Boric acid-catalyzed direct condensation of carboxylic acids with benzene-1,2-diamine into benzimidazoles

Article abstract of DOI:10.1002/hlca.201100064

The applicability of boric acid catalysis for the direct condensation of carboxylic acids with benzene-1,2-diamine to give 2-substituted benzimidazoles was investigated. It was found that catalytic amounts (5-10 mol-%) of boric acid efficiently promote the cyclocondensation of aliphatic carboxylic acids in refluxing toluene. In addition, the relatively neutral conditions allow the use of acid-sensitive substrates and give rise to specific transformations and selectivities that are not observed with some classical methods. Benzoic acids were found to be less reactive than aliphatic acids and thus require refluxing xylene for better efficiency. Phenylboronic acid was found to be inactive as a catalyst due to its rapid consumption by condensation with benzene-1,2-diamine to give a 2-phenylbenzodiazaborole. Copyright

Full text of DOI:10.1002/hlca.201100064

1860
Helvetica Chimica Acta – Vol. 94 (2011)
Boric Acid-Catalyzed Direct Condensation of Carboxylic Acids with
Benzene-1,2-diamine into Benzimidazoles
by Nenad Marasˇ and Marijan Kocˇevar*
ˇ
ˇ
University of Ljubljana, Faculty of Chemistry and Chemical Technology, Askerceva 5, SI-1000 Ljubljana,
Slovenia
(phone: þ 386-1-2419230; fax: þ 386-1-2419220; e-mail: marijan.kocevar@fkkt.uni-lj.si)
The applicability of boric acid catalysis for the direct condensation of carboxylic acids with benzene-
1,2-diamine to give 2-substituted benzimidazoles was investigated. It was found that catalytic amounts
(5 – 10 mol-%) of boric acid efficiently promote the cyclocondensation of aliphatic carboxylic acids in
refluxing toluene. In addition, the relatively neutral conditions allow the use of acid-sensitive substrates
and give rise to specific transformations and selectivities that are not observed with some classical
methods. Benzoic acids were found to be less reactive than aliphatic acids and thus require refluxing
xylene for better efficiency. Phenylboronic acid was found to be inactive as a catalyst due to its rapid
consumption by condensation with benzene-1,2-diamine to give a 2-phenylbenzodiazaborole.
Introduction. – The boric acid (B(OH)3)-catalyzed condensation of carboxylic
acids with amines and anilines is an elegant solution for the synthesis of amides; it
avoids the use of expensive coupling reagents or the intermediacy of acid chlorides. The
exceptional chemoselectivity and the applicability of this condensation has been
demonstrated on a wide spectrum of substrates (aliphatic or aromatic carboxylic acids,
primary and secondary amines, anilines, etc. [1]; B(OH)3-catalyzed esterifications were
also reported [2]). In most cases, 5 mol-% of B(OH)3 and the azeotropic removal of
H2O, employing toluene as the carrier, is efficient, although the less nucleophilic
anilines can require a larger amount of the catalyst. Arylboronic acids have been found
to generally catalyze this and related types of condensations even more efficiently [3].
Their catalytic activity was thought to be based on the carboxylic acid-activation
mechanism relying on the intermediate formation of a mixed carboxylicꢀboric acid
anhydride. The ability of boric acids to activate carboxylic acids, therefore, also seems
ideal for their cyclocondensation reactions, and it was thus the first option that we
considered in an attempt to prepare certain benzimidazoles directly from acid-sensitive
carboxylic acids.
The methods for the preparation of 2-substituted benzimidazoles based on direct
cyclocondensations of carboxylic acids with benzene-1,2-diamines are otherwise well
known [4], but nearly all of them require harsh conditions, consisting of either very high
temperatures or highly acidic media. Many carboxylic acids undergo condensation with
benzene-1,2-diamine at temperatures of ca. 2008, although some, like HCOOH, for
example, react at steam-bath temperatures [5]. Probably the most familiar method is
that developed by Phillips, which consists of refluxing the reactants in 4m aqueous HCl
[6]. In cases where this method fails, as is commonly the case when employing benzoic
ꢀ 2011 Verlag Helvetica Chimica Acta AG, Zꢁrich

Helvetica Chimica Acta – Vol. 94 (2011)
1861
acids, heating the reactants with polyphosphoric acid or other strongly dehydrating acid
reagents can give good results [7]. Even B(OH)3 at high temperatures (180 – 2008) has
been employed as a dehydrating medium [8]. But a literature search provided us with
only one report in which the use of B(OH)3 as an additive (catalyst) is briefly
mentioned, i.e., the condensation of three cholic acid derivatives with arene-1,2-
diamines in refluxing xylenes [9]. Long reaction times and ca. 70 – 80 mol-% of B(OH)3
were used to give products in moderate-to-excellent yields. On the other side, a related
cyclocondensation of carboxylic acids with 2-aminophenols to give benzoxazoles was
much more thoroughly studied earlier [10], but the use of 1 equiv. of B(OH)3 in
refluxing xylenes was proposed for the best results.
Since there was so little related information in the literature, we decided to study
the B(OH)3-catalyzed synthesis of 2-substituted benzimidazoles in order to better
understand the applicability of this method. This was deemed to be interesting,
particularly because of the importance of benzimidazoles in medicinal chemistry [4e]
and as ligands for coordination compounds [11].
Results and Discussion. – As the model reaction, we selected the condensation of
benzene-1,2-diamine (1a) and 2-phenylacetic acid (2a) to 2-benzylbenzimidazole
(4aa). The product 4aa is otherwise an active pharmaceutical ingredient due to its
vasodilator and spasmolytic activity, industrially produced under the name of Dibazol
or Bendazol. A stoichiometric mixture of 1a and 2a with 5 mol-% of B(OH)3 was
heated in toluene under reflux conditions, and a DeanꢀStark trap was used for the
azeotropic removal of H2O. Such a treatment led to a nearly quantitative conversion to
4aa after 20 h, and the product was isolated in a 90% yield after recrystallization
(Table 1, Entry 1). No N-monoacyl intermediate 3 could be observed when analyzing
the mixture with NMR spectroscopy at an intermediate reaction time. This is consistent
with the published observation that 2-aminoacetanilides are indeed intermediates in
the direct condensation of benzene-1,2-diamines with carboxylic acids, but that these
intermediates also readily cyclize to benzimidazoles upon heating in xylene, even in the
absence of a catalyst [12]. It has been reported that this cyclization occurs far below the
boiling point of xylene or slowly by heating neat at 50 – 608. Furthermore, acids greatly
enhance the rate of the reaction [13]. This suggests that the rate-limiting step in the
B(OH)3-catalyzed formation of benzimidazole 4aa is likewise the N-acylation to form
3aa rather than the subsequent cyclization.
In the absence of the catalyst, in boiling toluene, the reaction also proceeded, but
very slowly. After 12 h, ca. 35% of the theoretical amount of H2O distilled, and the
NMR spectroscopic analysis confirmed that the reaction conversion corresponds to the
same value. Upon further heating to reflux, the rate of sequestration of H2O in the
DeanꢀStark trap decreased to unpractical levels. Under the same conditions, but with
5 mol-% of B(OH)3 as a catalyst, it takes less than 2 h to collect 50% of the theoretical
amount of H2O. The catalytic activity of 5 mol-% of B(OH)3 in this case is rather
obvious, though not exceptionally high. Additionally, doubling the amount of B(OH)3
to 10 mol-% only increases the reaction rate by ca. 30% (i.e., it is completed in 15 h
instead of 20 h; Table 1, Entry 3).
Using xylene as the solvent increased the reflux temperature (b.p. 1408) and thus
considerably shortened the reaction time. The reaction was complete after just 8 h of

1862
Helvetica Chimica Acta – Vol. 94 (2011)
Table 1. Yields from the Condensations of 2-Phenylacetic Acid (2a), Benzoic Acid (2b), 4-Nitrobenzoic
Acid (2c), and 4-Methoxybenzoic Acid (2d) with Benzene-1,2-diamine (1a) under different reaction
conditions
Entry Diamine Carboxylic acid Solventa) Amount of B(OH)3 Time [h] Product Yield [%]
[mol-%]
1
2
3
4
5
6
7
8
9
1a
1a
1a
1a
1a
1a
1a
1a
1b
2a
2a
2a
2a
2b
2c
2c
2d
2a
Toluene
Toluene
Toluene
Xylene
Toluene
Toluene
Xylene
Xylene
Toluene
5
5
10
5
10
10
10
10
10
20
15
15
8
72
4aa
4aa
4aa
4aa
4ab
4ac
4ac
4ad
4ba
90
83
89
90
71
40
62
60
61
120b)
24
72
120b)
a) Reactions were conducted in boiling solvents using the DeanꢀStark trap for the removal of H2O.
b) Reaction was discontinued and worked up before reaching complete conversion.
reflux (Table 1, Entry 4). Deciding between longer reaction times and higher reaction
temperatures, we preferred to continue using toluene, as our goal was to develop a
method that is efficient under milder conditions so as to be suitable for sensitive
substrates.
The condensation of 1a with benzoic acid (2b) was very slow. It took 3 d in refluxing
toluene with 10 mol-% of B(OH)3 to obtain the theoretical amount of H2O. The
product, 2-phenylbenzimidazole (4ab), was obtained in a 71% yield after the
recrystallization (Table 1, Entry 5). In an attempt with 4-nitrobenzoic acid (2c) under
the same conditions, the reaction with 1a proceeded even more slowly and was still
incomplete after 5 d, when the mixture was worked up to give a 40% yield of 2-(4-
nitrophenyl)benzimidazole (4ac; Table 1, Entry 6). The lower reactivity of arenecar-
boxylic acids has also been noted in the application of the Phillips method [4c][4d].
Clearly, with these aromatic acids, refluxing xylene should be used as the solvent in
order to increase the reaction rate. The condensation of 4-nitrobenzoic acid (2c) in
xylene proceeded considerably faster, and a yield of 62% was obtained after 24 h

Helvetica Chimica Acta – Vol. 94 (2011)
1863
(Table 1, Entry 7). 4-Methoxybenzoic acid (2d) was even less reactive and required 3 d
in refluxing xylene for complete conversion. The benzimidazole 4ad was obtained in a
60% yield after recrystallization. This indicated that the electronic properties of the
para-substituent on benzoic acids are not uniformly related to the reaction rate, and
that the reactivity in this condensation to 2-phenylbenzimidazoles is an interplay of
different factors.
To evaluate the influence of the nucleophilicity and basicity of the diamine reactant,
we performed a reaction with 4-nitrobenzene-1,2-diamine (1b) as a poorer nucleophile.
Its reaction with 2-phenylacetic acid (2a) was extremely slow, even in the presence of
10 mol-% of B(OH)3. After 5 d of reflux in toluene, there was still some unreacted
starting material present, although the accumulation of H2O in the DeanꢀStark trap
practically stopped (85% of the theoretical amount was collected). Nevertheless, the
product 4ba was isolated in a 61% yield after recrystallization (Table 1, Entry 9).
The cyclocondensation of 1a with undec-10-enoic acid (2e) gave a 71% yield of 4ae
when using 10 mol-% of B(OH)3 after 36 h (Scheme 1). 2-(Phenylsulfanyl)acetic acid
(2f) reacted more rapidly and gave benzimidazole 4af in a 92% yield after 10 h of
reflux. On the other hand, even though the theoretical amount of H2O accumulated in
the DeanꢀStark trap, the (ꢀ)-(R)-phenylglycolic acid (2g) only gave a 38% yield of 4ag
after recrystallization. It appears that a lot of the phenylglycolic acid was consumed in
side reactions, yielding resinous materials. Furthermore, the product 4ag was racemic as
indicated by its optical inactivity. Similarly, the reaction with cinnamic acid 2h gave 4ah
with a yield of only 36%, even after a long reaction time. Interestingly, in both cases the
unidentified resinous side products did not disturb the purification, and the
benzimidazoles could be obtained in a pure form after a single recrystallization.
Scheme 1. Synthesis of Benzimidazoles from Various Carboxylic Acids and Benzene-1,2-diamine (1a)

1864
Helvetica Chimica Acta – Vol. 94 (2011)
The reaction with a carboxylic acid 2i containing the a-pinene moiety, which is
otherwise known to be susceptible to the WagnerꢀMeerwein type of rearrangements in
acidic media [14], also proceeded readily and without side reactions (Scheme 2). The
product 4ai was obtained in an 80% yield after 20 h reaction time. An attempt with a
more classical approach by refluxing 1a and 2i in 4m aq. HCl with 1,4-dioxane as a
cosolvent was not successful and led to a mixture of side products derived from 2i. The
carboxylic acid 2i used in these reactions was prepared via the ene reaction between
methyl acrylate and b-pinene (5; adapted from [15]), followed by hydrolysis.
Scheme 2. Application of the Method to an Acid Sensitive Substrate
Reactions with dicarboxylic acids were also considered. To determine the
discrimination between a reaction with one or both COOH groups, we condensed 1a
with succinic acid (2j) in a 1:1 molar ratio (Scheme 3). This gave a mixture of 3-
(benzimidazol-2-yl)propanoic acid (4aj) and 1,2-di(benzimidazol-2-yl)ethane (4ajj),
which were easily separated and isolated in 29 and 39% yields, respectively (based on
1a as the limiting reagent). This represents a 60 :40 molar ratio for the isolated products
4aj/4ajj. Clearly, in this case the selectivity for 4aj is not high, regardless of the
stoichiometric ratio of starting materials.
Scheme 3. Condensations of Succinic Acid with Benzene-1,2-diamine (1a)
With malonic acid (2k) and 2 equiv. of 1a, 2-methylbenzimidazole (4ak) was
obtained as the major product and 1,5-benzodiazepine 6 as the minor one (Scheme 4).
Interestingly, the conventional condensation methods like refluxing in 4m aq. HCl are
claimed, in various articles, to give as products either 1,2-di(benzimidazol-2-yl)-
methane [16], or more frequently, the 1,5-benzodiazepine 6 [4a][4d][6][17]. We
observed no di(benzimidazol-2-yl)methane formation among the products.
Initially, it was not clear at which stage the decarboxylation leading to 4ak occurred,
although it is well-known that malonic acid undergoes spontaneous decarboxylation at
higher temperatures. At first, we considered the possibility that B(OH)3 might catalyze
the decarboxylation. This was checked by refluxing 2k with 10 mol-% of B(OH)3 in

Helvetica Chimica Acta – Vol. 94 (2011)
1865
Scheme 4. Condensation of Malonic Acid with Benzene-1,2-diamine (1a)
toluene for 24 h. After this time, 2k remained undissolved, and no AcOH could be
detected by NMR spectroscopy of the mixture, thus indicating that the decarboxylation
either occurs only in the presence of the basic 1a, or it occurs at some later reaction
stage. The first hypothesis that decarboxylation requires the presence of a base was
checked by refluxing a mixture of 2k in toluene with 2 equiv. of N,N-dimethylaniline as
a nonreactive substitute of 1a (in the absence of B(OH)3). This indeed led to complete
decarboxylation of 2k to give AcOH in less than 3 h of reflux. A literature search
revealed that huge rate enhancements of the malonic acid decarboxylation had already
been observed when heating its solutions in quinoline, pyridine, N,N-dimethylaniline,
aniline and N,N-dimethylnaphthalen-1-amine (rates descending in this same order)
[18]. Based on these findings, we concluded that 2-methylbenzimidazole (4ak) must
have been formed as the main product due to the rapid decarboxylation of 2k to
AcOH. The reaction was induced by the presence of 1a, which thus acted at the same
time as a catalytic N-base and the substrate for the subsequent condensation. The
minor product 6 (cf. Scheme 4) is, therefore, produced by the reaction of 1a with 2k
before all of it is consumed in the decarboxylation.
An interesting selectivity was obtained when condensing (þ)-camphoric acid (2l)
with 2 equiv. of 1a (Scheme 5). This gave benzimidazole (þ)-4al as the main product in
75% yield. Neither the dibenzimidazolyl product 4all nor the other possible isomer 4al’,
the latter potentially resulting from a primary nucleophilic attack at the more sterically
hindered COOH group, could be unambiguously identified among the minor side
1
products in the mixture. The H-NMR chemical shift of the CH H-atom in 4al is
Scheme 5. Condensation of Camphoric Acid with Benzene-1,2-diamine (1a)

1866
Helvetica Chimica Acta – Vol. 94 (2011)
consistent with the connection to benzimidazol-2-yl rather than to the CO group, which
is the case in the isomer 4al’. The products 4al and 4al’ were otherwise already reported
to form in 45 and 33% yield, respectively, in the condensation of 2l with 1a, by using a
zeolite catalyst under microwave heating [19].
N-Acylbenzimidazoles are generally prone to nucleophilic attack at the CO group,
just like many other azolides [20]. We, therefore, checked the resistance of 4al toward
amine nucleophiles in order to better understand why it does not react further with 1a
to give the di(benzimidazolyl) product 4all, in analogy to the product 4ajj in the
reaction with succinic acid (2j). Heating 4al in a tenfold excess of morpholine in the
closed vessel of a microwave reactor at temperatures up to 2008 resulted in no reaction
at all (for the equipment used and methodology in general, see [21]). It is thus evident
that the reactivity of 4al more closely resembles that of classical amides rather than
typical azolides. The lack of reactivity, which we attributed to the steric hindrance
around the CO group, thus explains the absence in the formation of a di(benzimida-
zolyl) product 4all.
The reaction of 4-chlorobutanoic acid (2m) gave 3-(benzimidazol-2-yl)propan-1-ol
hydrochloride (4am · HCl) in a 56% yield, and the condensation produced only 1 equiv.
of H2O in the DeanꢀStark trap (Scheme 6). The hydrochloride was then transformed to
the free base 4am by treatment with aq. NaOH. The formation of an alcohol is not
surprising, as the acid 2m readily cyclizes to g-butyrolactone (2m’) in the presence of
bases, and this lactone should give the alcohol 4am upon formal condensation with 1a
or its monohydrochloride (formed in situ from the eliminated HCl). Indeed, this was
demonstrated by repeating the reaction using 2m’ and the monohydrochloride of 1a.
The reaction gave 4am · HCl in 81% yield. However, refluxing 2m’ with the free base 1a
under the same conditions resulted in no reaction and no substantial formation of H2O.
It thus appears that the equivalent of HCl eliminated from 2m also plays a crucial role
in the reaction with the lactone. In addition, the cyclization of 2m to 2m’ under the
applied reaction conditions was confirmed by refluxing 2m and 1a in toluene in the
absence of B(OH)3. The conversion of 2m to 2m’, as observed by 1H-NMR
spectroscopy of the mixture, was found to be practically quantitative after 1 h.
Scheme 6. Condensation of Benzene-1,2-diamine (1a) with 4-Chlorobutanoic Acid (2m) or g-Butyr-
olactone (2m’) Leading to the Same Product
The workup was, in most cases, very simple and required no chromatographic
separation. In fact, most products precipitated from the mixture in their crude form

Helvetica Chimica Acta – Vol. 94 (2011)
1867
upon cooling, and were simply filtered and recrystallized to give a pure isolated
material.
Since arylboronic acids have commonly been applied as catalysts in a variety of
amidations and esterifications [3], we decided to check the catalytic activity of
phenylboronic acid (PhB(OH)2; 7) in the synthesis of benzimidazoles. Surprisingly, the
use of 5 mol-% of 7 showed no observable catalytic activity in our model condensation
reaction of 1a with 2a in refluxing toluene, and the reaction conversion was the same as
in the control experiment with no catalyst. To explain the catalytic inactivity of
phenylboronic acid, we first performed its condensation with benzene-1,2-diamine (1a)
in refluxing toluene to obtain the benzodiazaborole 8 [22]. Compound 8 formed
rapidly under these conditions and crystallized directly from the mixture in a 92% yield
(Scheme 7). The analogous condensation of 1a with B(OH)3 was much slower and
required 4 d of reflux in order to collect less than 3 equiv. of H2O. In this reaction, H2O
formed rapidly at the beginning, and the NMR spectroscopic analysis of the mixture
indicated that 1a was being rapidly consumed, forming a mixture of products. However,
after 4 d the final condensation product 9 was isolated in 50% yield after
recrystallization from acetone. This trimeric diazaborole is otherwise better known
to form in the reactions of 1a with trialkyl borates, tris(dimethylamido)borane, or BCl3
in refluxing toluene or xylene [23], but a more thorough search showed that the
preparation from B(OH)3 was described in a patent [24].
Scheme 7. Condensation of Benzene-1,2-diamine (1a) with Phenylboronic Acid (7) or Boric Acid
We then investigated the reactivity of these diazaboroles toward 2-phenylacetic
acid (2a) and their potential catalytic activity in our model reaction. Refluxing
benzodiazaborole 8 with 2a without any catalyst resulted in no conversion to 4aa after
3 h of reaction time (Table 2, Entry 1). The analogous experiment using 5 mol-% of 8
as a catalyst gave approximately the same conversion to 4aa as did the control
experiments using 5 mol-% of PhB(OH)2 (7) or no catalyst (Table 2, Entries 2, 3, and 4,
resp.). Moreover, in the reaction between 1a and 2a using 5 mol-% of PhB(OH)2 (7),
the formation of 8 was observed. Under the same conditions, the comparative
experiment using 5 mol-% of B(OH)3, instead of 7, gave a 65% conversion to 4aa
(Table 2, Entry 5). In contrast to the PhB(OH)2-condensation product 8, the borazine 9
reacted with the 2-phenylacetic acid (2a) directly to give 4aa with a 22% conversion
after 3 h of reflux (Table 2, Entry 6). The low solubility of 9 in boiling toluene probably
has an influence on slowing down the rate of this reaction, but the low conversion could
also indicate that such a reaction pathway may not be part of the main catalytic cycle.
Compound 9 was also checked for potential catalytic activity, but using a 5/3 mol-% to

1868
Helvetica Chimica Acta – Vol. 94 (2011)
compensate for its trimeric structure. It did catalyze the model reaction (Table 2,
Entry 6), though with a slightly lower conversion (54%) when compared to the
B(OH)3-catalyzed reaction (65%). This observation also supported the possibility that
the catalytically most active boron species is not necessarily 9, but that this merely acted
as a precatalyst that was slowly transformed into a common catalytically active species.
Table 2. Conversions of 2-Phenylacetic Acid (2a) to 4aa after 3 h in Refluxing Toluene under Different
Reaction Conditions
Entry
Reactantsa)
Catalyst
Conversion [%]
1
2
3
4
5
6
7
8 þ 2a
–
–
11
10
10
65
22
54
1a þ 2a
1a þ 2a
1a þ 2a
1a þ 2a
9 þ 2a
5 mol-% of 8
5 mol-% of PhB(OH)2 (5)
–
5 mol-% of B(OH)3
–
5/3 mol-% of 9
1a þ 2a
a) A mixture of 2 mmol of each reactant with the specified catalyst were refluxed in 5 ml of toluene for
3 h; then, a sample of the mixture was evaporated and analyzed by 1H-NMR spectroscopy.
The catalytic inactivity of PhB(OH)2 (5) thus appears to be due to its rapid
consumption and the formation of the diazaborole 8, which also lacked catalytic
activity. This might be explained on the basis of the stability of the BꢀN bonds, which
are straightened by their partial double-bond character and the resonance-based
stabilization of the diazaborole system. It is conceivable that the compound 8
eventually forms ate complexes with nucleophiles [25], such as carboxylic acids or
carboxylate ions. Thus, the absence of any catalytic effect additionally confirms the
current opinion that such ate species cannot be considered to be active acylating
reagents. Most studies on the B(OH)3-catalyzed amidation indicate that the carboxylic
acid is most likely activated via the formation of various types of acyloxyboranes, the
formation of which would require an energetically unfavorable ring opening in the case
of the diazaborole 8. If so, the BꢀN bonds of the borazine ring in 9 must be much more
prone to ring opening through a nucleophilic attack. The reaction kinetics for the
hydrolyses of 8 were thoroughly studied [25]. This data could be useful as an indication
of the stability of 8 toward nucleophilic ring-opening reactions. However, for the
compound 9 only qualitative descriptions of its susceptibility toward hydrolysis and
methanolysis have been reported [14].
Since the results obtained during the synthesis of 8 and 9 indicated that the boric
acids are being rapidly consumed in condensations with benzene-1,2-diamine, we
propose that the main catalytic cycle involves diazaboroles as active species
(Scheme 8). The benzodiazaborole I has not been isolated, and it is reasonable to
expect it to be in a hydrolytic equilibrium with B(OH)3 and the trimeric 9 (similar to the
trimerizations of B(OH)3 to metaboric acid or arylboronic acids to boroxines [3g]). In
our proposal, the benzoborazole I coordinates with a carboxylic acid to give the
unstable ate complex II, which, by the elimination of H2O, decomposes to the
acyloxyborane III. Finally, the reaction of III with 1a regenerates I and forms the 2-

Helvetica Chimica Acta – Vol. 94 (2011)
1869
aminoanilide 3, which then spontaneously cyclizes to the corresponding 1H-benzimi-
dazole 4. An approach to III via a carboxylate anion reacting with I and subsequent
elimination of a HOꢀ anion might be another possibility. However, considering the
weak basicity of anilines and the lack of ion solvation, the concentration of carboxylate
anions must be extremely low in a nonpolar solvent like toluene [18]. Also, the HOꢀ
anion, in comparison with H2O, is less likely to eliminate in place of the carboxylate
anion from the intermediate ate complex without prior protonation, thus, in any case
making II or another protonated ate form a most likely intermediate in the ligand-
exchange process.
Scheme 8. The Proposed Main Catalytic Cycle for the Condensation of Benzene-1,2-diamines and
Carboxylic Acids
Interestingly, it was previously observed that, for the condensation of 4-phenyl-
butanoic acid with PhCH2NH2, various 1,3,2-dioxaborolan-2-ols are more effective
catalysts than B(OH)3 itself [3h]. An analogous catalytic cycle was proposed, and 1,3,2-
dioxaborolan-2-ols can be considered isosteric and isoelectronic to the 1,3,2-diazabor-
oles of type I.
One also has to consider the eventual possibility that benzimidazoles 4 might form
by the initial acyl group migration within III onto one of the N-atoms, followed by
further transformations toward 4. This alternative to the intermolecular acylation of 1a
with III is difficult to exclude, based on the current experimental data, particularly since
we found that the borazine 9 itself reacts with 2a. In this regard, it is known that acyclic
BꢀN bond-containing compounds like tris(dialkylamido)boranes react with carboxylic
acids to give amides, but it was also demonstrated that these reactions occur through
the initially formed acyloxyborane intermediates acylating the liberated amines [26].

1870
Helvetica Chimica Acta – Vol. 94 (2011)
Conclusions. – An evaluation of the B(OH)3-catalyzed cyclocondensation reaction
between carboxylic acids and benzene-1,2-diamine demonstrated that this method has
good practical applicability and was thus used to prepare a series of 2-substituted
benzimidazoles. Besides optimal atom economy, its specific advantage over the
otherwise classical use of strong dehydrating acids is also in the use of non-acidic
reaction conditions, allowing the application of acid-sensitive substrates (as in the case
of 4ai). These conditions gave rise to some specific transformations as well. It was thus
found that the cyclocondensation with malonic acid gives 2-methyl-1H-benzimidazole
(4ak) and 1,5-benzodiazepine-2,4-dione (9), whereas the condensation with camphoric
acid gives an excellent selectivity in the formation of the azolide 4al. The results from
the three benzoic acids applied indicate that they are less reactive and thus require the
use of xylene to achieve a higher reflux temperature for better conversions. Finally, we
also elucidated the nature of the catalytic cycle, demonstrating that B(OH)3 initially
reacts with benzene-1,2-diamine, forming catalytically active diazaboroles.
Experimental Part
General. All the reagents used were commercially available, with the exception of 2i, the preparation
of which is described below. The xylene used was a commercial mixture of isomers. Reactions were
monitored with TLC (SiO2). M.p.: uncorrected; Kofler micro hot stage. Optical rotations: Perkin-Elmer
241 MC Polarimeter; concentrations in g/100 ml. NMR Spectra: at 298; Bruker Avance DPX 300-MHz
spectrometer with TMS as the internal standard. The coupling constants JAX are given as an average
distance between spectral lines (in Hz), and these distances differ from each other for ca. 10%. MS: VG-
Analytical AutoSpec Q instrument. Elemental analyses (C, H, N): Perkin Elmer 2400 CHN Analyzer.
4-[(1R,5S)-6,6-Dimethylbicyclo[3.1.1]hept-2-en-2-yl]butanoic Acid (2i). To a stirred ice-cooled soln.
of (ꢀ)-(1S)-b-pinene (5; 20.43 g, 150 mmol) and methyl acrylate (13.49 g, 150 mmol) in CH2Cl2 (60 ml)
2,6-di(tert-butyl)-4-methylphenol (10 mg) was added as a polymerization inhibitor, followed by the
addition of anh. AlCl3 (1.0 g, 5 mol-%). The clear, yellowish soln. thus obtained was left standing for 48 h
at rt. Then, it was washed with H2O (80 ml), evaporated, H2O (50 ml) was added to the oily residue, and
this mixture was again evaporated to remove the unreacted starting materials and other volatiles with
H2O vapors. MeOH (100 ml) and NaOH (12 g, 300 mmol) were then added to the residue, and the
mixture was stirred on reflux for 3 h. Upon cooling, the mixture was diluted with H2O (500 ml), acidified
with conc. HCl(aq) to pH 3, and extracted with petroleum ether (PE; 300 ml). The PE extract was
washed with H2O (300 ml) and back-extracted into dil. aq. NH3 (1%, 500 ml). An emulsion can form
during extraction, but it breaks by mild heating. The aq. soln. was then washed with PE (150 ml),
acidified with aq. HCl to pH 3, and again extracted with PE (250 ml); the org. extract was dried (Na2SO4)
and evaporated in vacuo to give 2i (20.90 g, 67%). Slightly yellowish oil. [a]2D5 ¼ ꢀ25.2 (c ¼ 4.25, CH2Cl2).
IR (NaCl): 1708 (br.), 1433, 1413, 1382, 1365. 1H-NMR (300 MHz, CDCl3): 0.83 (s, 3 H); 1.13 (d, J ¼ 8.5,
1 H); 1.27 (s, 3 H); 1.62 – 1.76 (m, 3 H); 1.90 – 2.45 (m, 8 H); 5.21 (br. s, 1 H); 11.45 (br. s, 1 H). 13C-NMR
(75 MHz, CDCl3): 21.2; 22.2; 26.3; 31.3; 31.7; 33.7; 36.1; 37.9; 40.8; 45.7; 116.9; 147.2; 180.4. HR-ESI-MS:
209.1541 ([M þ H]þ, C13H21Oþ2 ; calc. 209.1542), 191, 141.
General Procedure for the Synthesis of Benzimidazoles. A mixture of benzene-1,2-diamine (1;
100 mmol), carboxylic acid 2 (100 or 50 mmol), and B(OH)3 (320 mg for 5 mol-%, or 640 mg for 10 mol-
%) in toluene or xylenes (100 ml) was stirred under reflux in a reaction flask equipped with a DeanꢀStark
trap for the specified time required for the approximately theoretical amount of H2O to accumulate in
the receiver (see Discussion). The mixture was then worked up as described below.
2-Benzyl-1H-benzimidazole (4aa). To the cooled heterogeneous mixture, PE (50 ml) was added.
The crude product was filtered and recrystallized from EtOH/H2O 2 :1 to give 4aa (18.80 g, 90%). Off-

Helvetica Chimica Acta – Vol. 94 (2011)
1871
white needles. M.p. 183 – 1868 ([27]: 1878). IR (KBr): 1536, 1401, 1456, 1426. 1H-NMR (300 MHz,
(D6)DMSO): 4.17 (s, 2 H); 7.05 – 7.17 (m, 2 H); 7.19 – 7.35 (m, 5 H); 7.37 – 7.58 (m, 2 H); 12.26 (br. s, 1 H).
2-Phenyl-1H-benzimidazole (4ab). To the cooled heterogeneous mixture, PE (50 ml) was added, and
the separated crude product was filtered off and recrystallized from EtOH/H2O 2 :1 to give 4ab (13.70 g,
71%). Off-white crystals. M.p. 2968 ([7]: 294.5 – 295.58). IR (KBr): 1477, 1463, 1443, 1410, 1374. 1H-NMR
(300 MHz, (D6)DMSO): 7.22 (m, 2 H); 7.45 – 7.73 (m, 5 H); 8.16 – 8.24 (m, 2 H); 12.91 (s, 1 H).
2-(4-Nitrophenyl)-1H-benzimidazole (4ac). The brown precipitate from the cooled mixture was
filtered off, washed with PE (50 ml), 1m aq. NaOH) (100 ml), H2O (2 ꢁ 100 ml), and MeCN/H2O (4 :1,
3 ꢁ 50 ml), and dried. This crude product was recrystallized from a minimal amount of DMF, washed
with MeOH (2 ꢁ 20 ml), and dried under vacuum to give 4ac (17.00 g, 62%). Brown crystalline powder.
M.p. 321 – 3238 ([28]: 321 – 3228). IR (KBr): 1636 (br.), 1606, 1515, 1435. 1H-NMR (300 MHz,
(D6)DMSO): 7.21 – 7.35 (m, 2 H), 7.55 – 7.80 (m, 2 H), 8.39 – 8.46 (m, 4 H), 13.28 (br. s, 1 H).
2-(4-Methoxyphenyl)-1H-benzimidazole (4ad). The brown precipitate from the cooled mixture was
filtered off, washed with PE (50 ml), 1m aq. NaOH (100 ml), and H2O (2 ꢁ 100 ml), and the crude
product was recrystallized from a mixture of EtOH/H2O to give 4ad (16.00 g, 61%). Brownish powder.
M.p. 231 – 2328 ([29]: 228.6 – 230.58). IR (KBr) 1612, 1502, 1476, 1452, 1437, 1254. 1H-NMR (300 MHz,
(D6)DMSO): 3.85 (s, 3 H); 7.12 (d, J ¼ 8.7, 2 H); 7.16 – 7.21 (m, 2 H); 7.57 (m, 2 H); 8.14 (d, J ¼ 8.7, 2 H);
12.74 (s, 1 H). 13C-NMR (75 MHz, (D6)DMSO): 55.3; 111.0; 114.3; 118.5; 121.7; 122.7; 128.0; 151.3;
160.6.
2-Benzyl-5-nitro-1H-benzimidazole (4ba). The precipitated crude product was filtered from the
cooled mixture and dissolved in warm CH2Cl2 (600 ml). The red insoluble material was removed by
filtration, and the filtrate was evaporated in vacuo. The remaining yellow solid was recrystallized from
EtOH/H2O to give 4ba (15.35 g, 61%). Yellowish powder. M.p. 184 – 1858 ([27]: 1848). IR (KBr): 1624,
1599, 1535, 1511, 1468, 1452, 1411, 1341. 1H-NMR (300 MHz, (D6)DMSO): 4.27 (s, 2 H); 7.22 – 7.38 (m,
5 H); 7.66 (d, J ¼ 8.9, 1 H); 8.08 (dd, J ¼ 2.2, 8.9, 1 H); 8.40 (d, J ¼ 2.2, 1 H); 12.95 (br. s, 1 H).
2-(Dec-9-en-1-yl)-1H-benzimidazole (4ae). The cooled homogenous mixture was washed with H2O
(300 ml), sat. aq. NaHCO3 (100 ml), and H2O (300 ml), and evaporated in vacuo to give a viscous oil,
which was crystallized from EtOH/H2O to give 4ae (15.10 g, 71%). Colorless needless. M.p. 106 – 107.58.
IR (KBr): 1641, 1621, 1591, 1541, 1452, 1422. 1H-NMR (300 MHz, CDCl3): 1.16 – 1.42 (m, 10 H); 1.80 –
1.91 (m, 2 H); 1.92 – 2.02 (m, 2 H); 2.94 (t, J ¼ 7.8, 2 H); 4.81 – 5.00 (m, 2 H); 5.69 – 5.83 (m, 1 H); 7.21,
7.54 (AA’XX’, JAX ¼ 3.0, 4 H); 9.85 (br. s, 1 H). 13C-NMR (75 MHz, CDCl3): 28.7; 29.0; 29.2; 29.4; 29.5;
29.5; 29.6; 33.9; 114.3; 114.7; 122.2; 138.8; 139.3; 156.0. HR-ESI-MS: 257.2030 ([M þ H]þ, C17H25N2þ ;
calc. 257.2018). Anal. calc. for C17H24N2 (256.38): C 79.64, H 9.44, N 10.93; found: C 79.92, H 9.58, N
10.81.
2-[(Phenylsulfanyl)methyl]-1H-benzimidazole (4af). To the cooled heterogenous mixture, PE
(50 ml) was added. The crude product was then filtered and recrystallized from EtOH/H2O 2 :1 to give
4af (22.0 g, 92%). Off-white needles. M.p. 131 – 1338 ([30]: 135 – 1368). IR (KBr): 1638, 1618, 1587, 1537,
1483, 1458, 1435. 1H-NMR (300 MHz, CDCl3): 4.39 (s, 2 H); 7.11 – 7.32 (m, 7 H); 7.51 – 7.54 (m, 2 H); 8.80
(br. s, 1 H).
(1H-Benzimidazol-2-yl)(phenyl)methanol (4ag). To the cooled heterogenous mixture, PE (50 ml)
was added. The crude product was then filtered and recrystallized from EtOH to give 4ag (8.50 g, 38%).
White crystalline powder. M.p. 206 – 2098 ([6]: 202 – 2038). [a]2D5 ¼ 0.0 (c ¼ 5.00, MeOH). IR (KBr):
1623br, 1455, 1435, 1292, 1276. 1H-NMR (300 MHz, (D6)DMSO): 5.92 (d, J ¼ 4.1, 1 H); 6.49 (d, J ¼ 4.1,
1 H); 7.08 – 7.16 (m, 2 H); 7.22 – 7.28 (m, 1 H); 7.30 – 7.37 (m, 2 H); 7.39 – 7.56 (m, 4 H); 12.35 (s, 1 H).
2-[(E)-2-Phenylethenyl]-1H-benzimidazole (4ah). To the cooled heterogeneous mixture, PE
(50 ml) was added. The crude product was then filtered and recrystallized from EtOH/H2O 6 :1 to
give 4ah (8.00 g, 36%). Colorless crystals. M.p. 200 – 2018 ([31]: 200 – 2018). IR (KBr): 1643, 1588, 1524,
1493, 1422. 1H-NMR (300 MHz, (D6)DMSO): 7.16 – 7.21 (m, 2 H); 7.24 (d, J ¼ 16.6, 1 H); 7.33 – 7.39 (m,
1 H); 7.41 – 7.48 (m, 2 H); 7.47 – 7.65 (m, 2 H); 7.65 – 7.72 (m, 3 H); 12.63 (s, 1 H).
2-{3-[(1R,5S)-6,6-Dimethylbicyclo[3.1.1]hept-2-en-2-yl]propyl}-1H-benzimidazole (4ai). This reac-
tion was performed on a 50-mmol scale. The mixture was washed with H2O (2 ꢁ 100 ml) and evaporated
in vacuo. The remaining brown resin was dissolved in AcOEt (100 ml), and 4-nitrobenzoic acid (8.4 g,
50 mmol) was added and dissolved by heating. Upon cooling, a solid precipitated, and this was vacuum-

1872
Helvetica Chimica Acta – Vol. 94 (2011)
filtered and washed with cold AcOEt (2 ꢁ 15 ml) to give 4ai complex with 4-nitrobenzoic acid (17.95 g,
80%). Brownish powder. M.p. 118 – 1218. IR (KBr): 1635, 1578, 1527, 1462, 1379, 1347. Part of the product
was quantitatively converted to the free base 4ai as a brownish resin, by adding it to 1m aq. NaOH,
extracting with AcOEt, washing with H2O, and evaporating. [a]2D5 ¼ 23.0 (c ¼ 1.00, CH2Cl2). IR (KBr):
1622, 1592, 1543, 1455, 1440, 1424. 1H-NMR (300 MHz, CDCl3): 0.72 (s, 3 H); 1.01 (d, J ¼ 8.5, 1 H); 1.15
(s, 3 H); 1.86 (td, J ¼ 5.7, 1.2, 1 H); 1.89 – 2.14 (m, 7 H); 2.21 (dt, J ¼ 5.6, 8.5, 1 H); 2.99 (t, J ¼ 7.2, 2 H);
5.10 (m, 1 H); 7.20, 7.56 (AA’XX’, JAX ¼ 3.0, 4 H); 11.28 (s, 1 H). 13C-NMR (75 MHz, CDCl3): 21.3; 26.2;
26.4; 29.3; 31.4; 31.8; 36.7; 38.0; 41.0; 45.8; 114.8; 116.9; 122.2; 138.9; 147.4; 155.9. HR-ESI-MS: 281.2025
([M þ H]þ, C19H25Nþ2 ; calc. 281.2018).
2,2’-Ethane-1,2-diylbis[1H-benzimidazole] (4ajj). The mixture was poured into 1m aq. NaOH
(150 ml), the lumps of solids were triturated, and the mixture was left stirring for 3 h. The mixture was
then vacuum-filtered, and the solid material was washed with H2O (50 ml). The aq. layer of the filtrate
was separated and kept for the isolation of 4aj. The solid material was dried, triturated in boiling DMF
(80 ml), cooled, vacuum-filtered, and washed with acetone (3 ꢁ 20 ml) to give 4ajj (5.10 g, 39%). Slightly
bluish powder. M.p. > 3608 ([32]: 4158). IR (KBr): 1622, 1590, 1542, 1452, 1433, 1276, 1223, 1152, 1033.
HR-ESI-MS: 263.1304 ([M þ H]þ, C16H15N4þ ; calc. 263.1297), 132. Due to poor solubility, the NMR
spectra were recorded of the soluble trifluoroacetate salt, formed by adding 2 equiv. of CF3COOH to the
sample: 1H-NMR (300 MHz, (D6)DMSO): 3.81 (s, 4 H); 7.45, 7.77 (AA’XX’, JAX ¼ 3.0, 8 H); 12.20 (br. s,
4 H). DEPT135-NMR ((D6)DMSO): 24.9 (CH2); 114.9 (CH); 125.7 (CH).
3-(1H-Benzimidazol-2-yl)propanoic acid (4aj). The cooled aq. phase of the filtrate was acidified to a
pH of ca. 5 with conc. aq. HCl, the precipitate was vacuum-filtered, washed with H2O (2 ꢁ 30 ml), and
dried under vacuum to give 4aj (5.50 g, 29%). Colorless needles. M.p. 225 – 2278 ([6]: 2248). IR (KBr):
1639, 1592, 1557, 1489, 1464, 1433, 1407. 1H-NMR (300 MHz, (D6)DMSO) 2.81 (t, J ¼ 7.2, 2 H); 3.06 (t,
J ¼ 7.2, 2 H); 7.12, 7.47 (AA’XX’, JAX ¼ 3.0, 4 H); 8.00 – 14.00 (br. s, 2 H). DEPT135-NMR ((D6)DMSO/
CF3COOH): 24.7 (CH2); 32.2 (CH2); 115.3 (CH); 122.0 (CH).
1H-1,5-Benzodiazepine-2,4(3H,5H)-dione (6). To the cooled heterogeneous mixture, PE (50 ml)
was added. The crude product was filtered off, washed with PE (2 ꢁ 30 ml), suspended in warm EtOH
(100 ml), and the insoluble material was filtered and washed with EtOH (3 ꢁ 15 ml) to give 6 (1.10 g,
13%). Grey powder. M.p. 349 – 3528, subl. ([17]: > 3508, subl.). IR (KBr): 1705, 1672, 1599, 1503, 1429,
1401. 1H-NMR (300 MHz, (D6)DMSO): 3.18 (s, 2 H); 7.14 (AA’BB’, 4 H); 10.37 (s, 2 H). 13C-NMR
(75 MHz, (D6)DMSO): 45.0; 122.2; 124.8; 129.7; 165.8.
2-Methyl-1H-benzimidazole (4ak). The above ethanolic filtrate was evaporated in vacuo, and the
residue was recrystallized from H2O to give 4ak (4.50 g, 65%). Colorless needles. M.p. 173 – 1758 ([7]:
177.0 – 177.78). IR (KBr): 1625 (br.), 1555, 1451, 1418, 1387, 1272. 1H-NMR (300 MHz, (D6)DMSO): 2.48
(s, 3 H); 7.07 – 7.13 (m, 2 H); 7.42 – 7.46 (m, 2 H); 12.16 (br. s, 1 H).
(6R,9S)-6,7,8,9-Tetrahydro-9,12,12-trimethyl-10H-[6,9]methanoazepino[1,2-a]benzimidazol-10-one
(4al). The mixture was diluted with AcOEt (100 ml), washed with 1m aq. NaOH (100 ml), H2O (2 ꢁ
100 ml), and 1m aq. HCl (100 ml), evaporated in vacuo, and the residue was recrystallized from hexane
to give 4al (9.50 g, 75%). Gray crystalline powder. M.p. 92 – 958 ([19]: 86 – 888). [a]2D5 ¼ þ55.0 (c ¼ 1.00,
CH2Cl2). IR (KBr): 1724, 1616, 1562, 1475, 1453, 1440. 1H-NMR (300 MHz, CDCl3): 0.84 (s, 3 H); 1.13 (s,
3 H); 1.34 (s, 3 H); 1.75 – 1.84 (m, 1 H); 2.04 – 2.10 (m, 2 H); 2.37 – 2.50 (m, 1 H); 3.31 (d, J ¼ 6.4, 1 H);
7.32 – 7.37 (m, 2 H); 7.66 – 7.70 (m, 1 H); 8.15 – 8.19 (m, 1 H). 13C-NMR (75 MHz, CDCl3): 13.9; 19.8;
23.2; 27.5; 34.2; 49.2; 49.5; 55.5; 115.2; 119.6; 124.8; 125.1; 131.3; 143.4; 159.1; 175.1. HR-ESI-MS:
255.1489 ([M þ H]þ, C16H19N2Oþ; calc. 255.1497). Anal. calc. for C16H18N2O (254.33): C 75.56, H 7.13, N
11.01; found: C 75.82, H 7.27, N 11.27.
The aq. NaOH and aq. HCl washings gave no isolable products upon neutralization.
3-(1H-Benzimidazol-2-yl)propan-1-ol (4am). a) From 2m According to the General Procedure. The
solvent was decanted from the solidified mass, which was dissolved in a minimum amount of boiling
EtOH. The hydrochloride that precipitated after standing for a few days in the refrigerator was filtered
off and washed with acetone (4 ꢁ 30 ml) to give, after drying under vacuum, 4am · HCl (12.00 g, 56%).
White crystalline powder. M.p. 150 – 152.58. IR (KBr): 1628, 1569, 1459, 1408. 1H-NMR (300 MHz,
(D6)DMSO): 1.99 – 2.09 (m, 2 H); 3.20 (t, J ¼ 7.6, 2 H); 3.51 (t, J ¼ 6.0, 2 H); 7.51, 7.77 (AA’XX’, JAX ¼ 3.0,
4 H); 14.00 (br. s, 2 H).

Helvetica Chimica Acta – Vol. 94 (2011)
1873
The hydrochloride was dissolved in H2O (100 ml) and basified with 10m aq. NaOH. The precipitated
free base was vacuum-filtered, washed with H2O (3 ꢁ 40 ml), and dried to give 4am (9.10 g, 51%). White
powder. M.p. 159 – 1618 ([33] 158 – 1608). IR (KBr): 1634 (br.), 1592, 1534, 1452, 1440, 1420.
b) From the Lactone 2m’. Compound 1a (5.69 g, 50 mmol) was dissolved in 2m aq. HCl (55 ml,
110 mmol), the soln. was then evaporated in vacuo, and the residue was thoroughly dried under vacuum.
The crystalline dihydrochloride was combined with a second equiv. of 1a (5.69 g, 50 mmol), g-
butyrolactone (2m’; 8.69 g, 100 mmol), B(OH)3 (620 mg, 10 mol-%), and toluene (100 ml). The resulting
mixture was stirred on reflux for 12 h in a reaction flask equipped with a DeanꢀStark trap and then
worked up, as described above, to give 4aj · HCl (17.20 g, 81%).
Malonic Acid Decarboxylation. A mixture of malonic acid (1.05 g, 10 mmol), N,N-dimethylaniline
(2.42 g, 20 mmol), and toluene (10 ml) was stirred at reflux. At first, a biphasic liquid-liquid mixture
formed, which with time and gas evolution became homogenous. After 3 h, a 0.200-ml sample of the
mixture was taken for 1H-NMR analysis in CDCl3. Then, to this same sample was added authentic AcOH
as a standard, and the sample was again analyzed to confirm the NMR signals attributed to the AcOH.
The conversion of malonic acid was found to be complete.
2,3-Dihydro-2-phenyl-1H-1,3,2-benzodiazaborole (8). A mixture of 1a (2.84 g, 25 mmol) and
PhB(OH)2 (7; 3.21 g, 25 mmol) in toluene (25 ml) was stirred for 2 h on reflux in a reaction flask
equipped with a DeanꢀStark trap. Upon cooling, the precipitated product was filtered and washed with
PE (3 ꢁ 15 ml) to give 8 (4.45 g, 92%) as gray crystals. M.p. 207 – 2088 ([34]: 206 – 2078). IR (KBr): 1602,
1484, 1425, 1362, 1272, 1175. 1H-NMR (300 MHz, (D6)DMSO): 6.82, 7.06 (AA’XX’, JAX ¼ 3.0, 4 H);
7.38 – 7.44 (m, 3 H); 7.87 – 7.93 (m, 2 H); 9.10 (s, 2 H).
7H,14H,21H-Bis[1,3,2]benzodiazaborolo[1’,2’:3,4;1’’,2’’:5,6][1,3,5,2,4,6]triazatriborinino[1,2-a][1,3,2]-
benzodiazaborole (9). A mixture of 1a (12.00 g, 110 mmol) and B(OH)3 (6.25 g, 25 mmol) in toluene
(100 ml) was stirred for 4 d on reflux in a reaction flask equipped with a DeanꢀStark trap. Upon cooling,
the precipitated crude product was filtered and recrystallized from acetone to give 9 (5.80 g, 50%). White
needles. M.p. 404 – 4068, with crystals changing form at 310 – 3508 ([35]: 405 – 4078). IR (KBr): 1611,
1
1543, 1510, 1481, 1460, 1282, 1254, 1022, 739. H-NMR (300 MHz, (D6)DMSO): 7.01 – 7.09 (m, 6 H);
7.22 – 7.28 (m, 3 H); 7.79 – 7.82 (m, 3 H); 9.23 (s, 3 H). 13C-NMR (75 MHz, (D6)DMSO): 111.3; 113.0;
118.6; 120.6; 134.1; 139.5.
We thank the Ministry of Higher Education, Science and Technology of the Republic of Slovenia and
the Slovenian Research Agency for financial support (No. P1-0230-0103) as well as EN ! FIST Centre of
ˇ
Excellence for partial support with infrastructure. Dr. B. Kralj and Dr. D. Zigon (Center for Mass
ˇ
Spectroscopy, ꢂJozef Stefanꢃ Institute, Ljubljana, Slovenia) are gratefully acknowledged for the mass-
spectrometric recordings.
REFERENCES
[1] P. Tang, Org. Synth. 2005, 81, 262; R. K. Mylavarapu, G. C. M. Kondaiah, N. Kolla, R. Veeramalla, P.
Koilkonda, A. Bhattacharya, R. Bandichhor, Org. Process Res. Dev. 2007, 11, 1065.
[2] T. A. Houston, B. L. Wilkinson, J. T. Blanchfield, Org. Lett. 2004, 6, 679; T. Maki, K. Ishihara, H.
Yamamoto, Org. Lett. 2005, 7, 5047.
[3] a) K. Ishihara, S. Ohara, H. Yamamoto, J. Org. Chem. 1996, 61, 4196; b) K. Ishihara, S. Ohara, H.
Yamamoto, Macromolecules 2000, 33, 3511; c) K. Ishihara, S. Kondo, H. Yamamoto, Synlett 2001,
1371; d) K. Ishihara, S. Ohara, H. Yamamoto, Org. Synth. 2002, 79, 176; e) T. Maki, K. Ishihara, H.
Yamamoto, Synlett 2004, 1355; f) T. Maki, K. Ishihara, H. Yamamoto, Org. Lett. 2005, 7, 5043; g) K.
Arnold, B. Davies, R. L. Giles, C. Grosjean, G. E. Smith, A. Whiting, Adv. Synth. Catal. 2006, 348,
813; h) T. Maki, K. Ishihara, H. Yamamoto, Tetrahedron 2007, 63, 8645; i) K. Arnold, A. S. Batsanov,
´
B. Davies, A. Whiting, Green Chem. 2008, 10, 124; j) K. Arnold, B. Davies, D. Herault, A. Whiting,
Angew. Chem., Int. Ed. 2008, 47, 2673; k) R. M. Al-Zoubi, O. Marion, D. G. Hall, Angew. Chem., Int.
Ed. 2008, 47, 2876; l) I. Georgiou, G. Ilyashenko, A. Whiting, Acc. Chem. Res. 2009, 42, 756; m) H.
Charville, D. Jackson, G. Hodges, A. Whiting, Chem. Commun. 2010, 46, 1813; n) K. Ishihara, in
ꢂBoronic Acidsꢃ, Ed. D. G. Hall, Wiley-VCH, Weinheim, 2005, pp. 377 – 409.

1874
Helvetica Chimica Acta – Vol. 94 (2011)
[4] a) J. B. Wright, Chem. Rev. 1951, 48, 397; b) P. N. Preston, Chem. Rev. 1974, 74, 279; c) M. R.
Grimmett, ꢂImidazole and Benzimidazole Synthesisꢃ, Academic Press, London, 1997, pp. 72 – 94;
d) K. Hofmann, in ꢂChemistry of Heterocyclic Compounds, Vol. 6.ꢃ, Ed. A. Weissberger, Wiley,
London, 1953, pp. 247 – 324; e) M. Alamgir, D. S. C. Black, N. Kumar, Top. Heterocycl. Chem. 2007,
9, 87.
[5] E. C. Wagner, W. H. Millett, Org. Synth. 1939, 19, 12.
[6] M. A. Phillips, J. Chem. Soc. 1928, 2393.
[7] D. W. Hein, R. J. Alheim, J. J. Leavitt, J. Am. Chem. Soc. 1957, 79, 427.
[8] L. N. Pushkina, S. A. Mazalov, I. Y. Postovskii, Zh. Obshch. Khim. 1962, 32, 2624; Chem. Abstr. 1963,
58, 53219; P. C. Shekar, S. Seshadri, Indian J. Chem. B 1985, 24, 477; G. Qian, J. Yu, Y. Cui, Z. Wang,
X. Fan, M. Wang, Chinese patent application 101270089, 2008; Chem. Abstr. 2008, 149, 471479.
´
ˇ
[9] T. T. H. Nguyen, J. Urban, E. Klinotova, J. Sejbal, J. Protiva, P. Drasar, M. Protiva, Collect. Czech.
Chem. Commun. 1995, 60, 257.
[10] M. Terashima, M. Ishii, Y. Kanaoka, Synthesis 1982, 484.
[11] C. J. Matthews, T. A. Leese, W. Clegg, M. R. J. Elsegood, L. Horsburgh, J. C. Lockhart, Inorg. Chem.
´
´
1996, 35, 7563; F. Tellez, H. Lopez-Sandoval, S. E. Castillo-Blum, N. Barba-Behrens, Arkivoc 2008,
(v), 245.
[12] C. H. Roeder, A. R. Day, J. Org. Chem. 1941, 6, 25.
[13] K. J. Morgan, A. M. Turner, Tetrahedron 1969, 25, 915.
[14] J. R. Hanson, ꢂWagner-Meerwein rearrangementsꢃ, in ꢂComp. Org. Synth.ꢃ, Vol. 3, Eds. B. M. Trost, I.
Fleming, Pergamon, Oxford, 1991, pp. 705 – 719.
[15] B. B. Snider, J. Org. Chem. 1974, 39, 255.
[16] H.-C. Yao, M.-M. Li, G.-S. Yang, Z.-J. Li, Y. Zhu, Inorg. Chim. Acta 2007, 360, 3959; A. S. Alpan, S.
´
´
Zencir, I. Zupko, G. Coban, B. Rethy, H. S. Gunes, Z. Topcu, J. Enzyme Inhib. Med. Chem. 2009, 24,
844; A. Kumar, H. K. Sinha, S. K. Dogra, Can. J. Chem. 1989, 67, 1200.
[17] R. L. Shriner, P. G. Boermans, J. Am. Chem. Soc. 1944, 66, 1810; E. M. Essassi, A. Lamkadem, R.
Zniber, Bull. Soc. Chim. Belg. 1991, 100, 277; A. Keita, F. Lazrak, E. M. Essassi, I. C. Alaoui, Y. K.
Rodi, J. Bellan, M. Pierrot, Phosphorus, Sulfur Silicon Relat. Elem. 2003, 178, 1541.
[18] G. Fraenkel, R. L. Belford, P. E. Yankwich, J. Am. Chem. Soc. 1954, 76, 15.
[19] J. Wang, T. Wu, X.-H. Fu, T. Sun, Yingyong Huaxue 2004, 21, 1233; Chem. Abstr. 2005, 143, 347097.
[20] H. A. Staab, H. Bauer, K. M. Schneider, ꢂAzolides in Organic Synthesis and Biochemistryꢃ, Wiley-
VCH, Weinheim, 1998.
ˇ
ˇ
ˇ
[21] N. Maras, S. Polanc, M. Kocevar, Tetrahedron 2008, 64, 11618; K. Kranjc, F. Perdih, M. Kocevar, J.
ˇ
Org. Chem. 2009, 74, 6303; K. Kranjc, M. Kocevar, Curr. Org. Chem. 2010, 14, 1050.
[22] R. L. Letsinger, S. B. Hamilton, J. Am. Chem. Soc. 1958, 80, 5411.
[23] R. J. Brotherton, H. Steinberg, J. Org. Chem. 1961, 26, 4632; H. Beyer, K. Niedenzu, J. W. Dawson, J.
Org. Chem. 1962, 27, 4701.
[24] R. J. Brotherton, L. L. Petterson, U. S. Pat. 3312737, 1967; Chem. Abstr. 1967, 67, 21613.
[25] T. Okuyama, K. Takimoto, T. Fueno, J. Org. Chem. 1977, 42, 3545.
[26] P. Nelson, A. Pelter, J. Chem. Soc. 1965, 5142; A. Pelter, T. E. Levitt, P. Nelsoni, Tetrahedron 1970,
26, 1539.
[27] B. N. Feitelson, R. Rothstein, J. Chem. Soc. 1958, 2426.
[28] M. Shen, T. G. Driver, Org. Lett. 2008, 10, 3367.
´
´
[29] G. Navarrete-Vazquez, H. Moreno-Diaz, S. Estrada-Soto, M. Torres-Piedra, I. Leon-Rivera, H.
´
˜
˜
Tlahuext, O. Munoz-Muniz, H. Torres-Gomez, Synth. Commun. 2007, 37, 2815.
´
[30] E. S. Milner, S. Snyder, M. M. Joullie, J. Chem. Soc. 1964, 4151.
[31] S. Roseman, J. Am. Chem. Soc. 1953, 75, 3854.
[32] W. Ried, J. Patschorke, Liebigs Ann. Chem. 1956, 599, 44.
[33] A. R. Freedman, D. S. Payne, A. R. Day, J. Heterocycl. Chem. 1966, 3, 257.
[34] E. Nyilas, A. H. Soloway, J. Am. Chem. Soc. 1959, 81, 2681.
[35] S. Trofimenko, J. Am. Chem. Soc. 1967, 89, 4948.
Received February 10, 2011