Hetero-arylboroxines: The first rational synthesis, X-ray crystallographic and computational analysis

Article abstract of DOI:10.1039/b804705g

A novel series of hetero-arylboroxines were synthesized and structurally characterized by X-ray diffraction, NMR and computational analysis. The solid-state structures of the hetero-arylboroxines represent the first report of AB₂-type hetero-arylboroxines.

Full text of DOI:10.1039/b804705g

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/dalton | Dalton Transactions
Hetero-arylboroxines: the first rational synthesis, X-ray crystallographic and
computational analysis†
Peter M. Iovine,*a Charles R. Gyselbrecht,a Emily K. Perttu,a Cole Klick,a Alexander Neuwelt,a Jason Loera,a
Antonio G. DiPasquale,b Arnold L. Rheingoldb and Jeremy Kuaa
Received 19th March 2008, Accepted 15th May 2008
First published as an Advance Article on the web 3rd June 2008
DOI: 10.1039/b804705g
A novel series of hetero-arylboroxines were synthesized and
structurally characterized by X-ray diffraction, NMR and
computational analysis. The solid-state structures of the
hetero-arylboroxines represent the first report of AB2-type
hetero-arylboroxines.
We began our investigation into hetero-arylboroxines armed
with two conclusions from our previous arylboroxine-based ther-
modynamic studies.18,19 First, arylboronic acids can be smoothly
and efficiently converted to 1 : 1 arylboroxine–ligand adducts.
Second, 1 : 1 arylboroxine–ligand adducts are thermodynamically
favored over 1 : 2 or 1 : 3 adducts. In an equilibrating mixture of two
arylboronic acids (e.g. A and B), if one of the arylboronic acids
contains an intramolecular ligand (denoted A in this example)
then there will be a thermodynamic preference for the AB2 hetero-
arylboroxine over the A2B or A3 arylboroxines as binding one
ligand is energetically more favorable than binding two or three.
In addition, the AB2 stoichiometry should dominate over the B3
arylboroxine because there is a thermodynamic preference for
ligated arylboroxine over unligated.
The synthesis of hetero-arylboroxines 3a–c is shown in Fig. 1.
The reaction conditions are straightforward and parallel known
procedures employing mild chemical dehydrating agents.20,21 Com-
mercially available ortho-(N,N-dimethylaminomethyl)phenyl-
boronic acid (1) was refluxed with two molar equivalents of
arylboronic acids (2a–c) in the presence of magnesium sulfate.
Fig. 1 shows the 19F NMR spectrum of the solid obtained after
concentrating the reaction of 1 and 2a.
Boroxines, sometimes termed boronic acid anhydrides or boroxins,
are the dehydration products of organoboronic acids (see Fig. 1).
Historically, there were two common themes in boroxine-related
research. First was the question of aromaticity in boroxine ring
compounds.1–4 The second line of investigation focused on the
propensity of arylboroxines to form Lewis acid–base adducts with
nitrogen-containing ligands.5–10 More recently, however, boroxine
research has been invigorated by a flurry of tantalizing papers.
Notably, the Yaghi group has synthesized and characterized
2D11 and 3D12 arylboroxine materials that function as perma-
nently porous organic frameworks with high thermal stabilities
and large surface areas. Additionally, arylboroxine ring motifs
have been important in the facile and reversible assembly of
arylboronic acid end-functionalized telechelic polymers,13 the
development of arylboroxine-based nonlinear optical materials,14
polymer electrolytes for lithium-ion battery applications,15 flame-
retardant materials,16 and the immobilization of oligofluorene
chromophores in networked matrices.17
In contrast to homo-arylboroxines, there are no published
reports (to our knowledge) of hetero-arylboroxines, i.e. aryl-
boroxine rings containing two or three different aryl substituents.
Developing rational methods to synthesize and control hetero-
arylboroxine structure will significantly expand the impact of
these boron-containing species in areas such as solid-state design,
porous organic materials and dynamic combinatorial chemistry. In
this paper we outline a novel synthetic approach toward the ratio-
nal synthesis of hetero-arylboroxines that relies on arylboroxine–
ligand interactions to control the substitution pattern on the
arylboroxine ring. A full computational analysis complements the
experimental work and further dissects the thermodynamics of
hetero-arylboroxine formation.
Based on both the 19F and 1H NMR data, a 9 : 1 : 1 molar
ratio of 3a : 4a : 2a was found. Although two arylboronic acid
building blocks can lead to four arylboroxine products, only two
such arylboroxine products, 3a and 4a, were detected by both 1H
and 19F NMR. The 19F resonances attributed to 4a and 2a were
made by comparison to authentic samples while the assignment
of 3a was made in the following manner. First, binding Lewis
basic species such as the N,N-dimethylaminomethyl group of
3a to arylboroxine rings decreases arylboroxine Lewis acidity
and generally results in upfield 19F NMR shifts relative to the
corresponding unligated arylboroxine.22 Therefore, the chemical
shift of the −110 ppm 19F resonance, assigned as 3a, is justified
based on the relative position of the resonances of 4a and 2a. In
addition, the chemical shift of the resonance assigned as 3a in
Fig. 1 correlates well with known mono-ligated fluoro-containing
arylboroxines such as 4a–pyridine.22 Although Fig. 1 shows only
arylboroxines 3a and 4a, two additional arylboroxine products are
possible. Compound 5a (see Fig. 4), an A2B hetero-arylboroxine,
would also have one unique 19F resonance just as in 3a. However,
integration data from the 1H NMR supports the formation of 3a
and not 5a. In addition, integration of the 1H NMR supports the
existence of 3a and not fluorine-free A3 homo-arylboroxine 6a. We
therefore conclude that arylboroxines 5a and 6a were not formed
in sufficient concentration to be measured by NMR.
aDepartment of Chemistry and Biochemistry, University of San Diego, 5998
Alcala Park, San Diego, CA, 92110, USA. E-mail: piovine@sandiego.
edu; Fax: +1 (1)619 260 2211; Tel: +1 (1)619 260 4028; Web: http://web.
mac.com/tetrakis
bDepartment of Chemistry and Biochemistry, University of California, San
Diego, La Jolla, CA, 92093, USA
† Electronic supplementary information (ESI) available: Experimental
details, NMR spectra, crystallographic experimentals for 3a–c, along
with details of the DFT computational methods (B3LYP/6-311+G*)
with implicit solvent, zero-point energy and thermal-enthalpy corrections.
CCDC reference numbers 671272–671274. For ESI and crystallographic
data in CIF or other electronic format see DOI: 10.1039/b804705g
Purification of the individual arylboroxine products was
not possible by standard bench-top methods such as column
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 3791–3794 | 3791
©

View Article Online
Fig. 1 The syntheses of hetero-arylboroxines 3a–c. The 19F NMR (CDCl3, 25 ◦C) is shown for the reaction of 1 and 2a and shows a 9 : 1 : 1 molar ratio
of 3a : 4a : 2a.
chromatography. Arylboroxines, including hetero-arylboroxines
3a–c, rapidly hydrolyze on silica gel and in wet polar solvents.
As solids, however, these arylboroxine-based materials are robust
and no degradation is observed over long periods under standard
atmospheric conditions. In an effort to corroborate the solution
NMR data and obtain the first known X-ray structure of a hetero-
arylboroxine, single crystals of 3a–c were grown by diffusion of
pentane into a 1,2-dichloroethane solution of the reaction mixture.
The structures of 3a–c are shown in Fig. 2.† In all three structures
the arylboroxine ring adopts a planar conformation and the AB2
˚
and the 1.84 A upper limit associated with arylboroxines having
three B–N bonds.27 It is important to note that despite the relative
lengthening of B–N bonds with increasing number of B–N bonds,
all the bond lengths cited above fall within the 1.5–1.8 A range
expected for strong B–N bonds.29
In the case of 3c, crystals from the same batch as those
used for the X-ray structure determination were isolated and
examined by 1H NMR. The upfield portion of the 1H NMR
spectrum for 3c is shown in Fig. 3a and highlights the benzylic
protons of the N,N-dimethylaminomethyl moiety. Upon addition
of 4-methoxyphenylboronic acid to a chloroform solution of 3c
˚
˚
stoichiometry is observed. The B–N bond lengths are 1.70 A for
˚
both 3a and 3b and 1.68 A for 3c. The average O–B1–O bond
angle (where B1 represents the sp3-hybridized boron atom in 3a,
3b and 3c, see Fig. 2) was 113.1◦ whereas the average O–B2–O
and O–B3–O bond angle (where B2 and B3 represent the sp2-
hybridized boron atoms in 3a, 3b and 3c) was 120.8◦. As expected,
coordination of B1 by the nitrogen ligand results in a lengthening
of the B1–O bond lengths (average B1–O bond length for 3a, 3b
˚
and 3c is 1.45 A) relative to the B2–O or B3–O bond lengths (the
2
˚
average sp B–O bond length is 1.37 A). A table summarizing key
X-ray crystallographic parameters can be found in the ESI.†
The B–N bond lengths in compounds 3a–c are typical for
arylboroxines having one ligand. Considering structures 3a–c
are the first reported hetero-arylboroxines, a comparative B–N
bond length analysis of 3a–c can only be done with published
homo-arylboroxines. The James group has reported an ortho-
(N-phenylmethylamino)-homo-arylboroxine containing one B–N
interaction26 while the Sotofte27 and Anslyn28 groups have reported
homo-arylboroxines with two and three B–N interactions. A
general trend emerges where B–N bond lengths increase with
increasing number of arylboroxine–nitrogen interactions. Bond
Fig. 3 1H NMR (CDCl3, 25 ◦C) spectra showing the room-temper-
ature scrambling of hetero-arylboroxine 3c upon addition of excess
4-methoxyphenylboronic acid (2b). Scrambling is observed in both the ben-
zylic (∼4.2 ppm) and methoxy (∼3.8 ppm) regions of the spectrum (panel
b). The peak marked with an asterisk (*) is residual 1,2-dichloroethane.
˚
lengths vary from 1.68 to 1.84 A with the shorter bond lengths
being associated with arylboroxines showing one B–N bond26
Fig. 2 X-Ray structures of 3a,23 3b24 and 3c25 with thermal ellipsoids at the 50% probability level.
3792 | Dalton Trans., 2008, 3791–3794 This journal is The Royal Society of Chemistry 2008
©

View Article Online
Table 1 Enthalpy for the formation of arylboroxines of the structural
type B3, AB2, A2B and A3 where A is arylboronic acid monomer 1 and B
represents non-coodinating arylboronic acid monomers 2a–c. Calculations
were performed using chloroform as an implicit solvent
The homo-arylboroxines B3 and A3 were calculated to be less
stable than the hetero-arylboroxines A2B and AB2, and their
formation is endothermic. Experimentally, small amounts of B3
are observed but not A2B. We think there are two reasons for the
difference between the computational and experimental results.
First, the ratio of B to A monomers is higher in experiment, thus
favoring the chemical potential of B over A. This is not taken into
account in the calculations. Second, there may be small amounts of
A2B formed but not detected by NMR. In any case, A2B crystals
were not isolable and AB2 is the dominant product from both
experiment and calculation. A3 is not observed experimentally and
is predicted to be much less stable from our calculations. Having
all three pendant groups on the same side of the ring is very
unfavorable (DH = +15.3 kcal mol−1). Alternatively, when two of
these pendant groups are on the same side and one opposite, only
two B–N bonds are formed as shown for 6a in Fig. 4.
Reaction
DH◦soln/kcal mol−1
3B → B3 (4a) + 3H2O
4.38
−2.56
2.06
−0.10
15.25
6.50
A + 2B → AB2 (3a) + 3H2O
2A + B → A2B (5aꢀ) + 3H2O
2A + B → A2B (5a) + 3H2O
3A → A3 (6ꢀ) + 3H2O
3A → A3 (6) + 3H2O
3B → B3 (4b) + 3H2O
4.09
A + 2B → AB2 (3b) + 3H2O
2A + B → A2B (5bꢀ) + 3H2O
2A + B → A2B (5b) + 3H2O
3B → B3 (4c) + 3H2O
−1.26
3.06
1.22
8.07
A + 2B → AB2 (3c) + 3H2O
2A + B → A2B (5cꢀ) + 3H2O
2A + B → A2B (5c) + 3H2O
−0.99
2.47
0.11
As the number of pendant N,N-dimethylaminomethyl groups
increases, the computed B–N bond distances also increase. For
AB2, A2B and A3 the average calculated distances are 1.76, 1.82
(see Fig. 3b), new peaks immediately appear in the benzylic (see
Fig. 3b), methoxy (see Fig. 3b) and aromatic (see ESI,† Fig.
S1) regions. These results and similar results with 3a and 3b
suggest that, at room temperature and in the presence of excess
4-methoxyphenylboronic acid, hetero-arylboroxine 3c scrambles
with 4-methoxyphenylboronic acid forming an assortment of new
homo- and hetero-arylboroxines.
Computational methods were used to examine the thermody-
namic profile of arylboroxine products derived from the reaction of
two different arylboronic acid monomers. As described earlier, “A”
is used to denote self-coordinating arylboronic acid 1 while “B”
denotes non-coordinating arylboronic acids 2a–c. The calculated
solution-phase enthalpies for the formation of A3, A2B, AB2 and
B3 arylboroxines are shown in Table 1.
˚
and 1.87 A respectively. When the pendant groups are bound on
˚
opposite faces, one B–N bond is on average 0.4 A shorter than
the other. The AB2 ring (3a) remains relatively planar although
having one tetrahedral boron results in a slight pucker with an
O–B–O–B dihedral angle of 4.3◦ (directed toward the pendant
group) compared with the planar 4a structure. With an additional
pendant group, the A2B ring still remains relatively planar (the
average dihedral is 2.7◦) if the amine groups are on opposite faces
(5a). However, significant puckering of the ring is observed for
amine groups on the same face of the boroxine ring (5aꢀ) and
the average dihedral is 17◦. The structural trends for varying
the A group are reproduced in the methoxy and acetyl series of
arylboroxines.
In conclusion, we have synthesized and characterized a se-
ries of AB2-type hetero-arylboroxines using a straightforward
experimental procedure. Both the experimental and DFT calcu-
lations suggest that equilibrating mixtures of arylboroxine ring
compounds can be biased by modulating the coordination envi-
ronment of the arylboroxine ring itself. The hetero-arylboroxine
structural motif represents an entirely untapped, yet accessible,
reservoir of structural diversity that may further expand the
utility of arylboroxines in areas such as crystal engineering,
boron-containing materials, dynamic combinatorial chemistry
and supramolecular chemistry.
As expected, AB2 (3a–c) was the most stable for all three
para substituents studied. Formation of all AB2 structures is
exothermic. A2B (5a–c) was the next lowest in stability from
our calculations. With two pendant N,N-dimethylaminomethyl
groups, two possible isomers can be formed: both pendants on
opposite sides of the ring (5a) or on the same side (5aꢀ). Both
structures are shown in Fig. 4. 5a is energetically more stable than
5aꢀ for two reasons: less steric hindrance from the amine’s methyl
groups, and a lower net dipole in a less polar solvent. DH is close
to zero for the formation of A2B.
Acknowledgements
This research was supported by awards from Research Corpo-
ration (P. M. I. and J. K.), the ACS PRF (40382-GB4, P. M. I;
46700-GB10, J. K.), and the Dreyfus Foundation (J. K.). NSF
MRI CHE-0417731 (P. M. I.) and NSF CAREER (P. M. I.) are
also acknowledged.
References
1 E. F. Archibong and A. J. Thakkar, Mol. Phys., 1994, 81, 557.
2 D. L. Coopera, S. C. Wright, J. Gerratt, P. A. Hyams and M. Raimondi,
J. Chem. Soc., Perkin Trans. 2, 1989, 719.
3 D. W. Lamb, R. I. Keir and G. L. D. Ritchie, Chem. Phys. Lett., 1998,
291, 197.
Fig. 4 Calculated structures of AB2, A2B and A3 arylboroxine rings.
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 3791–3794 | 3793
©

View Article Online
4 P. v. R. Schleyer, H. Jiao, N. J. R. van Eikema Hommes, V. G. Malkin
and O. Malkina, J. Am. Chem. Soc., 1997, 119, 12669.
5 M. A. Beckett, G. C. Strickland, K. S. Varma, D. E. Hibbs, M. B.
Hursthouse and K. M. A. Malik, Polyhedron, 1995, 14, 2623.
6 W. L. Fielder, M. M. Chamberlain and C. A. Brown, J. Org. Chem.,
1961, 26, 2154.
7 J. F. Mariategui and K. Niedenzu, J. Organomet. Chem., 1989, 369, 137.
8 J. M. Ritchey, Ph.D. Thesis, University of Colorado, 1968.
9 H. R. Snyder, M. S. Konecky and W. J. Lennarz, J. Am. Chem. Soc.,
1958, 80, 3611.
10 Q. G. Wu, G. Wu, L. Brancaleon and S. Wang, Organometallics, 1999,
18, 2553.
11 A. P. Cote, A. I. Benin, N. W. Ockwig, M. O’Keeffe, A. J. Matzger and
O. M. Yaghi, Science, 2005, 310, 1166.
22 P. M. Iovine, M. N. Fletcher and S. Lin, Macromolecules, 2006, 39,
6324.
23 Crystal data for 3a: C21H20B3F2NO3 (Mr = 404.81), 0.20 × 0.20 ×
0.20 mm, orthorhombic, space group Pbca (No. 61), a = 11.5420(10)
◦
◦
◦
−3
˚
˚
˚
A, b = 18.9350(15) A, c = 19.6870(17) A, a = 90 , b = 90 , c = 90 ,
3
˚
volume = 4302.5(6) A , Z = 8, calculated density = 1.250 Mg m
,
absorption coefficient = 0.092 mm−1, T = 298 K, 23 336 reflections
collected, 3806 reflections were independent, Rint = 0.0450. Final R
values were R1 = 0.0510, wR2 = 0.1180 [I > 2r(I)], R indices (all data)
R1 = 0.0697, wR2 = 0.1281, GOF = 1.058.
24 Crystal data for 3b: C23H26B3NO5 (Mr = 428.88), 0.20 × 0.16 ×
0.12 mm, monoclinic, space group P21/n (No. 14), a = 12.8312(13)
◦
◦
˚
˚
˚
A, b =◦12.0846(12) A, c = 15.7327(16) A, a = 90 , b = 112.252(2) ,
3
˚
c = 90 , volume = 2257.8(4) A , Z = 4, calculated density = 1.262
12 H. M. El-Kaderi, J. R. Hunt, J. L. Mendoza-Cortes, A. P. Cote, R. E.
Taylor, M. O’Keeffe and O. M. Yaghi, Science, 2007, 316, 268.
13 Y. Qin, C. Cui and F. Ja¨kle, Macromolecules, 2007, 40, 1413.
14 G. Alcaraz, L. Euzenat, O. Mongin, C. Katan, I. Ledoux, J. Zyss, M.
Blanchard-Desce and M. Vaultier, Chem. Commun., 2003, 2766.
15 Y. Yang, T. Inoue, T. Fujinami and M. A. Mehta, J. Appl. Polym. Sci.,
2002, 84, 17.
Mg m−3, absorption coefficient = 0.086 mm−1, T = 100(2) K, 9555
reflections collected, 4959 reflections were independent, Rint = 0.0263.
Final R values were R1 = 0.0506, wR2 = 0.1255 [I > 2r(I)], R indices
(all data) R1 = 0.0758, wR2 = 0.1378, GOF = 1.005.
25 Crystal data for 3c: C25H26B3NO5 (Mr = 452.90), 0.12 × 0.08 ×
0.04 mm, monoclinic, space group P21/n (No. 14), a = 13.0525(12)
◦
◦
˚
˚
˚
A, b =◦12.2950(11) A, c = 16.2138(14) A, a = 90 , b = 111.504(2) ,
3
˚
16 A. B. Morgan, J. L. Jurs and J. M. Tour, J. Appl. Polym. Sci., 2000, 76,
c = 90 , volume = 2420.9(4) A , Z = 4, calculated density = 1.243
1257.
Mg m−3, absorption coefficient = 0.084 mm−1, T = 208 K, 14 020
reflections collected, 4286 reflections were independent, Rint = 0.0351.
Final R values were R1 = 0.0616, wR2 = 0.1710 [I > 2r(I)], R indices
(all data) R1 = 0.0943, wR2 = 0.1975, GOF = 1.033.
26 L. I. Bosch, M. F. Mahon and T. D. James, Tetrahedron Lett., 2004, 45,
2859.
17 Y. Li, J. Ding, M. Day, Y. Tao, J. Lu and M. D’iorio, Chem. Mater.,
2003, 15, 4936.
18 J. Kua, M. N. Fletcher and P. M. Iovine, J. Phys. Chem. A., 2006, 110,
8158.
19 J. Kua and P. M. Iovine, J. Phys. Chem. A., 2005, 109, 8938.
20 J. Beckmann, D. Dakternieks, A. Duthie, A. E. K. Lim and E. R. T.
Tiekink, J. Organomet. Chem., 2001, 633, 149.
21 E. K. Perttu, M. Arnold and P. M. Iovine, Tetrahedron Lett., 2005, 46,
8753.
27 J. C. Norrild and I. Sotofte, J. Chem. Soc., Perkin Trans. 2, 2002, 303.
28 L. Zhu, S. H. Shabbir, M. Gray, V. M. Lynch, S. Sorey and E. V. Anslyn,
J. Am. Chem. Soc., 2006, 128, 1222.
29 H. Ho¨pfl, J. Organomet. Chem., 1999, 581, 129.
3794 | Dalton Trans., 2008, 3791–3794
This journal is
The Royal Society of Chemistry 2008
©