Intramolecular Allyl Transfer
664 672
2H), 8.46 (dd, 4J(H,H) 7.2, 3J(H,H) 1.0 Hz, 2H), 8.26 (td, 4J(H,H) 6.5,
3J(H,H) 1.0 Hz, 2H), 7.68 (td, 4J(H,H) 6.5, 3J(H,H) 1.0 Hz, 2H), 6.27
indicate that the allylation reaction can occur by two different
mechanisms. In both mechanisms [Ni0(bpy)2], formed chemi-
cally or electrochemically, is needed to effect the oxidative
addition to the allyl ether group of 1 and form an h3-
allylnickel(ii) complex. In a first mechanism, via a classical Ni-
mediated allylation reaction, the h3-allylnickel(ii) complex 7
can attack the aldehyde group to form the homoallylic alcohol
phenol product 2 via the proposed NiII intermediate 10.
In a more original pathway, the electrochemical reaction
involves the one-electron reduction of the h3-allylnickel(ii)
intermediate 7, followed by intramolecular allyl transfer from
reduced 8. The reduction of the h3-allylnickel(ii) complexes at
relatively low reduction potential (À1.2 V versus SCE) is
reported here for the first time for allylation reactions. The
electrochemical formation of reduced h3-allylnickel(ii) species
has only been recently reported in the cleavage of allyl
ethers.[16]
Theoretical studies on model compounds confirmed the
possibility of stabilized intermediates in both the chemical
and electrochemical mechanisms. However, calculations in-
dicate that under reductive electrochemical conditions, allyl
transfer to the carbonyl group should be facile.
The electrochemical allylation is catalytic in nickel, in
contrast to the stoichiometric chemical reaction. Moreover, in
the electrochemical process, NiII complexes can be used as the
starting materials, and the use of low-stability Ni0 species is
avoided.
(m, 1H); 3.78 (d, 3J(H,H) 7.2 Hz, 2Hsyn); 3.0 (d, 3J(H,H) 12.6 Hz,
À1
2Hanti); IR (KBr): nÄ 1602 (s, C N), 1446, 557 (aromatics), 847 cm (s,
PF); elemental analysis calcd (%) for C13H13F6N2NiP (400.89): C 38.94, H
3.27, N 6.98; found: C 38.6, H 3.60, N 6.90.
Reaction of [Ni(cod)2] with 2-allyloxybenzaldehyde (1) in the presence of
2,2'-bipyridine: A solution of 2,2'-bipyridine (139 mg, 0.89 mmol) and
2-allyloxybenzaldehyde (144 mg, 0.89 mmol) in distilled toluene (5 mL)
was added to [Ni(cod)2] (245 mg, 0.89 mmol) in a purged Schlenk flask. The
mixture was cooled to À788C for 5 min. After addition of toluene (5 mL)
the mixture was stirred for 3 h at room temperature. The initial yellow
solution became brown. A dark red solid precipitated after addition of
hexane (5 mL) and storing the solution for 24 h at 48C. The unstable solid
was filtered off under nitrogen and dried under reduced pressure. meff
3.22 mB; MS (FAB positive): m/z: 376 [Ni(bipy)(1)]. Neither satisfactory
elemental analysis nor 1H NMR data could be obtained because of the
instability and paramagnetic nature of the solid.
General procedure for one-compartment cell electrolyses: Preparation of
2-(1'-hydroxybut-3'-enyl)phenol (2): Freshly distilled DMF (20 mL),
nBu4N BF4 (10À2 m), [Ni(bipy)3]2(BF4À)2 (0.1 mmol) and the substrate 1
(1 mmol), prepared from the corresponding ortho-hydroxybenzaldehyde
by stirring at 508C with allyl bromide and potassium carbonate in DMF,
were introduced into the cell. The solution was stirred at room temperature
and electrolyzed at a constant current of 60 mA (current density of 0.2 A/
dm2, 5 15 V between the electrodes), up to total consumption of the
starting material (monitored by GLC analysis of aliquots), unless electrode
passivation occurred. After evaporation of the DMF under vacuum, the
crude mixture was hydrolyzed with 0.1m HCl saturated with NaCl up to
pH 1 2 and extracted with Et2O. The organic layers, dried over MgSO4,
were filtered and evaporated. The major product 2 was purified by column
chromatography on SiO2 with pentane/diethyl ether (90/10) as eluent and
analyzed by NMR and IR spectroscopy and mass spectrometry. Yield:
87%. To facilitate analysis, 2 was also converted into its methyl ether by
treatment with methyl iodide. 1H NMR (200 MHz, CDCl3, 258C, TMS):
d 7.90 (s, 1H; OH), 7.10 (m, 1H), 6.90 (m, 1H), 6.85 6.70 (m, 2H), 5.84
(dddd, 3J(H,H)trans 17.1, 3J(H,H)cis 10.5, 3J(H,H) 7.1, 3J(H,H) 7.1 Hz,
1H), 5.3 (dd, 3J(H,H)cis 10.5, 2J(H,H) 1.1 Hz, 1H), 5.25 (dd,
3J(H,H)trans 17.1, 2J(H,H) 1.1 Hz, 1H), 4.87 (dd, 3J(H,H) 7.6,
3J(H,H) 7.7 Hz, 1H), 2.90 (s, 1H; OH), 2.70 2.60 (ddd, 3J(H,H) 7.6,
3J(H,H) 7.1, 2J(H,H) 1.0 Hz, 1H), 2.65 2.55 (ddd, 3J(H,H) 7.7,
3J(H,H) 7.1, 2J(H,H) 1.0 Hz, 1H); 13C NMR (200 MHz, CDCl3, 258C,
TMS): 155.5, 134.0, 129.0, 127.2, 126.6, 119.9, 119.3, 117.3, 74.7, 42.2; MS:
À
Experimental Section
General: All solvents were dried and degassed by standard methods. DMF
was freshly distilled over calcium hydride before use in electrolyses.
Instrumentation and cells: 1H and 13C NMR spectra were recorded on a
Bruker AC-200 spectrometer. Infrared spectra were recorded as KBr disks
on a Nicolet 520 FT-IR spectrometer. Mass spectra were obtained with a
Finnigan MAT INCOS 500E spectrometer (GC-MS). Cyclic voltammetry
experiments and controlled-potential electrolyses were performed with
EG&G Model 362 equipment and were carried out at 208C with an SCE or
Ag/AgCl reference electrode, a graphite rod as working electrode for cyclic
m/z: 164 [M ], 146, 131, 121, 107, 91, 77, 65, 43 (100%); IR (KBr): 3366,
3071, 1235, 755 cmÀ1; HRMS calcd for C10H12O2: 164.083730; found:
164.083091.
General procedure for two-compartment cell electrolyses: Both compart-
voltammetry and
a platinum counterelectrode. Controlled constant-
À
ments were filled with a solution of nBu4N BF4 (1 g, 3 mmol) in DMF
(50 mL each) under inert atmosphere. The NiII complex (0.1 mmol) and 1
or benzaldehyde (0.1 mmol) were added to the cathodic compartment. The
electrolyses were run at 208C at the desired controlled potential and were
stopped when the current was negligible. The workup was as described
above; the reaction was monitored by GC.
intensity electrolyses were carried out by using a stabilized constant-
current supply (Sodilec, EDL 36.07). The single-compartment electro-
chemical cell was a cylindrical glass vessel (capacity 20 mL), equipped with
a magnesium rod anode (99.9% purity, immersed to 3 cm) and a stainless
steel grid cathode (apparent area: 20 cm2). In the two-compartment cell,
the two compartments were separated by sintered glass (no. 4); the anodic
compartment contained an Mg rod as anode, and the cathodic compart-
ment was equipped with a carbon fiber cathode and an SCE reference
electrode.
Methods of calculation: It is now generally accepted that DFT methods are
well suited to the study of transition metals complexes. Preliminary
calculations confirmed that the classical HF and post-HF ab initio methods
still remain too expensive in term of computational time. DFT methods
seem to be a good option for the study of systems in which a metal and a
large number of heavy atoms are involved. We chose the B3LYP method,
which uses the gradient-corrected functional of Becke[26] and the correla-
tion functional proposed by[27] Lee, Yang and Paar. Geometry optimiza-
tions were performed at the B3LYP/LanL2DZ[28] level. Thermodynamic
corrections[29] were applied to the optimized structures and led to the
differences in Gibbs free energy DG between the stationary points. To
refine the charge distributions for the atoms of the molecules, the charges
were calculated by using the natural population analysis (NPA) theory of
the natural bond orbital (NBO)[30] program. Solvent effects were not taken
into account in the calculations, since the aim was only to compare the
stability of the different conformations that can be adopted by the studied
complexes. A polar solvent would have certainly have slightly modified the
Tris(2,2'-bipyridine)nickel(ii) bis(tetrafluoroborate) was prepared accord-
ing to ref. [25].
(h3-Allyl)(2,2'-bipyridine)nickel(ii)
hexafluorophosphate:
[Ni(cod)2]
(430 mg, 1.5 mmol), prepared according to ref. [23], was introduced with
distilled toluene (5 mL) into a purged Schlenk flask, and 2,2'-bipyridine
(246 mg, 1.6 mmol) in toluene (1 mL) was added with a syringe. The
mixture was cooled À788C, allyl chloride (129 mL, 1.6 mmol) was added
and the mixture allowed to warm to ambient temperature with stirring. The
solution turned orange-red. NH4 PF6À (257 mg, 1.6 mmol) was added with
distilled THF (20 mL). After 1 h of stirring, the THF was evaporated and
distilled methanol (20 mL) was added. A white solid precipitated. The
mixture was stirred overnight and then filtered under nitrogen. The solid
was dried under reduced pressure. Yield: 85%. 1H NMR (250 MHz,
[D6]acetone, 258C, TMS): d 8.77 (dd, 4J(H,H) 6.5, 3J(H,H) 1.0 Hz,
Chem. Eur. J. 2002, 8, No. 3
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
0947-6539/02/0803-0671 $ 17.50+.50/0
671