A Simple 13C NMR Method for the Discrimination of Complex Mixtures of Stereoisomers: All Eight Stereoisomers of α-Tocopherol Resolved

Article abstract of DOI:10.1002/chir.22535

A simple one-dimensional ¹³C NMR method is presented to discriminate between stereoisomers of organic compounds with more than one chiral center. By means of this method it is possible to discriminate between all eight stereoisomers of α-tocopherol. To achieve this the chiral solvating agent (S)-(+)-1-(9-anthryl)-2,2,2-trifluoroethanol and the compound of interest were dissolved in high concentrations in chloroform-d, and the nuclear magnetic resonance (NMR) spectrum was recorded at a low temperature. The individual stereoisomers of α-tocopherol were assigned by spikes of the reference compounds. The method was also applied to six other representative examples.

Full text of DOI:10.1002/chir.22535

CHIRALITY 27:850–855 (2015)
A Simple 13C NMR Method for the Discrimination of Complex
Mixtures of Stereoisomers: All Eight Stereoisomers of α-Tocopherol
Resolved
1
2
PETER P. LANKHORST, THOMAS NETSCHER, AND ALEXANDER L. L. DUCHATEAU1
*
1DSM Biotechnology Center, DSM, Delft, Netherlands
2DSM Nutritional Products, Research and Development, Basel, Switzerland
ABSTRACT
A simple one-dimensional 13C NMR method is presented to discriminate be-
tween stereoisomers of organic compounds with more than one chiral center. By means of this
method it is possible to discriminate between all eight stereoisomers of α-tocopherol. To achieve
this the chiral solvating agent (S)-(+)-1-(9-anthryl)-2,2,2-trifluoroethanol and the compound of
interest were dissolved in high concentrations in chloroform-d, and the nuclear magnetic reso-
nance (NMR) spectrum was recorded at a low temperature. The individual stereoisomers of α-
tocopherol were assigned by spikes of the reference compounds. The method was also applied
to six other representative examples. Chirality 27:850–855, 2015. © 2015 Wiley Periodicals, Inc.
KEY WORDS: enantiomers; diastereoisomers; vitamin E; Pirkle’s alcohol; TFAE
Nuclear magnetic resonance (NMR) is a powerful method
for the determination of the stereochemical purity of organic
compounds.1 In the case of diastereomers the discrimination
between different diastereomers is without effort, provided
that the chiral centers are not too far separated. In the case
of enantiomers, these enantiomers can be converted into dia-
stereomers with a suitable chiral reagent, or a chiral solvating
agent can be used to create a chiral environment. Many chiral
solvating agents (CSAs) have been proposed and used with
success, and among those CSAs optically pure (S)-(+)-1-(9-
anthryl)-2,2,2-trifluoroethanol (TFAE, Pirkle’s alcohol) is
one of the most versatile compounds.2–4
However, it has been shown convincingly8 that short relaxa-
tion delays and standard broad band decoupling may be used
to obtain compound ratios, provided that the compounds have
similar structures, and carbon atoms with the same number
of attached protons are compared. The latter conditions are
fully met when ratios of stereoisomers are determined.
Tocopherols have three chiral carbon atoms, and therefore
eight different stereoisomers of α-tocopherol exist (Fig. 1).
These eight stereoisomers come as four pairs of enantiomers.
The discrimination between all eight stereoisomers of tocoph-
erol is an extremely challenging task.
In the past, several attempts have been undertaken to sepa-
rate the stereoisomers of (all-rac)-α-tocopherol by means of
chromatographic techniques. Partial separation was obtained
for the four pairs of enantiomers of (all-rac)-α-tocopherol as
trimethylsilyl (TMS) ether derivative by means of gas chroma-
tography (GC) on an achiral column.9 Baseline separation of
the four pairs of enantiomers was realized by analyzing (all-
rac)-α-tocopherol as the methyl ether derivative.10 By means
of chiral high-performance liquid chromatography (HPLC)
on a Chiralpak OT column, only two peaks were obtained for
(all-rac)-α-tocopherol as acetyl derivative.11 Separation of (all-
rac)-α-tocopherol on a Chiralcel OD-H column also resulted
in two peaks: 2R-stereoisomers (RRR, RRS, RSR, RSS) were
separated from 2S-stereoisomers (SRR, SRS, SSR, SSS).12 By
means of a Chiralpak OP(+) column, the acetyl derivative of
(all-rac)-α-tocopherol could be separated into four peaks: 2R-
stereoisomers constituted the first peak and 2S-stereoisomers
were separated into three peaks.13 Use of a chiral stationary
phase (CSP) in open-tubular electrochromatography resulted
in only two groups of α-tocopherol stereoisomers that were
separated.14 Two chromatography studies have been pub-
lished in which the eight stereoisomers of (all-rac)-α-tocoph-
erol could be separated by combination of a chiral HPLC and
1
Most work with CSAs has been done with H NMR, be-
cause 1H NMR is sensitive, a high resolution can be obtained,
and TFAE induces relatively large chemical shift differences.
However, in the case of spectra with broad multiplets, the
chemical shift differences between the R- and S-enantiomer
that can be obtained are often not sufficient. This problem be-
comes even more important when spectra are crowded,
exhibiting overlapping multiplets. Most studies with CSAs
deal with the discrimination of enantiomeric pairs, and it goes
without saying that if the compound studied contains more
than one chiral stereogenic center, and if each chiral carbon
adopts both possible configurations, this adds to the complex-
1
ity of the H NMR approach.
A simplification of such crowded spectra can be achieved
by the homonuclear decoupling (pure shift) approach,5,6 or
as recently proposed7 by means of 13C NMR. 13C NMR spec-
troscopy has long been neglected for enantiodiscrimination,
and only a few studies exist. This is probably mainly due to
the low sensitivity of 13C NMR. It has already been argued,7
that the sensitivity of 13C NMR has increased tremendously
due to the availability of stronger magnetic fields, and due
to probes with cryogenically cooled 13C coils, which makes
13C NMR a more suitable technique for enantiodiscrimination
than previously. At first sight, 13C NMR seems to be less suit-
*Correspondence to: Peter P. Lankhorst, DSM Biotechnology Center, DSM,
PO Box 1, 2600 MA Delft, Netherlands. E-mail: peter.lankhorst@dsm.com
Received for publication 8 July 2015; Accepted 27 August 2015
DOI: 10.1002/chir.22535
Published online 18 September 2015 in Wiley Online Library
(wileyonlinelibrary.com).
1
able than H NMR for quantitative evaluation of mixtures of
stereoisomers, because one has to avoid the nuclear
Overhauser effect (NOE) of protons, and long relaxation de-
lays are required to obtain truly quantitative spectra.
© 2015 Wiley Periodicals, Inc.

EIGHT STEREOISOMERS OF TOCOPHEROL RESOLVED
851
5 mm TCI probe. The low-temperature version of this probe was used,
which can be operated in a temperature range of 240–340 K.
Standard conditions for NMR data acquisition were as follows: Spec-
tra were recorded with 256 scans and a delay of 5 s between pulses in
order to avoid heating of the sample, resulting in an acquisition time
of 30 min. A time domain of 64 K data points was used, and FIDs were
zero-filled to 256 K data points. The sweep width was reduced to 85 ppm
and the aromatic part of the spectrum was filtered out. In this way a
high digital resolution was obtained of 0.1 Hz/pt. The low detection
limit of 1% for minor stereoisomers was achieved by recording 1 K scans
with a delay of 3 s between pulses, which resulted in a total acquisition
time of 90 min.
Before Fourier transformation a Gaussian filter was applied with param-
eters LB = À0.8, GB = 0.25. NMR spectra were also recorded at higher
temperatures and at lower concentrations. In these cases the parameter
GB was adapted to a maximum value of 0.7 in order to be able to observe
the rapidly decreasing peak separation.
Fig. 1. Structure of α-tocopherol. Chiral carbon atoms are C2, C4’ and C8’.
a GC method. In the first study,15 a derivatization was per-
formed to obtain the ethyl ether derivative of (all-rac)-α-to-
copherol, followed by two chromatographic separations: first
a chiral HPLC separation, which yielded two fractions. One
fraction contained the four (2R) stereoisomers, the other frac-
tion contained the four (2S) stereoisomers. Thus, each frac-
tion contains only diastereomers, and no pairs of
enantiomers, and in the next step each fraction could be sepa-
rated into the four constituent stereoisomers by means of achi-
ral capillary GC. In the second study,16 derivatization was
performed with dimethyl sulfate. The resulting methyl ether
derivatives of (all-rac)-α-tocopherol were injected on a chiral
HPLC column, yielding five peaks corresponding to the 2S
(SRS+SSR+SSS+SRR) and RSS, RRS, RRR, RSR stereoiso-
mers. The 2S stereoisomer fraction was collected with a frac-
tion collector and analyzed by achiral GC.
Synthetic Procedures for Enriched Stereoisomers
General. All reactions were performed under an argon atmosphere.
Room temperature was 20–24 °C. All reagents and chemicals not men-
tioned separately were obtained from commercial suppliers and were
used without further purification, if not stated individually. Tetrahydrofu-
ran (THF) was freshly distilled from Na/ benzophenone under argon. For
transfer of solvents, reagents (if liquid), or their solutions, syringes, can-
nulas, and rubber septa were used. The reactions were monitored by thin-
layer chromatography (TLC) using glass plates coated with silica gel
60 F254, thickness 0.25 mm (Merck, Germany); spots were visualized
by UV-light and by spraying with a solution of ammonium molybdate / ce-
rium sulfate in water / sulfuric acid, and subsequent heating. Flash chro-
matography was performed using silica gel 60, 0.040–0.063 mm (Merck)
with max. 0.2 bar gauge pressure argon. The identity and purity of inter-
mediates were checked by using standard spectroscopic and physico-
It has been shown17,18 that the four pairs of enantiomers
can be discriminated by 13C NMR, albeit after the overnight
accumulation of NMR signal in a 10-mm tube.
The aim of the present study was to find conditions to dis-
criminate between all eight stereoisomers of α-tocopherol.
MATERIALS AND METHODS
General NMR Procedures
chemical methods and compared to literature data.
A schematic
representation of the procedure for the synthesis of enriched α-
tocopherol stereoisomers is given in Figure 2 and experimental proce-
dures are detailed below.
(all-rac)-Isophytol, (all-rac)-3-methyl-1,2-cyclopentanediol, and (S)-(+)-
1-(9-anthryl)-2,2,2-trifluoroethanol (TFAE) were purchased from Sigma-
Aldrich (St. Louis, MO).
All-racemic tocols: (all-rac)-α-tocopherol is a commercial product of
DSM Nutritional Products, Switzerland; (all-rac)-β-tocopherol, (all-rac)-γ-
tocopherol, (all-rac)-δ-tocopherol, and (all-rac)-tocol were synthesized in-
house at F. Hoffmann-La Roche in Basel on lab scale.
Samples were dissolved in CDCl3. Approximately 20 mg of compound
of interest and 60 mg of TFAE were dissolved in 0.6 ml of solvent, unless
otherwise stated. These concentrations correspond to molar concentra-
tions of 77 mM and 362 mM, respectively.
Trolox. (R)-Trolox, content 99.9% (GC), enantiomeric excess (ee) >
99.8% (GC); (S)-Trolox, content 99.9% (GC), ee > 99.9% (GC); both com-
pounds were prepared by optical resolution in the kilo-lab of F.
Hoffmann-La Roche, Basel, according to a published procedure.19
Hexahydrofarnesol. Two samples of the following stereoisomeric com-
position have been prepared according to published procedures.20,21
Sample 1: 3R7R 93.45%, 3R7S 5.81%, 3S7R 0.75%, 3S7S 0%; sample 2:
3R7R 0.12%, 3R7S 6.22%, 3S7R 1.36%, 3S7S 92.30%. The stereoisomeric
NMR spectra were recorded on a Bruker (Billerica, MA) Avance III
700 MHz NMR spectrometer, equipped with a cryogenically cooled
Fig. 2. Schematic procedure for the synthesis of enriched α-tocopherol stereoisomers.
Chirality DOI 10.1002/chir

852
LANKHORST ET AL.
composition was determined by using the acetal method (after oxidation
of the primary alcohol to the corresponding aldehyde) as described in the
literature.22,23
to room temperature and the supernatant filtered through a syringe filter
(0.45 μm) into an oven-dried two-necked round-bottomed flask. The re-
maining Mg turnings were washed with THF (5 mL) and the supernatant
was filtered through a syringe filter (0.45 μm) and added to the solution
obtained above. The combined solutions of the Grignard reagent
(14.51 g yellowish liquid) were stored at room temperature in a flask
sealed by rubber septa. Titration of the solution (sec. butanol, o-
phenanthroline method25) gave a 73% yield (0.35 M solution).
3,4-Dihydro-6-benzyloxy-2,5,7,8-tetramethyl-2H-1-benzopyran-2-
yl]methanol. To a suspension of NaH (7.42 g, 97%, dry, Aldrich, Mil-
waukee, WI, 300.0 mmol, 3.0 equiv.) in THF (200 mL) the solution of
Trolox (25.06 g, 99.9%, 100.0 mmol) was added dropwise at 26 °C under
stirring. Additional THF (150 mL) was added to facilitate stirring. The
beige suspension was heated to reflux (oil bath 90 °C) under stirring
for 2 h. Benzyl bromide (26.18 g, 98%, 300.0 mmol, 3.0 equiv.) was added,
and the mixture further stirred at reflux temperature for 2.5 h. The mix-
ture was cooled to 0 °C, and LiAlH4 (15.65 g, 97%, 400.0 mmol, 4 mol
equiv.) was added carefully in portions. After heating for 30 min at reflux
temperature (oil bath 95 °C), the mixture was stirred at 23 °C for 16 h. At
0 °C small portions of ice were added slowly and carefully until foaming
ceased, and 2 N HCl (750 mL) was added to the gray suspension. The or-
ganic phase was separated, and the aqueous phase extracted with diethyl
ether (250 mL, twice). The organic phases were washed separately with
1 N HCl and sat. NaCl solution (100 mL each), combined, dried over so-
dium sulfate, filtered, evaporated (20 mbar, 40 °C), and dried (30 min,
0.022 mbar, 24 °C). The 36.34 g red-yellow oil was dissolved in n-pentane/
diethyl ether (85 mL/ 15 mL). After crystallization (À20 °C, 16 h) and
washing with cold n-pentane/ diethyl ether (9:1, 75 ml) the crystals were
dried (0.022 mbar, 24 °C): 28.76 g (88%) off-white crystals. From the
mother liquor, additional 4.79 g yellow oil was obtained. An analytically
pure sample (colorless crystals, 98.1%) was obtained from the crystals
by recrystallization from the same solvent mixture.
6-Benzyloxy-α-tocopherol26–28
. The triflate (3.52 g, 97.7%, 7.5 mmol)
and a magnetic stirrer bar were placed in an oven-dried 100 mL two-
necked round-bottomed flask. After threefold evacuation followed by
purging with argon, THF (4 mL) was added, and the flask placed in a
cooling bath of À20 °C. The Grignard solution in THF (0.35 M, 32.1 mL,
11.25 mmol, 1.5 equiv.) was dropped in, followed by the solution of
Li2CuCl4 in THF29 (3 mL, 0.1 M, 0.3 mmol, 4 mol%). After stirring for
48 h at À20 to 24 °C additional Li2CuCl4 in THF (3 mL, 0.1 M, 0.3 mmol,
4 mol%) was added. After overall 95 h (TLC control; SiO2, n-hexane/
toluene 7:3, benzyl-tocopherol Rf 0.38, triflate Rf 0.14) the reaction mix-
ture was poured onto ice (25 g). The organic phase was separated and
washed subsequently with 2 N H2SO4, sat. NaHCO3 and NaCl solutions
(50 mL each). The water phases were extracted with diethyl ether
(100 mL, three times), and the combined organic extracts dried over so-
dium sulfate, filtered, evaporated (20 mbar, 40 °C), and dried (2 h, high
vacuum, room temperature): 6.39 g black solid. Column chromatography
(SiO2, n-hexane/toluene 7:3) delivered, after evaporation and drying
(20 mbar, 50 °C/0.021 mbar, 23 °C), 2.42 g (purity 97.3% by NMR, yield
60%) benzyl-tocopherol as a yellowish oil, and in a second fraction 0.53 g
(yield 15%) triflate as a purple-colored solid.
3,4-Dihydro-6-methoxy-2,5,7,8-tetramethyl-2H-1-benzopyran-2-yl]
methyl trifluoromethane-sulfonate. Triflic anhydride (19.75 g,
11.5 mL, 70.0 mmol, 1.4 equiv.) was added dropwise during 20 min to
the stirred solution of the alcohol (16.93 g, 96.4%, 50.0 mmol) and 2,6-
lutidine (8.20 g, 8.90 mL, 98%, 75.0 mmol, 1.5 equiv.) in dichloromethane
(240 mL) at À30 °C.24 After stirring for 1 h at this temperature, the mix-
ture was poured onto 2 N H2SO4 of 0 °C (100 mL). The organic phase
was separated and washed with sat. NaHCO3 and NaCl solutions
(100 mL each). The water phases were extracted with dichloromethane
(100 mL), and the combined organic extracts dried over sodium sulfate,
filtered, evaporated (20 mbar, 40 °C), and dried (1 h, 0.022 mbar, 24 °C).
The 24.73 g slightly beige crystals were dissolved in diethyl ether
(200 mL), treated with active carbon (1.0 g), filtered through Dicalite,
and the solution evaporated under reduced pressure to ~40 g. n-Pentane
(30 mL) was added, and crystallization (room temperature, then À20 °C,
16 h) gave, after filtration, washing (cold n-pentane / diethyl ether 1:1,
100 mL) and drying (2 h, 0.021 mbar, 24 °C) 17.33 g (97.7% by NMR,
yield 73.8%). The colorless crystals should be stored at À20 °C in order
to avoid decomposition. From the mother liquor, additional 4.84 g yel-
lowish crystals (yield 21%) were obtained.
α-Tocopherol. α-Tocopheryl benzyl ether (2.36 g, 97.3%, 4.4 mmol),
EtOAc (30 mL) and 5% Pd/C (1.17 g, 50 wt%) were placed in an autoclave
(50 mL, glass, gassing stirrer). After threefold evacuation followed by
purging with argon, and 3-fold purging with hydrogen gas (5 bar) and
subsequent pressure release, the autoclave was pressurized with hydro-
gen gas (10 bar), and the stirrer (1000 rpm) was started. After 9 min, hy-
drogen uptake was completed. After pressure release the reaction
mixture was filtered, and the catalyst residue washed with additional
EtOAc (~4 g). Evaporation of the filtrate (20 mbar, 40 °C) and drying
(0.1 mbar, 23 °C, 2 h) gave 1.89 g slightly yellowish oil, which was puri-
fied by column chromatography (SiO2, n-hexane/EtOAc 9:1), evaporated
(20 mbar, 40 °C) and dried (0.021 mbar, 23 °C, 2 h): 1.80 g almost color-
less oil, purity 97.6% by quant. GC, yield 92.6%.
Three samples of the following stereoisomeric composition were pre-
pared. Reference sample 1 (“RRR”): 2R,4’R,8’R 93.40%, 2R,4’R,8’S 5.81%,
2R,4’S,8’R 0.75%, 2R,4’S,8’S 0%; reference sample 2 (“SRR”): 2S,4’R,8’R
93.40%, 2S,4’R,8’S 5.81%, 2S,4’S,8’R 0.75%, 2S,4’S,8’S 0%; reference sample
3 (“SSS”): 2S,4’R,8’R 0.12%, 2S,4’R,8’S 6.21%, 2S,4’S,8’R 1.36%, 2S,4’S,8’S
92.21%.
Hexahydrofarnesylmagnesium bromide. The mixture of hexahy-
drofarnesol (11.67 g, 97.9%, 50.0 mmol) and HBr (63% in water, 41.7 g,
325 mmol, 6.5 equiv.) was heated under stirring to 120 °C (oil bath 130 °C)
for 5 h. Completion of the conversion was checked by TLC control (SiO2, n-
hexane/EtOAc 9:1): alcohol Rf 0.07, bromide Rf 0.65. After cooling to room
temperature the mixture was poured onto ice-water (50 mL) and extracted
with n-hexane (75 mL, twice). The combined organic phases were washed
with sat. NaHCO3 and NaCl solutions (50 mL each), dried over sodium sulfate,
filtered, evaporated (20 mbar, 40 °C), and dried (1 h, 0.022 mbar, 24 °C):
13.93 g yellowish oil which was filtered through SiO2 (100 g). The fractions
containing the bromide were evaporated (20 mbar, 40 °C) and dried
(0.021 mbar, 24 °C). The colorless oil (13.24 g) was distilled (0.27 mbar,
140 °C, Kugelrohr oven): 12.96 g colorless liquid (98.9% by NMR, yield 88%).
Grignard magnesium turnings (469 mg, Alfa Aesar 99.8%, 19.25 mmol,
2.75 equiv.) were placed in an oven-dried 25 mL two-necked round-
bottomed flask and heated in an oil bath to 90 °C. THF (4 mL) was added,
and the solution of bromide (2.06 g, 98.9%, 7.0 mmol) in THF (8 mL) was
added dropwise under magnetic stirring for 30 min. After 6 h heating un-
der reflux (TLC control: no bromide detectable) the mixture was cooled
RESULTS AND DISCUSSION
Method Development
1H NMR spectra are shown in the Supporting Information
(S1). It is clear that spectra are too crowded to extract informa-
tion on the stereochemical composition. Therefore, the dis-
crimination of all eight stereoisomers was attempted with a
13C NMR measurement, which is outlined below. The full
13C NMR spectrum is shown and its assignment according
to previous results17,18 is shown in Supporting Information S2.
In the abovementioned previous NMR work17,18 separate
13C NMR signals for each of the four pairs of enantiomers
for C3’ and C2’ were detected. NMR spectra were recorded
in acetone as solvent, but, unfortunately, acetone is not suit-
able for separation of signals of enantiomers with the aid of
the chiral solvating agent TFAE, because acetone is too polar.
Enantiodiscrimination requires the use of an apolar solvent
such as chloroform or benzene. In deuterated chloroform
Chirality DOI 10.1002/chir

EIGHT STEREOISOMERS OF TOCOPHEROL RESOLVED
853
(CDCl3) the abovementioned signals show overlap with other
peaks, but in this solvent the signal of C14’, the methyl group
on the chiral carbon C4’ appears in a relatively uncrowded
part of the spectrum. Figure 3 shows that (R,R,R)-α-tocoph-
erol gives, as expected only one signal for C14’, the all-
racemic tocopherol gives rise to four different NMR signals,
and in the presence of TFAE each of these four signals is split
into two peaks, yielding eight peaks in total, which represent
the eight stereoisomers of α-tocopherol.
It can be seen from Figure 3 that all peaks have shifted to
high-field in the presence of TFAE, which is due to the mag-
netic anisotropy of the anthryl moiety of TFAE. It is due to the
different magnitudes of these effects (different association
constant between TFAE and each stereoisomer, and different
spatial arrangement of tocopherol carbons relative to the
anthryl ring) that the carbon signals of each stereoisomer
shift to a different extent, which enables the separation of
peaks.
similar way. Only the magnitude of the separation is slightly
different. It is clear that full resolution is obtained at 260 K
or 250 K. In fact, the separation of peaks of enantiomers is
even better at lower temperatures (240 K), but some extra
line-broadening was observed, and moreover at temperatures
below 250 K separation between the signals of two enantio-
mers is even too large, and the peaks numbers 2 and 3 partly
overlap. The same is happening with peaks 6 and 7. Thus,
250 K emerged as the optimal temperature to achieve the
separation of signals. At this temperature the line width of
the signals at 10% of the peak-height is 1.3 Hz. It is shown in
Figure 4 that the separation of signals is small but more than
sufficient to obtain baseline resolution (Δδ = 2.1 Hz).
For other compounds the optimal temperature can be dif-
ferent (see Supporting Information S3–S9).
The importance of the second-mentioned condition, the
concentration of both compounds, is also demonstrated in
Figure 4.
The spectra shown in Figure 3 were obtained at 250 K and
the concentration of α-tocopherol was 77 mM and TFAE was
added in 4–5-fold molar excess. These two conditions are ex-
tremely important for the successful application of the CSA
method. In Figure 4 the temperature dependence of the sep-
aration between the signals of one pair of enantiomers is
shown (peak no. 5 and peak no. 6 in Fig. 3). The distance be-
tween the peaks of other enantiomeric pairs behaves in a
The solution of α-tocopherol with TFAE was diluted 2×, 4×,
and 8×. In this way the ratio between the two compounds
remained the same, i.e., 4–5-fold molar excess of TFAE. From
Figure 4 it can be seen that at lower concentrations the sepa-
ration between signals of enantiomers rapidly decreases, and
it is concluded that the highest concentration achievable
must be employed. It should be noted that higher concentra-
tions of TFAE are not compatible with the low temperature of
250 K, because crystallization starts to occur, which should be
avoided, because it negatively affects the resolution of the
NMR spectrum, and moreover, the desired condition of 4-fold
molar excess of CSA is no longer fulfilled.
Assignment of Individual Stereoisomers
Samples of the individual stereoisomers were synthesized
from stereochemically pure or enriched building blocks.
Each reference compound contained one major and one mi-
nor stereoisomer and a third stereoisomer in even lower
quantity. The composition of the reference compounds was
derived from the known composition of the reactants used
in the synthesis of the α-tocopherol stereoisomers and is
shown in Table 1.
All spectra shown in the previous paragraphs were ob-
tained in a total acquisition time of 0.5 h. In order to detect
small quantities of minor stereoisomers, a sufficient signal-
to-noise ratio is needed. Significant signal-to-noise improve-
ment was obtained by increasing the total acquisition time.
We show in Figure 5 the spectrum of reference sample 1
(main component R,R,R stereoisomer) containing 5.8% of R,
R,S and 0.75% of R,S,R.
Fig. 3. Part (0.3 ppm) of the carbon spectrum of α-tocopherol. A: (R,R,R)-α-
tocopherol, B: all-racemic α- tocopherol, C: all-racemic α-tocopherol + (S)-
TFAE. Signals marked with "x" are due to the methyl group C14’.
It can be seen from Figure 5 that with a total acquisition
time of 1.5 h (1024 scans, 3 s interpulse delay) 1% of a minor
compound can be detected in a sample of a single stereoiso-
mer of α-tocopherol. The amount of R,R,S was determined
by setting the integral of the R,R,R signal to 93.4% (expected
value). A value of 5.8% was found for R,R,S, which is in
TABLE 1. Stereochemical composition of the enriched stereo-
isomers (%)
Reference sample
Major
2nd
3rd
1
2
3
RRR 93.4
SRR 93.4
SSS 92.2
RRS 5.8
SRS 5.8
SRS 6.2
RSR 0.75
SSR 0.75
SSR 1.4
Fig. 4. Chemical shift difference between peak 5 and 6 as a function of tem-
perature and of concentration of the sample. ◊ undiluted, □ 2× diluted, Δ 4× di-
luted, + 8× diluted.
Chirality DOI 10.1002/chir

854
LANKHORST ET AL.
Fig. 5. NMR spectrum of reference sample 1. RRR-α-tocopherol with minor
Fig. 7. NMR spectrum of reference sample 3. SSS-α-tocopherol with minor
amounts of SRS- and SSR-α-tocopherol. Amounts detected by NMR 6.6% and
1.7%, respectively. Signal at 19.615 ppm is due to C15’. Expected concentra-
tions see Table 1.
amounts of RRS- and RSR-α-tocopherol. Amounts detected by NMR 5.8% and
1.2%, respectively. Signal at 19.615 ppm is due to C15’. Expected concentra-
tions see Table 1.
excellent agreement with the expectation. For the R,S,R
stereosiomer we find 1.2%, which is slightly different from
the expectation due to the fact that the peak of R,S,R is just
at the detection limit of the present method. Therefore, quan-
tification at this level is less precise, but the impurity is clearly
detected.
The NMR spectra of reference samples 2 and 3 are shown
in Figures 6 and 7, respectively. Also in these two cases there
is excellent agreement between the expected concentrations
of impurities and the experimental concentrations.
With the samples enriched in individual stereoisomers, as-
signment of the eight peaks is an easy task. In fact, only two
enriched stereoisomer samples are sufficient to assign all
eight peaks. At this point some explanation is needed. Be-
cause the temperature-dependent behavior of the racemate
in the presence of TFAE was studied, there is no doubt as
to which pair of peaks is due to a pair of enantiomers. It fol-
lows that if one peak is assigned, automatically the peak of
its enantiomer is assigned as well. Furthermore, each refer-
ence sample contains two well-detectable stereoisomers with
Fig. 8. A: all-racemic α-tocopherol. B: same with spike of RRR- (and RRS-)
α-tocopherol. C: with spike of SRR- (and SRS-) α-tocopherol. D: with spike of
SSS- (and SRS-) α-tocopherol.
Fig. 6. NMR spectrum of reference sample 2. SRR-α-tocopherol with minor
amounts of SRS- and SSR-α-tocopherol. Amounts detected by NMR 6.2% and
2.2%, respectively. Signal at 19.615 ppm is due to C15’. Expected concentra-
tions see Table 1.
Fig. 9. Structures of additional compounds possessing three chiral centers
and successful observation of eight stereoisomers.
Chirality DOI 10.1002/chir

EIGHT STEREOISOMERS OF TOCOPHEROL RESOLVED
855
7. Pérez-Trujillo M, Monteagudo E, Parella T. 13C NMR spectrosopy for the
differentiation of enantiomers using chiral solvating agents.. Anal Chem
2013;85:10887–10894.
a characteristic ratio; the third stereoisomer is not useful for
this purpose. As a consequence of these two conditions, with
each spike eventually four peaks could be assigned, and thus
two spikes assigned all eight stereoisomers. The third spike
merely served for confirmation of the results obtained
through the first two spikes.
8. Otte DAL, Borchmann ED, Lin C, Weck M, Woerpel KA. 13C NMR spec-
troscopy for the quantitative determination of compound ratios and poly-
mer end groups. Org Lett 2014;16:1566–1569.
9. Slover HT, Thompson RH. Chromatographic separation of the stereoiso-
mers of α-tocopherol. Lipids 1981;16:268–275.
10. Cohen N, Scott CG, Neukom C, Lopresti RJ, Weber G, Saucy G. Total syn-
thesis of all eight stereoisomers of α-tocopheryl acetate. Determination of
their diastereoisomeric and enantiomeric purity by gas chromatography.
Helv Chim Acta 1981;64:1158–1173.
The results of the spiking experiments are shown in
Figure 8.
With similar conditions it is also possible to observe eight
different stereoisomers of other compounds containing three
chiral carbon atoms (structures shown in Fig. 9). The spectra
of (all-rac)-β-tocopherol, (all-rac)-γ-tocopherol, (all-rac)-δ-to-
copherol, (all-rac)-tocol, (all-rac)-isophytol, and (all-rac)-3-
methyl-1,2-cyclopentanediol are shown in the Supporting In-
formation S3–S9. The interaction of the compound of interest
with TFAE requires the presence of an aliphatic or aromatic
hydroxy (OH) group. If such a group is not present, the reso-
lution of enantiomers seems to be insufficient.
11. Yamagushi H, Itakura Y, Kunihiro K. Analysis of the stereoisomers of
alpha-tocopheryl acetate by HPLC. Iyakuhin Kenkyu 1984;15:536–540.
12. Mazzini F, Betti M, Netscher T, Galli F, Salvadori P. Configuration of the
vitamin E analogue garcinoic acid extracted from garcinia kola seeds. Chi-
rality 2009;21:519–524.
13. Ueda T, Ichikawa H, Igarashi OJ. Determination of α-tocopherol stereoiso-
mers in biological specimens using chiral phase high-performance liquid
chromatography. Nutr Sci Vitaminol 1993;39:207–219.
14. Chen J-L, Hsieh K-H. Nanochitosan crosslinked with polyacrylamide as the
chiral stationary phase for open-tubular capillary electrochromatography.
Electrophoresis 2011;32:398–407.
CONCLUSION
15. Vecchi M, Walther W, Glinz E, Netscher T, Schmid R, Lalonde M, Vetter
W. Chromatographische Trennung und quantitative Bestimmung aller
acht Stereoisomeren von α-Tocopherol. Helv Chim Acta 1990;73:
782–789.
16. Klaczkow G, Anuszewska E. Determination of the stereoisomers of race-
mic alpha-tocopherol in pharmaceutical preparations by high-
performance liquid chromatography and gas chromatography. Acta Pol
Pharm 2008;65:715–721.
In summary, we have shown that the discrimination of all
eight stereoisomers of α-tocopherol, as well as six other rep-
resentative examples possessing three chiral centers, by
means of 13C NMR is possible, provided that the right exper-
imental conditions are applied. The important conditions are:
low temperature (250 K) and the highest concentration
achievable at the low temperature (80 mM of α-tocopherol
and 4–5-fold molar excess of TFAE). In the case of α-
tocopherol, all individual stereoisomers were identified. It is
expected that this 13C NMR method using the
abovementioned experimental conditions is applicable to
most chloroform soluble organic compounds bearing a hy-
droxyl group.
17. Bremser W, Vogel GM. Determination of the enantiomeric purity of α-
tocopherol by 13C NMR spectroscopy. Org Magn Reson 1980;14:155–156.
18. Brownstein S, Burton GW, Hughes L, Ingold KU. Chiral effects on the
carbon-13 resonances of alpha-tocopherol and related compounds. A novel
illustration of Newman’s "rule of six.". J Org Chem 1989;54:560–569.
19. Walther W, Vetter W, Vecchi M, Schneider H, Müller RK, Netscher T.
(S)-Trolox™ methyl ether: a powerful derivatizing reagent for the GC de-
termination of the enantiomers of aliphatic alcohols. Chimia 1991;45:
121–123.
SUPPORTING INFORMATION
20. Netscher T, Scalone M, Schmid R. Chapter 4. Enantioselective hydroge-
nation: towards a large-scale total synthesis of (R,R,R)-α-tocopherol. In:
Blaser H-U, Schmidt E editors, Asymmetric catalysis on industrial scale.
Wiley-VCH: Weinheim, Germany; 2004. p 71–89.
Additional supporting information may be found in the
online version of this article at the publisher’s web-site.
21. a Wang A, Wüstenberg B, Pfaltz A. Enantio- and diastereoselective hydro-
genation of farnesol and O-protected derivatives: stereocontrol by chang-
ing the C=C bond configuration. Angew Chem 2008;120:2330–2332;
b Angew Chem Int Ed 2008;47:2298–2300.
22. Knierzinger A, Walther W, Weber B, Netscher T. A practical method for
the stereochemical analysis of acyclic terpenoid carbonyl-compounds.
Chimia 1989;43:163–164.
23. Knierzinger A, Walther W, Weber B, Müller RK, Netscher T. Eine neue
Methode zur stereochemischen Analyse offenkettiger terpenoider
Carbonyl-Verbindungen. Helv Chim Acta 1990;73:1087–1107.
24. Netscher T, Bohrer P. Formation of sulfinate esters in the synthesis of
triflates. Tetrahedron Lett 1996;37:8359–8362.
LITERATURE CITED
1. Wenzel TJ. Discrimination of chiral compounds using NMR spectroscopy.
Wiley Interscience: Hoboken, NJ; 2007.
2. Pirkle WH, Sikkenga DL, Pavlin MS. Nuclear magnetic resonance deter-
mination of enantiomeric composition and absolute configuration of γ-
lactones using chiral 2,2,2-trifluoro-1-(9-anthryl)ethanol. J Org Chem
1977;42:384–387.
3. Pirkle WH, Hoover DJ. NMR chiral solvating agents. Top Stereochem
1982;13:263–327.
4. Weisman GR. Nuclear magnetic resonance analysis using chiral solvating
agents. In: Morrison JD editor, Asymmetric Synthesis 1. Academic Press:
New York; 1983. p 153–196.
25. Watson SC, Eastham JF. Colored indicators for simple direct titration of
magnesium and lithium reagents. J Organomet Chem 1967;9:165–168.
26. Netscher T. Sulfonate leaving groups for nucleophilic substitution reac-
tions – improved structures and procedures. Recent Res Dev Org Chem
2003;7:71–83.
5. Chaudhari SR, Suryaprakash N. Chiral discrimination and the measure-
ment of enantiomeric excess from a severely overcrowded NMR spec-
trum. Chem Phys Lett 2013;555:286–290.
27. Fouquet G, Schlosser M. Improved carbon-carbon linking by controlled
copper catalysis. Angew Chem 1974;86:50–51.
28. Schlosser M. Rezepte und Zutaten für kontrollierte CC-Verknüpfungen
6. Pérez- Trujillo M, Castanar L, Monteagudo E, Kuhn LT, Nolis P, Virgili A,
Thomas Williamson R, Parella T. Simultaneous 1H and 13C NMR
enantiodifferentiation from highly-resolved pure shift HSQC spectra.
Chem Commun 2014;50:10214–10217.
mit Organometall-Reagentien. Angew Chem 1974;86:751–788.
Chirality DOI 10.1002/chir