Communication
ChemComm
might release hydroxide ions more gradually as compared to
potassium hydroxide due to its lower solubility.26
Having optimized the reaction conditions, we next investi-
gated whether we could verify the central hypothesis of this
work, namely that the two unsymmetrically substituted
reagents can undergo a site-selective cross-coupling reaction.
In an attempt to observe a single regioisomer, we used
unsymmetrically substituted biaryls and performed an annula-
tion with alkyl 1,2-bisboronic pinacol esters (Scheme 2). Excitingly,
introduction of a Me- or F-substituent on the biaryl substrate
indeed led to a single regioisomer (3b–3i). Additionally, an
electron-poor mono CF3-substituted biaryl and an electron-rich
methoxy-substituted biaryl also gave the corresponding products
with high regioselectivity (3j, 3k). We rationalize this selectivity with
the different reaction rates of the reactive centers in both bifunc-
tional reagents (11 vs. 21 C–Bpin, C–Br vs. C–Cl). Initially, a C–C
bond between the primary C–Bpin and Ar–Br is formed, followed
by the intramolecular cyclization between the secondary C–Bpin
and Ar–Cl.
Biaryls with a symmetric substitution pattern (3l, 3m)
were also well tolerated. An attempt to cyclize 2,20-dibromo-1,
10-binaphtalene did not afford any annulated product. The ability
to rotate freely around the biaryl bond might therefore be crucial
to the efficiency of the second coupling step. Furthermore, an
extension to heterocyclic biaryls proved unsuccessful.
Turning our focus to the alkyl 1,2-bisboronic pinacol ester
scope, sterically more encumbered substituents showed good
compatibility with the annulation reaction (3n–s). A dinucleophile
containing a primary chloride (3s) and a eugenol-derivative (3t)
were efficiently converted to the annulated product. Unfortu-
nately, nitrile, ester or ketone functionalities were not tolerated.
Similarly, double secondary, ethane- or methane-derived dinu-
cleophiles did not yield cyclized products. Finally, the model
reaction could be scaled up to a 5 mmol scale yielding 57%
(0.76 g) of the annulated product 3a.
Intrigued by the high efficiency and selectivity of the trans-
formation, we monitored the model reaction more closely. The
chlorinated reaction intermediate 4 could be isolated when
stopping the reaction after three hours (Scheme 3A). In con-
trast, the corresponding brominated intermediate was not
observable. NMR studies showed that compound 4 exists as a
mixture of two atropoisomers that are interconvertible at
elevated temperature (see ESI,† for details). Using the inter-
mediate 4 as a starting material in our reaction, the annulated
product was observed in high yield (Scheme 3A). Interestingly,
using potassium hydroxide as a base gave a comparable con-
version (77%, see ESI†) to the cyclized product in this single-
step, intramolecular coupling reaction. The intramolecular
nature of the coupling between the secondary boronic pinacol
ester and the aryl chloride may be key to facilitate the challenging
secondary cross-coupling. To substantiate this hypothesis, an
intermolecular control reaction was performed. As expected, both
reactions using the corresponding homobenzylic boronic pinacol
ester and either bromobenzene or chlorobenzene did not yield
any of the cross-coupled product (see ESI,† for details). Usually,
the Pd-catalyzed Suzuki–Miyaura cross-coupling of aliphatic,
Scheme 1 Optimization of the annulation reaction using alkyl 1,2-bisboronic
pinacol ester 1a and biaryls 2 to yield 9,10-dihydrophananthrene 3a.
the octyl-1,2-bisboronic pinacol ester 1a only gave trace amounts
of the cyclized product, further highlighting the challenges in
developing such a process. A sequence of experiments allowed us
to find a promising Pd2(dba)3/IPr HCl (1 : 4) catalytic system in
THF which gave 35% of 9-hexyl-9,10-dihydrophenanthrene 3a in
the presence of KOH (Scheme 1 and Table 1, entry 1). We next
explored the reactivity of other electrophilic coupling partners in
the cyclization reaction. 2,20-Dichlorobiphenyl 2c gave a cleaner
reaction and a similar amount of annulated product compared to
the dibromide 2b (entry 1 and 2). In contrast, the diiodide 2d gave
a diminished product yield (entry 3). 2,20-Biphenol derived ditriflate
2e and dimesylate 2f were unreactive and starting materials were
recovered (entry 4 and 5). Combining these preliminary observa-
tions, we envisioned that optimizing conditions for 2-bromo-20-
chloro biphenyl 2a may set the stage for the subsequent develop-
ment of a fully site-selective annulation process.20,21
Excitingly, using 2-bromo-20-chloro biphenyl 2a, the yield
increased to 48% (entry 6). Further optimization showed that
the use of barium hydroxide as a base under more dilute
reaction conditions was crucial to increase the yield of the
double Suzuki–Miyaura to 76% (entry 12). The availability of
hydroxides plays a pivotal role in the efficiency of the Suzuki–
Miyaura cross-coupling as discussed by Amatore and Jutand,22
as well as Hartwig.23 Based on these studies, we postulate that
changing the counterion of the hydroxide-source may increase
the activity of the Pd-catalyst during transmetalation due to
a weaker binding affinity of the barium-compared to the
potassium-counterion.24,25 Furthermore, barium hydroxide
Table 1 Optimization of reaction conditionsa
Entry
Base (x equiv.)
2a–f X1, X2
Solvent/H2O (y : z)
3ab [%]
1
2
3
4
5
6
7
8
KOH (3)
KOH (3)
KOH (3)
KOH (3)
KOH (3)
KOH (3)
Ba(OH)2 (2)
Ba(OH)2 (2)
Ba(OH)2 (2)
Ba(OH)2 (2)
—
Br, Br
Cl, Cl
I, I
THF (10 : 1)
THF (10 : 1)
THF (10 : 1)
THF (10 : 1)
THF (10 : 1)
THF (10 : 1)
THF (1 : 1)
Dioxane (1 : 1)
H2O
35
33
6
4
0
48
70
31
53
76
0
OTf, OTf
OMs, OMs
Br, Cl
Br, Cl
Br, Cl
Br, Cl
Br, Cl
Br, Cl
Br, Cl
9
10c
11c
12cd
THF (4 : 1)
THF (4 : 1)
THF (4 : 1)
Ba(OH)2 (2)
0
a
Reaction conditions: 1a (0.1 mmol),
2 (0.1 mmol), Pd2(dba)3
(0.005 mmol), IPr HCl (0.02 mmol), base (0.2–0.3 mmol), THF/H2O
b
(0.1 mol Lꢀ1), 100 1C, 24 h. GC yield using n-tetradecane as internal
c
d
standard. THF/H2O (0.04 mol Lꢀ1). Without Pd2(dba)3 and IPr HCl.
This journal is © The Royal Society of Chemistry 2021
3910 | Chem. Commun., 2021, 57, 3909–3912