reasons). Instead in these cases we only obtain the diarylated
product. The calculations show that the most probable
aluminim-mediated reaction proceeds by a zwitterionic inter-
mediate via a transition state in which the H–C–C–O dihedral
angle is ~Ϫ180Њ, i.e. the incoming aromatic compound
approaches the activated carbonyl bond with the C–H bond to
be broken in an anti fashion.
Experimental and theoretical
General methods
All reactions were carried out using anhydrous solvent and
flame-dried glasware using standard Schlenk techniques under
N2. Commercially available compounds were used without
further purification. Solvents were dried according to standard
procedures. Purification of the products, when necessary, was
carried out by flash chromatography (FC) using Merck silica
gel 60 (230–400 mesh). TLC was performed using Merck silica
gel 60 F254 plates and visualized with blue stain. Optical
rotations were measured on a Perkin-Elmer 241 polarimeter.
1H NMR and 13C NMR spectra were recorded at 400 and 100
MHz respectivly, using CDCl3 as the solvent, and are reported
in ppm downfield from CDCl3 (δ = 7.25) for 1H NMR and
relative to the central CDCl3 resonance (δ = 77.0) for 13C NMR.
The enantiomeric excess (ee) of the products were determined
by HPLC.
Fig. 6 The lowest transition-state structure for the formation of a
Friedel–Crafts intermediate of type B by reaction of pyridine-2-
carbaldehyde 1a with N,N-dimethylaniline 2a starting from complex I.
They observed higher reactivity for 1a compared to the corre-
sponding 3- and 4-isomers, 1b and 1c, respectively, of pyridine-
carbaldehyde, and proposed that this could indicate a possible
coordination of the metal by the pyridine nitrogen atom,
because they at the same time observed no reactivity of benz-
aldehyde under similar conditions. Focusing on the hydroxy
product (in this product the proton has been transferred from
the aromatic C–H bond to the oxygen atom of the activated
carbonyl bond) in the reaction of the pyridine-2-carbaldehyde
General procedure
The appropriate BINOL ligand (0.2 mmol) was placed in the
reaction flask and was subjected to vacuum for 1 h. The flask
was then refilled with N2 and solvent (2.5 ml) was then added.
A commercially available heptane solution of AlMe2Cl (0.2
mmol) was added by microsyringe resulting in gas evolution
(CH4) and white precipitate. The reaction mixture was stirred at
room temperature for 1 h. If needed the reaction mixture was
cooled to the appropriate temperature and the pyridinecarb-
aldehyde (1.05 equiv.) and aromatic compound (1.0 equiv.) were
added resulting in a brownish color. The reaction was left stir-
ring giving a clear orange solution. The mixture was poured on
to 10 mL of NH4Cl (aq. sat.) and additional solvent was added.
The phases were separated and the aq. phase was extracted with
CH2Cl2 (10 ml) and EtOAc (10 ml). The combined organic
phases were dried over Na2SO4 and the solvent was removed.
The product was purified by column chromatography on
silica. (Rf(product): 0.25 in 5% MeOH–CH2Cl2) The BINOL
ligand and excess starting material were eluted using ether–
petroleum ether (30%) after which the product was collected
using 5% MeOH in CH2Cl2.
we can observe a stabilisation by chelation of 17.7 kcal molϪ1
.
The fact that bidentate coordination is not possible for the
3- and the 4-isomers and the fact that only the 2-isomer
gives the hydroxy product while the 3- and 4-isomers react fur-
ther to give a diarylated product, 3a and 3b, respectively, leads
us to believe that (i) chelation could be important for reaction
to stop at the alcohol stage, thereby making it possible upon
aqueous workup to isolate the hydroxy product 3a, and/or (ii)
that the increased reactivity for 1b and 1c is due to the lower
LUMO energy of these compounds compared to 1a.
Summary
We have found that it is possible to mediate an enantio-
selective Friedel–Crafts reaction of pyridine-2-carb-
aldehyde with N,N-dimethylaniline by the application of chiral
Lewis acids. The reactions proceed with moderate to good
yields depending on the reaction conditions, and moderate
enantioselectivity could be obtained by applying an optic-
ally active ligand ((R)-BINOL) to the catalyst system. The
3- and 4-isomers of the pyridinecarbaldehyde also react with
(4-Dimethylaminophenyl)pyridin-2-ylmethanol 3a. 1H NMR
δ 8.54 (d, J = 5.0 Hz, 1H), 7.59 (dt, J = 7.7 Hz, J = 1.7 Hz,
1H), 7.20 (d, J = 8.7 Hz, 2H), 7.12–7.22 (m, 2H), 6.68 (d, J = 8.7
Hz, 2H), 5.67 (d, J = 3.6 Hz, 1H), 5.06 (d, J = 3.6 Hz, 1H), 2.91
(s, 6H). 13C NMR δ 177.5, 150.3, 147.6, 136.6, 131.1, 128.1,
122.1, 121.3, 112.5, 74.7, 40.6. HPLC (Daicel Chiralcel OD,
hexane–i-PrOH = 90 : 10, flow rate = 0.5 ml minϪ1) tr = 33.9
min (major), tr = 41.1 min (minor). HRMS for C14H16N2O
(Naϩ): calcd 251.1160, found 251.1161.
The reactions of 4-methoxyphenol 2b and 1,3-dimethoxy-
benzene 2c with N,N-dimethylaniline in CH2Cl2 in the presence
of (R)-4a-AlCl were performed according to the general
procedure. The data for for the products are available from the
literature.9
N,N-dimethylaniline under similar conditions giving
a
diarylated product, while benzaldehyde showed no reactivity.
The investigation of a mono- vs. bidentate coordination of the
aluminium catalyst both experimentally and computationally
leads us to believe that the bidentate coordination is not
necessary for the reaction of 1a and 2a to proceed. The compu-
tational results point to monodentate coordination making
the lowest energy reaction path. Thus, we believe that it is the
electron-withdrawing effect of the pyridine ring that enhances
the reactivity of the pyridinecarbaldehydes as compared to e.g.
benzaldehyde. This is consistent with the fact that we also
observe reactions of 1b and 1c with 2a in the presence of an
aluminium catalyst. However, a bidentate coordination might
be important for the reaction to stop at the stage of the hydroxy
product. We see a stabilisation of the product by the bidentate
coordination; moreover we observe no hydroxy product in the
reactions of 1b or 1c (in which the coordination of the Lewis
acid to the nitrogen atom is not possible for geometrical
The reactions of pyridine-3- and -4-carbaldehyde, 1b and
1c, respectively, with N,N-dimethylaniline were performed
according to the general procedure. Compound 3b: 43% con-
1
version after 24 h. H NMR δ 8.25 (s, 2H), 7.41 (m, 1H), 7.19
(m, 1H), 6.94 (d, J = 9.2 Hz), 6.66 (d, J = 9.2 Hz), 5.33 (s, 1H),
1
2.93 (s, 12H). Compound 3c: 50% conversion after 24h. H
J. Chem. Soc., Perkin Trans. 1, 2001, 854–860
859