Exploration of 1,2-phenylenediboronic esters as potential bidentate catalysts for organic synthesis

Article abstract of DOI:10.1007/s00706-012-0898-y

Recently, we described the first inverse electron demand Diels-Alder reaction of 1,2-diazene catalyzed by a bidentate Lewis acid. Herein we investigate 1,2-phenylenediboronic esters as potential catalysts for this transformation offering higher stability and easier handling than the currently used boranthracene derivatives. Different 1,2-phenylenediboronic esters were prepared and their ability to form bidentate coordination complexes with phthalazine was analyzed. Although a 1:1 complex was observed, X-ray analysis revealed binding only in a monodentate fashion.

Full text of DOI:10.1007/s00706-012-0898-y

Monatsh Chem (2013) 144:531–537
DOI 10.1007/s00706-012-0898-y
ORIGINAL PAPER
Exploration of 1,2-phenylenediboronic esters as potential
bidentate catalysts for organic synthesis
•
Samuel L. Bader Simon N. Kessler
•
•
Jennifer A. Zampese Hermann A. Wegner
Received: 1 October 2012 / Accepted: 4 December 2012 / Published online: 31 January 2013
Ó Springer-Verlag Wien 2013
Abstract Recently, we described the first inverse elec-
tron demand Diels–Alder reaction of 1,2-diazene catalyzed
by a bidentate Lewis acid. Herein we investigate 1,2-
phenylenediboronic esters as potential catalysts for this
transformation offering higher stability and easier handling
than the currently used boranthracene derivatives. Different
1,2-phenylenediboronic esters were prepared and their
ability to form bidentate coordination complexes with
phthalazine was analyzed. Although a 1:1 complex was
observed, X-ray analysis revealed binding only in a
monodentate fashion.
high sensitivity of this scaffold towards water and oxygen
requires special care in handling during preparation and
application. Therefore, we investigated in this report the
suitability of 1,2-bisboronic esters as bidentate Lewis acid
catalysts, as they show a much higher stability, which
would greatly facilitate their practical use.
Results and discussion
Preparation of bidentate Lewis acidic boronic esters
Keywords Boron ꢀ Catalysis ꢀ Coordination chemistry ꢀ
Lewis acid ꢀ NMR spectroscopy
Our approach to access phenyl-1,2-bisboronic esters relied on
tetrachloro-1,2-bisborobenzene(8) as a synthetic intermediate,
which can be obtained by metal exchange from the corre-
sponding 1,2-bis(trimethylsilyl)benzene (7) (Scheme 2) [8].
An iron-catalyzed Grignard method, developed by our group,
allowed the synthesis of 1,2-bis(trimethylsilyl)benzenes on a
large scale in moderate yield [9, 10]. The final transformation
to the boronic ester 10 involved exchange of the chlorides in 8
with 1,2-diols 9 according to the literature method (Scheme 2)
[11].
Introduction
The development of new catalytic systems for chemical
transformations is of increasing importance in view of the
shortage of natural resources. Recently, we presented a
bidentate Lewis acid [1] as a catalyst for the inverse
electron demand Diels–Alder (IEDDA) reaction of 1,2-
diazines (Scheme 1) [2, 3]. A bidentate coordination mode
to the 1,2-diazene was found to be crucial for catalytic
activity. In case of a monodentate interaction, the diazine
does not undergo the desired IEDDA reaction [4, 5].
Currently, our bidentate Lewis acid catalyst is based on
the boranthracene scaffold 6 combining the optimal
geometry with a high Lewis acidity [6, 7]. However, the
The results of the preparation of various boronic esters
10 are summarized in Table 1. The mediocre yields are
rationalized by the difficulty to purify the key intermediate
8 because of its high sensitivity. Impurities as well as their
reaction products had to be removed from the final product.
All purifications were done under nitrogen, as some of the
boronic esters might be sensitive towards water and oxy-
gen. Purification by crystallization was troublesome for the
chlorinated compounds 10c and 10d because they are
poorly soluble in most organic solvents. For the same
reasons analysis by solution-phase methods (e.g., NMR
spectroscopy) was only partially successful. However,
elemental analysis confirmed the purity of 10d.
S. L. Bader ꢀ S. N. Kessler ꢀ J. A. Zampese ꢀ H. A. Wegner (&)
Department of Chemistry, University of Basel,
Basel, Switzerland
e-mail: hermann.wegner@unibas.ch
123

532
S. L. Bader et al.
Scheme 1
R
R
N
N
B
B
5
1
N2
B
=
B
B
B
B
6
B
N
B
N
N
R
N
B
R
2
4
R
3
R
R
R
HO
HO
R
O
9
TMS
TMS
BCl2
BCl2
1M BCl3 in heptane (2.5 eq.)
DCE, 100 °C, 5 d
B
O
O
CHCl3, 0 °C
B
O
7
8
10
Scheme 2
Properties of bidentate Lewis acidic boronic esters
molecule can, with approximations, be correlated by NMR
spectroscopy. To rank the Lewis acidity of all synthesized
boronic esters, the 11B NMR shifts were measured and
collected in Fig. 1 [12]. Unfortunately, the solubility of
compound 10d was too low to allow the determination of
its chemical shift. The fluorinated compound 10e showed
the highest chemical shift and is estimated to be the
strongest Lewis acid in the series. As expected, the Lewis
acidity correlated with the electron-withdrawing ability of
the diol (Fig. 1).
As the application of the prepared Lewis acids as bidentate
catalysts was of special interest, the mode of coordination
(monodentate vs. bidentate) was evaluated. For this reason,
the interaction of the prepared bidentate Lewis acids with
substrates such as phthalazine (1) was investigated, mainly
with bis-boronic ester 10b.
NMR studies
Interestingly, compound 10c is an exception in this
collection. The negative inductive effect of the double
chloro-substituted catechol should lead to a larger electron-
withdrawing effect compared to the unsubstituted catechol
10b. Nevertheless, compound 10c is the most high-field-
shifted in this row of Lewis acids. Additionally, the
chemical shifts of the protons next to the carbon–boron
The Lewis acidity of the 1,2-phenylenediboronic esters is
one of the crucial parameters determining their ability to
form coordination bonds. In similar structures the acidity
depends mainly on the electron density at the Lewis acidic
centers, which can be altered by electron-withdrawing
substituents. The electron density at specific atoms in a
123

1,2-Phenylenediboronic esters
533
bond were studied, allowing a comparison with the 1H
NMR data of the octachlorinated compound 10d. The
results displayed the same trend, whereas an increase in
chlorine substitution leads to a larger high-field shift. This
behavior might originate either from a change in the
paramagnetic constant or the positive mesomeric effect of
the chlorine substituents. However, stronger electron-
withdrawing substituents on the aromatic backbone are of
special interest for the future design of more powerful
Lewis acids.
Table 1 Synthesized 1,2-phenylenediboronic esters
R
R
HO
R
O
B
9
BCl2
BCl2
HO
O
O
CHCl3, 0 °C
B
O
8
10
Entry
1
Diol
Product
Yield/ %
38
Complex formation experiments
10a
The determination of chemical shifts gave a crude estimate
of the strength of the Lewis acids 10. However, it is neither
able to give a quantitative result nor to describe the com-
plex formation ability. These parameters were studied in
complexation experiments with Lewis bases such as
phthalazine (1). Lewis acid 10b was titrated with phthal-
azine (1) and the chemical shifts were recorded by 1H
NMR spectroscopy (Fig. 2, spectrum 1 represents boronic
ester 10b and spectrum 13 phthalazine). As the coordina-
tion results only in a high-field shift of the signals of 10b
and not in the development of new resonances, the com-
plex formation has to be fast compared on the NMR
timescale. Therefore, the observed shift is the average
between the shift of the free acid and those of the formed
complexes [13].
HO
HO
9a
2
10b
48
HO
HO
9b
3
10c
46
HO
Cl
Cl
1
The H NMR signal at d = 7.65 ppm from compound
10b was chosen for further analysis, as it shows the
strongest shift without any overlapping with other peaks.
The bidentate nature of the used Lewis acids as well as the
Lewis bases allows not only 1:1 complexes. To determine
the stoichiometry of the complex, a Job plot of the mea-
sured shifts was calculated. The plot had a maximum at
around 0.48 indicating a 1:1 complex (Fig. 3) [14].
HO
9c
4
10d
36
Cl
HO
HO
Cl
Cl
The equilibrium constant (K) was calculated from the
concentration of the reaction partners in the equilibrium
1
based on the H NMR signals of the boronic ester 10b at
8.11 and 7.65 ppm. The averaged result gives an equilib-
rium constant of K = 6,936 dm3/mol. As this analysis
relies only on two data points it certainly includes a sub-
stantial statistical error. Nevertheless, it shows clearly that
a rather strong complex is formed.
Cl
9d
5
10e
54
CF3
CF3
CF3
CF3
Vapor pressure osmometry (VPO)
HO
The stoichiometry of the complex was further investigated
by VPO measurements. An equimolar solution of the
Lewis acid 10b (M = 314 g/mol) and phthalazine
(M = 130 g/mol) was analyzed and the average molar
weight of the complex (theoretical M for 1:1 complex,
HO
9e
123

534
S. L. Bader et al.
Fig. 1 11B NMR (in CD2Cl2)
shifts of synthesized bidentate
boronic Lewis acids
444 g/mol) was found to be 388 g/mol. The increased
average molar mass compared to the non-complexed Lewis
acid is an additional proof of the complex formation in
solution. Additionally, the VPO measurement excluded the
occurrence of any kind of agglomerate and oligomer.
(1) (CCDC 903492) were obtained from methylene chlo-
ride and subjected to X-ray analysis (Fig. 4).
For 10b, both boron atoms are trigonal planar as
expected, revealing sp2 hybridization. One of the boronic
esters lies in the same plane as the aromatic backbone. This
observation might be evidence for an interaction between
the different systems resulting in a higher bond order of the
carbon(2)–boron(2) bond, which would be reflected in a
decrease of its length; however, both carbon–boron bonds
have about the same length, so this does not seem to be the
case. This arrangement could also be explained by an
interaction between boron(1) and oxygen(3). Such an
interaction is, however, improbable, as boron(1) shows no
evidence of sp3 hybridization and lies on one line with the
O(3)–C(13) bond. The planar geometry is most likely due to
the crystal packing and might not be significant in solution.
Even more interesting is the structure of the complex
from Lewis acid 10b and phthalazine (Fig. 5). Most
importantly, the formation of two coordination bonds
between the boronic centers and the 1,2-diazenes is not
observed. The X-ray structure shows clearly only one
X-ray analysis
The Job plot determined the Lewis acid to substrate ratio of
the complex; however, it gives no information about the
number of centers involved in this coordination. One of the
most accurate ways to investigate the structure of a mole-
cule (at least in the solid state) is the measurement of an
X-ray diffraction structure. Suitable crystals of Lewis acid
10b (CCDC 903493) and its 1:1 complex with phthalazine
0.1
0.05
0
0
0.5
1
Phthalazine
Xbase
Fig. 3 Job plot with the measured data (cross) and the fitted bell-
shaped curve (line)
Fig. 2 1H NMR titration of boronic ester 10b with phthalazine (1)
123

1,2-Phenylenediboronic esters
535
connection between boron(1) and nitrogen(1). Boron(1)
shows sp3 hybridization; the bond to nitrogen(1) is just
slightly longer than the other three bonds of boron(1)
indicating the high strength of the formed bond. The reason
why only one bond is formed may originate from a decrease
of the acidity caused by a reduction of the electron-with-
drawing effect on boron(2) through the first coordination to
boron(1). Additionally, the close proximity of boron(2) and
oxygen(1) indicates an additional interaction between these
atoms, which results in a further weakening of the Lewis
acidity. The sp2 geometry of boron(2) is slightly distorted
towards oxygen(1) allowing perfect overlap with the empty
p orbital. Moreover, the oxygen(1) lies in one plane with
B(1), B(2), C(1), and C(2). Interestingly, the monodentate
complex does not exhibit any symmetry, which is in con-
trast to the observations by NMR spectroscopy. As already
mentioned the formation of the complex is fast compared on
the NMR timescale; the exchange of the different positions
might be even faster. Therefore, the NMR showed the
average of the different dynamic complexes.
refluxed and freshly distilled from sodium benzophenone
ketyl under nitrogen. Technical grade solvents for extrac-
tions and chromatography were distilled once before usage.
NMR solvents were obtained from Cambridge Isotope
Laboratories (Andover, MA, USA). Air-sensitive reactions
were set up in dry glassware that had previously been
heated to 200 °C and dried in several evacuation–flush
cycles or just flushed with nitrogen. All manipulations of
air-sensitive chemicals were either done by Schlenk tech-
niques (liquids) or in a nitrogen-filled glovebox (solids).
Bruker DPX-NMR (400 MHz) (also 19F-NMR, 376
MHz) and Bruker BZH-NMR (250 MHz) instruments were
1
used to measure the H NMR spectra. Two-dimensional
spectra were recorded on a Bruker Avance 500 (500 MHz)
as well as 13C NMR (100.6 MHz) and 11B NMR (160
MHz). Chemical shifts (d) are reported in parts per million
(ppm) relative to residual solvent peaks or tetramethylsil-
ane; coupling constants (J) are given in Hertz (Hz). The
measurements were done at room temperature (rt). The
multiplicities are reported as s singlet, d doublet, t triplet, m
multiplet, and dd doublet of a doublets. For multiplets only
the chemical shift (ppm) of the center is reported. FT-IR
spectra were recorded on a Shimadzu FTIR-8400S. The
compounds were measured tel quel through a Specac
Golden Gate ATR sampling system. Mass spectra were
recorded on a Finnigan MAT 95Q instrument for electronic
ionization (EI). GC–MS analysis was performed on a
Hewlett Packard 5890 series II gas chromatography system
with a Macherey–Nagel OPTIMA 1 Me2Si column
(25 m 9 0.2 mm 9 0.35 m), at 1 cm3/min He flow rate
(split = 20:1) using a Hewlett Packard 5971 mass-selec-
tive detector (EI 70 eV). Elemental analyses (C, H) were
conducted using a Perkin-Elmer 240; the results were
In summary, different 1,2-phenylenediboronic esters
were prepared and their binding properties with 1,2-diazine
were investigated. Analysis via NMR revealed a 1:1
complex. However, only one boron atom interacts with one
nitrogen of the 1,2-diazene, as no bidentate coordina-
tion was observed by X-ray analysis. These studies give
valuable information about the future design of bidentate
Lewis acids as suitable catalysts for organic synthesis.
Future efforts will evaluate the potential of the bidentate
1,2-phenylenediboronic esters in this context.
Experimental
All chemicals were used as received from Acros, Alfa
Aesar, Fluka, Fluorochem, or Sigma-Aldrich without any
prior purification. Dry solvents for the reactions were
purchased from Fluka or Biosolve. THF was continuously
Fig. 5 Crystal structure of the complex of Lewis acid 10b and
phthalazine (1)
Fig. 4 Crystal structure of compound 10b
123

536
S. L. Bader et al.
and cooled in an ice bath. A solution of 1.39 g 1,2-
bis(dichloroboryl)benzene (8, 5.80 mmol, 1.00 eq.) in
8 cm3 chloroform was slowly added within 20 min. The
mixture was stirred for 2 h at 0 °C and then warmed to rt
overnight. After that, all volatile parts were removed by
vacuum distillation and the residue was dried in high
vacuum. The resulting solid was dissolved in 25 cm3 hot
toluene and cooled to rt overnight. The mixture was then left
unstirred for 8 h. The supernatant liquid was removed and
replaced by the same quantity of toluene. This heat-up and
replace procedure was repeated four times. Then, all volatile
parts were removed by distillation in vacuum. The resulting
solid was suspended in 30 cm3 chloroform and filtered. The
solid was dried in high vacuum to yield 1.20 g (46 %) of
found to be in good agreement ( 0.6 %) with the calcu-
lated values. VPO for molar mass calculations was
measured on a Knauer K-7000 vapor pressure osmometer.
1,2-Di(1,3,2-dioxaborolan-2-yl)benzene
(10a, C10H12B2O4)
In a dry round-bottom flask 1.19 g 1,2-bis(dichlorobo-
ryl)benzene (8, 4.95 mmol, 1.00 eq.) was dissolved in
10 cm3 chloroform and cooled in an ice bath. Within
25 min 0.61 g ethylene glycol (9a, 0.55 cm3, 9.9 mmol,
2.0 eq.) was slowly added. The mixture was stirred for 2 h
at 0 °C and then allowed to warm to rt and stirred
overnight. After 12 h the solvent and the hydrogen chloride
were removed by distillation in vacuum. The resulting solid
was dried in high vacuum, mixed with 3 cm3 hexane, and
heated to the boiling point. The emulsion was slowly
cooled to 30 °C and stirred at this temperature overnight.
First crystals emerged and the mixture was cooled to 0 °C.
The liquid part was removed and the yellowish crystals
were dried in high vacuum overnight to yield 0.41 g
(38 %) of the title compound. 1H NMR (400 MHz,
CDCl3): d = 7.69 and 7.43 (m, AA0XX0, 4H), 4.37 (s,
8H, HO-CH2-CH2-O) ppm; 13C NMR (100 MHz, CDCl3):
d = 133.8, 129.9, 66.6 ppm (carbon atoms bound to boron
atoms were not observed); 11B NMR (160 MHz, CD2Cl2):
d = 32.22 ppm; GC–MS (70 eV): m/z = 218 (M?, 90),
188 (100).
1
the title compound as a gray solid. H NMR (400 MHz,
THF-d8): d = 8.00 and 7.61 (m, AA0XX0, 4H), 7.44 (s, 4H,
H-catechol-Cl2) ppm; 13C NMR (100 MHz, THF-d8):
d = 148.9, 134.8, 130.7, 125.7, 114.0 ppm (carbon atoms
bound to boron atoms were not observed); 11B NMR
(160 MHz, CD2Cl2): d = 31.72 ppm; MS (70 eV):
m/z = 452 (100), 450 (M?, 81).
1,2-Bis(4,5,6,7-tetrachlorobenzo[d][1,3]dioxaborol-2-
yl)benzene (10d, C18H4B2Cl8O4)
In a dry round-bottom flask 2.75 g 3,4,5,6-tetrachlorocate-
chol (9d, 11.1 mmol, 2.00 eq.) was suspended in 20 cm3
chloroform and cooled in an ice bath. A solution of 1.33 g
1,2-bis(dichloroboryl)benzene (8, 5.55 mmol, 1.00 eq.) in
8 cm3 chloroform was slowly added within 25 min. The
mixture was stirred for 2 h at 0 °C and then warmed to rt
overnight. After that, all volatile parts were removed by
vacuum distillation and the residue was dried in high
vacuum. The resulting solid was suspended in 30 cm3 hot
toluene and cooled to rt overnight. The mixture was then
left unstirred for 8 h. The supernatant liquid was removed
and replaced by the same quantity of toluene. This heat-up
and replace procedure was repeated four times. The
resulting suspension was filtrated and washed with
10 cm3 toluene. The solid was dried in high vacuum to
yield 1.19 g (36 %) of the title compound as a gray solid.
The product is barely soluble and due to the high reactivity
only aprotic solvents can be used. Therefore, all of the
following analyses were done on a saturated suspension. 1H
NMR (400 MHz, THF-d8): d = 7.88 and 7.50 (m,
AA0XX0, 4H) ppm; 13C NMR (100 MHz, THF-d8):
d = 146.9, 133.6, 129.8, 124.7, 115.2 ppm (carbon atoms
bound to boron atoms were not observed); MS (70 eV):
m/z = 589 (100), 586 (M?, 38).
1,2-Di(benzo[d][1,3]dioxaborol-2-yl)benzene
(10b, C18H12B2O4)
In a dry round-bottom flask 0.55 g 1,2-bis(dichlorobo-
ryl)benzene (8, 2.3 mmol, 1.0 eq.) was dissolved in 7 cm3
chloroform and cooled in an ice bath. Solid catechol (9b,
0.51 g, 4.6 mmol, 2.0 eq.) was slowly added, the mixture
was stirred for 2 h at 0 °C, and then warmed to rt
overnight. After 16 h the solvent and the hydrogen chloride
were removed by distillation in vacuum. The resulting solid
was dissolved in 35 cm3 hot acetonitrile, hot filtrated, and
washed with 10 cm3 hot acetonitrile. The filtrate was
cooled to rt overnight and the crystals formed were filtered,
washed with 5 cm3 cold acetonitrile, and dried to yield
1
0.35 g (48 %) of the title compound as white crystals. H
NMR (400 MHz, CDCl3): d = 8.08 and 7.65 (m, AA0XX0,
4H, H-aryl), 7.25 and 7.12 (m, AA0XX0, 4H, H-catechol)
ppm; 13C NMR (100 MHz, CDCl3): d = 148.9, 135.4,
131.2, 123.2, 113.0 ppm (carbon atoms bound to boron
atoms were not observed); 11B NMR (160 MHz, CD2Cl2):
d = 32.86 ppm; GC–MS (70 eV): m/z = 314 (M?, 100).
1,2-Bis(5,6-dichlorobenzo[d][1,3]dioxaborol-2-yl)benzene
(10c, C18H8B2Cl4O4)
1,2-Bis[4,4,5,5-tetrakis(trifluoromethyl)-1,3,2-dioxaboro-
lan-2-yl]benzene (10e, C18H4B2F24O4)
In a dry round-bottom flask 2.08 g 4,5-dichlorocatechol (9c,
11.6 mmol, 2.00 eq.) was suspended in 10 cm3 chloroform
In a dry round-bottom flask 0.733 g 1,2-bis(dichlorobo-
ryl)benzene (8, 3.06 mmol, 1.00 eq.) was dissolved in
123

1,2-Phenylenediboronic esters
537
7.5 cm3 chloroform and cooled in an ice bath. A solution of
2.04 g hexafluoro-2,3-bis(trifluoromethyl)-2,3-butandiol
(9e, 6.12 mmol, 2.00 eq.) in 4 cm3 chloroform was slowly
added within 15 min. The mixture was stirred for 2 h at
0 °C and then warmed to rt overnight. After 16 h all
volatile parts were removed by distillation in vacuum and
the residue was dried in high vacuum. The resulting solid
was dissolved in 9 cm3 hot hexane and slowly cooled to rt.
No crystals emerged and the solution was lyophilized to get
1.26 g (54 %) of the title compound as a white solid in a
References
1. Wegner HA, Kessler SN (2012) Synlett 699–705
2. Kessler SN, Wegner HA (2010) Org Lett 12:4062
3. Kessler SN, Neuburger M, Wegner HA (2011) Eur J Org Chem
3238–3245
4. Tu¨rkmen YE, Montavon TJ, Kozmin SA, Rawal VH (2012) J Am
Chem Soc 134:9062
5. Uno H, Okada S-I, Suzuki H (1991) Tetrahedron 47:6231
6. Lorbach A, Bolte M, Lerner H-W, Wagner M (2010) Chem
Commun 46:3592
7. Lorbach A, Bolte M, Li H, Lerner H-W, Holthausen MC, Ja¨kle F,
Wagner M (2009) Angew Chem Int Ed 48:4584
8. Kaufmann D (1987) Chem Ber 120:901
9. Bader SL, Kessler SN, Wegner HA (2010) Synthesis 2759–2762
10. Lorbach A, Reus C, Bolte M, Lerner H-W, Wagner M (2010)
Adv Synth Catal 352:3443
11. Bettinger HF, Filthaus M (2007) J Org Chem 72:9750
12. Nozaki K, Yoshida M, Takaya H (1995) Angew Chem Int Ed
33:2452
13. Hu¨nig S, Kraft P (1990) Chem Ber 123:895
14. Farcasiu D, Lukinskas P, Ghenciu A, Martin R (1999) J Mol
Catal A Chem 137:213
1
purity of 70 % (1H NMR). H NMR (400 MHz, CDCl3):
d = 8.10 and 7.75 (m, AA0XX0, 4H) ppm; 13C NMR
(100 MHz, CDCl3): d = 138.2, 133.6, 122.0, 119.1 ppm
(carbon atoms bound to boron atoms were not observed);
11B NMR (160 MHz, CD2Cl2): d = 33.52 ppm; 19F NMR
(376 MHz, CDCl3): d = 70.08 (s) ppm; MS (70 eV):
m/z = 762 (M?, 69), 215 (100).
Acknowledgments H.A.W. is indebted to the Fonds der Chemis-
chen Industrie for a Liebig fellowship.
123