass spectrum
of a sample of 4 dissolved in acetone (Fig. 6b), recorded
prior to heating under the conditions described above, shows
that the monomer (m/z 238) is the dominant species present
in this sample. We are therefore certain that the higher oligo-
mers observed in Figure 7 are representative of the sample
composition and this distribution of compounds is a result
of sample heating and not the result of dehydration occurring
during the recording of the spectrum.
to 700 ꢁC at a rate of 50 ꢁC minꢀ1
.
Continued heating from 200 to 700 ꢁC, leads only to a further
mass loss, due to sample decomposition.
O
B
O
B
4
HO
OH
6. Solution phase reactivity
– (n+1)H2O
We have demonstrated that 10-hydroxy-10,9-boroxophe-
nanthrene is capable of reacting with oxygen nucleophiles,
such as alcohols with loss of water to form covalent adducts,
which retain the boroxoaromatic ring. We were therefore in-
terested to explore the solution phase formation of oligomers
by compound 4. Potentially, reaction with bifunctional
O
B
O
B
O
B
O
B
O
B
O
B
HO
OH
O
O
n
13

L. M. Greig et al. / Tetrahedron 63 (2007) 2391–2403
2397
Figure 9. (a) The 160.4 MHz 11B NMR spectrum of 4 recorded at 293 K in
d6-acetone. (b) The 160.4 MHz 11B spectrum of a 1:1 solution of 4 and 17 in
d6-acetone. Both spectra were recorded using B(OMe)3 in d6-acetone as an
external reference (d¼0.0). The poor signal-to-noise ratio in (b) is a result of
the poor solubility of the adduct formed by 4 and 17 in d6-acetone.
insoluble in solvents such as chloroform and acetonitrile,
these experiments were performed in d6-acetone. The sol-
vent used in these experiments was dried thoroughly over an-
hydrous CaSO4 prior to use and samples were left for a period
of 24 h in the presence of anhydrous CaSO4 prior to analysis.
Disappointingly, in the cases of the reaction between 4
1
and 14, 15 and 16, 300 MHz H NMR spectra, recorded
in d6-acetone showed only resonances characteristic of the
unreacted starting materials. In the case of the reaction be-
tween 4 and 17, 11B NMR spectroscopy indicated that there
were significant differences (Fig. 9) between the starting
material 4 and the product of the reaction between 4 and 17.
Figure 8. Calculated (AM1) structures of the 2+2 macrocycles formed from
the reaction between: (a) 4 and 14, and (b) 4 and 15.
Whilst the 11B NMR spectrum of 4 shows only a single res-
onance at d +9.7, the addition of 17 to the solution results in
the appearance of a second signal much further upfield at
d ꢀ16.2. The large upfield shift of the 11B resonance sug-
gests that the boron centre has undergone rehybridization
from trigonal (sp2) to tetrahedral (sp3). It therefore seems
likely that the second resonance arises from the formation
of a structure such as 18 or 19.
nucleophiles, such as benzenedimethanols 14 and 15, could
result in the formation of linear hetero(oligomers), of the
type illustrated in Figure 1. However, semi-empirical elec-
tronic structure calculations suggested that macrocyclic
structures were also accessible, with the 2+2 macrocycles
being the most stable structures. Semi-empirical electronic
structure calculations, using the AM1 method, reveal that
both the macrocycle formed from two molecules of 4 and
two molecules of 1,3-benzenedimethanol 14 (Fig. 8a) and
that formed from two molecules of 4 and two molecules of
1,4-benzenedimethanol 15 (Fig. 8b) are relatively strain free.
O
B
O
O
B
O
O
B
O
B
O
O
O
O
NH2
H2N
NH2
H2N
H2N
NH2
18
19
HO
OH HO
OH
14
16
15
17
Analysis of the sample by MALDI-TOF mass spectrometry
confirmed the identity of the new adduct as 18—signals are
observed at m/z 352 and 294 corresponding to the molecular
ion of 18 and the corresponding monoadduct incorporating 4
and one molecule of 17. The limited ability of 4 to form ad-
ducts with alcohols and other species was troubling given the
ability of the corresponding monofunctional boroxoaro-
matic 3 to react readily with benzylic alcohols. In order to
HO
NH2
HO
OH
In order to test these possibilities, equimolar solutions of
4 with a series of nucleophiles, 14, 15, 16 and 17, were pre-
pared, and analyzed by H and11B NMR spectroscopies
and MALDI-TOF mass spectrometry. Since 4 is relatively
1

2398
L. M. Greig et al. / Tetrahedron 63 (2007) 2391–2403
assess whether this lack of reactivity was a result of an inap-
propriate choice of nucleophile or whether it was a property
of compound 4 itself, we repeated the experiments described
above using the monofunctional boroxoaromatic 3 in place
of compound 4. The results of these experiments are reveal-
ing. Boroxoaromatic 3 reacts rapidly and selectively with the
benzenedimethanols 14 and 15 in d6-acetone to form the 2:1
adducts 20 and 21, respectively, which were characterized
by 1H and11B NMR spectroscopies and MALDI-TOF
mass spectrometry. The exclusive observation of a signal
at m/z 494 in solutions where there is an excess of 3 indicates
that the 2:1 adduct is by far the dominant species. The 2:1
adducts 20 and 21 have 11B chemical shifts of d +9.0 and
d +9.6, respectively, indicating that the boron centre is still
trigonal in these compounds.
O
B
O
B
O
O
O
20
O
B
O
B
O
Figure 10. (a) Packing of compound 22 in the solid state. The outline of the
unit cell is shown as a solid line. (b) Packing of compound 22 along the
b axis in the solid state. (c) Molecular structure of compound 22 determined
by single crystal X-ray diffraction. Carbon atoms are coloured green, boron
atoms are mauve, oxygen atoms are red and hydrogen atoms are light grey.
21
The more flexible propane-1,3-diol 16 can also form either a
1:1 or a 2:1 adduct with boroxoaromatic 3. In this case, how-
ever, the 1:1 adduct can be stabilized by the interaction of the
second OH group with the boron centre resulting in a six-
membered chelate ring at boron and rehybridization of this
centre to a tetrahedral (sp3) geometry. Boroxoaromatic 3 re-
acts rapidly with propane-1,3-diol 16 in d6-acetone to form
the corresponding 2:1 adduct 22. Characterization by
MALDI-TOF mass spectrometry suggested that only the
2:1 adduct was present in the sample—a signal correspond-
ing to adduct 22 was observed at m/z 432 whereas no signal
was observed at m/z 254 corresponding to the 1:1 adduct 23.
11B NMR spectroscopy did, however, suggest that the 1:1
adduct 23 was present in solution—in addition to the reso-
nance expected for the 2:1 adduct at d +9.0 (sp2 boron), an
additional resonance at d ꢀ0.4 was observed, indicating
that a boron species was present in solution, which had
undergone some rehybridization at boron, as would be
expected for the 1:1 adduct 23.
Slow evaporation of a solution of 4 and 16 in d6-acetone gave
single crystals suitable for analysis26 by X-ray diffraction.
Structures revealed that these crystals were of the 2:1 adduct
22. In the solid state, compound 22 is packed (Fig. 10a) in
a series of paired stacks, which are related by translation
in the a and c directions. In these pairs of stacks, the mole-
cules of 22 adopt a conformation (Fig. 10b) in which the
two aromatic planes are essentially at right angle to each
other. The molecules stack along the b direction with the
direction of adjacent stacks being opposite in sense giving
rise to the pairing. It is noteworthy that within the molecular
structure (Fig. 10c) of 22, the syn arrangement of the exo-
cyclic bond—in this case the C–O bond—with respect to
the endocyclic B]O bond in the boroxoaromatic ring is
observed strictly, even although this requires the linking
of the trimethylene spacer to adopt a gauche–gauche confor-
mation.
O
B
O
B
We have described previously dynamic exchange in esters of
boroxoaromatic 3. This dynamic behaviour is also present in
compound 22. When compound 22 is dissolved in d6-ace-
tone at room temperature and 1 equiv of 1,2-ethanediol
and 1 equiv of 1,4-butanediol are added (Fig. 11), rapid ex-
change occurs and a new equilibrium position is established
almost immediately. Analysis of the reaction mixture by
MALDI-TOF mass spectrometry and 11B NMR reveals the
presence of compound 22 (m/z 432, d (11B) +9.5), 24 (m/z
418, d (11B) +9.7) and 25 (m/z 446, d (11B) +9.9) in approx-
imately equal proportions.
O
O
22
O
B
O
O
23

L. M. Greig et al. / Tetrahedron 63 (2007) 2391–2403
2399
Figure 11. Treatment of a sample of 22 in acetone (mass spectrum A) with 1,2-ethanediol and 1,4-butanediol results in rapid re-equilibration at room temper-
ature to give mixture of compounds 22, 24 and 25 (mass spectrum B). Mass spectrum B was recorded 10 min after mixing the diols with compound 22.
7. Conclusions
atmosphere and collected by distillation. Toluene and xylene
˚
were distilled over sodium under N2 and stored over 4 A
The formation of 1:2 adducts between 3 and a range of diols
and the dynamic behaviour of these species demonstrates the
utility of boroxoaromatics in dynamic covalent chemistry.
The facilesynthesis of compound 4 demonstrates that bifunc-
tional boroxoaromatics required for the assembly of more
complex superstructures is possible and that such compounds
are stable entities, which can be handled easily. However, the
lack of reactivity exhibited by 4 towards bifunctional nucleo-
philes in solution seriously limits its potential utility in con-
structing unusual or complex assemblies through dynamic
selection processes. The reasons for this lack of reactivity
are probably two fold. Firstly, the solubility of 4 in relatively
non-polar solvents, when compared to 3, is rather poor. This
poor solubility must hamper the formation of covalent
adducts by 4. Secondly, and more importantly, placing the
two boroxoaromatic rings within the same conjugated system
has a detrimental effect on their reactivity. The formation of
one ester by 4 will perturb the electronic structure of the sys-
tem sufficiently to prevent formation of a second ester by 4. It
is therefore necessary to modify our design of bifunctional
boroxoaromatics to remove the two boron containing rings
from direct conjugation with each other. Such molecules are
current synthetic targets in our laboratory.
molecular sieves. Hexane and dichloromethane (CH2Cl2)
were dried by heating under reflux over calcium hydride and
distilled under N2. All alkyllithium reagents were titrated
against N-pivaloyl-o-toluidine. All other solvents and re-
agents were used as received. Thin-layer chromatography
(TLC) was performed on aluminium plates coated with
Merck Kieselgel 60 F254. Developed plates were scrutinized
under a UV lamp, and, where necessary, stained with iodine
or PMA dip to aid identification. Column chromatography
was performed using Kieselgel 60 (0.040–0.063 mm mesh,
Merck 9385). Microanalyses (CHN) were carried out at
the University of St. Andrews. Infra-red spectra were re-
corded on a Perkin Elmer 1605 FTIR spectrometer, using
samples prepared as KBr discs or as thin films.
8.2. NMR spectroscopy
1H nuclear magnetic resonance (NMR) spectra were re-
corded on a Bruker Avance 300 (300.1 MHz), Varian Gemini
2000 (300.0 MHz) or Varian UNITYplus (500.1 MHz) spec-
trometer using the deuterated solvent as the lock and the
residual solvent as the internal reference in all cases. 13C
NMR spectra using the PENDANT sequence were recorded
on a Bruker Avance 300 (75.5 MHz) spectrometer or on a
Varian Gemini 2000 (75.5 MHz) spectrometer using broad-
8. Experimental section
8.1. General
1
band H decoupling. 11B NMR spectra were recorded on a
Varian UNITYplus (160.4 MHz) spectrometer using the
deuterated solvent as the lock and trimethyl borate as the ex-
ternal reference. All coupling constants are quoted to the
nearest 0.1 Hz.
Diethyl ether (Et2O) and tetrahydrofuran (THF) were dried
by refluxing with sodium-benzophenone under a N2

2400
L. M. Greig et al. / Tetrahedron 63 (2007) 2391–2403
8.3. Mass spectrometry
residue was taken up in CH2Cl2 (40 cm3), water (30 cm3)
added, and the aqueous layer extracted with CH2Cl2
(3ꢂ20 cm3). The combined organic extracts were dried
(MgSO4), and evaporated under reduced pressure. The crude
product was purified by column chromatography (Et2O/hex-
ane, v/v, 2:3; Rf¼0.34) to afford 8 (5.5 g, 89%) as yellow
needles.
Electron impact mass spectrometry (EIMS) and high-resolu-
tion mass spectrometry (HRMS) were carried out on either
a VG PROSPEC or a VG AUTOSPEC mass spectrometer.
Electrospray mass spectrometry (ESMS) was recorded on
a VG PLATFORM mass spectrometer. Matrix assisted laser
desorption ionization time-of-flight (MALDI-TOF) spectra
were recorded on either a Kratos Kompact Maldi III or a
Micromass Tof Spec 2E spectrometer. A nitrogen laser
(337 nm, 85 kW peak laser power, 4 ns pulse width) was
used to desorb the sample ions. The instrument was operated
in positive linear time-of-flight mode, and results from 50
laser shots were signal averaged to give one spectrum. The
matrixused was asuspensionof cobalt ina methanol/glycerol
mixture and was applied to the plate surface in 1-mL aliquots
and allowed to dry before deposition of the analyte solution
on to the matrix.
Found: C, 57.87; H, 3.96%. Calcd for C20H17O2I: C, 57.71;
H, 4.12%; mp: 90.2–91.7 ꢁC; dH (300 MHz; CDCl3) 7.43
3
(4H, d, J¼7.4 Hz, Ar), 7.21–7.34 (6H, m, Ar), 7.07–7.17
3
(1H, m, Ar), 6.46 (2H, d, J¼8.3 Hz, Ar) and 5.09 (4H, s,
CH2); dC (75.5 MHz; CDCl3) 159.1 (Ar, quat), 137.1 (Ar,
quat), 130.1 (Ar, CH), 128.9 (Ar, CH), 128.2 (Ar, CH),
127.4 (Ar, CH), 106.4 (Ar, CH), 80.2 (Ar, quat) and 71.4
(CH2); m/z (EI) 416 (M+, 12%), 91 (100); HRMS: Found:
416.0284; Calcd for C20H17O2I: 416.0273.
8.6.2. 2,6-Benzyloxybiphenyl 9. Method A. A solution
of phenyl boronic acid (90 mg, 0.72 mmol) in EtOH
(2 cm3) was added to 2,6-benzyloxy-1-iodobenzene (0.2 g,
0.48 mmol) and Pd(PPh3)4 (0.08 g, 0.072 mmol) in dry,
degassed DME (20 cm3). This addition was followed by
treatment of the reaction mixture with Na2CO3 (0.16 g,
1.54 mmol) in water (5 cm3). The mixture was heated under
reflux for 3 h, and then allowed to cool. CH2Cl2 (20 cm3)
was added, and the mixture filtered through a short pad of
CeliteÒ. The filtrate was dried (MgSO4) and concentrated
under reduced pressure to afford a brown oil. Column chro-
matography (Et2O/hexane, v/v, 2:3; Rf¼0.36) afforded 9
(30 mg, 16%) as colourless plates.
8.4. Thermal studies
Melting points were determined on an Electrothermal 9200
melting point apparatus and are uncorrected. All melting
points are quoted to the nearest 0.5 ꢁC. Thermogravimetric
analysis (TGA) experiments (typical sample mass ca.
10 mg) were performed on a Perkin Elmer TGA 6 Thermo-
gravimetric Analyzer, using nitrogen as the sample purge gas
or a RheoTherm TG1000M+, using argon as the purge gas.
Differential scanning calorimetry (DSC) experiments (typi-
cal sample mass ca. 5 mg) were carried out using a Perkin
Elmer Pyris 1 Differential Scanning Calorimeter under he-
lium purge or a Perkin Elmer DSC 7 under nitrogen purge.
Method B. A solution of bromobenzene (1.5 g, 9.6 mmol) in
dry diethyl ether (20 cm3) was added in small portions to
magnesium turnings (0.23 g, 9.6 mmol), at such a rate so
as to maintain a gentle reflux. After the addition was com-
plete, the reaction mixture was heated under reflux for a fur-
ther hour to ensure completion of the reaction. The solution
of phenyl magnesium bromide was added slowly to a stirred
suspension of 2,6-benzyloxybiphenyl (2.0 g, 4.8 mmol) and
[Ni(dppe)Cl2] (0.1 g, 0.19 mmol) in dry diethyl ether
(20 cm3). The mixture was heated under reflux for 5 h, and
after cooling, diethyl ether (20 cm3) and water (20 cm3)
were added. The aqueous layer was extracted with diethyl
ether (3ꢂ20 cm3), the combined organic extracts dried
(MgSO4) and concentrated to afford a yellow oil, which
was purified by column chromatography (CH2Cl2/hexane,
v/v, 1:2; Rf¼0.26) to afforded 9 (0.77 g, 44%) as small
colourless plates.
8.5. Powder X-ray diffraction studies
Ambient temperature powder X-ray diffractograms were
recorded on a Siemens D5000 diffractometer operating in
transmission mode using Ge monochromatized Cu Ka1
radiation. The data were recorded for 2q in the range 4 to
40ꢁ in steps D(2q)¼0.02ꢁ and collection times varied. Vari-
able temperature powder X-ray diffraction experiments were
performed on a Siemens D5005 diffractometer equipped
with an Auton Paar HTK 1200 heating stage.
8.6. Computational studies
All semi-empirical (AM1) and Hartree–Fock ab initio elec-
tronic structure calculations were performed using either
SPARTAN (Version 5.0, Wavefunction, Inc., Irvine, CA,
€
1997) or Jaguar (Version 4.1, Release 53, Schrodinger,
Inc., Portland, OR, 2001) on a Silicon Graphics Octane 2
workstation.
Found: C, 85.50; H, 5.89%. Calcd for C26H22O2: C, 85.22;
H, 6.05%; mp: 87–88 ꢁC; dH (300 MHz; CDCl3) 7.42–7.51
(4H, m, Ar), 7.22–7.38 (12H, m, Ar), 6.73 (2H, d,
3J¼8.3 Hz, Ar) and 5.05 (4H, s, CH2); dC (75.5 MHz;
CDCl3) 157.2 (Ar, quat), 137.8 (Ar, quat), 135.1 (Ar,
quat), 131.5 (Ar, CH), 129.0 (Ar, CH), 128.7 (Ar, CH),
127.9 (Ar, CH), 127.8 (Ar, CH), 127.2 (Ar, CH), 127.0
(Ar, CH), 123.0 (Ar, quat), 107.2 (Ar, CH) and 70.9
(CH2); m/z (EI) 366 (M+, 21%), 91 (100); HRMS: Found:
366.1629; Calcd for C26H22O2: 366.1620.
2-Iodoresorcinol 7 was prepared by a literature27 method. All
other known compounds were prepared by methods that
differ significantly from those reported previously.
8.6.1. 2,6-Benzyloxy-1-iodobenzene 8. 2-Iodoresorcinol
(3.5 g, 14.8 mmol), benzyl bromide (3.5 cm3, 29.7 mmol),
potassium carbonate (5.1 g, 37.1 mmol) and 18-crown-6
(0.2 g, 0.74 mmol) were stirred in dry acetone (120 cm3),
and the mixture heated under reflux for 15 h. After cooling,
the solvent was removed in vacuo to yield a brown solid. The
8.6.3. 2,6-Dimethoxyiodobenzene 11. tert-Butyllithium
in hexanes (10 cm3, 20 mmol, 1.2 mol equiv) was added

L. M. Greig et al. / Tetrahedron 63 (2007) 2391–2403
2401
dropwise to a stirred solution of 1,3-dimethoxybenzene
(2.5 g, 18 mmol) in THF (75 cm3) under N2 at 0 ꢁC and
the mixture was stirred at this temperature for a further
30 min. A solution of I2 (4.6 g, 18 mmol) in THF (30 cm3)
was added dropwise whilst maintaining the temperature of
the mixture at 0 ꢁC. The mixture was then allowed to warm
to room temperature. After 2 h, ethyl acetate (20 cm3) was
added and the mixture washed with sodium sulfite solution
(2ꢂ40 cm3, saturated aqueous solution). The organic extract
was washed with aqueous NaCl solution, dried (MgSO4)
and concentrated under reduced pressure to afford a pale
yellow solid. Recrystallization (CHCl3/hexane) afforded 2,6-
dimethoxyiodobenzene (3.0 g, 63%) as colourless plates.
(MgSO4) and evaporated to dryness to afford a pale brown
powder. Flash column chromatography (CH2Cl2; Rf¼0.41)
gave 5 (620 mg, 89%) as small colourless needles.
Found: C, 77.23; H, 5.34%. Calcd for C12H10O2: C, 77.40;
H, 5.41%; mp: 116–118 ꢁC (lit.30 mp: 118–119 ꢁC); IR
nmax (cmꢀ1) 3402br (OH); dH (300 MHz; CDCl3) 7.54–
7.59 (2H, m, Ar), 7.40–7.49 (3H, m, Ar), 7.15 (1H, t,
3J¼8.1 Hz, Ar), 6.58 (2H, d, 3J¼8.1 Hz, Ar) and 4.91 (2H,
br s, OH); dC (75.5 MHz; CDCl3) 153.4 (Ar, quat), 131.0
(Ar, CH), 130.8 (Ar, quat), 130.2 (Ar, CH), 129.6 (Ar,
CH), 129.2 (Ar, CH), 115.5 (Ar, quat) and 107.8 (Ar, CH);
m/z (EI) 186 (M+, 100%), 168 (15) and 139 (12); HRMS:
Found: 186.0688; Calcd for C12H10O2: 186.0681.
Found: C, 36.52; H, 3.36%. Calcd for C8H9IO2: C, 36.39; H,
3.44%; mp: 103.2–104.7 ꢁC (lit.28 mp: 99–100 ꢁC); IR nmax
(cmꢀ1) 2840 m (OMe); dH (300 MHz; CDCl3) 7.26 (1H, t,
3J¼8.1 Hz, Ar), 6.50 (2H, d, 3J¼8.1 Hz, Ar) and 3.88 (6H,
s); dC (75.5 MHz; CDCl3) 159.5 (Ar, quat), 129.8 (Ar,
CH), 104.1 (Ar, CH), 77.6 (Ar, quat) and 56.5 (OMe,
CH3); m/z (EI) 264 (M+, 100%), 249 (10), 234 (5), 221
(50) and 206 (15); HRMS: Found: 263.9652; Calcd for
C8H9O2I: 263.9647.
8.6.6. 5,9-Dihydroxy-5,9-dibora-4,10-dioxopyrene 4. Di-
chloroborane–dimethylsulfide complex (0.5 cm3, 4 mmol)
was added in small portions to a cooled, stirred solution of
boron trichloride in hexane (1 M, 4 cm3, 4 mmol). A white
precipitate (BCl3$Me2S) formed immediately upon addi-
tion. The mixture was stirred with cooling for a further
10 min, and then allowed to warm to room temperature
and stirred for 1 h. The resulting solution of BHCl2 in hexane
was added dropwise to a stirred solution of 2,6-dihydroxy-
biphenyl (0.3 g, 1.6 mmol) in dry toluene (20 cm3) at room
temperature. After stirring at room temperature for 2 h,
AlCl3 (90 mg, 0.6 mmol) was added and the mixture heated
under reflux overnight. After cooling, water (20 cm3) was
added carefully, and the aqueous layer extracted with diethyl
ether (3ꢂ20 cm3). Combined organic extracts were dried
(MgSO4) and concentrated under vacuum to afford a pale
brown solid. Recrystallization (CH2Cl2/hexane or aceto-
nitrile) afforded 4 (330 mg, 87%) as small colourless prisms.
8.6.4. 2,6-Dimethoxybiphenyl 12. Bromobenzene (1.6 g,
10.2 mmol) in dry diethyl ether (20 cm3) was added in small
portions to magnesium turnings (0.3 g, 12.3 mmol) under
N2, at such a rate so as to maintain a gentle reflux. After
addition was complete, the mixture was stirred under reflux
for a further hour to ensure complete reaction. The ethereal
solution of phenylmagnesium bromide was allowed to cool
before it was added dropwise to a stirred solution of 2,6-di-
methoxyiodobenzene (1.3 g, 4.92 mmol) and [Ni(PPh3)2Cl2]
(64 mg, 2 mol %) in dry diethyl ether (20 cm3) under N2.
The resulting black mixture was stirred under reflux for
24 h. After cooling, water (50 cm3) was added, and the aque-
ous layer extracted with ether (2ꢂ20 cm3). Combined
organic extracts were dried (MgSO4) and evaporated under
reduced pressure. Column chromatography (Et2O/hexane,
v/v, 1:4; Rf¼0.33) afforded 2,6-dimethoxybiphenyl (0.8 g,
76%) as small colourless prisms.
Found: C, 60.44; H, 3.47%. Calcd for C12H8B2O4: C, 60.61;
H, 3.39%; mp: >400 ꢁC (decomp.); IR nmax (cmꢀ1) 3350br
(OH), 1370s (OB); dH (300 MHz; (CD3)2CO) 8.45 (2H, s,
3
3
OH), 8.39 (2H, d, J¼7.4 Hz, Ar), 7.68 (1H, t, J¼7.4 Hz,
Ar), 7.43 (1H, t, 3J¼8.2 Hz, Ar) and 7.04 (2H, d, 3J¼8.2 Hz,
Ar); dC (75.5 MHz; (CD3)2CO) 152.9 (Ar, quat), 145.3 (Ar,
quat), 138.0 (Ar, CH), 130.1 (Ar, CH), 128.0 (Ar, CH) and
113.0 (Ar, CH); m/z (EI) 238 (M+, 100); HRMS: Found:
238.0816; Calcd for C12H8B2O4: 238.0838.
Found: C, 78.34; H, 6.64%. Calcd for C14H14O2: C, 78.48;
H, 6.59%; mp: 87.0–88.1 ꢁC (lit.29 mp: 83–85 ꢁC); IR nmax
(cmꢀ1) 2838 m (OMe); dH (300 MHz; CDCl3) 7.48–7.29
3
(6H, m, Ar), 6.69 (2H, d, J¼8.5 Hz, Ar) and 3.76 (6H, s,
Acknowledgements
Me); dC (75.5 MHz; CDCl3) 157.9 (Ar, quat), 134.4 (Ar,
quat), 131.1 (Ar, CH), 128.9 (Ar, CH), 127.9 (Ar, CH),
127.0 (Ar, CH), 119.7 (Ar, quat), 104.4 (Ar, CH) and 56.1
(OMe, CH3); m/z (EI) 214 (M+, 100%), 199 (16), 184 (24)
and 128 (18).
We thank the Engineering and Physical Science Research
Council and the University of St Andrews for financial
support of this work.
References and notes
8.6.5. 2,6-Dihydroxybiphenyl 5. A solution of 2,6-dimeth-
oxybiphenyl (0.8 g, 3.7 mmol) was stirred in dry CH2Cl2
(20 cm3) at ꢀ78 ꢁC under N2. Boron tribromide (1 M solu-
tion in CH2Cl2, 7.5 cm3, 7.5 mmol) was added dropwise
via a syringe and the mixture allowed to warm slowly to
room temperature. After re-cooling to 0 ꢁC, MeOH (10 cm3)
was added dropwise to quench any remaining BBr3, and the
solvents were removed on a water aspirator. Water (20 cm3)
and CH2Cl2 (20 cm3) were added to the residue and the
aqueous layer was extracted with CH2Cl2 (3ꢂ20 cm3).
Combined organic extracts were washed with water, dried
1. (a) Greig, L. M.; Philp, D. Chem. Soc. Rev. 2001, 30, 287–302;
(b) Philp, D.; Stoddart, J. F. Angew. Chem., Int. Ed. 1996, 35,
1155–1196; (c) Lawrence, D. S.; Jiang, T.; Levett, M. Chem.
Rev. 1995, 95, 2229–2260; (d) Lindsey, J. S. New J. Chem.
1991, 15, 153–180; (e) Whitesides, G. M.; Mathias, J. P.;
Seto, C. T. Science 1991, 254, 1312–1319.
2. (a) Jahn, T. R.; Radford, S. E. FEBS J. 2005, 272, 5962–5970;
(b) Gruebele, M. Comptes. Rend. Biol. 2005, 328, 701–712; (c)
Kumar, T. K. S.; Yu, C. Acc. Chem. Res. 2004, 37, 929–936.

2402
L. M. Greig et al. / Tetrahedron 63 (2007) 2391–2403
3. Klug, A. Angew. Chem., Int. Ed. Engl. 1983, 22, 565–582.
4. Stryer, L. Biochemistry, 3rd ed.; W. H. Freeman: New York,
NY, 1988; pp 852–853.
13. The first definition of the term predisposition can be found in:
Comprehensive Supramolecular Chemistry; Atwood, J. L.,
€
Davies, J. E. D., MacNicol, D. D., Vogtle, F., Eds.; Elsevier:
5. For some recent examples, see: (a) MacLachlan, M. J. Pure
Appl. Chem. 2006, 78, 873–888; (b) Kotch, F. W.; Raines,
R. T. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 3028–3033; (c)
Mancin, F.; Rampazzo; Tecilla, P.; Tonellato, U. Chem.—Eur.
J. 2006, 12, 1844–1854; (d) Zeckert, K.; Hamacek, J.;
Senegas, J. M.; Dalla-Favera, N.; Floquet, S.; Bernardinelli,
G.; Piguet, C. Angew. Chem., Int. Ed. 2006, 44, 7954–7958;
(e) Melkko, S.; Sobek, J.; Guarda, G.; Scheuermann, J.;
Dumelin, C. E.; Neri, D. Chimia 2005, 59, 798–802; (f)
Hahn, U.; Gonzalez, J. J.; Huerta, E.; Segura, M.; Eckert,
J. F.; Cardinali, F.; de Mendoza, J.; Nierengarten, J. F.
Chem.—Eur. J. 2005, 11, 6666–6672; (g) ten Cate, M. G. J.;
Reinhoudt, D. N.; Crego-Calama, M. J. Org. Chem. 2005, 70,
8443–8453; (h) Schmittel, M.; Kalsani, V. Top. Curr. Chem.
2005, 245, 1–53; (i) Stoddart, J. F. Pure Appl. Chem. 2005,
77, 1089–1106; (j) Beck, J. B.; Ineman, J. M.; Rowan, S. J.
Macromolecules 2005, 38, 5060–5068; (k) Keizer, H. M.;
Sijbesma, R. P. Chem. Soc. Rev. 2005, 34, 226–234; (l)
Piguet, C.; Borkovec, M.; Hamacek, J.; Zeckert, K. Coord.
Chem. Rev. 2005, 249, 705–726.
6. (a) Zafar, A.; Hamilton, A. D. Current Challenges in
Large Supramolecular Assemblies; Tsoucaris, G., Ed.;
Kluwer Academic: London, 1999; pp 67–81; (b) Yang, J.;
Marendaz, J.-L.; Geib, S. J.; Hamilton, A. D. Tetrahedron
Lett. 1994, 35, 3665–3668; (c) Seto, C. T.; Whitesides, G. M.
J. Am. Chem. Soc. 1993, 115, 905–916; (d) Garcia-Tellado,
F.; Geib, S. J.; Goswami, S.; Hamilton, A. D. J. Am. Chem.
Soc. 1991, 113, 9265–9269; (e) Lehn, J.-M.; Mascal, M.;
DeCian, A.; Fisher, J. J. Chem. Soc., Chem. Commun. 1990,
479–480.
New York, NY, 1996; Vol. 9, p 218. Sanders defines
(Ref. 11d) the term predisposition as a strong conformational
or structural preference expressed by the building block once
incorporated into a larger structure.
14. (a) Ipaktschi, J.; Hosseinzadeh, R.; Schlaf, P. Angew. Chem.,
Int. Ed. 1999, 38, 1658–1660; (b) Ipaktschi, J.; Hosseinzadeh,
R.; Schlaf, P.; Dreiseidler, E.; Goddard, R. Helv. Chim. Acta
1998, 81, 1821–1834.
15. (a) Cantrill, S. J.; Chichak, K. S.; Peters, A. J.; Stoddart, J. F.
Acc. Chem. Res. 2005, 38, 1–9; (b) Chichak, K. S.; Cantrill,
S. J.; Pease, A. R.; Chiu, S. H.; Cave, G. M. W.; Atwood,
J. L.; Stoddart, J. F. Science 2004, 304, 1308–1312.
16. (a) Saur, I.; Scopelliti, R.; Severin, K. Chem.—Eur. J. 2006, 12,
1058–1066; (b) Buryak, A.; Severin, K. Angew. Chem., Int. Ed.
2005, 44, 7935–7938; (c) Saur, I.; Severin, K. Chem. Commun.
2005, 1371–1473; (d) Christinat, N.; Scopelliti, R.; Severin, K.
Chem. Commun. 2004, 1158–1159; (e) Severin, K. Chem.—
Eur. J. 2004, 10, 2565–2580; (f) Vial, L.; Sanders, J. K. M.;
Otto, S. New J. Chem. 2005, 29, 1001–1003; (g) Corbett,
P. T.; Sanders, J. K. M.; Otto, S. J. Am. Chem. Soc. 2005,
127, 9390–9392; (h) Corbett, P. T.; Tong, L. H.; Sanders,
J. K. M.; Otto, S. J. Am. Chem. Soc. 2005, 127, 8902–8903;
(i) Leclaire, L.; Vial, L.; Otto, S.; Sanders, J. K. M. Chem.
Commun. 2005, 1959–1961.
17. Comina, P. J.; Philp, D.; Kariuki, B. M.; Harris, K. D. M. Chem.
Commun. 1999, 2279–2280.
18. (a) Harris, K. D. M.; Kariuki, B. M.; Lambropoulos, C.; Philp,
D.; Robinson, J. M. A. Tetrahedron 1997, 53, 8599–8612;
(b) Robinson, J. M. A.; Kariuki, B. M.; Philp, D.; Harris,
K. D. M. Tetrahedron Lett. 1997, 38, 6281–6284.
19. Robinson, J. M. A.; Philp, D.; Kariuki, B. M.; Harris, K. D. M.
J. Chem. Soc., Perkin Trans. 2 2001, 2166–2173.
7. (a) Smith, V. C. M.; Lehn, J.-M. Chem. Commun. 1996, 2733–
2734; (b) Stang, P. J.; Chen, K.; Arif, A. M. J. Am. Chem. Soc.
€
1995, 117, 8793–8797; (c) Kramer, R.; Lehn, J.-M.; Marquis-
20. Greig, L. M.; Kariuki, B. M.; Habershon, S.; Spencer, N.;
Johnston, R. L.; Harris, K. D. M.; Philp, D. New J. Chem.
2002, 701–710.
21. Stanforth, S. P. Tetrahedron 1998, 54, 263–303.
22. Thomsen, I.; Torssell, K. B. G. Acta Chem. Scand. 1991, 45,
539–542.
Rigault, A. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 5394–
5398; (d) Hutin, M.; Frantz, R.; Nitschke, J. R. Chem.—Eur. J.
2006, 12, 4077–4082; (e) Hutin, M.; Schalley, C. A.;
Bernardinelli, G.; Nitschke, J. R. Chem.—Eur. J. 2006, 12,
4069–4076; (f) Nitschke, J. R.; Schultz, D.; Bernardinelli, G.;
ꢀ
Gerard, D. J. Am. Chem. Soc. 2004, 126, 16538–16543; (g)
23. Banzatti, C.; Carfagna, N.; Commisso, R.; Heidempergher, F.;
Pegrassi, L.; Melloni, P. J. Med. Chem. 1988, 31, 1466–
1471.
Schultz, D.; Nitschke, J. R. Angew. Chem., Int. Ed. 2006, 45,
2453–2456.
8. Rowan, S. J.; Cantrill, S. J.; Cousins, G. R. L.; Sanders,
J. K. M.; Stoddart, J. F. Angew. Chem., Int. Ed. 2002, 41,
898–952.
9. Tam-Chang, S.-W.; Stehouwer, J. S.; Hao, J. J. Org. Chem.
1999, 64, 334–335.
24. X-ray diffraction studies on crystals of 4, grown by slow
evaporation of an acetonitrile solution, were performed using a
Bruker SMART/Mo system. The structure was solved by direct
methods, the non-hydrogen atoms were refined (Ref. 31) with
anisotropic displacement parameters, hydrogen atoms bound
˚
10. James, T. D.; Samankumara Sandanayake, K. R. A.; Shinkai, S.
Angew. Chem., Int. Ed. 1996, 35, 1911–1922.
to carbon were idealized and fixed (C–H 0.95 A). Crystal
data for 4; C12H8B2O4, colourless prism, 0.18ꢂ0.1ꢂ0.1 mm,
11. (a) Rowan, S. J.; Brady, P. A.; Sanders, J. K. M. Angew. Chem.,
Int. Ed. 1996, 35, 2143–2145; (b) Rowan, S. J.; Brady, P. A.;
Sanders, J. K. M. Tetrahedron Lett. 1996, 37, 6013–6016; (c)
Brady, P. A.; Sanders, J. K. M. Chem. Soc. Rev. 1997, 26,
327–336; (d) Rowan, S. J.; Hamilton, D. G.; Brady, P. A.;
Sanders, J. K. M. J. Am. Chem. Soc. 1997, 119, 2578–2579;
(e) Rowan, S. J.; Sanders, J. K. M. Chem. Commun. 1997,
1407–1408; (f) Rowan, S. J.; Sanders, J. K. M. J. Org. Chem.
1998, 63, 1536–1546.
M¼237.80,
tetragonal,
c¼4.0964(2) A, a¼90 b¼90 g¼90ꢁ, F(000)¼1952,
a¼32.1413(9)
b¼32.1413(9)
˚
3
˚
U¼4231.8(3) A , T¼293(2) K, space group I41/a, Z¼16,
m(Mo Ka)¼0.108 mmꢀ1, 2qmax¼46.54ꢁ. Of 1527 measured
data, 994 were unique reflections (Rint 0.0391) to give R1
[I>2s(I)]¼0.0327 and wR2¼0.1018. Crystallographic data
(excluding structure factors) have been deposited with the
Cambridge Crystallographic Data Centre as supplementary
publication no. CCDC 607600. Copies of the data can be
obtained free of charge on application to CCDC, 12 Union
Road, Cambridge CB2 1EZ, UK [fax: +44 1223 336 033;
e-mail: deposit@ccdc.cam.ac.uk].
12. (a) Ercolani, G.; Mandolini, L.; Mencarelli, P.; Roelens, S.
J. Am. Chem. Soc. 1993, 115, 3901–3908; (b) Ercolani, G.
J. Phys. Chem. B 1998, 102, 5699–5703.

L. M. Greig et al. / Tetrahedron 63 (2007) 2391–2403
2403
25. A related bis-boroxoaromatic compound, 5,12-dihydroxy-
5,12-dibora-6,13-dioxodibenz[a,h]anthracene, has been reported
as showing similar behaviour. Unfortunately, no structural
details were provided: Offenhauer, R. D.; Rodewald, P. G.
J. Polym. Sci., Part B 1968, 6, 573.
26. X-ray diffraction studies on crystals of 22, grown by slow
evaporation of an acetonitrile solution, were performed using a
Bruker SMART/Mo system. The structure was solved by direct
methods, the non-hydrogen atoms were refined (Ref. 31) with
anisotropic displacement parameters, hydrogen atoms bound
[I>2s(I)]¼0.1345 and wR2¼0.4164. Crystallographic data
(excluding structure factors) have been deposited with the
Cambridge Crystallographic Data Centre as supplementary
publication no. CCDC 607601. Copies of the data can be
obtained free of charge on application to CCDC, 12 Union
Road, Cambridge CB2 1EZ, UK [fax: +44 1223 336 033;
e-mail: deposit@ccdc.cam.ac.uk].
27. Weitl, F. L. J. Org. Chem. 1976, 41, 2044–2045.
28. Anderson, J. C.; Pearson, D. J. J. Chem. Soc., Perkin Trans. 1
1998, 2023–2029.
˚
to carbon were idealized and fixed (C–H 0.95 A). Crystal data
for 22; C27H22B2O4, colourless prism, 0.14ꢂ0.1ꢂ0.1 mm,
29. Saa, J. M.; Martorell, G.; Garcia-Raso, A. J. Org. Chem. 1992,
57, 678–685.
30. Joschek, H.-I.; Miller, S. I. J. Am. Chem. Soc. 1966, 88,
3269–3272.
31. Sheldrick, G. M. SHELXTL, version 6.10, Program for
Crystal Structure Solution; Bruker AXS: Madison, WI,
2002.
M¼432.07,
monoclinic,
a¼19.301(3)
b¼5.2214(7)
ꢁ
˚
c¼22.837(3) A, a¼90 b¼107.316(4) g¼90 , F(000)¼904,
3
˚
U¼2197.2(5) A , T¼293(2) K, space group P21/c, Z¼4,
m(Mo Ka)¼0.085 mmꢀ1, 2qmax¼46.58ꢁ. Of 3103 measured
data, 1414 were unique reflections (Rint 0.0439) to give R1