Werner Bonrath et al.
COMMUNICATIONS
less than 1.5% with the best catalysts) when 2 was reacted with
1 in the presence of catalytic amounts of imides 4 under the
conditions specified in the experimental section. Under those
conditions, the conversion of 2 is around 99% and the yield of
(all-rac)-a-tocopherol (3) is in the range of 94%. Results
obtained from experiments with NH-acidic catalysts are
presented in Table 1.
selectivity of the overall reaction. Compared to other syntheses
of 3,[9] it was found that the ratio of benzopyran to -furan was
higher than 200, and the amount of phytadienes was in the
range of a few percent only.
In further experiments the recovery of the NH-acidic cata-
lysts was investigated. In polar aprotic solvents within a temper-
ature range from 50 to 100 8C, the catalyst could be recovered
ten times without detectable loss of activity and selectivity.
With regard to the synthesis of 3 we recommend the use of
perfluorinated NH-acidic imides 4 as excellent catalysts. Their
principal advantage when applied in biphasic solvent systems is
based on the combination of activity and selectivity of
homogeneous catalysts with heterogeneous solvent systems.
Table 1. Reaction of
1 and 2 in various solvents using NH-acidic
compounds of the type (Rf1SO2)(Rf2SO2)NH (4).
4
Rf1
Rf2
solvent[a]
yield 3 (%)[b]
a
a
b
b
c
d
d
d
e
f
g
g
h
i
k
m
n
o
CF3
CF3
CF3
CF3
C2F5
CF3
CF3
CF3
C2F5
C3F7
C6F5
C6F5
CF2CF2CF2
C4F9
C4F9
CF3
C8F17
C4F9
CF3
CF3
toluene (50)
EC (40)/hept. (50)
toluene (50)
EC (40)/hept. (50)
toluene (50)
89.5
91.0
89.6
94.0
90.2
83.8
87.4
92.6
93.6
94.5
87.2
84.9
94.5
94.0
93.3
86.4
85.6
91.1
C4F9
C4F9
C4F9
C6F5
C6F5
C6F5
C2F5
C3F7
C6F5
C6F5
Experimental Section
diethyl ketone (50)
g-but.lac (40)/hept. (50)
PC (40)/hept. (50)
EC (40)/hept. (50)
EC (40)/hept. (50)
EC (40)/hept. (50)
PC (40)/hept. (50)
EC (40)/hept. (50)
PC (40)/hept. (50)
PC (40)/hept. (50)
PC (40)/hept. (50)
PC (40)/hept. (50)
PC (40)/hept. (50)
General
[c]
[c]
Trimethylhydroquinone (1), toluene, heptane, propylene carbonate, ethyl-
ene carbonate, and (CF3SO2)2NH (4a) were purchased from Fluka and (C2F5
SO2)2NH (4e) and [4H]1,3,2-dithiazine-4,4,5,5,6,6-hexafluorodihydro-
1,1,3,3-tetroxide (4h) from K. F. Meyer (Fussgˆnheim). Isophytol was
obtained from Teranol AG and (C4F9SO2)2NH (4i) from Bayer AG
(Leverkusen). All the compounds listed above were used without further
purification. All solvents and liquid reagents were degassed by three freeze-
thaw cycles before use.
The catalytic reactions were carried out in a batch reactor under an argon
atmosphere using Schlenk techniques. Gas chromatographic analyses (GC)
were carried out on a HP 5890 apparatus equipped with an autosampler and
a capillary column Macherey-Nagel type Optima 5 (30 m  0.32 mm). The
compounds elute at detected retention times of 10.5 min (1), 13.1 min (2),
and 23.7 min (3). The crude product was analyzed by GC with an internal
standard.
C4F9
C6F5
C8F17
C8F17
C8F17
[a]
In mL, g-but.lac g-butyrolactone, hept. heptane, EC ethylene
carbonate, PC propylene carbonate.
[b]
[c]
Yields are based on 2, determined by GC analysis of the crude product.
Mixture of n/iso.
The key findings are:
Preparation of Imides 4
Maximum yields of (all-rac)-a-tocopherol are obtained in
polar aprotic solvents.
For the synthesis of imides 4 not being commercially available, literature
procedures[11,12] were followed and slightly adapted to the individual
compounds (cf. Scheme 2). The products were obtained after high-vacuum
sublimation or short-path distillation from conc. H2SO4. Their air-sensitivity
requires strict handling under an argon atmosphere. Aryl derivative 4g was
isolated from the water-insoluble ammonium salt after use of cation
exchange resin (Amberlite IR-A 120). All compounds were fully charac-
terized by 1H/19F NMR, IR, MS, microanalysis, and mp.
Best yields are obtained in two-phase solvent systems.
The optimal number of C-atoms for alkyl-substituted imides
4 is between two and four.
The cycloaliphatic imide 4h is among the best catalysts.
In 4, the optimal perfluoroaliphatic substituents are more
efficient than pentafluorophenyl substituents.
Mixed alkyl-/aryl-substituted catalysts 4 are better than
aryl-/aryl-substituted ones.
In addition, we found that the yield is strongly dependent on
the solvent polarity, corroborating results from earlier work.[7]
The main by-products in this reaction are phytadienes and
furan derivatives, which have already been characterized by
Yamamoto and coworkers in the scandium(III) trifluorome-
thanesulfonate-catalyzed condensation of 1 and 2.[15]
Compared to the results of zinc chloride/Br˘nsted acid- or
BF3-catalyzed reactions, we found good yields and selectivities.
Using ZnCl2/HCl or BF3 as catalysts, (all-rac)-a-tocopherol (3)
could be obtained in yields of approximately 80% at best.[16]
Another advantage of the new procedure is the low amount of
catalyst (as low as 0.1 mol %) needed.
Typical Procedure for the Catalytic Preparation of 3
To a mixture of 1 (7.69 g, 50.0 mmol) and the catalyst (0.1 mol %, based on 2)
in 40 mL ethylene carbonate and 50 mL heptane (Table 1), 2 (10.0 g,
11.9 mL, 33.0 mmol) was added at 100 8C during 20 min. The reaction
mixture was stirred for an additional 30 min at reflux temperature
monitored by thin layer chromatography. After total conversion, (all-rac)-
a-tocopherol (3) was isolated from the reaction mixture by cooling to 60
80 8C, separation of the catalyst by decanting (phase separation) and
distilling off the non-polar solvent. Excess 1 (contained in the ethylene
carbonate phase) may be re-used. Unambiguous identification of products
was made by comparison of GC retention times and spectroscopic measure-
ments (NMR) with authentic samples. In the case of 4b (in EC/heptane)
crude 3 was transformed (pyridine, Ac2O) to the more stable acetate
derivative. After distillation (250 8C/10À1 mbar) (all-rac)-a-tocopheryl ace-
tate was obtained as pale yellow oil; yield: 93.2% (based on 2).
It must be pointed out that an additional advantage of this
type of imide catalysts is documented by an extremely high
38
Adv. Synth. Catal. 2002, 344, 37 39