Dearomatizing and Derivatizing a Mesityl Group on Boron by One-Pot Photoisomerization and [4+2] Diels–Alder Addition

Article abstract of DOI:10.1002/chem.201903534

Boron compounds having a conjugated chelate backbone (N,C-chelate or C,C-chelate) and two mesityl substituents on boron have been found to undergo a facile one-pot transformation/reaction with dienophiles, which leads to the dearomatization of one mesityl ring and its [4+2] Diels–Alder addition with the dienophile. Photochemical activation is the key in this transformation of the aryl ring.

Full text of DOI:10.1002/chem.201903534

DOI: 10.1002/chem.201903534
Full Paper
&
B-Mesityl Dearomatization
Dearomatizing and Derivatizing a Mesityl Group on Boron by
One-Pot Photoisomerization and [4+2] Diels–Alder Addition
Billy Deng, Xiang Wang, and Suning Wang*[a]
Abstract: Boron compounds having a conjugated chelate
backbone (N,C-chelate or C,C-chelate) and two mesityl sub-
stituents on boron have been found to undergo a facile
one-pot transformation/reaction with dienophiles, which
leads to the dearomatization of one mesityl ring and its
[4+2] Diels–Alder addition with the dienophile. Photochemi-
cal activation is the key in this transformation of the aryl
ring.
Introduction
Arene dearomatization is a powerful approach for generating
structurally complex alicyclic building blocks in total synthe-
sis.[1a] Among the myriad of dearomatization strategies,[1c,2c]
dearomatization by cycloaddition has the advantage of high
atom efficiency, and significantly increases the structural com-
plexity in a single step by forming bridged or bicyclic products
from simple arenes.[1] The Diels–Alder (D-A) [4+2] reaction is
perhaps the most widely used and versatile mode of cycload-
dition in synthetic chemistry.[2a,b] However, despite its impor-
tance, D-A reactions of non-activated monocyclic arenes (e.g.,
benzene, toluene, mesitylene) continue to pose a synthetic
challenge due to their exceptional stability conferred by the ar-
omaticity.[1b] Typical D-A reactions of non-activated monocyclic
arenes require high temperatures/pressures or transition metal
catalysts,[3] or activated dienophiles.[4] Transition-metal-mediat-
ed dearomatization has been extensively explored in recent
decades, whereby simple arenes are either h6- or h2-coordinat-
ed to transition metals, enabling their reactivity towards elec-
trophiles in D-A additions under relatively mild conditions.[3]
Recently, certain boron chelate compounds bearing two aryl
substituents on boron have been found to readily undergo
quantitative photoisomerization, dearomatizing one aryl group
to a cyclodienyl unit that remains h2-coordinated to the boron
atom. The product could be described as a boratanorcaradiene
(BNCD), a 7-borabicyclo[4.1.0]hepta-2,4-diene, or a borirane-
fused cyclohexadiene (e.g., 1a–3a in Scheme 1).[5] Compared
to the chelate starting materials (e.g., 1–3 in Scheme 1), which
are thermally stable and chemically inert, the species 1a–3a
are much higher in energy (by ca. 30 kcalmolÀ1) and their bor-
irane-cyclohexadienyl unit is highly electron-rich and reactive
Scheme 1. Sequential photoisomerization and [4+2] D-A reaction of com-
pounds 1–3.
towards oxygen.[5a] We therefore considered the possibility of
taking advantage of this photoisomerization to derivatize aryl
substituents in molecules such as 1–3. It was envisaged that
this might be achieved by a one-pot process involving initial
generation of the BNCD unit by photoisomerization, followed
by a [4+2] D-A reaction with an appropriate alkyne or alkene,
as illustrated in Scheme 1. To our delight, we found the photo-
chemically generated species 1a–3a to readily react with elec-
trophilic alkynes/alkenes in a highly stereoselective manner
under mild conditions, quantitatively affording the correspond-
ing [4+2] D-A cycloadducts. The unique features of this ap-
proach include the absence of transition metal catalysts as well
as the mild reaction conditions. The key activation step of this
reaction is driven by light, a form of renewable energy. Fur-
thermore, while examples of [4+2] cycloadditions of alkynes[6]
and azides[7] to boracyclopolyenes have been reported previ-
ously, they involve highly reactive boroles. To the best of our
knowledge, cycloaddition reactions involving species such as
BNCDs have been hitherto unknown.
Herein, we demonstrate a series of stepwise, light-assisted,
catalyst-free [4+2] cycloadditions of BNCDs with dienophiles
under ambient conditions, affording a series of air-stable, borir-
ane-fused D-A adducts. This new light-driven dearomatization
and functionalization of arenes further enriches the metal-free
bond-activation chemistry facilitated by boron.[8]
[a] B. Deng, Dr. X. Wang, Prof. Dr. S. Wang
Department of Chemistry, Queen’s University
Kingston, Ontario, K7L 3N6 (Canada)
E-mail: sw17@queensu.ca
Supporting information and the ORCID identification number(s) for the
author(s) of this article can be found under:
https://doi.org/10.1002/chem.201903534.
Chem. Eur. J. 2019, 25, 1 – 8
1
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
appreciable reactivity towards 1a–3a at ambient temperature.
Because 1a and 2a either undergo rapid thermal reversion to
1 and 2, respectively, or transform to other species at high
temperatures,[5a,b] we restricted the temperature of the reaction
with dienophiles to ambient or 308C. Compound 3a shows
high thermal stability and does not thermally revert back to
3.[5c] Nonetheless, apart from DMAD, none of the dienophiles
in Scheme 3 react with 3, even at elevated temperature (e.g.,
1008C).
Results and Discussion
One-pot syntheses of 1b–3b and 1c
Boranes 1–3 were obtained according to previously reported
procedures.[5a–c] They are air-stable and thermally stable. Upon
photoirradiation, they can be quantitatively converted into
their corresponding dark-coloured isomers. The one-pot proce-
dure used in the present investigation (Scheme 2) involves ini-
The 1H NMR spectra in Figure 1 demonstrate the quantitative
one-pot transformation of 1 into 1b. Irradiating a solution of 1
in benzene (0.04m) at 365 nm for about 12 h resulted in quan-
titative conversion into 1a. After the addition of excess DMAD
to the solution of 1a and heating the mixture at 308C over-
night, the D-A product 1b was quantitatively generated. The
DMAD starting material (99%, Aldrich) contained unidentified
impurities, which could be readily separated from the product
by column chromatography on neutral activated alumina. The
one-pot conversions of 2 into 2b and of 3 into 3b proceeded
in a similar manner (see the Supporting Information), except
that for 3 excitation at 300 nm was necessary owing to its con-
siderably blue-shifted absorption bands compared to those of
1 and 2. Interestingly, compared to those with 1a and 2a, the
D-A reaction of 3a with DMAD to produce 3b was much
faster, reaching completion within a few hours at 308C. Con-
sidering the strong s-donating NHC chelate and the high elec-
tron density on the borirane ring in 3a, the strong electron-do-
nating substituent on boron accelerates D-A reactions. Com-
pounds 1b–3b proved to be unstable on silica or basic alumi-
na columns, but could be easily purified/isolated on neutral
alumina with only a small amount of decomposition. 1b and
2b were obtained as orange crystals, and 3b as light-yellow
crystals, from a solvent mixture of hexanes/ethyl acetate. 1b–
Scheme 2. Synthetic routes to compounds 1b–3b and 1c and their isolated
yields.
tial photochemical conversion of 1–3 to their respective dark
borirane isomers 1a–3a in J-Young NMR tubes, in benzene or
toluene, as we have described previously.[5] This photochemical
conversion can be readily tracked by NMR spectroscopy. Upon
full conversion to 1a–3a, various dienophiles were examined
for their reactivity towards the cyclohexadienyl units therein.
Among the alkyne dienophiles we examined, dimethyl acetyl-
enedicarboxylate (DMAD) proved to be the most reactive
toward 1a–3a, showing appreciable reactions at ambient tem-
perature, which may be attributed to its relatively high electro-
philicity toward the electron-rich dienyl unit. Other alkyne di-
enophiles, such as those shown in Scheme 3, did not show any
Figure 1. 1H NMR spectra showing the clean photochemical conversion of 1
into 1a (the green highlighted peaks) and the quantitative formation of 1b
(the blue highlighted peaks) after the addition of ca. 40 equiv. of DMAD (the
yellow highlighted peak at d=3.06 ppm along with associated impurity
peaks) to a 0.04m solution of 1a in C6D6.
Scheme 3. Dienophiles that failed to react with compounds 1–3.
&
&
Chem. Eur. J. 2019, 25, 1 – 8
www.chemeurj.org
2
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
3b proved to be air-stable and their solutions could be heated
at 908C overnight under inert conditions without undergoing
any changes.
3a, as befits saturated borirane structures.[5,10] Similarly, com-
pound 1c shows an 11B NMR peak at d=À8.5 ppm.
The structures of 1b–3b and 1c were first determined by
comprehensive 2D NMR spectroscopic analysis, which estab-
lished that the addition of the dienophile to the cyclohexa-
dienyl ring proceeds exclusively from the exposed face of the
latter, instead of the cavity between the mesityl and cyclohexa-
dienyl rings, owing to the far lower steric hindrance. The re-
sulting 3-boratricyclo[3.2.2.02,4]nona-6,8-diene has the endo
stereogeometry, following the typical [4+2] D-A reaction path-
way.[2] The protons of the Med and Mee groups as well as Ha on
the D-A addition unit of 1b show long-range couplings with
one of the methyl groups (Mef) of the mesityl unit, as shown
by the 2D NOESY spectrum in Figure 3. Similar coupling pat-
Among the various alkene dienophiles (Scheme 3) that we
examined, only maleic anhydride (MA) showed D-A reactivity
with the borirane molecule 1a. The reaction of 1a with
5 equiv. of MA reached completion in about 20 min at room
temperature, quantitatively affording the D-A product 1c, as
shown in Figure 2. 1c could not be purified by column chro-
Figure 2. 1H NMR spectra showing the quantitative formation of 1c (the
blue highlighted peaks) by the reaction of 1a with 5 equiv. of maleic anhy-
dride (MA, the yellow highlighted peaks, 0.03m) in C6D6 at room tempera-
ture.
Figure 3. 2D H,H NOESY of 1b showing the long-range coupling of the pro-
tons of Mef on the mesityl unit with those of the bridged ring (mixing time
0.4 s, recorded in C6D6).
matography. When the reaction of 1a with MA was repeated
at a 1:1.05 ratio under the same conditions, it reached comple-
tion in about 2 h, and without further purification, compound
1c could be isolated in 42% yield as orange crystals by recrys-
tallization. Compounds 1b–3b and 1c were fully characterized
terns were also observed in the 2D NOESY spectra of 1c and
2b (see the Supporting Information). The signal of ortho-H1 on
the phenyl ring of 1b is unusually downfield shifted, appearing
as a doublet at dꢀ9.2 ppm (Figure 1). The corresponding H1
protons in compounds 2b and 3b show similar chemical shifts
(see the Supporting Information). One possible explanation for
this is the formation of an H bond between the H1 atom and
one of the carbonyl oxygen atoms. This is indeed confirmed
by the crystal structures of 1b–3b. The borirane species 1a is
known to undergo slow enantiomer interconversion in solution
through a 1,3-sigmatropic BÀC bond migration, based on the
observation of a chemical exchange peak in the NOESY spec-
trum between the two methyl groups on the sp3 carbon
atoms of the borirane unit.[5a] Interestingly, similar enantiomer
interconversion was not observed for 1b, possibly because of
its highly congested nature.
1
by H, 13C, and 11B NMR, HRMS, and single-crystal X-ray diffrac-
tion analysis (see the Experimental Section and the Supporting
Information for details).
We also examined unsymmetrical molecules,[9] such as
BMes(Ar) chelate compounds, for possible selective aryl activa-
tion by the same approach. However, the borirane-cyclohexa-
dienyl units generated from the Ar rings of the unsymmetrical
molecules underwent efficient thermal isomerization that effec-
tively competed with the intermolecular D-A addition reac-
tions. We attempted low-temperature photoisomerization of
the unsymmetrical boron compounds in order to slow down
the undesired thermal isomerization and achieve effective D-A
additions. However, only thermal isomerization products were
observed, and no [4+2] D-A reactivity was detected for the un-
symmetrical boron compounds.
The formation of four new stereogenic centres in 1c is evi-
1
dent from its H-COSY NMR spectrum (see the Supporting In-
Structures of 1b–3b and 1c
formation). The clean formation of one stereoisomer shown by
NMR for 1b–3b and 1c also supports the high stereoselectiv-
ity of the adduct formation. The absence of a highly deshield-
ed doublet for the ortho-H proton on the phenyl ring in 1c in-
Compounds 1b–3b display distinct 11B NMR peaks at d=À5.7,
À6.3, and À16.0 ppm, respectively (recorded in C6D6), which
are downfield-shifted by a few ppm relative to those of 1a–
Chem. Eur. J. 2019, 25, 1 – 8
www.chemeurj.org
3
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
dicates that the carbonyl groups of the anhydride unit are ori-
ented away from the phenyl ring, with exclusive formation of
the endo adduct.[11]
Table 1. Selected bond lengths (ꢁ) and angles (8) for 1b–3b and 1c.
Bond lengths and angles
The crystal structures of 1b/1c and 2b/3b are shown in Fig-
ures 4 and 5, respectively. Selected bond lengths and angles
are shown in Table 1. The two C=C double bonds in the cyclo-
addition unit (C4=C5 1.322–1.326 ꢁ, C10=C11 1.341–1.345 ꢁ) of
1b–3b have typical olefinic bond lengths. The C5=C6 double
bond in the cycloaddition unit of 1c also has a typical C=C
bond length of 1.324(3) ꢁ. The crystal structures show that one
of the CO2Me groups (C14 in 1b, C12 in 2b and 3b) is close to
the phenyl ring and oriented approximately perpendicular
with respect to the C=C bond. Indeed, the carbonyl oxygen
atom of this group forms an H bond with an H atom of the
phenyl ring (that bound to C17 in 1b and 2b, and to C18 in
3b), as evidenced by the separation distance of 2.18–2.20 ꢁ
between the oxygen atom and the H atom. This observation is
consistent with the deshielded chemical shift of the H1 proton
of 1b shown in Figure 1. The crystal structures confirmed the
stereogeometry established from the NMR data. The borirane
ring remains intact in all four structures. 1b–3b all have a rela-
tively long BÀC bond in the borirane ring (B1ÀC1=1.664(3),
1.695(3), and 1.6976(17) ꢁ for 1b–3b, respectively). Compared
to the previously characterized borirane ring in 3a, D-A adduct
formation increases the BÀCPh bond length in 3b by about
0.013 ꢁ, which may be ascribed to the increased steric conges-
tion about the B atom. In the adducts 1b–3b, the bicy-
clo[2.2.2]octa-2,5-diene unit is approximately perpendicular to
B1ÀC1
B1ÀC2
B1ÀN1
B1ÀC27
C1ÀC2
C2ÀC3
C3ÀC4
C4ÀC5
C1ÀC6
C5ÀC6
1.664(3)
1.582(3)
1.600(3)
1.605(3)
1.553(3)
1.562(3)
1.529(3)
1.322(3)
1.590(3)
1.513(3)
C3ÀC11
1.521(3)
1.550(3)
1.341(3)
57.09(14)
58.80(14)
64.11(15)
116.28(19)
132.04(19)
107.77(19)
128.09(18)
C6ÀC10
C10ÀC11
C2-B1-C1
C2-C1-B1
C1-C2-B1
C2-B1-N1
C27-B1-C1
N1-B1-C27
B1-C1-C6
1b
B1ÀC1
B1ÀC2
B1ÀN1
B1ÀC25
C1ÀC2
C1ÀC4
C2ÀC3
C2ÀC7
C3ÀC11
C3ÀC6
C4ÀC5
1.636(3)
1.602(3)
1.574(3)
1.607(3)
1.567(3)
1.563(3)
1.543(3)
1.525(3)
1.558(3)
1.508(3)
1.518(3)
C4ÀC10
1.581(3)
1.324(3)
1.539(3)
57.89(12)
59.99(12)
62.12(12)
109.84(16)
112.05(15)
120.44(14)
124.29(16)
C5ÀC6
C10ÀC11
C2-B1-C1
C2-C1-B1
C1-C2-B1
N1-B1-C25
N1-B1-C1
C4-C1-B1
C3-C2-B1
1c
B1ÀC1
B1ÀC2
B1ÀC26
B1ÀN1
C1ÀC2
C2ÀC3
C3ÀC4
C1ÀC6
C5ÀC6
C4ÀC5
1.695(3)
1.578(3)
1.597(3)
1.568(3)
1.551(3)
1.570(3)
1.519(3)
1.592(3)
1.508(3)
1.324(3)
C6ÀC11
1.546(3)
1.341(3)
C10ÀC11
C2-C1-B1
C1-C2-B1
N1-B1-C26
N1-B1-C1
C16-C1-B1
C6-C1-B1
B1-C2-C3
57.99(12)
65.57(13)
106.93(16)
109.31(16)
117.63(16)
122.80(16)
120.33(16)
2b
B1ÀC1
B1ÀC2
B1ÀC23
B1ÀC31
C1ÀC2
C1ÀC6
C2ÀC3
C3ÀC4
C4ÀC5
C5ÀC6
C3ÀC11
C6ÀC10
1.6976(17)
1.5941(17)
1.5969(17)
1.6092(18)
1.5316(16)
1.5813(16)
1.5705(17)
1.5268(17)
1.3265(18)
1.5138(17)
1.5195(17)
1.5517(17)
C4ÀC5
1.3265(18)
1.5517(17)
1.3353(19)
1.3353(19)
55.35(7)
58.89(7)
65.76(8)
121.35(10)
124.08(10)
119.53(10)
108.33(10)
C6ÀC10
C10ÀC11
C10ÀC11
C2-B1-C1
C2-C1-B1
C1-C2-B1
C2-B1-C23
C2-B1-C31
C3-C2-B1
C23-B1-C31
3b
Figure 4. Crystal structures of 1b and 1c with a partial labelling scheme and
35% thermal ellipsoids. H atoms are omitted for clarity. The mesityl group
and part of the dienophile are shown as sticks.
the borirane ring. The cyclohexene bridge (pink) (e.g., in 1b,
C3-C4-C5-C6) is bent away from the bridgehead (C3-C11-C10-
C6 in 1b) towards the mesityl unit.
The structure of 1c (Figure 4) shows that the maleic anhy-
dride unit in the adduct is indeed oriented away from the
phenyl ring without any intramolecular H-bond interactions.
This preferred orientation is likely dictated by the need to min-
imize steric interactions between the phenyl ring and the an-
hydride unit.
Absorption spectra of 1b–3b and 1c
Figure 5. Crystal structures of 2b and 3b with a partial labelling scheme
and 35% thermal ellipsoids. H atoms are omitted for clarity. The mesityl
group and part of the dienophile are shown as sticks.
The absorption spectra of 1b–3b and 1c, along with those of
1a–3a, are shown in Figure 6. For 1b, 1c, and 2b, the first
&
&
Chem. Eur. J. 2019, 25, 1 – 8
www.chemeurj.org
4
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
Figure 6. Absorption spectra of 1a–3a, 1b–3b, and 1c recorded in benzene
(normalized with respect to the first low-energy band). Inset: photographs
showing the solution colours of these compounds in benzene.
Figure 7. HOMOs and LUMOs of 1c and 1b–3b with an iso-contour of 0.03
for all surfaces.
low-energy absorption band appears at lmax ꢀ420 nm, which is
responsible for the yellow-orange colour of these three com-
pounds. In contrast, the first low-energy absorption band of
3b is hypsochromically shifted by about 60 nm relative to
those of 1b and 2b, making it pale-yellow. Compared to the
cyclohexadienyl borirane parent molecules 1a, 2a, and 3a,
which are much more intensely coloured, with the first absorp-
tion band appearing at around 600 nm for 1a and 2a and at
454 nm for 3a, the first absorption bands of the D-A borirane
species 1b–3b and 1c are sharply hypsochromically shifted.
This difference is caused by the unsaturated dienyl unit directly
attached to the borirane ring in 1a–3a, which significantly
contributes to the HOMO level, along with the borirane ring,
raising the HOMO energy level such that these compounds
display reactivity toward oxygen.[5] In 1b–3b/1c, D-A addition
converts the unsaturated sp2 carbon atoms attached to the
borirane ring into sp3 carbon atoms, which presumably greatly
reduces the contribution of the D-A unit to the HOMO level
and the electron density on the borirane unit, leading to a sta-
bilized HOMO and an increased HOMO–LUMO gap. To further
understand the electronic properties of 1b–3b/1c, TD-DFT
computational studies were performed at the B3LYP/6–311G(d)
level of theory.[12] HOMO–LUMO diagrams for 1b–3b/1c are
shown in Figure 7.
of their HOMO–LUMO gaps. For 2b and 3b, although there
are significant contributions from the borirane ring to the
HOMO, the contributions from the D-A adduct unit to the
HOMO are very small, leading to a widening of their HOMO–
LUMO gaps relative to those of 2a and 3a. In contrast to the
instability of 1a–3a toward oxygen, the relatively deeper
HOMOs of 1b–3b and 1c (stabilized by ca. 1 eV) make these
compounds stable to air. Furthermore, unlike their precursors
1a–3a, which are either thermally or photochemically active,
compounds 1b–3b and 1c do not show any thermal or pho-
toreactivities upon heating to 1008C or irradiation at around
lmax of their first low-energy absorption band.
Conclusions
Photochemical transformation of an aromatic mesityl group on
boron has been established as an effective method for activat-
ing it toward facile Diels–Alder addition reactions with dieno-
philes in a one-pot manner. This type of photo-initiated aryl ac-
tivation has been found to be a general reactivity for BMes2
units with different diaryl chelate backbones. The clean photo-
chemical conversion of the BMes2 chelate compounds into the
corresponding cyclohexadienyl-fused boriranes is the key step
in this type of aryl activation.
Consistent with their absorption spectra, the calculated
HOMO–LUMO gaps of 1b and 1c are significantly larger
(3.47 eV and 3.43 eV, respectively) than that of their precursor
1a (2.32 eV).[5a] TD-DFT data indicate that the vertical excitation
to the S1 state for 1b and 1c is dominated by the HOMO to
LUMO transition with an energy of around 2.85 eV, approxi-
mately 0.90 eV higher than that of 1a. Similar trends are also
found for 2b vs 2a and 3a vs 3b (see the Supporting Informa-
tion). As shown in Figure 7, the LUMO levels for 1b–3b and 1c
are mainly localized on their chelate backbones. Unlike in 1a–
3a, in which the HOMO is dominated by the borirane and the
cyclohexadienyl ring,[5a] the key contributions to the HOMOs of
1b and 1c stem from the mesityl ring and the C=C bonds in
the D-A adduct, which are clearly responsible for the widening
Experimental Section
General
Unless stated otherwise, all reactions were carried out under an
inert atmosphere of dry nitrogen in sealed J-Young NMR tubes
(conventional DURANꢂ glass). Starting reagents were purchased
from chemical suppliers and used without further purification.
Deuterated benzene (C6D6) was purchased from Cambridge Iso-
tope Laboratories and purged with dry nitrogen and stored over
molecular sieves prior to use. NMR spectra were recorded at 298 K
using the following spectrometers: Bruker Avance 400, Neo 500,
Avance 600, or Neo 700. Abbreviations: s=singlet, d=doublet,
dd=doublet of doublets, t=triplet, td=triplet of doublets, m=
Chem. Eur. J. 2019, 25, 1 – 8
www.chemeurj.org
5
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
multiplet, High-resolution mass spectra were recorded on
a
141.03, 140.90, 137.40, 136.92, 134.74, 134.31, 132.22, 131.16,
129.04, 128.20, 128.06, 127.92, 127.10, 123.90, 123.17, 54.65, 52.50,
51.87, 51.44, 23.90, 23.89, 21.30, 20.32, 20.27, 16.33, 11.15 ppm;
11B NMR (128 MHz, C6D6): d=À6.32 ppm; HREI-MS: calcd for
C34H36BNO4S [M]+: 565.2458; found: 565.2473.
Thermo Fisher Scientific MALDI LTQ Orbitrap XL mass spectrome-
ter. Exact masses were calculated based on the predominant com-
bination of natural isotopes. Photoisomerization reactions of all
compounds were performed in sealed J-Young NMR tubes using a
Rayonet photochemical reactor at 298 K. UV/Vis spectra were re-
corded on a Varian Cary 50 spectrophotometer.
Synthesis of 3b: Compound 3 was synthesized according to a pre-
vious literature procedure.[5c] Compound 3 (4.0 mg, 9 mmol) was
solubilized in C6D6 (0.51 mL, c=0.017m) and irradiated with UV
light at 300 nm for 3 h to fully convert it into 3a. DMAD (0.1 mL,
0.82 mmol) was then added to the reaction mixture in a single por-
tion. After heating in an oil bath at 308C for 2 h, the clear light-
yellow solution containing the product was concentrated in vacuo
and purified by column chromatography on neutral alumina elut-
ing with hexanes/ethyl acetate (9:1), affording a light-yellow resi-
due of 3b. Crystals of 3b were grown by slow evaporation of the
volatiles from a concentrated solution in hexanes/dichloromethane
(9:1) and washed with hexanes to afford the pure compound
Synthesis and characterization
Synthesis of 1b: Compound 1 was synthesized according to a pre-
vious literature procedure.[5a] 1 (15 mg, 37 mmol) was dissolved in
C6D6 (0.90 mL, c=0.04m) and irradiated at ambient temperature
with UV lamps at 365 nm for 8 h to fully convert colourless 1 into
dark-blue 1a, as monitored by 1H NMR spectrometry. DMAD
(0.1 mL, 0.82 mmol) was then added in a single portion to the solu-
tion of 1a. After heating overnight in an oil bath at 308C, the solu-
tion turned yellow and NMR spectra indicated nearly quantitative
formation of 1b. The clear yellow solution containing the product
was concentrated in vacuo and purified by column chromatogra-
phy on neutral alumina eluting with hexanes/ethyl acetate (9:1), af-
fording a yellow-orange residue. Crystals of 1b were grown by
slow evaporation of the volatiles from a concentrated solution in
hexanes/ethyl acetate (5:1) and subsequently washed with hexanes
1
(4.4 mg, 84%). H NMR (700 MHz, C6D6): d=9.39 (d, J=8.0 Hz, 1H;
o-Cbz-H), 7.68 (d, J=8.0 Hz, 1H), 7.65–7.61 (m, 1H; CBz-N-(Ph)C=
CÀH), 7.61 (t, J=8.0 Hz, 1H; m-3-Cbz-H), 7.05 (t, J=8.5 Hz, 1H; p-
CBz-H, overlapped with a C6D6 satellite peak), 6.85 (s, 1H; Mes-H),
6.84–6.80 (m, 2H; BzIm-H), 6.73 (s, 1H; Mes-H), 6.44–6.40 (m, 1H;
CH3-N-(Ph)C=CÀH), 5.27 (t, J=2.0 Hz, 1H; cyclohexene-sp2-CÀH),
4.43 (d, J=2.2 Hz, 1H; cyclohexene-sp3-CÀH), 3.60 (s, 3H; -OCH3),
3.41 (s, 3H; -OCH3), 2.75 (s, 3H; o-Mes-CH3), 2.49 (s, 3H; N-CH3),
2.29 (s, 3H; p-Mes-CH3), 2.00 (s, 3H; cyclohexene-CH3), 1.86 (s, 3H;
o-Mes-6-CH3), 1.49 (d, 4J=1.59, 3H; allylic CH3), 0.88 ppm (s, 3H;
BCC-CH3); 13C NMR (176 MHz, C6D6): d=170.01, 165.91, 162.37,
150.67, 145.78, 142.71, 141.30, 137.18, 136.60, 136.60, 135.58,
133.70, 132.71, 132.37, 130.18, 128.45, 128.20, 128.06, 127.92,
126.13, 124.14, 123.74, 123.09, 118.67, 114.45, 110.28, 56.46, 53.63,
51.87, 51.44, 31.50, 30.23, 24.86, 23.75, 21.91, 21.31, 20.70,
19.76 ppm; 11B NMR (225 MHz, C6D6): d=À15.96 ppm; HREI-MS:
calcd for C38H39BN2O4 [M]+: 598.3003; found: 598.3001.
1
to afford the pure product (19 mg, 93%). H NMR (400 MHz, C6D6):
d=9.19 (d, J=8.3 Hz, 1H; o-Cbz-H), 8.18 (dd, J=6.5, 1.7 Hz, 1H; o-
Py-H), 7.64 (t, J=7.5 Hz, 1H; m-3-Cbz-H), 7.58 (d, J=8.2 Hz, 1H; m-
5-Cbz-H), 7.39 (d, J=8.5 Hz, 1H; m-5-Py-H), 6.99 (t, J=7.6 Hz, 1H;
p-Cbz-H, overlapped with Mes-H), 6.95 (s, 1H; Mes-H, overlapped
with p-Cbz-H), 6.82 (s, 1H; Mes-H), 6.56 (td, J=8.3 Hz, 1.6 Hz, 1H;
p-Py-H), 5.88 (t, J=6.8 Hz, 1H; m-3-Py-H), 5.18 (s, 1H; cyclohexene-
sp2-CÀH), 4.42 (d, 4J=2.0 Hz, 1H; cyclohexene-sp3-CÀH), 3.60 (s,
3H; -OCH3), 3.38 (s, 3H; -OCH3), 2.66 (s, 3H; o-Mes-2-CH3), 2.27 (s,
3H; p-Mes-CH3), 2.10 (s, 3H; o-Mes-6-CH3), 2.07 (s, 3H; cyclohex-
ene-CH3), 1.52 (d, 4J=1.06 Hz, 3H; allylic CH3), 0.62 ppm (s, 3H;
BCC-CH3); 13C NMR (126 MHz, C6D6): d=169.49, 165.57, 160.24,
151.70, 148.61, 147.82, 147.61, 145.80, 145.64, 141.40, 140.24,
137.12, 136.20, 134.33, 131.61, 130.39, 127.98, 127.88, 127.69,
127.49, 126.28, 125.78, 122.94, 120.95, 119.93, 54.80, 52.52, 51.47,
51.04, 29.88, 23.68, 23.16, 20.92, 20.56, 20.05, 16.58 ppm; 11B NMR
(128 MHz, C6D6): d=À5.77 ppm; HREI-MS: calcd for C35H36BNO4
[M]+: 545.2737; found: 545.2738.
Synthesis of 1c: Compound 1a (8.8 mg, 21.8 mmol, 30 mm, 0.5 mL
of C6D6) was generated in the same manner as described for the
synthesis of 1b. Maleic anhydride (2.20 mg, 22 mmol) was added to
the reaction mixture in a single portion. After 24 h at room tem-
perature, the clear yellow solution was concentrated in vacuo and
then extracted with hexanes to afford a yellow residue of 1c. Crys-
tals of 1c were grown by slow evaporation of the solvent from a
concentrated solution in hexanes and washed with cold hexanes
Synthesis of 2b: Compound 2 was synthesized according to a pre-
vious literature report.[5b] 2 (22 mg, 51 mmol) was dissolved in C6D6
(2 mL, c=0.025m) and irradiated with UV light at 365 nm for
7 days to fully convert it into 2a. DMAD (0.1 mL, 0.82 mmol) was
then added in a single portion to the solution of 2a. After heating
overnight in an oil bath at 308C, the clear yellow solution contain-
ing the product was concentrated in vacuo and purified by
column chromatography on neutral alumina eluting with hexanes/
ethyl acetate (9:1), affording a yellow residue of 2b. Crystals of 2b
were grown by slow evaporation of the volatiles from a concen-
trated solution in hexanes/ethyl acetate (5:1) and washed with
hexanes to afford the pure compound (28 mg, 94%). 1H NMR
(700 MHz, C6D6): d=9.23 (d, J=8.5 Hz, 1H; o-Cbz-H), 7.56 (td, J=
7.8, 7.2, 1.3 Hz, 1H; m-3-Cbz-H), 7.49 (dd, J=7.8, 1.1 Hz, 1H; m-5-
Cbz-H), 6.95 (s, 1H; Mes-H), 6.86 (t, J=7.5 Hz, 1H; p-CBz-H), 6.84 (s,
1H; Mes-H), 6.75 (d, J=1.4 Hz, 1H; thiazole-H), 5.28 (t, J=2.0 Hz,
1H; cyclohexene-sp2-CÀH), 4.45 (d, J=2.2 Hz, 1H; cyclohexene-sp3-
CÀH), 3.62 (s, 3H; -OCH3), 3.38 (s, 3H; -OCH3), 2.66 (s, 3H; o-Mes-
2-), 2.27 (s, 3H; p-Mes-CH3), 2.16 (s, 3H; o-Mes-6-CH3), 2.08 (s, 3H;
1
to afford the pure compound (42%). H NMR (700 MHz, C6D6): d=
8.10 (d, J=6.4 Hz, 1H; o-Py-H), 7.96 (d, J=8.4 Hz, 1H; o-Cbz-H),
7.67 (d, J=8.2 Hz, 1H; m-5-Cbz-H), 7.43 (d, J=8.6 Hz, 1H; m-5-Py-
H), 7.29–7.26 (m, 1H; m-2-Cbz-H), 7.01 (t, J=7.6 Hz, 1H; p-Cbz-H),
6.88 (s, 1H; Mes-H), 6.75 (s, 1H; Mes-H), 6.57 (t, J=7.8 Hz, 1H; p-
Py-H), 5.89 (t, J=6.8 Hz, 1H; m-2-Py-H), 4.82 (t, J=1.9 Hz, 1H; cy-
clohexene-sp2-CÀH), 3.63 (d, J=8.1 Hz, 1H; exo-H), 3.23 (t, J=
2.6 Hz, 1H; cyclohexene-sp3-CÀH), 2.81 (dd, J=8.1, 3.2 Hz, 1H; exo-
H), 2.37 (s, 3H; o-Mes-CH3), 2.24 (s, 3H; p-Mes-CH3), 2.16 (s, 3H; cy-
clohexene-CH3), 1.87 (s, 3H; o-Mes-CH3), 1.46 (d, J=1.6 Hz, 3H; al-
lylic CH3), 0.05 ppm (s, 3H; BCC-CH3); 13C NMR (176 MHz, C6D6): d=
173.95, 173.09, 146.02, 145.51, 141.75, 140.21, 139.54, 136.07,
136.01, 135.14, 130.17, 129.96, 128.59, 128.20, 128.06, 127.92,
127.41, 122.62, 121.16, 120.45, 49.92, 49.09, 44.32, 43.36, 23.64,
23.56, 23.30, 21.83, 21.23, 13.97 ppm; 11B NMR (225 MHz, C6D6): d=
À8.70 ppm; HRESI-MS: calcd for C33H32BNO3 [M+1]: 502.2548;
found: 502.2536.
4
4
cyclohexene-CH3), 1.42 (d, J=1.7 Hz, 3H; allylic CH3), 1.16 (d, J=
1.2 Hz, 3H; thiazole-CH3), 0.59 ppm (s, 3H; BCC-CH3); 13C NMR
(176 MHz, C6D6): d =169.90, 165.83, 162.79, 159.64, 148.53, 147.59,
&
&
Chem. Eur. J. 2019, 25, 1 – 8
www.chemeurj.org
6
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
´
[5] a) Y. L. Rao, H. Amarne, S. Bin Zhao, T. M. McCormick, S. Martic, Y. Sun,
R. Y. Wang, S. Wang, J. Am. Chem. Soc. 2008, 130, 12898–12900; b) Y. L.
Rao, H. Amarne, L. D. Chen, M. L. Brown, N. J. Mosey, S. Wang, J. Am.
Chem. Soc. 2013, 135, 3407–3410; c) Y. L. Rao, L. D. Chen, N. J. Mosey, S.
Wang, J. Am. Chem. Soc. 2012, 134, 11026–11034; d) K. Nagura, S. Saito,
R. Frçhlich, F. Glorius, S. Yamaguchi, Angew. Chem. Int. Ed. 2012, 51,
7762–7766; Angew. Chem. 2012, 124, 7882–7886; e) Y.-L. Rao, C. Hçrl,
H. Braunschweig, S. Wang, Angew. Chem. Int. Ed. 2014, 53, 9095–9096;
Angew. Chem. 2014, 126, 9243–9243.
X-ray crystallographic analysis
Crystal data for 1b–3b and 1c were collected on a Bruker D8 Ven-
ture diffractometer with an Mo target (l=0.71073 ꢁ) at 180 K.
Data were processed on a PC with the aid of the Bruker SHELXTL
software package[13] and corrected for absorption effects. The crys-
tals of 1b, 1c, and 2b belong to the triclinic space group P-1,
whereas that of 3b belongs to the monoclinic space group P21/n.
All non-hydrogen atoms were refined anisotropically. The positions
of hydrogen atoms were calculated and their contributions were
included in the structural factors. Full details of the crystal data,
collection parameters, and results of analyses are provided in the
Supporting Information. Crystal data for 1b–3b and 1c have been
deposited with the Cambridge Crystallographic Data Centre. CCDC
1943914 (1b), 1943915 (2b), 1943916 (3b), and 1943917 (1c) con-
tain the supplementary crystallographic data for this paper. These
data can be obtained free of charge from the Cambridge Crystallo-
graphic Data Centre.
[6] a) C. Fan, W. E. Piers, M. Parvez, R. McDonald, Organometallics 2010, 29,
5132–5139; b) J. J. Eisch, J. E. Galle, J. Am. Chem. Soc. 1975, 97, 4436–
4437; c) J. J. Eisch, J. E. Galle, B. Shafii, A. L. Rhelngold, Organometallics
1990, 9, 2342–2349; d) H. Kelch, S. Kachel, J. Wahler, M. A. Celik, A.
Stoy, I. Krummenacher, T. Kramer, K. Radacki, H. Braunschweig, Chem.
Eur. J. 2018, 24, 15387–15391; e) J. J. Baker, K. H. M. Al Furaiji, O. T.
Liyanage, D. J. D. Wilson, J. L. Dutton, C. D. Martin, Chem. Eur. J. 2019,
25, 1581–1587.
[7] a) H. Braunschweig, C. Hçrl, L. Mailꢄnder, K. Radacki, J. Wahler, Chem.
Eur. J. 2014, 20, 9858–9861; b) S. A. Couchman, T. K. Thompson, D. J. D.
Wilson, J. L. Dutton, C. D. Martin, Chem. Commun. 2014, 50, 11724–
11726.
Computational study
[8] a) A. R. Jupp, D. W. Stephan, Trends Chemistry 2019, 1, 35–48; b) M. A.
Lꢅgarꢅ, G. Bꢅlanger-Chabot, R. D. Dewhurst, E. Welz, I. Krummenacher,
B. Engels, H. Braunschweig, Science 2018, 359, 896–900; c) M.-A.
Lꢅgarꢅ, M.-A. Courtemanche, E. Rochette, F. G. Fontaine, Science 2015,
349, 513–516; d) D. W. Stephan, G. Erker, Angew. Chem. Int. Ed. 2015,
54, 6400–6441; Angew. Chem. 2015, 127, 6498–6541; e) A. J. Warner, A.
Churn, J. S. McGough, M. J. Ingleson, Angew. Chem. Int. Ed. 2017, 56,
354–358; Angew. Chem. 2017, 129, 360–364; f) K. Yuan, S. Wang, Org.
Lett. 2017, 19, 1462–1465; g) S. Kirschner, S.-S. Bao, M. K. Fengel, M.
Bolte, H.-W. Lerner, M. Wagner, Org. Biomol. Chem. 2019, 17, 5060–
5065; h) E. C. Neeve, S. J. Geier, I. A. I. Mkhalid, S. A. Westcott, T. B.
Marder, Chem. Rev. 2016, 116, 9091–9161.
DFT and TD-DFT calculations were performed using the Gaussian
09 suite of programs[12] and computational facilities provided by
Compute Canada. Geometry optimizations and vertical excitations
of all compounds were obtained at the B3LYP/6-31g(d) level of
theory, using crystal structural data as the initial geometrical pa-
rameters.
Acknowledgements
[9] S. K. Mellerup, C. Li, T. Peng, S. Wang, Angew. Chem. Int. Ed. 2017, 56,
6093–6097; Angew. Chem. 2017, 129, 6189–6193.
We thank the Natural Sciences and Engineering Research
Council of Canada for financial support.
[10] a) H. Braunschweig, C. Claes, A. Damme, A. Deißenberger, R. D. Dew-
hurst, C. Hçrl, T. Kramer, Chem. Commun. 2015, 51, 1627–1630; b) P. Bis-
singer, H. Braunschweig, K. Kraft, T. Kupfer, Angew. Chem. Int. Ed. 2011,
50, 4704–4707; Angew. Chem. 2011, 123, 4801–4804; c) C. Balzereit, C.
Kybart, H. J. Winkler, W. Massa, A. Berndt, Angew. Chem. Int. Ed. Engl.
1994, 33, 1487–1489; Angew. Chem. 1994, 106, 1579–1581; d) S. W.
Denmark, K. Nishide, A. M. Faucher, J. Am. Chem. Soc. 1991, 113, 6675–
6676; e) T. R. McFadden, C. Fang, S. J. Geib, E. Merling, P. Liu, D. P.
Curran, J. Am. Chem. Soc. 2017, 139, 1726–1729.
[11] a) Z. Goldschmidt, S. Antebi, H. E. Gottlieb, D. Cohen, U. Shmueli, Z.
Stein, J. Organomet. Chem. 1985, 282, 369–381; b) F. K. Brown, K. N.
Houk, D. J. Burnell, Z. Valenta, J. Org. Chem. 1987, 52, 3050–3059; c) E.
Vogel, H. Gꢃnther, Angew. Chem. Int. Ed. Engl. 1967, 6, 385–401; Angew.
Chem. 1967, 79, 429–446.
[12] Gaussian 09, Revision A.02, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E.
Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, G. A. Pe-
tersson, H. Nakatsuji, X. Li, M. Caricato, A. Marenich, J. Bloino, B. G. Jane-
sko, R. Gomperts, B. Mennucci, H. P. Hratchian, J. V. Ortiz, A. F. Izmaylov,
J. L. Sonnenberg, D. Williams-Young, F. Ding, F. Lipparini, F. Egidi, J.
Goings, B. Peng, A. Petrone, T. Henderson, D. Ranasinghe, V. G. Zakrzew-
ski, J. Gao, N. Rega, G. Zheng, W. Liang, M. Hada, M. Ehara, K. Toyota, R.
Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H.
Nakai, T. Vreven, K. Throssell, J. A. Montgomery, Jr., J. E. Peralta, F.
Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov,
T. Keith, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C.
Burant, S. S. Iyengar, J. Tomasi, M. Cossi, J. M. Millam, M. Klene, C.
Adamo, R. Cammi, J. W. Ochterski, R. L. Martin, K. Morokuma, O. Farkas,
J. B. Foresman, D. J. Fox, Gaussian, Inc., Wallingford CT, 2016.
Conflict of interest
The authors declare no conflict of interest.
Keywords: arene ligands · boriranes · dearomatization · Diels–
Alder addition · photoisomerization · synthetic methods
[1] a) S. P. Roche, J. A. Porco, Angew. Chem. Int. Ed. 2011, 50, 4068–4093;
Angew. Chem. 2011, 123, 4154–4179; b) W. C. Wertjes, E. H. Southgate,
D. Sarlah, Chem. Soc. Rev. 2018, 47, 7996–8017; c) F. C. Pigge, in Arene
Chemistry (Ed.: J. Mortier), Wiley, Hoboken, 2015, pp. 399–423.
[2] a) F. Fringuelli, A. Taticchi, The Diels–Alder Reaction: Selected Practical
Methods, Wiley, Chichester, 2002; b) K. C. Nicolaou, S. A. Snyder, T. Mon-
tagnon, G. Vassilikogiannakis, Angew. Chem. Int. Ed. 2002, 41, 1668–
1698; Angew. Chem. 2002, 114, 1742–1773; c) C. X. Zhuo, W. Zhang,
S. L. You, Angew. Chem. Int. Ed. 2012, 51, 12662–12686; Angew. Chem.
2012, 124, 12834–12858.
[3] a) M. F. Semmelhack, A. Chlenov, Transition Metal Arene p-Complexes in
Organic Synthesis and Catalysis, Springer, Heidelberg, 2004; b) A. R.
Pape, K. P. Kaliappan, E. P. Kꢃndig, Chem. Rev. 2000, 100, 2917–2940;
c) B. K. Liebov, W. D. Harman, Chem. Rev. 2017, 117, 13721–13755;
d) M. S. Bailey, B. J. Brisdon, D. W. Brown, K. M. Stark, Tetrahedron Lett.
1983, 24, 3037–3040; e) M. D. Chordia, P. L. Smith, S. H. Meiere, M.
Sabat, W. D. Harman, J. Am. Chem. Soc. 2001, 123, 10756–10757;
f) W. D. Harman, Chem. Rev. 1997, 97, 1953–1978.
[13] SHELXTL, version 6.14; Bruker AXS, Madison, WI, 2000–2003.
[4] a) Y. Inagaki, M. Nakamoto, A. Sekiguchi, Nat. Commun. 2014, 5, 3018;
b) X. Li, S. J. Danishefsky, J. Am. Chem. Soc. 2010, 132, 11004–11005;
c) F. Fringuelli, O. Piermatti, F. Pizzo, L. Vaccaro, Eur. J. Org. Chem. 2001,
439–455.
Manuscript received: August 2, 2019
Accepted manuscript online: September 5, 2019
Version of record online: && &&, 0000
Chem. Eur. J. 2019, 25, 1 – 8
www.chemeurj.org
7
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
FULL PAPER
&
B-Mesityl Dearomatization
B. Deng, X. Wang, S. Wang*
&& – &&
Dearomatizing and Derivatizing a
Mesityl Group on Boron by One-Pot
Photoisomerization and [4+2] Diels–
Alder Addition
Dearomatization of boron-bound
dienophiles leads to clean conversion of
the aryl ring to a 3-boratricy-
clo[3.2.2.02,4]nona-6,8-diene (see graph-
ic).
arenes: One-pot photochemical activa-
tion of an aryl ring on boron and subse-
quent [4+2] Diels–Alder addition with
&
&
Chem. Eur. J. 2019, 25, 1 – 8
www.chemeurj.org
8
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!