oxide was energetically more favoured than coordination with
the phenol moiety.[15,33] Hence, the activation of the glycerol
boronic ester by the phenoxy group to form the “ate complex”
was also considered presumed? to be the earliest step of this
reaction. A decrease in both electronic and free energies for
the ate complex formation was determined, demonstrating the
stabilisation of the species by interaction of the zwitterionic
iminium and the boron species through BÀO bond formation,
and corroborated by the 0.5–0.6 Wiberg indices (WI) deter-
mined for the BÀO bond in ATE1 and ATE2. The stabilisation
of the six-membered ring ate complex ATE1 is higher than the
stabilisation of ATE2 by 4 kcalmolÀ1 when compared with the
initial set of reagents. Inspection of the conformations of the
ate complexes reveals the tetrahedral character of the boron
atom, forcing the five-membered cyclic boronate ester moiety
to adopt a puckered conformation in ATE2 and a chair confor-
mation in ATE1.
In the determined energy profile, the ate complex needs to
adopt a different conformation, placing the migratory substitu-
ent closer to the sp2 carbon of the iminium. Such conforma-
tions (ATE1’ and ATE2’) are only <1 kcalmolÀ1 higher in
energy than the previous ATE1 and ATE2 conformations. The
energy required to achieve product formation starting from
either the five-membered or the six-membered boronate ester
was shown to be very similar. Indeed, there is no energetic dif-
ference between TS1 and TS2, and the energy barrier totals
20.2 kcalmolÀ1 for TS1 and 18.1 kcalmolÀ1 for TS2.
The transition states obtained have similar features, with in-
termediate geometries between the second conformation of
the ate complex (ATE1’ and ATE2’) and the tertiary amines
(TA1 and TA2). In both cases, it is clear that the BÀC(Ph) boron
bond is being broken (d=1.71 ꢁ, WI=0.54 for TS1 and d=
1.47 ꢁ, WI=0.65 for TS2), whilst a new CÀC(Ph) bond is start-
ing to form (d=1.97 ꢁ, WI=0.49 for TS1 and d=2.00 ꢁ, WI=
0.46 for TS2). The long distance and small WI value associated
with the recently formed CÀC bond suggest an early transition
state, which is even more pronounced in the five-membered
cyclic boronate ester pathway (TS2).
Conclusions
We have demonstrated that glycerol is an effective medium
for the Petasis borono–Mannich (PBM) reaction. Alkylamino-
phenols containing tertiary amines, allyl derivatives, and 2-sub-
stituted pyridines, as well as 2H-chromenes, can be prepared
in glycerol in good yields. In some cases, glycerol outper-
formed ethanol as a solvent for this reaction, and with 2-pyri-
dinecarbaldehyde as the carbonyl component, glycerol was
found to be comparable with acetonitrile. The results of a com-
parative mechanistic study of the reaction suggested that the
participation of glycerol-derived boronic esters is competitive
with the mechanism via free boronic acid.. Overall, the report-
ed results show for the first time the possibility of using glycer-
ol in the PBM reaction, and the examples described here can
be considered to be new entries in the compendium of reac-
tions that revalorise waste materials generated by the biodiesel
industry.
Experimental Section
General Procedure: A long, capped test tube containing a magnet-
ic stirrer was charged with the appropriate boronic acid (1.5 equiv)
and pure glycerol (1.0 mL). The boronic acid was left to dissolve for
5 min at 50, 80, or 908C. The aldehyde (0.41 mmol) was then
added, and the reaction was stirred for 2 min at the same tempera-
ture, followed by addition of amine (1.5 equiv). The reaction was
stirred at that temperature until complete consumption of the al-
dehyde as monitored by thin-layer chromatography (TLC) or for
48 h at the longest. After cooling to RT, the reaction was quenched
by the addition of water (1.0 mL) and saturated aq NaHCO3
(1.0 mL), and then extracted with Et2O (3–5ꢂ5 mL) until the extract
contained no further product as determined by TLC. The combined
organic layers were concentrated in vacuo, and the crude product
was purified by flash chromatography on silica gel (EtOAc/hexane).
Supporting Information: Detailed experimental procedures, spec-
troscopy data for the synthesized compounds, computational
methods, and the energy profile for the Petasis borono–Mannich
(PBM) reaction with phenylboronic acid.
A mechanism where phenylboronic acid was considered as
the boron-reacting species was also calculated (see Supporting
Information). A Gibbs energy barrier of only 3.1 kcalmolÀ1
higher than the one determined for path 2 was calculated.
Continuum models are limited to account for electrostatic
interactions. However, solute–environment interactions,[34] such
as hydrogen bonding between the intervenient species and
glycerol, can be anticipated to have a determinant role in
these mechanisms. Assuming that such interactions have the
same effect in both mechanisms (i.e., involving the boronic
ester or the boronic acid), and considering the small difference
in their energetic barriers and the similar yields obtained for
the reaction of glycerol phenylboronic esters in presence and
absence of water (Table 2, entries 5 and 6), it is likely that both
mechanisms are competitive in glycerol.
Acknowledgements
The authors acknowledge CSC–IT Center for Science Ltd, Finland
for the allocation of computational resources, the support of
Prof. Luis F. Veiros (Universidade Tꢁcnica de Lisboa, Portugal) in
the DFT study, and Mrs. Pꢂivi Joensuu (University of Oulu, Fin-
land) for analysing the HRMS data.
Keywords: amines
· boron · glycerol · multicomponent
reactions · sustainable chemistry
[1] a) N. R. Candeias, F. Montalbano, P. M. S. D. Cal, P. M. P. Gois, Chem. Rev.
993–995; d) H. Jourdan, G. Gouhier, L. Van Hijfte, P. Angibaud, S. R. Piet-
ꢀ 2014 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemistryOpen 2014, 00, 1 – 9
&
7
&
ÞÞ
These are not the final page numbers!