close

Вход

Забыли?

вход по аккаунту

?

Gas Chromatographic Separation of Enantiomers on Optically Active Metal-Complex-Free Stationary Phases. New Analytical Methods (24)

код для вставкиСкачать
Volume 23
-
Number 10
October 1984
Pages 747-830
International Edition in English
j\lrthud.l
Gas Chromatographic Separation of Enantiomers
on Optically Active Metal-Complex-Free Stationary Phases**
New Analytical
By Volker Schurig”
If a stationary phase A employed in gas chromatography possesses a chemical affinity for
substance B, which is to be separated, then the retention behavior is not only determined by
the normal physical equilibrium between the gas and liquid phases but also by the chemical
equilibrium A + B + A B . If A and B are chiral and A is present in optically active form
while B is a racemic mixture, then it is possible to achieve a gas chromatographic enantiomer resolution without the isolation of diastereomers: the energetically different diastereomeric associates AR B, and ARB, are formed rapidly and reversibly. This enantiospecific
resolution principle was first demonstrated in 1966 by the quantitative resolution of racemic
amino acid derivatives on optically active peptide phases in analogy to the well-known stereospecificity of enzymes. The anchoring of the chiral resolving agent to thermally stable
polysiloxanes together with the employment of high resolution capillary columns and the
use of appropriate derivatization strategies has led to the development of enantiomer resolution into a routine modern method for many classes of substances. The demonstration of
enantiospecificity in the gas chromatographic separation process is of fundamental interest,
and its systematic study can result in a significant contribution to the understanding of the
molecular mechanism of “chiral recognition”. The gas chromatographic separation of enantiomers has also proven to be an accurate and sensitive method for the determination of
the enantiomeric composition of natural products and products of enantioselective transformations (asymmetric syntheses, “chiral pool” transformations, kinetic resolutions, biomimetic reactions) and for the quantification of racemization, e. g. in the synthesis and hydrolysis of peptides. In any research program devoted to the phenomenon of chirality, the gas
chromatographic separation of the enantiomers of volatile compounds constitutes an indispensable modern instrumental technique.
1. Introduction
The phenomenon of optical activity is one of the most
fascinating manifestations of living matter which has in[*I Prof. Dr. V. Schurig
Institut fur Organische Chemie der Universitat
Auf der Morgenstelle 18, D-7400 Tubingen (FRG)
I**]The separation of enantiomers on optically active metal-complex stationary
phases (“complexation gas chromatography”) will not he discussed here.
Anaew. Chem. Inl. Ed. Engl.23 (1984) 747-765
spired the chemist to many imitations in the laboratory
ever since the founding of stereochemistry. The resolution
of synthetic racemic mixtures by methods analogous to
those of Pusteur in his pioneering investigations is of undiminished importance both as a preparative method for
the isolation of Dure enantiomers and as an analvtical tool
for the determination of enantiomeric compositions. The
availability of high-performance methods of enantiomer
analysis has become urgently necessary in order to keep
0 Verlag Chemie GmbH, 0-6940 Weinheim, 1984
0570-0833/84/lfJ10-0747 $ 02.50/0
141
pace with the development of highly enantioselective reactions“’ (asymmetric synthesis, kinetic resolution, “chiral
pool” syntheses, enzymatic transformations) and to cope
with the requirements of mechanistic investigations.
Enantiomer resolution by gas chromatography constitutes a technique that allows the exact determination of enantiomer composition, the determination of enantiomer
purity up to 99.9%, and the assignment of absolute configurations employing subnanogram quantities for many
groups of compounds[’]. Furthermore, the method is suitable for the investigation of the phenomenon of chiral recognition and can in this way contribute to an understanding of structure/activity relationships and the enantiospecific interaction of chiral selectors and select and^[^"].
The resolution of enantiomeric mixtures by gas chromatography can be performed in two
a) Indirect method: conversion of the enantiomers into
diastereomeric derivatives by chemical reaction with a n
auxiliary, enantiomerically pure, chiral resolving agent
and subsequent gas chromatographic separation of the
resulting diastereomers on an achiral stationary phase.
b) Direct method: Gas chromatographic separation of the
enantiomers on a chiral stationary phase containing an
auxiliary resolving agent of high (but not necessarily
complete) enantiomeric purity.
An alternative method, when the chiral agent is mixed
with the mobile gas phase, has also been described[’“].
These processes rest on the use of a chiral non-racemic
auxiliary agent and hence constitute variants of the classical method of enantiomer resolution via diastereomeric
salts according to Pasteur. Chromatographic resolution of
enantiomers without the employment of a chiral auxiliary
agent is, in analogy to the construction of a perpetual motion machine, impossible according to the present state of
scientific knowledge. Reports of enantiomer resolution in
achiral systems seem either ~ n r e l i a b l e [ ’ ~ ~require
~ ‘ ~ , a rigorous examination[5d1,or they d o in fact involve a chiral
agent, e.g. as internal standardf6!
While method (a) involves the formation and isolation
of diastereomers before separation, method (b) involves the
rapid and reversible diastereomeric interaction between optically active selector and racemic selectands. A prerequisite for method (a) is the presence of a reactive chemical
function in the enantiomers which allows quantitative
reaction with an optical active reagent. Racemization of
the sample on derivatization or kinetic resolution at incomplete conversion obscures the original enantiomer ratio. The diastereomeric ratio to be determined can also be
altered by accidental fractionation during derivatization,
work-up and chromatographic analysis (injection, detection). Systematic errors always occur if the chiral agent is
not present in complete enantiomeric purity or is racemized during the derivatization reaction. Such sources of
error can be detected by a combination of methods (a) and
(b), i.e. by employing a chiral phase which resolves all four
configurational isomers.
The direct resolution of enantiomers by method (b) is of
necessity more difficult in that, apart from the chiroptical
parameters, all physical and chemical properties of optical
748
antipodes are identical in an achiral environment. If the
gas chromatographic retention volume V’ of the selectand
is expressed as a product of the vapor pressure p o and the
activity coefficient y
V‘
y
then the resolution of enantiomers ( R and S ) , which is expressed in terms of the separation factor a,
can only take place as a result of a difference in the vapor
pressure or the activity coefficient[’]. While the presence of
an optically active component in the mobile phase would
appear to alter the vapor pressures p R and ps, the employment of a chiral stationary phase can lead to a discrimination of YR and yR because of the differing stabilities of the
diastereomeric associates of selector and selectand.
Method (b) is characterized by the introduction of chernical selectivity into the gas chromatographic separation
process, whereby chiral recognition is caused by the molecular association between the racemic selectand and the
optically active selector in the stationary phase.
An optically active stationary phase is of itself no guarantee for an effective enantiomer resolution. On using a
high resolution capillary column, a quantitative resolution
is achieved only when the difference between the free enthalpy of diastereomeric association for the enantiomers is
at least 10 cal/mol (22°C). The success of enantiomer resolution should always be verified by control experiments.
Peak coalescence, when a racemic phase is used, and peak
inversion of nonracemic mixtures on employment of chiral
phases of the opposite configuration are unequivocal criteria of enantiomer resolution[’].
This progress report will only be concerned with the direct method (b) of gas chromatographic enantiomer resolution. At present two well-developed methods are available
which are complementary to each other:
-
-
Resolution of enantiomers on optically active amino acid
and peptide selectors via hydrogen b ~ n d i n g [“1.~ , ~ ,
Resolution of underivatized enantiomers on optically active metal chelates by coordination interaction (“complexation gas chromatography”)[“].
The first reproducible and quantitative gas chromatographic enantiomer resolution was achieved in 1966 by GilAv, Feibush and. Charles-Sigler at the Weizmann Institute
of Science[”]. Gil-Av had previously investigated the resolution of racemic amino acids by the indirect method[’31
and now sought to employ the chiral agent not for derivatization but as an additive to the stationary phase. The contemporary state of the art of gas chromatographic enantiomer resolution should not blind us to the initial difficulties with which the method was beset: “with a state of frust r a t i ~ n ” [ In
~ ~retrospect
.
Gil-Av may be
“A considerable number of workers all over the world
have tried to solve the problem of the resolution of antipodes by this method (i.e. gas chromatographically) without any success. When we turned to this research topic alA n g r w . Chem. l n t .
Ed. Engl. 23 (1984) 747-765
most all the workers in the field had given u p the idea as
impracticable. And not only that: because many reports of
positive results had proved to be in error, the topic had
fallen into disrepute”.
“After long and painstaking effort, which was accompanied by many a disappointment, we finally discovered a
solution to the problem. The core of the matter was to find
a liquid for wetting the walls, that in some respects imitated an enzyme. ... To our surprise it turned out that
amino acids ... retained to a certain extent the ability of
the larger enzyme structures to distinguish between leftand right-handed molecules ... (cf. Fig. 1). We were thus
able to add a n important new method for the study of optically active molecules to those developed by Pasteur and
other workers.”
The demonstration of enantiospecificity in a gas chromatographic system as a counterpart to the already known
liquid chromatographic racemate r e s o l ~ t i o n “ ~was
~ ” of
~ itself remarkable, particularly since the resolution occurred
quantitatively (cf. Fig. 1). Furthermore, the manifold possibilities for its application, which result from the advantages of the gas chromatographic method (high performance, sensitivity, reproducibility, precision, simple detection), were realized immediately. The ability to resolve
minute quantities of amino acid enantiomers coincided
with the intensive efforts to demonstrate the presence of
biogenic materials in interstellar space, in meteorites and
in lunar samples from the NASA Apollo moon program.
The cooperation of Gil-Av with biophysicists and peptide
chemists at the University of Houston, Texas, E. Bayer, J .
Oro, W. Parr, and W. A. Koenig, produced important advances which, with the employment of thermally stable
chiral polysiloxane phases, laid the foundations for the
rapid development of gas chromatographic enantiomer resolution as a routine, modern, high-performance method[%‘01.
In 1969 Gil-Au and Schurig attempted to extend the area
of application of method (b) for gas chromatographic enantiomer resolution to other classes of compounds. Chiral
unsaturated hydrocarbons, ethers, esters, ketones etc. are
not resolvable or only resolvable with difficulty into their
enantiomers via formation of diastereomeric salts. Thus
the enantiospecific coordination interaction of racemic selectands with optically active organometallic selectors was
regarded as an important complementary possibility for
gas chromatographic chiral recognition because of the va-
R
S
7
0
t“h
1
-
2
3
Fig. 2. Gas chromatographic rrsolulioii 01 riicum1c 3-mt‘lhylcyclopentene o n
an optically active metal chelate by enantiospecific coordination interaction
181.
riety and variability of metal complex structures and of optically active ligands.
By analogy with the resolution of racemic trans-cyclooctene via crystallization of a diastereomeric platinum-ncomplex by Cope et al.[’811,
it was decided to test whether a
racemic olefin could be resolved gas chromatographically
into its enantiomers without the isolation of diastereomers,
on the basis of the diastereomeric interaction of the enantiomers with a chiral metal complex. This principle, whose
practical realizability has at one time been doubted[”’, was
first realized in 1977, as a result of improvements in gas
chromatographic instrumentation, with demonstration of
the resolution of racemic 3-methylcyclopentene on dicarbonylrhodium-3-trifluoroacetyl IR-camphorate[81,a selector-selectand system that was first conceived in 1971r20,2’1
(cf. Fig. 2):
Although this successful experiment demonstrated that
gas chromatography possesses the necessary thermodynamic and kinetic prerequisites for chiral recognition by
coordination interaction, this resolution principle has not
yet been generalized for olefins. The method, however, was
extended to oxygen, nitrogen, and sulfur-containing compounds by the employment of chiral bivalent transition
metal bischelates and led to a rapid extension of complexation gas chromatography for the resolution of enantiomers[’’*221(see footnote ‘**I on p. 747).
The methods available today for the quantitative gas
chromatographic resolution of enantiomers offer many
possibilities for their application and investigation. It is
appropriate here to differentiate between the information
which is accessible from the various peak parameters[22’
(Table 1).
Table 1. Gas chromatographic parameters in enantiomer resolution.
0
8
16
24
Peak parameter
Definition
Peak retention
A thermodynamic measure of the selective interaction
between the selector and the selectand ( K , -AGO)
A thermodynamic measure of the enantiomer discrimination between the enantiomerically pure selector and
the
racemic
selectand
(“chiral
recognition”)
(-AR,.S(AG’))
A quantitative measure of the enantiomeric composition
of the selectand (‘enantiomeric excess’ ( e . e . )[2])
Assignment of the absolute configuration of the selectand by correlation of the order of elution and the molecular configuration
A kinetic measure of the enantiomerization of the selectand during resolution (AG’)
Peak separation
Peak ratio
Peak assignment
32 min
Fig. I . Gas chromatographic resolution of a racemic amino acid mixture according to Gil-Au 1141.
Angew. Chem. I n ( . Ed. Enyl. 23 (1984) 747-765
Peak coalescence
(second kind)
749
The peak retention of the selectand is determined by the
(physical) partition equilibrium between the gas and liquid
phases and by the (chemical) association equilibrium
which arises when the selector exhibits chemical affinity
towards the selectand. In the case of complexation gas
chromatography the equilibrium constant for association
K and, for variable temperature studies, the Gibbs-Helmholtz parameters AH“ and A S ” are easily accessible from
relative retention data[231.
The peak separation of racemic selectands on chiral selectors is exclusively the result of the differences between
the free enthalpies of association, -AR,s(AGo). Although
there should be no doubt that there will always be a difference in the free enthalpies of association for each chiral
combination of selector and selectand, practical results are
only anticipated with the employment of optimized systems that exhibit effective chiral recognition. Gas chromatography is a n effective method of acquiring thermodynamic data of enantiomeric discrimination. In this respect,
the question of the proportionality of AGO and AR,,(AGo)
is of interest. The intuitive assumption that an effective enantiomer discrimination is only to be expected via a strong
chemical interactions turns out to be false and has been
disproved in many instances. The impeding of molecular
association, e.g. by steric effects, seems indeed to be advantageous for the discrimination of configurational isomers. Analogously to enzyme models, enantiomeric discrimination in chromatographic systems has been interpreted by Dalglzesh in terms of “three point interaction”
between selector and selectand (“three point rule”)[”! The
three point theory requires three effective bonds between
selector and selectand for a favorable interaction. Later on,
this model had to be altered so as to require a minimum of
three stereochemically important interactions for enantiomer differentiati~n[~].
Along with bonding interactions
(hydrogen bonds, charge transfer interactions, coordination) these can also include nonbonding interactions (steric
hindrance). Thus, in principle, it ought to be possible to resolve a nonfunctionalized racemic alkane by host-guest interaction with a chiral clathrate forming compound by
physical inclusion. The recently observed gas chromatographic resolution of racemic pinenes on an a-cyclodextrin
inclusion compound points in this interesting direction[”].
In contrast to enzymes, chiral catalysts, and chiral NMR
shift reagents, it is only possible for chromatography to
distinguish between enantiomers (“enantiomorphy”), but
not between enantiotopic groups or faces (“enantiotopy”).
For this reason enantiospecific selector/selectand systems
are only suitable model systems for the specificity of enzymes with respect to enantiomers or of receptors with respect to the enantiomers of pharmaceuticals, pheromones,
odorants etc. A comparison of enantiospecificity is then
only possible in a thermodynamic sense. The realm of enzymatic reactions, i.e. the reaction of prochiral educts to
yield chiral products, has no analogue in a gas chromatographic experiment, since the prochiral substrate cadnot be
discriminated nor can a chemical reaction occur which
leads to a chiral product.
The ratio of the peak areas is a direct measure of the ratio of the amounts of the enantiomers. The measured enantiomer ratio corresponds to the actual enantiomer ratio of
750
the sample when all the chemical, physical and analytical
manipulations, including injection and detection, are carried out in an achiral environment. A falsification of the
enantiomer ratio, if any, would only be possible as a consequence of an enantioselective decomposition or an irreversible adsorption of the sample by the optically active
stationary phase during chromatographic elution. The advantages of the gas chromatographic determination over
other methods for the determination of the enantiomeric
purity e.e. (enantiomeric excess)]
e.e. =
~
R-S
x 100%
R+S
have recently been discussed in detail by Schurigtzl.Typical
advantages are sensitivity of detection, high accuracy, and
reproducibility. The method is also independent of the chiroptical properties of the sample (e.g., the magnitude of the
specific rotation) and optically active impurities d o not interfere. Gas chromatography is particularly suitable for
two limiting cases: for the detection of small enantiomeric
excesses (e.g. during the investigation of the accumulation
of optical activity in the laboratory under abiotic conditions) and for the proof of very high enantiomeric purity
(e.g., for the determination of traces of enantiomers during
investigations of the enantiospecificity of enzymes, during
kinetic resolution experiments and enantioselective synthesis). When the separation factor is high, gas chromatography can even detect 0.1% of an enantiomer. Such a contamination corresponds, for an enantioselective reaction
from a prochiral precursor (e.e. =99.8%), to a difference in
free enthalpy of the diastereomeric transition state of
“only” AAGf = 4 kcal/mol at 18”C[21.
The derivatization-free resolution of enantiomers by
complexation gas chromatography is particularly suitable
for analytical applications, since this method allows the
determination of the enantiomeric composition directly in
the vapor phase without the necessity for substrate isolation. Thus, the course of an asymmetric reaction can be
followed by repeated determinations without interrupting
it, and, for example, alterations of enantioselectivity by autocatalysis or kinetic resolution of the product can be easily recognized[26’.
The peak assignment requires a knowledge of the molecular configuration of the selectand. Since gas chromatographic enantiomer resolution depends on the enantiospecific association between selector and selectand, it ought to
be possible to formulate rules for substrates that belong to
homologous series, which correlate the retention time with
the absolute c ~ n f i g u r a t i o n [ ~ ’ ~The
~ * ~ absolute
.
configuration can, if necessary, be determined with only
g of
substance e.g. for an enzymatic reaction by the co-injection
of a reference compound of known chirality. In the case of
an enantiomerically pure sample or reference substance, an
aliquot of the other enantiomer (or of the racemate) must
be coinjected for purposes of peak identification. The determination of the configuration of the selectand by gas
chromatography is independent of the chiroptical properties of the selector, which, in principle, is not required to
be either enantiomerically pure or of known configuration.
A n y e w . Chem. Int. Ed. Engl. 23 (1984) 747-765
A peak coalescence (of thefirst kind) is observed, as mentioned, if a racemic mixture is employed instead of the optically active selector. It is possible in this way to distinguish achiral diastereomers and other isomers from enantiomers, since only enantiomers coalesce. A peak coalescence (of the second kind) occurs during the chromatographic resolution when configurationally labile enantiomers invert. “Enantiomerization” of the selectand leads to an interconversion peak profile, from which kinetic parameters
of inversion, as a reversible first order reaction, can be calc ~ l a t e d [ ~ Whereas
~ , ~ ~ ’ . at slow inversion a plateau, which is
produced by the inverting molecules, appears between the
terminal peaks, a rapid configurational interconversion destroys resolution.
Because of its high resolving power gas chromatography
is particularly suitable for enantiomer analysis, while for
liquid-chromatographic enantiomer resolution the emphasis is on the preparative a s p e ~ t ‘ ~ ’ - ~However,
~’.
when the
separation factor is large, a > 1.3, gas chromatography may
also be employed for the preparative resolution of enantiomers in the mg to g sample range. The isolation of pure
enantiomers is suitable for chiroptical investigations, for
biological tests or, in the case of chiral radiochemicals, for
labeling experiments. In principle, it is possible to obtain
100% enantiomerically pure enantiomers even though the
selector may not be enantiomerically pure.
Gas chromatographic methods of enantiomer resolution
(see footnote [**’ on page 747) and their applications and
mechanisms will be discussed in the following section.
opment of high sensitivity detection systems that the conditions were met which led to the development of tailored
selector/selectand systems and their optimization to high
resolution performance. Mainly those enantiospecific resolution systems that yield quantitative (baseline) resolutions
of racemic mixtures under the instrumental conditions employed will be reviewed in the following discussion.
2.2. The Resolution of Derivatized Selectands on
Optically Active Amino Acid and Peptide Selectors via
Enantiospecific Hydrogen Bonding
In 1966 Gil-Av, Feibush, and Charles-Sigler described the
first reproducible resolution of 18 racemic, protein amino
acids as their N-trifluoroacetyl(TFA) alkyl esters in glass
capillaries coated with optically active N-TFA-L-isoleucine
dodecyl ester 1 or N-TFA-phenylalanine cyclohexyl
The design of this enantiospecific selector/selectand system was based on the idea of biomimetically
imitating the stereospecific peptide enzyme interaction employing simple amino acid entities as model substances. In
order to achieve this it was thought necessary to employ a
high-performance capillary column to amplify the minute
chiral recognition effect. This fundamental discovery by
Gil-Au et al. was the starting point of a fruitful development, particularly by the research groups of GiI-Av, Bayer,
and Kdnig, from a supposed curiosity with rather archaiclooking chromatograms~’21
into a highly developed routine
method for the resolution of the enantiomers of many important classes of compounds by gas chromatograp h ~ [’O1.~This
~ ~ development
.
was inspired by the multiplic2. Gas Chromatographic Enantiomer Resolution
ity
of
possible
applications
on
the one hand and by the invia Hydrogen Bonding
terest in the mechanism of chiral recognition-as a model
for the enantiospecific enzyme substrate interaction-on
2.1. Pioneering Investigations
the other.
The first claims of the resolution of racemic substrates
Gil-Av et al. ascribed the resolution to the rapid reversi(2-butano1, 2-bromobutane) on optically active stationary
ble molecular association of the enantiomers with the
phases (starch, diethyl tartrate/AI,O3) were made by Karaasymmetric molecules of the optically active stationary
In order to increase the various steric and pogounis and L i p p ~ l d ‘ ~Other
~ ’ . authors were unable to repeat
the resolutions, which were only marginal a n y ~ a y [ ~ o~ r, ~ ’ ] lar interactions between the diastereomeric associates,
both the derivatization strategy for the amino acids to be
were caused, in the case of 2-bromobutane, by dehydroharesolved and the structure of the stationary phase were varlogenation to E- and Z-2-butene ( I : 1) in the injection
ied systematically. The results have been summarized in
In spite of considerable instrumental shortcomtwo detailed review a r t i c l e ~ [ ~ ’ . ~ ’ ~ .
ings, later investigations by Karagounis and Lerr~perle[~’]
Scheme 1 includes the structures of the most important
pointed the way to interesting future developments, e.g.
optically active stationary phases for the enantiomeric repreparative racemate resolution using polarimetry as an
solution of amino acid derivatives and other nitrogen-conenantiospecific detection system or the use of chiral metal
taining racemates via hydrogen bonding.
coordination compounds ( A - ( C ~ ( e n ) ~ ) B r ~ / A as
l ~ oan~ )enThe “second generation” phases 2-4 introduced by
antiospecific stationary phase. An “inverse” system,
Gil-Av
and Feibush exhibit better enantiospecificity than
namely the gas chromatographic resolution of volatile, rathe amino acid phase 1 because of the presence of an adcemic chromium(r1r) tris-hexafluoroacetylacetonate on
ditional amido function capable of hydrogen bonding‘43pulverized d-quartz was also investigatedL3*].It has since
501 The first dipeptide phase tested, N-TFA-L-valyl-L-Valproved impossible to establish with certainty whether or
ine cyclohexyl ester 2, already proved capable of the prepnot a marginal degree of enantiomer resolution was
arative resolution of racemic N-TFA-alanine tert-butyl esachieved in these experiments. Attempts to resolve racemic
ter on a packed 2-m ~ o l u m n ‘ ~ ’The
’ . arsenal of dipeptide
alcohols (2-alkanols, 3-menthanols) on various optically
has been greatly enlarged by many investigaactive stationary phases (( +)-2-octyl sebacate, (+)-dimetionS[52-661.
thy1 tartrate, peracetylated sucrose) were without successL35, 391
A systematic investigation has been made of the influence of the N a n d C-terminal amino acids of N-TFA-OIt was only with the introduction of high performance
cyclohexyl derivatives of the dipeptides Val-Va1[671and
capillary columns for gas chromatography and the develAnyew. Chem. Int. Ed. Engl. 23 (1984) 747-765
751
iPr
amino acid phase
dipeptide phase
iPr
also permit preparative resolutions on packed colu m n ~ ~Racemic
~ ~ , ~p-~ and
~ . y-amino acid derivativesr7']
and derivatized 1,2- and 2,l-amino alkanols can also be
separated on diamide phases 3. Further applications of
diamide phases 3 concern a diamide-diamide selector/selectand system, i.e. the resolution of racemic N-TFAamino acid tert-butylamides on the N-lauroyl-tert-butylamides of optically active a-amino acids[751.a-Methyl-aamino acid amides and a-halogenocarboxylic acid amides
have also been resolved on diamide phases 3[76,771.
A concise summary of the synthesis and application of diamide
phases in gas chromatographic enantiomer resolution is
4
3, R' = z.B. n-C,,H,,
The N,N'-carbonyl-bis(amino acid ester) phase 4 (incorrectly named "ureide" phase in the original literature) is
particularly suitable for the resolution of racemic N-TFAsubstituted amines RR'HC-NH2[431. These phases contain
two structurally equivalent N H- and CO-functions capable
n - ~ 1 1 ~ , H3 - ~=- ~ - C -5~
of hydrogen bonding in close proximity to the asymmetric
? H3
center. The influence of the consistency of the "ureide"
a m i d e phase
phases (liquid, liquid crystalline, solid) on the separation
factor has been investigated in great detai1179-821.
Stationary
Scheme I.Representative structures of optically active amino acid and amide
stationary phases for the resolution of enantiomers via hydrogen bonding.
phases in the solid state exhibited large separation factors
a, but also a considerable amount of peak broadening,
particularly for the amino acid enantiomers that were
eluted as the second peak[7y1.In spite of the anomalous
peak shapes, the high separation factors a. that could be
P h e - L e ~ [ ~ ~ . "on
' ] the order of elution of N-TFA-amino
acids. An expected peak inversion was observed with the
achieved with the mesophases permitted quantitative resomirror image phases L-Val-L-Val and D-Val-D-Val. On
lutions on short packed columns'"'. N,N'-Carbonyl-bis(Ldiastereomeric phases, e.g. L-Val-L-Val as compared to Dvaline isopropyl ester) 4 and the D-kUCine analogue have
Val-L-Val or L-Val-D-Val, peak inversion only occurred on
been employed in the smectic mesophase for the resolution
changing the configuration of the N-terminal amino acid.
of the racemic amines (N-perfluoroacylated l-phenylethylHence the enantiospecificity of dipeptide phases is largely
amines)[821.The close temperature range lying 10-20" begoverned by the N-terminal amino acid. This was also conlow 100°C seems disadvantageous.
firmed by the finding that the dipeptide phase Gly-L-Val
It was demonstrated by means of phase 5 (N-acylated
possesses significantly inferior separating properties than
chiral I-(1-naphthy1)ethylamine) that the presence of an
~-Val-Gly[~'].
Accordingly the employment of tripeptide
amide function and a single asymmetric center is sufficient
phases, e.g., N-TFA-L-valyl-L-valyl-L-valine
isopropyl ester
for the enantiospecific resolution of racemic amides, e.g.
did not bring about any improvement in the separation
N-acylamino acid esters, a-methyl-a-amino acid esters, aliphatic N-acylamines and a-substituted carboxylic acid
factor but merely decreased the volatility as compared to
amide~['~I.Optically active stationary phases have also
dipeptide p h a ~ e s [ ~ ~ The
. ~ ' l .diamide phases 3 developed by
Gil-Au and F e i b ~ s h are
~ derived
~ ~ ~ from
~ ~ the
~ peptide
~ ~ - ~ ~ been
~
synthesized by incorporation of a chiral acyl component in 5 , e.g. (lR,3R)-tran.~-chrysanthernicacid 6, O-lauphases 2 by the removal of the C-terminal amino acid,
roylmandelic acid, and N-lauroylpr~line~'~~~~~,
which allow
which does not invoke any significant contribution to enantiomer discrimination. Phases of type 3 contain two
H3C CH3
thermally stable amide functions capable of hydrogen
bonding, but only one asymmetric center, and are more enH3C,.)iOOH
antiospecific than type 2 phases. A systematic investigaH
tion has been made of the influence of the alkyl residues
H3C
R ' and R2 of the valine diamide phase 3 on the separation
the quantitative resolution of underivatized (+)-menthol,
factors of N-TFA-a-amino acid alkyl esters(5o! The alkyl
nitrile compounds, cis and trans-chrysanthemic acid
residues R' and R2 influence the resolution because of
6[86.871,
macrolide lactones1881
as well as the improved resotheir steric requirement, whereby R2 plays a dominating
lution of racemic N-TFA amines and carboxylic acid tertrole. It was found to be advantageous if R1 constitutes the
butylamides.
long chain residue necessary for thermal stability (nIn order to improve temperature stability Oi et al. incorC , 1H23).A decrease in the resolution in the order tert-butyl
porated the chiral structure elements shown in Scheme 1
> cyclooctyl > neopentyl was observed for R2. The therinto a triazine skeleton 7 via the N-terminud*'] and remally stable (up to 190OC) and less volatile phases 3,
solved, for instance, proline as the tert-butylamide[9n1, diR2= 1,I-dimethylhexade~yl~~~~
are very promising resolupeptides (e.g. all four configurational isomers of N-TFAtion phases for racemic amino acid derivatives; one phase
alanylalanine isopropyl ester)[''], O-acyl-a-hydroxycarboxis commercially a~ailable'~''.
The high separation factors
or n-CZ1H4,
R' s e e text
diamide phase
"ureide" phase
Qo
F I =
152
Ang(,w. Cliem. 1nf. Ed. Engl. 23 (1984) 747-76s
c f. Scheme 3
D i r e c t s y n t h e s i s -+ .--, .+
I
I
ylic acid esters["] and underivatized a l ~ o h o l s l into
~ ~ ] their
antipodes.
The observation that nitrogen-free racemic O-acyl-a-hydroxycarboxylic acid esters could be resolved into their antipodes via hydrogen b ~ n d i n g ~ ~ ~ led
. ~ "to
. ~the
~ ' ,development of oxygen-containing, amide-free stationary phases,
e.g. di-( -)-menthy1 (+)-tartrate for the resolution of racemic N-acylamino acid esters, N-acylamines and carboxylic acid a m i d e ~ l ~The
~ ] . results confirmed previous findings of a low degree of chiral recognition between racemic
alcohols and didodecyl (+)-tartrate in a gas chromatographic
Nitrogen-free 2-hydroxycarboxylic
acids were also resolved as 0-TFA alkyl esters on S-rnandelic acid S-a-phenylethylamide[96~971.
The disadvantage of
these phases is their low thermal stability. The resolution
of racemic, underivatized alcohols on deactivated glass capillary columns, coated with optically active polypropylene glycol (PPG) is remarkable in spite of the low separation factors ~ b t a i n e d l ' ~ , ~ ~ ] .
o=C O=C-NHtRu
p o l y m e r A, 8
( = Chirasil-Val)
I
HN-*C-H
iPr
Y
0-c-Cl
10
p o l y m e r R, 11
2.3. Optically Active Polymeric Siloxane Stationary
Phases
The employment of low molecular weight stationary
phases 1-5 (Scheme 1 ) is limited at high temperatures by
thermal decomposition, racemization, and column bleeding. This precludes the resolution of many less volatile racemates in this manner. Only carefully purified amide
phases 3 which are modified with long-chain substituents
(e.g. R ' = 1,l-dimethylhexadecyl, R2=n-CzlH4,) can be
employed u p to 200°C due to their low polarity and volatility17']. The lower molecular weight stationary phases tend
to crystallize at lower temperatures.
A new type of enantiospecific stationary phase was developed by Frank, Nicholson, and Bayer, where the amino
acid amides which form the basis of the low molecular
weight diamide phase 3 were attached via the free amino
function to the carboxyl group of a polymeric matrix, a
statistical copolymer of dimethylsiloxane and (2-carboxypropy1)methyl siloxane, having a high viscosity['0"-'021.
This stratergy enabled the chemical selectivity of the chiral
component L-valine-tert-butylamide to be combined with
the high thermal stability and involatility of organic polysiIo~anes"~~].
The optically active polymeric siloxane 8 (polymer A,
Scheme 2), which is available c ~ m m e r c i a l l y ~
under
' ~ ~ ~the
name Chirasil-Val@,is characterized by high thermal stability and low column bleeding and, as a consequence, allows the resolution of enantiomers in temperature programs ranging from 70-240°C. The copolymer component dimethylsiloxane, whose proportion can be varied between wide limits, serves to increase the viscosity and to
regulate the mean distance between the chiral side groups;
the highest separation factors were observed for a separation of about eleven dirnethylsiloxane units"01. The resolvAnqew. Chem. Int. Ed. Enql. 23 (1984) 747-765
::
I
C=O
n-CSHll-C-HN~~-II
i$r
Scheme 2. Entry to optically active polysiloxanes for gas chromatographic
enantiomer resolution.
ing properties deteriorate with closer spacing because the
functionalized polysiloxane tends to crystallize due to the
interaction between neighboring peptide side chains. If the
spacing is increased, then the cooperative effect between
two chiral selectors for enantiomeric discrimination of the
selectand begins to decrease. Incidentally, the mean distance bears no relationship to the actual molecular statistical distribution which has been calculated iteratively"05!
This analysis revealed that not all of the chiral centers can
contribute to the enantiomer resolution"05! It ought to be
153
mentioned that Chirasil-Val 8 contains an additional chiral center in the (racemic) side chain.
The synthesis of 8 is outlined in Scheme 3[Io6].The coating of deactivated glass capillary columns with 8 enabled
the rapid baseline resolution of all 17 racemic protein amino acids for the first time as their N-pentafluoropropionyl
isopropyl esters['07.'"'I (see Fig. 3).
7%
CI-$i-H
i.1
y 3 3
HzFiC16
+ CH,=CH-COOR
7H3
7H3
Cl-Si-CH,-CH-COOR
c1
17
Hi0
(OH')
Me
I
-Si-0I
CH,
I
Me-CH
kOOH
18
h'Ie-$H
COOTI
19
+
O=C-NHfRu
H,N-C-H
DCC
8
f
ipr
Scheme 3 . Four-step synthesis of 8 (Polymer A). I . Synthesis of the dichlorosilane monomer 17. 2. Synthesis of the carboxyalkylmethylsiloxane 18. 3.
Synthesis of the copolymer 19. 4. Coupling of 19 with Val-NHtBu in presence of dicyclohexylcarbodiimide (DCC) to give S.-Derivatives of other
a m i n o acids and peptides can react analogously with 19.
Even the analysis of the involatile amino acids Met, Phe,
Glu, Tyr, Orn, Lys and Trp does not present any difficulties after suitable derivatization. The analysis of racemic
drug metabolites is also possible employing pre-treated
glass and, more recently, fused silica capillaries which
have been coated with Chirasil-Val 8[108,1091.
The range of
applicability of 8 extends from the resolution of the racemates of aryl glycol esters, atropisomeric binaphthyl derivatives 20['"'], an underivatized diketone (threo-3,4-diphenyl-2,5-hexanedione, 21)[1051,alkyl 2-hydroxycarboxylates[""~"'], the nerve gas Soman 22 (four isomers)["'], amino acid s u l f ~ x i d e s " and
~ ~ ~the
, methyl ester of phosphinotricin 23['751
which, like 23, possesses an asymmetric phosphorus atom as well as a chiral carbon atom.
20
&Me,
vv
LU
/
CqF;
*
Me
*I
O=P-CH2-CH2-CH-COOH
I
OMe
I
23
NH2
The high thermal stability and the low tendency to column bleeding of Chirasil-Val 8 is of prime importance for
combined gas chromatography-mass ~ p e c t r o ~ c o p y [ ' ~ ' ~ .
The incorporation of phenyl groups['"] and polar modifiers (cf. 24) in 8 allows the tailoring of the enantiospeci154
ficity and polarity of the chiral phase such that all protein
amino acids can be resolved quantitatively without peak
overlap["41. (S)- or (R)-a-phenylethylamine and (S)- or
(R)-I-( 1-naphthy1)ethylamine have also been employed as
additional chiral components as the amide substituent of
L-valine (or L-leucine). The diastereomeric polysiloxane
with L-valine-(R)- I -(1-naphthyl)ethylamide moieties was
able to accomplish a partial resolution of the 0-TFA derivative of the alcohol 1 -cycl~propylethanol["~~.
Furthermore, the amine (R)-1-(I -naphthyl)ethylamine has been
coupled to a polysiloxane and employed to resolve the
phenyl or methylurethane derivatives of alcohols into their
enantiomers for the first
Verzele et al. have described a preparatively simple route
to chiral polysiloxanes[' '', ' 16]. Thus, the cyano groups of
commercially available polysiloxanes of type 9['031, e.g.
OV-225 or Silar-1OC are hydrolyzed and, after conversion
into the acid chloride 10, the carboxyl groups are coupled
with the free amino function of ~-valine-tert-butylamide[~~l
to yield polymeric diamide 11 (Polymer B, Scheme 2). KOnig et al. prepared a variant of Polymer B from the commercially available XE-60 polysiloxane 12, e.g. the diamide 14 (Polymer C, Scheme 2), which, in analogy to modified Chirasil-Val["41, contains ( R ) - or (S)-I-phenylethylamine as an additional chiral component in the amide function of the L-valine selectori"7-119! In the case of a diastereomeric selector the relative configurations of the chiral
centers is decisive for an optimal enantiomer discrimination and this must be elucidated empirically. Both polysiloxanes 14 (Polymer C ( I ,R) and (L,S)) have been synthesized and their complementary resolving power has been
compared by means of two capillary columns, each packed
with diastereomeric phasc1l2''.Thus, polymer C (L,R) is especially suitable for the resolution of N-TFA-amines and
N,O-bis(TFA)-amino alcohols, while polymer C (L,S)can
be employed for the resolution of racemic N-TFA-amino
acid isopropyl esters[9,12')-1221 . Polymer C (L,S)has also
proved valuable for the separation of constitutional and
configurational isomers of carbohydrates[' 17, 'Ix1.
Polymer
C (L,S)has become known as XE-60-~-valine-S-a-phenylethylamide['231.
Another synthetic strategy consists of the reduction of
the cyano groups of OV-225 9 with LiAlH4 in diethyl ether
and reaction of the resulting amine with the carboxy function of N-acyl-L-valine (or N-acyl-L-leucine) to form the
polymeric diamide 16 (Polymer D, Scheme 2)["91. While in
Polymers A-C the macromolecular residue formally constitutes the acyl component, in Polymer D it formally functions as the amide component. Polymer D is particularly
suitable for the resolution of racemic N,0-bis-acylamino
alcohols and N-acyl amines["9r.
One advantage of the direct synthesis of chiral polysiloxanes of the type 8 lies in the manifold possibilities for
variation of copolymer composition, e.g. in the deliberate
inclusion of polar modifiers, as in 24["41.
Commercial polysiloxmes with varying amounts of cyanopropyl groups were employed for the synthesis of polymers B-D" 15,117,1191 which, after hydrolysis and coupling with established chiral selectors, yielded polysiloxanes with similar properties to those of Chirasil-Val 8"'l or
its variants" "I. However, the disadvantage of these polA n y c w . Chem. Int.
Ed. Engl. 23 (1984) 747-765
ymer analogues is that the strongly acid or basic conditions necessary for hydrolysis of the cyano groups may
also bring about some depolymerization, so that the relationship between the copolymers is uncontrollably altered
and hence polarity and viscosity are affected"431.The acid
chloride group in 10 could also cause problems in the preparation of Polymer B, and the depolymerizing properties
of LiAIH, must be taken into account in the synthesis of
Polymer D[1431.
+OOR'
OH
24
Figures 3--5
typical gas chromatograms featuring analytical enantiomer resolution. Table 2 provides
Fig, 5. Enantiomer resolution of racemic n-hydroxy acid esters on ChirasilVal 8 at 50°C [ I lo]. R'=CH(C2H5),.
Fig. 3. Resolution of enantiomers of racemic protein amino acids as N-pentafluoropropionyl isopropyl esters on Chirasil-Val 8 [ 1751. OH- and SH-functions were likewise protected with the pentafluoropropionyl group.
m
a,
C
m
u
m
-ax
m
c
m
m
C
c
a
u
m
u
S
0
0
.C
.-C
E
E
T
mI
N
N
f
x
5
5
a,
0)
m f
! 5:
In
I
mm
S
l
C
m
c
C
m
a
0 0
0
E
E E
2
N
N N
0
._
T
information concerning the classes of compounds, derivatization method, and optically active polysiloxane employed in the gas chromatographic resolution of enantiomers via hydrogen bonding.
m
C
.t.t
??
2.4. Derivatization Strategies
c
N
I
LO
30
20
time (min)
10
Fig. 4. Enantiomer resolution of racemic amines RMeHC-NH2 as their isopropyl u r e a on Polymer C 14 (L,S) [125].
Angrw. Chem. Inl. Ed. Engl. 23 (1984) 747-765
0
The gas chromatographic resolution of enantiomers via
hydrogen bonding must almost always be preceded by a
derivatization of the racemic selectand (cf. Table 2 and
[ l37,138]
). The derivatization is necessary, on the one hand,
to transform involatile and/or polar substrates (e.g. amino
hydrogen bonds and increasing the chiral recognition of
the selector/selectand system. In this way the introduction
of a chemically stable amide function can lead to the formation of additional hydrogen bonds with the optically ac-
I55
Table 2. Resolution of enantiomers by gas chromatograhy on chiral polymers A-D (see Scheme 2). A selection of classes of compounds and derivatization.
Class of compound
Derivatization
n-Amino acids
N-Methyl-tr-amino acids
Amino alcohols
Amines RR'HC-NH2
Amines RR'HC-NH2
Hexobes, pentoses
Hexoses
Cyclic polyols
a- Hydroxycarboxylic acids
a-Hydroxycdrboxylic acids
13-Hydroxycdrboxyl acids
Phenyl glycols
Atropisomeric binaphthyls
Sulfur-containing amino acids
Sulfoxides (methionine-S-oxide)
cec- Alcohols
I,2-Diola, 1.3-diols
Ketones
Halocarboxylic acids
Pharmaceuticals
N-perfluoroacyl alkyl ester
N-alkylureido alkyl ester
N . 0-bis-(trifluoroacetyl)
N-isopropylcarbamoyl
N-trifluoroacetyl
0-trifluoroacetyl glycoside
boronate
0-trifluoroacetyl
0-carbamoyl alkyl ester
alkyl ester
N-(tert-butylcarbamoyl tert-butylamide
bis(pentafluor0propionate)
bis(pentafluoropr0pionate)
N-pentafluoropropionyl methyl ester
none (perfluoracyl alkyl ester)
urethanes
cyclic carbonates
syn. anti-oximes
amides
various
tive selector. The design of an appropriate derivative can
also lead to greater detector selectivity or sensitivity (e.g.
by element-selective detectors).
The derivatization process should fulfill the following
criteria:
1. quantitative reaction;
2. absence of side and rearrangement reactions;
3 . exclusion of racemization.
For quantitative analysis, the yield of the derivative and
its separation from solvent and derivatizing reagents must
be reproducible. It is particularly necessary that derivatization reactions proceed without racemization, if the enantiomeric composition ("enantiomeric excess", e.e.) is to be
determined gas chromatographically. The proof of absence
of racemization is often not easy to perform.
The introduction of thermostable chiral polymer phases
has widened the scope of derivatization strategies, since
less volatile derivatives can now be analyzed. Thus, the
often low separation factors can be compensated at high
temperatures by increasing the enantiospecificity between
the derivatized selectand and the selector with the advantage of shorter analysis times. Gas chromatographic derivatization methods are well established; recently, derivatization has been automated for quantitative amino acid
analysis~
I 39. I401.
The classical method of derivatization of a-amino acids
is their conversion into N-perfluoroacyl alkyl esters; OHand SH-functions are likewise protected with perfluoroacyl g r o ~ p s ~The~tert-butyl
~ ~ esters
~ ~ exhibit
~ ~the"
best properties as far as enantiomer resolution is concerned. However, since these are only accessible with difficulty, the isoalkyl esters (e.g. isopropyl esters) are preferred. Trifluoroacetyl (TFA), pentafluoropropionyl (PFP)
and heptafluorobutyryl (HFB) are employed as perfluoroacyl residues[54.'421. TFA derivatives exhibit higher separation factors, while PFP derivatives are more volatile.
The derivatization of a-hydroxycarboxylic acids was
first performed by introducing amide functions, e.g. as the
756
Polymer
Reference
A,B,C,D
[IOO, 115, 116, 120, 1361
[ 1241
[109, 118, 1191
[I251
[ I 18, 119, 1261
[117, 118, 1271
[ 1281
[I261
[ I l l , 1251
[I101
[ 1241
[105, 109, 1751
[ l o & 1751
[I871
[147, 1751
[114, 1291
[I301
[I311
[I881
[lo& 109, 116, 132-1351
c(LS)
C(L,R),D
c(LS)
C (L,R),D
c( L a
A
C(L,R)
A, C ( L S )
A
C @S), A
A
A
A
A
A, C (LS)
C(L,R)
C(L,S)
A
A,B,C@,R)
0 - P F P cyclohexylamide for lactic acid['''', or as the O-carbamoylalkyl ester["
Later resolutions were also
achieved with nitrogen-free hydroxycarboxylic acids, e.g.
as 0-acyl
'''1
or even as free alkyl esters[92, 1 1 0 , 1 1 I ] . Th e peak tailing caused by the presence of a
free hydroxy group could be suppressed by employing a
fused silica capillary column[""1. The conversion of the
easily racemizing mandelic acid into the 1-ethylpropyl
ester (10% HC1, llO"C, 1 h) proceeded without racemization. In contrast a considerable degree of racemization was
observed in the case of mandelic acid during the threestage derivatization of 2-hydroxycarboxylic acids to O-acylated amidesl"O1.The conversion of the hydroxy groups of
2-hydroxycarboxylic acids into urethane by isocyanate is
neither quantitative nor free from by-product formation.
The resolution of acylated secondary alcohols~'4~88~1141
IS
'
difficult. An improvement is achieved by chromatography
on optically active polymer stationary phases as the urethane
1291. Phosgene has recently been suggested as a derivatizing agent for 1,2- and 1,3-diols, a-amino alcohols, a-hydroxy acids, and N-methylamino
while isocyanates (see Scheme 4) can be employed to derivatize both functional groups of hydroxy
and amino acids s i m u l t a n e ~ u s l y ~ ' ~The
~ ~ . isocyanate
method is particularly advantageous for the derivatization
of N-methyl-a-amino acids 30, since on acylation these
react with complete racemization to yield alkylidene oxazolidin-5-0nes~'~~~.
Oximes have been employed for the resolution of ra~ cemic
~ ~ ketones[1311.
~ ~ ~ The
~ occurrence
~ ~ ~ of~syn-~ and
. anti-isomers
may be regarded as a complication. Interconversion of the
isomers during separation, which takes place in the case of
2,4-dinitrophenylhydrazones of carbonyl compounds,
leading to characteristic interconversion peak profiles"461,
has not been observed for oximes. Racemization has also
been excluded for enantiomerically pure camphor and fenchone during derivatization to their o ~ i r n e s [ ~ ~ ' l .
For practical purposes enantiomer resolution without
resorting to derivatization procedures, as practised in comAngew. Chem. Int. Ed. Engl. 23 (1984) 747-765
P
0
R"
R-&H-OH
2
9OH R
-
2
28
8''
R'-CH-O-S-NHR
0
HE-NH-CH-C02H
?
RNH-c-O-CH-CH,-C,
I
R'
30
F: T
40
RNH-C-N-CH-C\
I
CH,
NHR
40
NHR
Scheme 4. lsocyanates (R = Me, iPr, IBu, Ph) as derivatizing reagents for a- 25 and p-hydroxycarboxylic acids 29,2-alkylcarboxylic acids 26 (each IO0"C. 60 min), amines 27 (20"C, 30 min),sec-alcohols 28 (IOO"C, 10 min), and N-methyl-namino acids 30 (IOO"C, 60 min) [144].
plexation gas chromatography". ''I, is advantageous. Only
in isolated instances have underivatized selectands, e.g. 1 phenyl-2,2,2-trifluor0ethanol~~~~,
3,4-diphenyl-2,5-hexanedione 21 [lo5],s u l f ~ x i d e s [ 'barbiturates['321
~~~,
and photodimers of c y c l o a I k e n o n e ~ [ ' ~been
~ ~ , resolved on dipeptide
phases and chiral polysiloxanes. The separation factors are
often very small since no suitable functions are available
for strong hydrogen bond formation. Complexation gas
chromatography, which was designed following the realization that many racemic selectands d o not possess functions suitable for derivatization (e.g. ethers, esters, lactones, unsaturated hydrocarbons) is particularly suitable
for the resolution of underivatized alcohols and
(see footnote ["I on page 747).
2.5. Thermodynamic Parameters of
contributions to retention, the decisive chemical contribution of the enantiomer differentiation - A(AGo) is generally underestimated when using Equation (2). When a selector is used that is only partially optically enriched then
knowledge of the degree of enantiomeric purity ( e x . ) can
be employed to obtain the maximum separation factor
valid for quantitative enantiomeric purity by mathematical
The temperature dependence of the separation factor
a,,, can be employed to calculate the Gibbs-Helmholtz
parameters -AR,S (AH") and AK,S(ASo)of chiral recognition according to Eq. (3) and (4):
(3)
(4)
Gas Chromatographic Enantiomer Resolution
Gas chromatography retention parameters are an important source of data for the calculation of thermodynamic
association constants of molecular complexation (see the
literature cited in [231 and
The difference in the free
enthalpy of association for an enantiomer pair-AR,S(AGo) can be calculated from the difference in the
association constants for the antipodes of the selectand
and the chiral selector according to Eq. (1) ( R refers arbitrarily to the later and S to the earlier eluting enantiomer,
respectively):
The difference between the free enthalpies of association
of selectands, which are resolved via hydrogen bonding
with dipeptide and polysiloxane stationary phases, can be
calculated from the difference in retention of the enantiomers via the separation factor amax~so~80~x'1:
- AK,..(AC0)=RTln(tL/t.k)=
RTlna,,,
(2)
The separation factor a,;,, represents the relationship between the net retention time t' (measured from the air or
methane peak, f M =dead time) of the more strongly and
more weakly retained enantiomer. Since the net retention
time is determined by both the physical and the chemical
Angew. Chem. I n t . Ed. Engl. 23 (1984) 747-765
According to Equation (3) a plot of R Ina,,,, against 1/Tis
linear, whose slope is the difl'erence between the enthalpy
of association and whose intercept is the difference of the
entropy of association of the highly dilute enantiomers in
the stationary phase. The lower limit for enantiomer differentiation, which still yields complete peak separation in a
high-performance capillary column can be estimated as 10
cal/mol for - AH,S(AGo) corresponding to a separation
factor a = 1.01 at 25 OC['].
Values of -A(AGo) in the range 40-580 cal/mol were
measured for the gas chromatographic resolution of N TFA-2-amino alkanes on the "ureide" phase 4 (Scheme
1)[801. As would be expected a linear relationship was observed between I n a and 1/T for the resolution of the enantiomers of N-perfluoroacylnorleucine alkyl esters on the
dipeptide phase N-TFA-phenylalanyl-L-leucinecyclohexyl
esters[471.The Gibbs-Helmholtz parameters of the enantiomer discrimination between racemic N-TFA-alanine
and leucine esters and the diamide phase 3, R ' =n-CIIH23,
R2 = CH2tBu, have been determined""]. The highest values
for the selectand N-TFA-leucine ( I -ethylpropyl ester) were
A(AH")= - 1.61 I kcal/mol, A(AGo)= -338 cal/mol, and
A(ASo)= -3.32 cal mol-' K ' at I10"C.
When chiral polysiloxanes are employed as stationary
phases for enantiomer resolution the physical contribution
of the achiral polymer network to the net retention time t'
cannot be ignored. Otherwise, the actual contribution of
the chiral side chain to enantiospecificity as calculated
157
from the ratio tk and t.4 will be underestimated. The separation of the physical and chemical contribution to the retention time can be attempted by employing a polysiloxane
reference column (e.g. SE 30)[“”].
The temperature dependence of the retention times of
racemic aromatic diol diesters has been studied in order to
determine the thermodynamic data for enantiomer discrimination on Chirasil-Val 8 (Polymer A)[’051.Enthalpy
parameters were obtained from the approximation equation (5) (t‘= net retention time, tM =dead time):
For gas chromatographic enantiomer resolutions reported
up to now, a decrease in the separation factor a has always
been observed on increasing the temperature. Since the
separation factors are linked to changes in the free enthalpy of association, and the latter depends on the interplay between enthalpy change, entropy change, and temperature [Eq. (6) according to the Cibbs-Helmholtz equation, at very high temperatures the separation factors may,
in principle, increase again with simultaneous reversal of
the order of elution[’051.
Thus, the determination of the Cibbs-Helmholtz parameters provides a criterion for a more precise understanding
of the mechanisms of chiral recognition. Gas chromatography constitutes an important tool (which has only been
employed in isolated instances) for a deeper insight into
the mechanisms and origins of enantiospecificity, which is
a common phenomenon in living matter, with preference
for the image over the mirror image.
2.6. Mechanistic Aspects of
Gas Chromatographic Enantiomer Resolution via
Hydrogen Bonding
The design of the first selector/selectand system for the
gas chromatographic resolution of enantiomers on the basis of simple amino acid entities was inspired by the
known stereospecificity of enzymes[41.The observed enantiomer discrimination was attributed to the formation of
diastereomeric associates between the racemic selectand
and the optically active s e l e c t ~ r [ ’ *In~ their
~ ~ ~ first
.
interpretation Feibush and Gil-Au suggested that association complexes were formed rapidly and reversibly by hydrogen
bonding between carbonyl and amide functions[441
(Scheme 5).
Depending on the configurations of the selector and the
selectands, complexes possessing differing stabilities are
produced during combination to diastereomeric associates.
This thermodynamic expression of the enantiomer discrimination, caused by chemical selectivity, depends on a
rapid establishment of the chemical equilibrium. This condition is fulfilled for hydrogen bonding even when preequilibria are i n v ~ l v e d ~ ’ ~ ~ ’ .
A kinetic control of the enantiomeric resolution process
can also be envisaged and could be recognized, on the one
hand, by peak broadening and, on the other, by dependence of the retention time on the flow rate of the mobile
phase.
Such effects have apparently been observed when liquid-crystalline stationary phases were employed1x’.821.
The
sharp peak elution observed in the overwhelming number
of cases allows the exclusion of kinetic effects.
While only two hydrogen bonds can be formed in the association of N-TFA-amino acid alkyl esters with phase 1,
phases 2-4 can enter into a further association (due to
the presence of an additional amido function (cf. Scheme
1).
The order of elution ol‘ amide and amino acid derivatives on “ureide” phases has been interpreted by Feibush et
al.[271by employing the Crum model, in which the steric requirement of the substituents ( I > m > s) at the asymmetric
center is employed to explain enantiospecificity.
Enantiospecific interactions other than hydrogen bonds
can also lead to enantiomer discrimination. Thus Stolting
and K0nig[641observed a racemate resolution of N-TFAproline esters on N-TFA-I.-prolyl-L-proline cyclohexyl ester as stationary phase, a selector/selectand system which
does not possess amide hydrogens capable of hydrogen
bonding.
The original model of a dimeric association complex,
comprising three hydrogen bonds[471,was abandoned by
Beitler and FeibushL5’I in a systematic investigation of the
variations of residues R ’and R2 in the diamide phase 3.
More recent models take into account the common knowledge of the structure and conformation of pep tide^'^,^'^.
Assuming a planar trans-conformation 31 for the amide
function, diamides at high dilution can form a planar fivemembered ring (C, unit) or a puckered seven-membered
ring (C, unit) by intramolecular hydrogen bonding (see
32).
---_c
“
______
3
However, under the conditions of a gas chromatographic experiment oligomerization by intermolecular hydrogen
bonding with the formation of “Cs-C5”, “C7-C7” and
“C,-C,” associates is assumed in the molten diamide (cf.
Fig. 6).
The P-pleated sheet structure of peptides constitutes a
typical example of this interaction. According to Figure 7
associates of ring size 5 + 5 = 10 and 7 7 = 14 are formed
when the chains are arranged parallel; antiparallel arrangement leads to associates of ring size 5 + 7 = 7 + 5 = 12.
Numerous spectroscopic and structural investigations on
+
Scheme 5 . Hydrogen bonding between selector and selectand.
758
Anp’w. Chem. Inr. Ed Engl. 23 (1984) 747-765
.
Q+$p
a
L-
0-pleated s h e e t
&-helix
R-
1
9
t
0
=
Fig. 6. Bifunctional coordination possibilities Ibr hydrogen bonding in peptides [105].
N-acyl-a-amino acid alkylamides have confirmed these association mechanisms. It is important to note that (under
the conditions of the gas chromatographic experiment) the
preferred structure of the diamide in the crystalline state is
largely preserved in the melt.
L.L'
L, D '
Fig. 8. The mechanism of association between amino acid derivatives and
diamide phases [SO].
A
the association of the N-TFA isopropyl ester of an L-a-amino acid with the diamide N-acyl-L-leucine-tert-butylamide
and analogues (Fig. 9)[781.Here too the differing steric situation of the residue R and the isobutyl group, which are aligned parallel in the case of amino acids of the same configuration, leads to differing stability for the associate and is
assumed to be responsible for chiral recognition. All amino acid derivatives which possess a primary amino group
(except proline) are resolved into their enantiomers by the
same mechanism since the elution of D- and L-form is
identical in all cases (see Section 2.7).
B
Fig. 7. Ij-Pleated sheet structure of polypeptides (A: parallel chains, B: antiparallel chains) [78]. The numbers signify ring size.
According to Beitler and F e i b u ~ l i ' ~a ~selectand
~,
molecule associates with a diamide selector via two hydrogen
bonds forming a 5 5 = 10 ring in an analogous manner to
the P-pleated sheet structure of polypeptides (cf. Fig. 8).
The differing positions of the isopropyl group of valine in
the diamide phase and the amino acid residue R3 of the selectand in the diastereomeric associates L,L' and L,D' may
explain the observed enantiospecificity, whereby it has
been demonstrated experimentally that the L,L' combination is more stable than the L,D' one'501.Later, an association by means of 7 7 = 14 rings was postulated to explain
+
+
Angew. Chem. Int. Ed. Engl. 23 (1984) 747-76.5
Fig. 9. Association of N-TFA amino acid isopropyl esters and a diamide of
type 3 (781.
This model is able to explain the following experimental
res~Its[~~"J:
1. The diamide phase N-lauroyl-N-methyl-L-leucine-tertbutylmethylamide, which does not possess any free am7 59
ide functions, does not exhibit any enantiospecificity
for amino acid derivatives. This underlines the importance of hydrogen bonding for enantiospecificity.
2. Proline is intrinsically incapable of forming more than
one hydrogen bond and, hence, is very difficult to resolve into its enantiomers. The transformation of the
carboxy group to an amide 33[75,901,
however, allows an
effective resolution of the enantiomers because of the
presence of a C, unit[78"'.
1
Abb. 10. Hydrogen bond formation along a 5-A translation axis in ( R ) - N lauroyl-I-( 1Lnaphthyl)ethylamine ( R ) - 5 [ 150).
3. Only the N-terminal amino acid of dipeptide phase 2[441
is capable of forming a "C," unit with associating selectands. Indeed, the C-terminal amino acid does not participate in enantiomer d i s c r i m i n a t i ~ n [ ~ ' . ~ ~ ~ .
According to K~ppenhOjer"~~]
the observed enantiospecificity of Chirasil-Val 8 for bifunctional oxygen-containing selectands containing no amide functions can, according to Figure 6 , on1 occur with a conformation of the valinamide selector in which both amide groups are directed towards the substrate (a-helix type), while a p-pleated sheet
structure is to be expected for nitrogen-containing selector~['~'].
Since models only reflect the major part of an experimental finding, other association mechanisms can also
come into play. For example, association by acyclic hydrogen bonding, van der Waals interactions, and dipole dipole interactions can also be considered. The multiplicity
of possible enantiospecific selector/selectand interactions
is clearly demonstrated by the resolvability of N-TFA-proline esters and N - a c y l a r n i n e ~ [ ~of~ ~2-hydroxycarboxylic
,
acid alkyl esters[961,and a-chlorocarboxylic acid a m i d e ~ [ ~ ~ ]
on diamide phases.
Apart from I : 1 association complexes there is also the
possibility of 1 :2 interaction, whereby the selectand,
Figure 11 illustrates the postulated p-pleated sheet strucwhich is employed at great dilution, is sandwiched beture of an associate between Chirasil-Val and intercalated
tween two molecules of the undiluted selector which is
O-pentafluoropropiony1-~-lactic
acid c y c l o h e ~ y l a m i d e [ ' ~ ~ ~ ,
present in large excess in the stationary phase. Such interwhich is characterized by the formation of the maximum
calation interactions have been postulated for amide
number of hydrogen bonds. According to this model for
phases of structure 5 (Scheme l)lB3].According to Figure
the L,L' combination the bulky isopropyl groups of 8 and
10 the selector (R)-N-lauroyl-1-(1-naphthy1)ethylamine
the methyl groups of the lactic acid derivative are stacked
crystallizes with the formation of intermolecular hydrogen
in layers one above the other and stabilize the associate by
bonds along a 5-A translation axis['501. The parallel arrangement of the aromatic rings and the linearity of the alkyl side chains which a r t also retained in the melt are important structural features. The constitutionally and configurationally equivalent selectand (R)-N-TFA-1-(I-naphthy1)ethylamine fits ideally into this structure, while the intercalation of its enantiomer is more difficult. Hence, the
gas chromatographically observed enantiospecificity can
readily be attributed to the differing stability of the diastereomeric intercalation aggregates, which has been additionally confirmed by calculations of energy datafx3.I5O1.
The interpretation of the enantiospecificity of optically
active polysiloxanes (Polymers A-D) must take into account both the (statistical) distance between the chiral side
chains and the preferred conformation of the amino acid
selector. According to statistical evaluation of valine residues in polypeptides and theoretical calculations valinamide exists preferentially in a p-pleated sheet type structure or a-helical form with a preference for the latter conFig. 11. P-Pleated sheet structure 01' an associate of Chirasil-Val 8 (Polymer
formation['05.''3.
A) with 0-pentafluoropropionyl-L-lactic
acid cyclohexylamide [IOS].
760
Angew. Chem. Int. Ed. Engl. 23 (1984) 747-76.5
van der Waals forces[Io8].The L,D' combination does not
allow of such a layer formation. The model also takes into
account the importance of the main polysiloxane chain
which separates the L-valinamide side chains.
These models assume that different conformations of
the valinamide selector are required by nitrogen-containing and nitrogen-free selectands. This implies either that
Chirasil-Val exists in various conformations or that a preferred conformation is induced during the "diastereotropic"" I 11 approach of selector and selectand.
2.7. The Correlation between
Absolute Configuration and Retention Properties
The large degree of agreement between the retention
properties of individual enantiomers of selectands with respect to their absolute configuration within an homologous series constitutes one of the significant characteristics
of gas chromatographic selector/selectand systems. The
correlation of absolute configuration with order of elution
has acquired a place of great importance in enantiomer
analysis. The basis for such correlations is the overwhelming analogy between the separation mechanism for structurally similar substrates on selectors of defined chirality.
Early on Gil-Av et a1.['21demonstrated that all the D-NTFA-a-amino acid esters investigated eluted before their
L-enantiomers on stationary phase 1 , if the selector possesses the L-configuration. As expected the order of elution is reversed if the mirror image D-stationary phase is
employed (peak inversion). Likewise, the D-derivatives are
eluted before L-N-TFA-amino acid esters on L-valine
diamide phases regardless of whether the diamide is present in low molecular weightlSo1or polymeric form (Polymers A, B, C, Scheme 2)['00,115,1201
. Th e order of elution of
p and y-amino acid derivatives is, however, the reverse of
that for a-amino acids o n N-lauroyl-L-valine (l-pentylhexyI)amide'sol.
A systematic investigation of the retention behavior of
N-TFA amines and N-TFA a, and y amino acid esters on
the "ureide" phase 4 revealed[271that the order of elution
and the separation factors depend on the steric requirements of the substituents at the asymmetric carbon atoms
of the selectands. The substituents were ordered according
to their effective size without taking any account of their
chemical function in analogy to Cram's model. If the substrate model is viewed from the asymmetric carbon atom in
the direction of the nitrogen atom, then the substituents
are arranged in decreasing size in either a clockwise or
counterclockwise direction. All enantiomers whose substituents are arranged clockwise in this classification are
eluted as the second peak on the L,L' phase. The size of the
separation factor a increases with increasing size difference of the substituents. This rule, which correlates the
configuration with the order of elution, has been confirmed by many observations: Since the size of the substituents R and COZR'is reversed, D-N-TFA-alanine alkyl esters are eluted before the L-enantiomers while L-N-TFAleucine methyl esters are eluted before the D-iSOmerS; the
separation factors become smaller as the effective size difference between R and COZR' decreases or become proAngew. Chem. Int. Ed. Engl. 23 (1984) 747-765
portionally larger with a corresponding increase[271.The
correlation between configuration and retention behavior
in gas chromatography has enabled a controversy to be resolved without resort to chiroptical methods[281concerning
the absolute configuration of 2-amino-3-methylbutane and
2-amino-4-methyl butane.
Derivatized amines, alcohols, ketones, hydroxycarboxylic acids and amino acids also exhibit consistent retention
properties with respect to their absolute configuration on
optically active p o l y s i l o ~ a n e s ~ ' In
~ ' ~analogy
~ ~ ~ . to the differing separation mechanisms for nitrogen-containing and
nitrogen-free selectands on Chirasil-Val (Polymer A), 2-hydroxycarboxylic acids exhibit varying retention relationships depending on the derivatization strategy. Thus the Lenantiomers of 2-hydroxycarboxylic acid alkyl esters are
eluted before those with the D-configuration on Chirasil-LVal, while the order of elution is reversed for the corresponding amides o r urethanes" I I 'I.
Needless to say, the correlation between molecular configuration and retention behavior of selectands in conjunction with the absolute configuration of the selector constitutes an important criterion for the mechanistic consideration of chiral recognition.
"3
2.8. Applications of the Gas Chromatographic Method
to the Resolution of Enantiorners
The suitability of gas chromatography for sensitive and
precise determinations of enantiomer compositions[21was
recognized by several research groups after the first successful quantitative separation of amino acid derivatives
and employed for many types of investigations, e.g. the
demonstration of unnatural amino acids in bacterial cell
walls and in peptide antibiotics, the monitoring of amino
acid purity during peptide synthesis, the determination of
the degree of racemization during peptide hydrolysis, the
dating of paleontological and archeological finds by exploiting the time-dependence of amino acid racemization,
the search for biogenic amino acids in extraterrestrial material, and the amplification of optical activity under
abiotic conditions14, 10,14. 151. 1751
Thus a D / L ratio of u p to 2.33 was observed gas chromatographically for alanine from the cell walls of staphyloand isovaline (2-amino-2-methylbutyric acid)
from various peptide antibiotics could be assigned to the
R - c o n f i g u r a t i ~ n ~During
' ~ ~ ~ . the hydrolysis of the dodecapeptide (Leu-Ala),, which had been synthesized by the
MerrifieId solid state methodL's31,less than 0.1% of unnatural amino acid antipodes could be detected, whereas the
racemization of L-asparagine during the hydrolysis of the
heptadodecapeptide desamidosecretin amounted to up to
2%[1531.
By gas chromatography, it was also possible to demonstrate precisely the high optical purity of the amino
acids of peptides synthesized according to the four component condensation reaction of Ugi[1541.
Gas chromatographic methods have also been employed for the systematic stereochemical analysis of various peptide antibiotics (antiamoebin I, emerimicin, alamethicin I , 11)"58.159~1601.
The racemization of amino acids in living cells, which
d o not exhibit renewal processes, is of importance in the
761
study of ageing phenomena. The racemization of amino
acids in dead, fossilized materials can be employed as a
dating parameter. Knowledge of the effect of temperature,
time, humidity, and other environmental conditions on the
rate of inversion of natural amino acids is hence of great
importance for gerontological, geochronometric (paleotemperature profiles), and archeometric investigations.
As a reversible first order reaction the rate constant for
the inversion of amino acids can be determined from the
following equation when using optically pure starting materials[’6’l,
where L and D are the concentrations of the enantiomers
determined gas chromatographically by peak area comparison at time t. As expected when plotted against
2t. ln[(L+ D)/(L- D)] yields a straight line passing through
zero whose gradient corresponds to the rate constant k .
There is a great interest in the detection of organogenic
material in interstellar space. If there were to be any preference of the image over the mirror image, it is to be expected from terrestrial evidence that this would involve
biogenic material. In the case of meteorites the demonstration of racemic amino acids is an indication of either
abiotic origins or of very great age, while the detection of
L-amino acids can be attributed to terrestrial contamination. Investigations of the Murray and Murcheson meteorite and comparison with urban earth samples suggest considerable terrestrial contamination with L-amino acids[’621.
No D,L-amino acids could be demonstrated gas chromatographically at the detection limit of 0.1 ppm in the Apollo
moon samples[1631.Incidentally, this result eliminates the
possibility of a terrestrial contamination of the carefully
isolated samples from the lunar Sea of Tranquility. An interesting archeometric investigation was concerned with
the determination of the D,L-asparagine ratio in parchment
from the Dead Sea Scrolls, which provided information
concerning the date of the degradational transformation of
the collagen to gelatine“64. 1651.
The gas chromatographic method of determining enantiomer composition is particularly suited to investigations
concerned with the laboratory simulation of the origin of
optical activity. Since only slight deviations from the racemic composition would be expected in such cases, the
unbiased employment of polarimetric measurements may
lead to grave misjudgements. In a gas chromatographic
Bonner et al. were
analysis employing diamide phase 3[’661,
able to demonstrate small amounts of enantiomer enrichment (up to e.e. 2.5%) in the abiotic kinetic resolution of
l e ~ c i n e [ ’ ~It~ was
~ . demonstrated in preliminary experiments that the gas chromatographic method employing
high performance capillary columns and digital peak integration yielded an accuracy and precision of 0.03%-0.7%
absolute error and a standard deviation of 0.03%-0.6%
with leucine mixtures of known enantiomeric composition“661.Furthermore, the results were verified by the use
of stationary phases with the opposite chirality. This strategy allows the recognition of a falsification of the analyti162
cal data by co-eluting (achiral) impurities, since a peak inversion only takes place with optical isomers[’671.
The application of gas chromatographic enantiomer
analysis in the field of enantioselective synthesis has recently been the subject of a detailed review12’. It was
pointed out that this reliable method finds astonishingly
little employment in work claiming “quantitative enantiomeric purity”. The determination of enantiomeric purity
by polarimetry applying the criterion of optical purity may
involve errors[21and the employment of nuclear resonance
methods with the aid of chiral shift reagents usually allows
the determination of enantiomeric purity to about e.e.
97%.
The hydrogenation of dehydroamino acids 34 to amino
acids in the presence of optically active rhodium complexes is an important asymmetric homogeneously catalyzed, reaction”681.
34 R-CH=C-COOII
I
NHR‘
-
“I*] R-CH,-CH-COOH
Hz
I
NHR’
Gas chromatography has only been employed occasionally for the precise determination of the enantiomeric yield
of the amino acids so produced[l”~I7O1. The enantiomeric
yield for N-acetylphenylalanine, which has been underestimated in polarimetric determinations because the maximum optical rotation reported in the literature is too high,
has been corrected by gas chromatographic methods and
the correct specific rotation a,,;,, obtained by extrapolationI171. 1721. Fo r investigations concerning the homogeneously catalyzed, enantioselective synthesis of dipeptides by
“double asymmetric induction”[173. the chiral stationary phase described by Oi et al.[”] can be employed to determine not only their diastereomeric composition but also
the enantiomeric purity.
The introduction of thermally stable polymeric stationary phases (Polymers A-D, cf. Scheme 2) into gas chromatographic enantiomer analysis[17s1,and their commerof opening up a broad
c i a l i ~ a t i o n [ ’ ~had
~ ~ ’the
~ ~ effect
~,
range of potential applications for a wide range of users.
The suitability of the polymeric phases for GC-MS coupling due to their low rate of bleeding is worthy of special
mention[’761. The employment of the SIM (selected ion
monitoring) technique can bring about a further considerable increase in sensitivity and substrate selectivity.
Frank, Nicholson, and B a ~ e r “have
~ ~ ] developed a complementary approach to amino acid analysis“391401 by
means of the “enantiomer labeling” method. Here the unnatural D-aminO acid is employed as an internal standard[177.178, I791 f or the quantification of the L-enantiomer.
Since the enantiomers possess identical (nonchiroptical)
properties in an achiral environment, the enantiomeric ratio is neither affected by sample manipulation (isolation,
derivatization, fractionation) nor by gas chromatographic
manipulations (dilution of sample, injection, splitting, detection). Thermal or catalytic decomposition, loss of substance, incomplete isolation, and semiquantitative derivatization are all without effect on the analytical result. It is
unnecessary to calibrate the detector, but detector response must be linear over a wide concentration range, a
Angcw. Chem. Int. Ed. Engl. 23 (1984) 747-765
requirement that is fulfilled by flame ionization detector~['~~].
The method of enantiomer labeling presupposes a precise knowledge of the enantiomeric purity of both the sample and the standard, which must be determined separate1y1'7k1.The concentration of a racemate can also be measured by this method by employing an enantiomerically
pure standard, just as can an enantiomerically pure substrate with a racemic standard. The method has been proposed for the precise determination of the chemical yield
in enantioselective syntheses[*], whereby a known quantity
of optically active standard is admixed with the reaction
mixture (or an aliquot therefrom) and the amount of product is calculated from the change in enantiomeric composition before and after addition. This method is useful for
enzymatic reactions, for instance, where the amounts of
reaction products are so small that they cannot be isolated
without losses.
The employment of optically active polysiloxane phases
for the resolution of the enantiomers of various classes of
compound (see Table 2) has led to important applications,
e.g. investigations of amino acid racemization during peptide synthesis and hydrolysis. Frank, Woiwode, Nicholson
and Buyer"801investigated the rate of acid-catalyzed racemization o f all the proteinogenic amino acids and differentiated between the rates of inversion of bound and free
amino acids. Furthermore, with the aid of deuterium labeling and coupled GC-MS"761it was possible to distinguish
between the amount of unnatural amino acid originally
present and the amount produced on hydrolysis"s'l since
the incorporation of deuterium is only to be expected on
inversion of configuration.
The investigations of biological fluids, forensic chemistry, pharmaceuticals and therapeutics are important
fields for the application of gas chromatographic enantiomer analysisI132, 133, 175, 176, 182-1851, as is also the configurational analysis of polysaccharides['861.
It is to be expected that, in the future, gas chromatography will become established as one of the standard methods for the reliable enantiomeric analysis of volatile chiral
compounds.
3. Summary and Outlook
Enantiospecific selector/selectand systems have been
described which allow a quantitative gas chromatographic
enantiomer resolution. Advances in glass and fused silica
capillary column technology in combination with the development of thermally stable chiral polysiloxane phases
and the employment of improved derivatization strategies
has led to a n extension of gas chromatographic enantiomer
resolution to many classes of compounds. The polymeric
phases are especially suitable for temperature programming and for the coupling of gas chromatography and
mass spectroscopy. The method of "enantiomer labeling"
permits the reliable quantification of enantiomers in biological fluids, of chiral pharmaceuticals, in forensic chemistry and of the products of enantioselective transformations. Gas chromatography provides unequivocal results in
Angew. Chem. Int. Ed. Engl. 23 (1984) 747-765
the determination of the enantiomeric composition (e.e.,
enantiomeric excess) particularly in the boundary regions
of slight enantiomeric excess (deviations from racemic
composition) and high enantiomeric purity (detection of
traces of the other enantiomer).
The large body of data on enantiomer resolution by gas
chromatography allows a deliberate choice of selector/selectand systems for the detailed investigation of the molecular mechanisms of chiral recognition and of the correlation of the influence of the structural factors of selector
and selectand on the enantiospecificity. The order of elution is constant within wide limits for analogous and homologous selectands. For this reason gas chromatography
can be employed for the determination of absolute configuration without knowledge of chiroptical properties. The
use of optically active phases of opposite configuration
can confirm or exclude the enantiomeric origin of a peak
separation.
For future development in chromatographic enantiomer
separation it seems likely that GC and LC (HPLC) will
complement each other and each method will find its own
particular range of applications. Whereas preparative separation is of interest in LC, the main application of GC
will be in enantiomer analysis. GC can be employed to resolve the enantiomers of many components simultaneously
within a short time (e.g. amino acid analysis) and complications associated with the nature of the mobile phase are
largely unknown. GC readily lends itself to automation.
It is surprising that all the GC systems known today are
characterized by small separation factors in comparison
with those obtained in LC. This, however, is not a practical
disadvantage since together with the high separating performance, particularly of capillary columns, it allows in
this way acceleration of the analysis and substances in
mixtures can be resolved without peak overlap. The present enantiospecific selector/selectand systems of gas chromatography, however, seem quite unsuitable as models for
stereospecific enzymes, Hence, the optimization of such
systems remains a stimulating aim for the future. In particular the employment of alternative principles of chiral recognition, that are known in nature, will increase in importance. Mention should be made of enantiomer resolution
o n phases with host-guest interaction in chiral cavities via
molecular inclusion, which has already been initiated employing cyclodextrin phases.
Besides hydrogen bonds, other intermolecular forces, or
combinations thereof, can also be considered for the resolution of enantiomers by gas chromatography, e.g. charge
transfer interactions and ionic and dipole-dipole attractions. The resolution of enantiomers by complexation gas
chromatography based on the enantiospecific coordination
interaction often yields high separation factors.
Preparative enantiomer resolution is possible by gas
chromatography, too. The amounts isolated are small,
however, but are suitable for the determination of chiroptical properties or for the correlation of configuration or determination of biological activity.
As far as the employment of commercially available enantiospecific stationary phases is concerned, the user must
distinguish between the factors availability, performance,
reproducibility and stability. The inspection of chromato-
763
grams, in so far as they are available, often provides decisive information.
The author was in the fortunateposition to witness decisive
developments in the methods described here ‘on site’ in the
individual laboratories. V. S . wishes to express his sincere
thanks to Professor E. Gil-Av, Dr. B. Feibush, Professor E.
Bayer, Dr. H . Frank, G. J . Nicholson, Dr. B. Koppenhoefer
and Prof. W. A . Konig for valuable discussions and advice,
which constituted an important stimulation for this manuscript. Support of this work bv the Fonds der Chernischen Industrie and the Deutsche Forschungsgemeinschaft is gratefully acknowledged.
Received: June 16, 1984 [A 510 IE]
German version: Angew. Chem. 96 (1984) 733
Translated by Dr. Frank Hampsan, Saarbriicken
111 J. D. Morrison in J. D. Morrison: Asymmetric Synthesis, Bd. 1, Academic Press, New York 1983, p. I .
121 V. Schurig in [I], p. 59.
(31 a) This nomenclature has been introduced, in analogy with the cybernetic terms operator and operand, for the chiral separating agent and
the racemic sample, to be resolved in a chromatographic system [3b].
Selector is in general understood t o he an (essentially) enantiomerically
pure material which is able to distinguish the enantiomers of the (essentially) racemic substrate by chiral recognition; h) F. Mike;, PhD Thesis,
Weizmann Institute of Science, Rehovot 1975.
141 E. Gil-Av, J . Mol. Euol. 6 (1975) 131.
[5] a) P. D. Maestas, C. J. Morrow, Tetrahedron Lett. 1976, 1047; b) J.
GaaI, J. Inszedy, J . Chromatogr. 102 (1974) 375; c) W. Sczepaniak, W.
Ciszewska, Chromatographia 15 (1982) 38; 18 (1984) 221; d) E. Bayer,
Chromatographia 18 (1984) 220.
[6] K. C. Cundy, P. A. Crooks, J. Chromatogr. 281 (1983) 17.
[7] C. H. Lochmiiller, R. W. Souter, J . Chromatogr. 113 (1975) 283.
[XI V. Schurig, Angew. Chem. 89 (1977) 113; Angew. Chem. Int. Ed. Engl.
I 6 (1977) 110.
191 W. A. Konig, J . High Resolut. Chromatogr. Chromatogr. Commun. 5
(1982) 588.
[lo] E. Bayer, G. J. Nicholson, H. Frank in C. W. Gehrke, K. C. Kuo, R. W.
Zumwalt: Amino Acid Analysis by Gas Chromatograph). CRC Press,
Boca Raton, FL, in press.
[ I I] V. Schurig, Chromatographia 13 (1980) 263.
[I21 E. Gil-Av, B. Feibush, R. Charles-Sigler, Telrahedron Lett. 1966,
1009.
1131 E. Gil-Av, D. Nurok, Adu. Chrornatogr. 10 (1974) 99.
[14] E. Gil-Av: Modell, Bericht aus Rehauot, Weizmann Institute o f Science
1976, p. 25.
[I51 V. Prelog, W. Wieland, Helu. Chim. Acta 27(1944) 1127.
[I61 L. H. Klemm, D. Reed, J. Chromatogr. 3 (1960) 364.
(171 L. H. Klemm, K. B. Desai, J. R. Spooner, Jr., J . Chromatogr. 14 (1964)
300.
[I81 A. C. Cope, C. R. Gannelin, H. W. Johnson, T. V. Van Auken, H. J. S.
Winkler, J. Am. Chem. Soc. 85 (1963) 3275.
[I91 V. Schurig, R. C. Chang, A. Zlatkis, E. Gil-Av, F. MikeS, Chromatographia 6 (1973) 223.
[20] V. Schurig, E. Gil-Av, Chem. Commun. 1971. 650.
1211 E. Gil-Av, V. Schurig, Anal. Chem. 43 (1971) 2030.
[22] V. Schurig, W. Biirkle, J . Am, Chem. SOC.104 (1982) 7573.
[23] V. Schurig, R. C. Chang, A. Zlatkis, B. Feihush, J . Chromatogr. 99
(1974) 147.
[24] C. E. Dalgliesh, J . Chem. SOC.1952, 3940.
[25] T. Kokielski, D. Sybilska, J. Jurczak, J . Chromatogr. 280 (1983) 131.
[26] H. B. Kagan, H. Mimoun, C. Mark, V. Schurig, Angew. Chem. 91
(1979) 51 1 ; Angew. Chem. Int. Ed. Engl. 18 (1979) 485.
[27] B. Feibush, E. Gil-Av, T. Tamari, J . Chem. Soc. Perkin Trans. 2 1972,
1197.
[28] H. Ruhinstein, B. Feibush, E. Gil-Av, J . Chem. SOC.Perkin Trans. 2
1973, 2094.
(291 W. Biirkle, H. Karfunkel, V. Schurig, J . Chromatogr. 288 (1984) 1.
I301 1. S . Krull, Adu. Chromatogr. 16 (1978) 175.
[31] G. Blaschke, Angew. Chem. 92 (1980) 14; Angew. Chem. Inf. Ed. Engl.
19 (1980) 13.
[32] V. A. Davankov, Adu. Chromatogr. 18 (1980) 139.
[33] G. Karagounis, G . Lippold, Naturwissenschaften 46 (1959) 145.
[34] N. A. Goeckner, Diss. Abstr. 19 (1959) 3127.
[35] G. Goldberg, W. A. Ross, Chem. lnd. (London) 1962, 657.
[36] C. B. Coleman, G . D. Cooper, J. F. ODonnell, J . Org. Chem. 31 (1966)
975.
164
(371 G. Karagounis, E. Lemperle, Fresenius Z . Anal. Chem. 189 (1962)
131.
(381 R. E. Severs, R. W. Moshier, M. L. Morris, Inorg. Chem. 1 (1962)
966.
[39] P. J. Porcaro, V. D. Johnston, Anal. Chem. 33 (1961) 1748.
[40] E. Gil-Av, B. Feibush, R. Charles-Sigler in A. B. Littlewood: Gas Chromatography 1966, lnstitute of Petroleum, London 1967, p. 227.
[41] C . H. Lochmiiller, R. W. Souter, J . Chromatogr. 113 (1975) 283.
[42] R. H. Liu, W. W. Ku, J. Chromatogr. 271 (1983) 309 (see Table 1 therein).
[43] B. Feihush, E. Gil-Av, J. Gas Chromatogr. 5 (1967) 257.
[44] B. Feibush, E. Gil-Av, Tetrahedron 26 (1970) 1361.
[45] B. Feihush, E. Gil-Av, T. Tamari, Isr. J . Chem. 8 (1970) 50.
(461 S. Nakaparksin, P. Birell, E. Gil-Av, J. Oro, J. Chromatogr. Sci. 8
(1970) 177.
(471 W. Parr, C. Yang, E. Bayer, E. Gil-Av, J . Chromafogr. Sci. 8 (1970)
591.
[48] E. Gil-Av, B. Feihush, US-Pat. 3494105; Jpn. Pat. 567267.
[49] B. Feihush, Chern. Commun. 1971, 544.
[50] U. Beitler, B. Feibush, J. Chromatogr. 123 (1976) 149.
[51] E. Gil-Av, B. Feibush, Tetrahedron Lett. 1967, 3345.
[52] W. Konig, W. Parr, H. A. Lichtenstein, E. Bayer, J. Oro, J. Chromatogr.
Sci. 8 (1970) 183.
[53] W. Parr, C. Yang, E. Bayer, E. Gil-Av, J. Chromatogr. Sci. 8 (1970)
591.
1541 W. Parr, C. Yang, J. Pleterski, E. Bayer, J. Chromafogr. 50 (1950) 510.
[55] W. Parr, J . Pleterski, C. Yang, E. Bayer, J . Chromatogr. Sci. 9 (1971)
141.
[56] W. Parr, Y. Howard, Chromutographia 4 (1972) 162.
[57] W. Parr, Y. Howard, J . Chromarogr. 66 (1972) 141.
[58] W. Parr, Y . Howard, J . Chromatogr. 67 (1972) 227.
1591 W. Pam, Y. Howard, J . Chromatogr. 71 (1972) 193.
[60] W. Parr, Y. Howard, Anal. Chem. 45 (1973) 71 I .
[61] F. Andawes, R. Brazell, W. Parr, A. Zlatkis, J . Chramatogr. 112 (1975)
197.
[62] W. A. Konig, G . J. Nicholson, Anal. Chem. 47 (1975) 951.
[63] W. A. Konig, Chromatographia 9 (1976) 72.
(641 K. Stoking, W. A. Konig, Chromatographia 9 (1976) 331.
[65] W. A. Konig, K. Stoking, K. Kruse, Chromatographia 10 (1977) 444.
[66] 1. Abe, T. Kohno, S . Musha, Chromatographia 11 (1978) 393.
[67] S. Weinstein, G. Jung, E. Gil-Av, Proc. Annu. Meet. lsr. Chem. SOC.41
(1971) 220.
[68] J. Corhin, J. Rhoad, L. Rogers, Anal. Chem. 43 (1971) 327.
[69] E. Gil-Av, B. Feibush in Y . Wolman: Peptides 1974, Proc. 13th European Peptide Symp., Jerusalcm 1974, p. 279.
(701 R. Charles, U. Beitler, B. Feibush, E. Gil-Av, J. Chromatogr. 112 (1975)
121.
1711 Supelco, Inc.: Bulletin 765F. Bellefonte, PA 1976, S . 1.
[72] R. Charles, E. Gil-Av, J . Chromatogr. 195 (1980) 317.
[73] S.-C. Chang, R. Charles, E. Gil-Av, J . Chromatogr. 235 (1983) 87.
1741 B. Feihush, B. Altman, E. Gil-Av, Proc. Annu. Meet. Isr. Chem. SOC.41
(1971) 196.
[75] S.-C. Chang, R. Charles, E. Gil-Av, J . Chromatogr. 202 (1980) 247.
[76] S.-C. Chang, R. Charles, E. Gil-Av, J . Chromatogr. 238 (1982) 29.
[77] S.-C. Chang, E. Gil-Av, R. Charles, J . Chromatogr. 289 (1984) 53.
I781 a) E. Gil-Av, R. Charles, S.-C. Chang in [lo]; b) B. Feibush, A. Balan,
B. Altman, E. Gil-Av, J . Chrm. SOC.Perkin Trans. 2 1979, 1230.
[79] J. A. Corhin, L. B. Rogers, Anal. Chem. 42 (1970) 974.
(801 C. H. Lochmiiller, J. M. Harris, R. W. Souter, J. Chromatogr. 72 (1972)
405.
[XI] C. H. Lochmiiller, R. W. Souter, J . Chromatogr. 87 (1973) 243; 88
(1974) 41.
1821 S. Suzuki, T. Hobo, K. Watabe, H. Ishikawa, Bunseki Kagaku 30 (1981)
497.
[83] S. Weinstein, B. Feihush, E. Gil-Av, J . Chromatogr. 126 (1976) 97.
(841 N. Oi, H. Kitahara, Y. Inda, J . Chromatogr. 213 (1981) 137.
[85] N. Oi, H. Kitahara, Y . Inda, J . Chromatogr. 237 (1982) 297.
[86] N. Oi, T. Doi, H. Kitahara, Y . Inda, J . Chromatogr. 239 (1982) 493.
(871 N. Oi, H. Kitahara, T. Doi, J . Chromatogr. 254 (1983) 282.
[88] N. Oi, R. Takai, H. Kitahara, J . Chromatogr. 256 (1983) 154.
[89] N. Oi, H. Takeda, H. Shimada, DOS 2720995 (November 24, 1977).
(901 N. Oi, M. Horiha, H. Kitahara, J. Chromatogr. 202 (1980) 299.
[YI] N. Oi, M. Horiba, H. Kitahara, H. Shimada, J. Chromatogr. 202 (1980)
302.
[92] N. Oi, H. Kitahara, M. Horiha, T. Doi, J . Chromatogr. 206 (1981)
143.
[93] N. Oi, T. Doi, H. Kitahara, Y . Inda, J . Chromatogr. 208 (1981) 404.
(94) N. Oi, H. Kitahara, T. Doi, J. Chromatogr. 207 (1981) 252.
[95] G. Berred, J. Bourdon, J. Dreux, R. Longeray, M. Moreau, P. Schifter,
Chromatographia 12 (1979) 150.
1961 W. A. Konig, S. Severs, U. Schulze, Angew. Chem. 92 (1980) 935; Angew. Chem. Int. Ed. Engl. JY (1980) 910.
(971 W. A. Konig, S . Severs, J . Chromatogr. 200 (1980) 189.
Angew. Chem. lnt. Ed. Engl. 23 (1984) 747-765
1981 W. Kirmse, R. Siegfried, J. Am. Chem. SOC.105 (1983) 950.
[99] F. Scheidt, H. Raskopf, LaborPraxis. April 1984, p. 332.
[IOO] H. Frank, G. J. Nicholson, E. Bayer, 3. Chromatogr. Sci. 15 (1977)
174.
[loll E. Bayer, H. Frank, DOS 2740019 (22. 3. 1979).
[I021 E. Bayer, H. Frank, US-Pat. 4387206 (1983).
1103) J. K. Haken, J. Chromafogr.300 (1984) I .
[I041 Chrompack News I0 (1983) 6.
IlOS] B. Koppenhdfer, Dissertation, Universitat Tiibingen 1980.
11061 E. Bayer, H. Frank, ACS Symp. Ser. 121 (1980) 341.
1107) G. J. Nicholson, H. Frank, E. Bayer, J . High Resolut. Chromatogr.
Chromatogr. Commun. 2 (1979) 41 1.
[I081 H. Frank, G. J. Nicholson, E. Bayer, Angew. Chem. 90 (1978) 396; Angew. Chem. int. Ed. Engl. 17 (1978) 363.
11091 H. Frank, G. J. Nicholson, E. Bayer, J. Chromarogr. 146 (1978) 197.
[ I 101 B. Koppenhofer, H. Altmendinger, G. J. Nicholson, E. Bayer, J. Chromafogr. 260 (1983) 63.
[II I] H. Frank, J. Gerhardt, G. J. Nicholson, E. Bayer, J. Chromafogr. 270
(1983) 159.
[I 121 H. P. Benschop, C. A. G. Konings, L. P. A. De Jong, J. Am. Chem. Soc.
103 (1981) 4260.
[I131 D. W. UIff Pittsburgh Conf. on Anal. Chem. and Appl. Spectrosc., Atlantic City 1982, p. 791.
[I 141 D. Thumm, Dissertation, Universitat Tiibingen 1980.
11151 T. Saeed, P. Sandra, M. Verzele, J . Chromatogr. 186 (1980) 611.
[ I161 T. Saeed, P. Sandra, M. Verzele, J. High Resolut. Chromatogr. Chromafogr. Commun. 3 (1980) 35.
[ I 171 W. A. Konig, I. Benecke, H. Bretting, Angew. Chem. 93 (1981) 688; Angew. Chem. Int. Ed. Engl. 20 (1981) 693.
[I181 W. A. Kdnig, 1. Benecke, S. Sievers, J. Chromalogr. 217 (1981) 71.
[I191 W. A. Konig, 1. Benecke, J. Chromatogr. 209 (1981) 91.
11201 W. A. Konig, S. Sievers, I. Benecke: Proc. 4. Int. Symp. Capillary Chromatogr.. Huthig, Heidelberg 1981, p. 703.
(I211 W. A. Konig in P. Schreier: Analysis of Volatiles. d e Gruyter, Berlin
1984, p. 77.
[I221 W. A. Konig in E. J. Elving, J. Wineforder, I. Kolthoff, Adu. Anal. Ser.,
in press.
[I231 Chrompack News 9 (1982) 4.
11241 W. A. Konig, 1. Benecke, N. Lucht, E. Schmidt. 1. Schulze, S. Sievers, J .
Chromatogr. 279 (1983) 5 5 5 .
(1251 W. A. Konig, I. Benecke, S. Severs, 1. Chromarogr. 238 (1982) 427.
[I261 W. A. Konig, I. Benecke, J. Chromafogr. 269 (1983) 19.
[I271 1. Benecke, E. Schmidt, W. A. Konig, J. High Resolul. Chromatogr.
Chromatogr. Commun. 4 (1981) 553.
[I281 A. L. Leavitt, W. R. Sherman, Methods Enzymol. 89 (1982) 3.
[I291 W. A. Konig, W. Francke, 1. Benecke, J. Chromatogr. 239 (1982) 227.
[I301 W. A. Konig, E. Steinbach, K. Ernst, Angew. Chem. 96 (1984) 516; Angew. Chem. Jnr. Ed. Engl. 23 (1984) 527.
[I311 W. A. Konig, 1. Benecke, K. Ernst, J. Chromarogr. 253 (1982) 267.
[I321 W. A. Konig, K. Ernst, J. Chromatogr. 280 (1983) 135.
[I331 W. A. Konig, K. Ernst, J. Vessman, J . Chromatogr. 294 (1984) 423.
[I341 J. H. Liu, W. W. Ku, Anal. Chem. 53 (1981) 2180.
(1351 J. H. Liu, W. W. Ku, J. T. Tsay, M. P. Fitzgerald, S. Kim, J. Forensic
Sci. 27 (1982) 41.
[I361 R. Liardon, S. Ledermann, J. High Resolut. Chromalogr. Chromalogr.
Commun. 3 (1980) 475.
[ 1371 D. R. Knapp: Handbook of Ana/ytical Deriuatization Reactions, Wiley,
New York 1979.
[I381 W. F. Lindner in R. W. Frei, J. F. Lawrence: Chemical Deriuatization.
Analytical Chemistry, Plenum, New York 1982, p. 145.
[I391 J. Gerhardt, Dissertation, Universitat Tubingen 1984.
[I401 H. Frank, J. Gerhardt, G. J. Nicholson, E. Bayer, Fresenius 2. Anal.
Chem. 317 (1984) 688.
(1411 S. L. MacKenzie, D. Tenaschuk, J. Chromatogr. 171 (1979) 195.
[I421 S. L. MacKenzie, D. Tenaschuk, J. Chromatogr. 173 (1979) 53.
11431 H. Frank, E. Bayer, private communication.
[I441 I. Benecke, W. A. Konig, Angew. Chem. 94 (1982) 709: Angew. Chem.
I n t . Ed. Engl. 21 (1982) 709; Angew. Chem. Suppl. 1982, 1605.
[I451 W. A. Konig, U. Hess, Justus Liebigs Ann. Chem. 1977, 1087.
Angew. Chem. Int. Ed. Engl. 23 (1984) 747-765
[I461 V. P. Uralets, J. A. Rijks, P. A. Leclercq, J. Chromatogr. 194 (1980)
135.
11471 E. Kiistrrs, Dissertation, Universitat Tubingen 1983.
I1481 E. Anklam, W. A. Konig, P. Margaretha, Tetrahedron Lett. 24 (1983)
5851.
11491 R. L. Laub, C. A. Wellington, M a / . Assoc. 2 (1979) 171.
I1501 S. Weinstein, L. Leiserowitz, E. ( 3 - A v , J. Am. Chem. Sac. I01 (1980)
2768.
{IS11 W. A. Bonner in D. C. Walker: Origin oJOptical Actiuity in Nature. Elsevier, Amsterdam 1979, p. 5 .
I1521 E. Gil-Av, R. Z. Korman, S. Weinstein, Biochim. Biophys. Acra 211
(1970) 101.
11531 E. Bayer, E. Gil-Av, W. A. Konig, S. Nakaparksin, J. Oro, W. Parr, J.
Am. Chem. SOC.92 (1970) 1738.
[IS41 1. Ugi, lnlra-Sci. Chem. Rep. 5 (1971) 229.
[ISs] R. Charles, B. Feibush, E. Gil-Av in [69], p. 93.
[IS61 M. Waki, J. Meienhofer, J. Am. Chem. SOC.99 (1977) 6075.
11571 H. Bruckner, G. J. Nicholson, G. Jung, K. Kruse, W. A. Konig, Chromafographia 13 (1980) 209.
11581 R. C. Pandey, J. C. Cook, Jr., K. L. Rinehdrt, Jr., J. Am. Chem. SOC.99
(1977) 5203,5205. 8469.
[IS91 W. A. Kiinig, W. Loeffler, W. H. Meyer, R. Uhmann, Chem. Ber. 106
(1973) 816.
11601 A. Hasenbohler, H. Kneifel, W. A. Konig, H. ZBhner, H. J. Zeiler, Arch.
Microbiol. 99 (1974) 307.
(1611 S. Nakaparksin, E. Gil-Av, J. Oro, Anal. Biochem. 33 (1970) 374.
(1621 .I.Oro, S. Nakaparksin, H. Lichtenstein, E. Gil-Av, Nature (London)
230(1971) 107.
[I631 J. Oro, W. S. Updegrove, J. Gibert, J. McReynolds, E. Gil-Av, J. tbanez, A. Zlatkis, D. A. Flory, R. I>. Levy, C. Wolf, Science 167 (1970)
765.
11641 S. Weiner, Z. Kustdnovich, E. Gil-Av, W. Traub, Nature (London) 287
(1980) 5785.
[I651 M. Levin: Modell. Berich, ous Rehouot, Weizmann Institute of Science
1981, p. 3.
[I661 W. A. Bonner, M. A. Dort, J. J. Flores, Anal. Chem. 46 (1974) 2104.
[I671 W. A. Bonner, N. E. Blair, J. Chromatogr. I69 (1979) 153.
[I681 H. B. Kagan, Pure Appl. Chem. 43 (1975) 401.
11691 I . M. Brown, B. A. Murrer, Tefrahedron Leif. 21 (1980) 581.
[I701 J. W. Scott, D. D. Keith, G. Nix, Jr., D. R. Parrish, S. Remington, G. P.
Roth, J. M. Townsend. D. Valentine, Jr., R. Ydng, J. Org. Chem. 46
(1981) 5086.
[171] T. P. Dang, J.-C. Poulin, H. B. Kagan, J. Organomet. Chem. 91 (1975)
105.
[I721 G. Gelhard, H. B. Kagan, R. Stern, Tetrahedron 32 (1976) 233.
[I731 J.-C. Poulin, D. Meyer. H. B. Kagan, C. R. dead. Sci. Ser. C 2 9 1 (1980)
69.
11741 1. Ojima, T. Suzuki, Tetrahedron Left. 21 (1980) 1239.
[I751 E. Bayer, 2. Naturforsch. 838 (19x3) 1281.
I1761 H. Frank, G. J. Nicholson, E. Bayer, J. Chromatogr. 146 (1978) 197.
(1771 W. A. Bonner, J. Chromatogr. Sci. I 1 (1973) 101.
(1781 H. Frank, G. J. Nicholson, E. Bayer, J. Chromafogr.167 (1978) 187.
(1791 N. E. Blair, W. A. Bonner, J. Chromatogr. 198 (1980) 185.
11801 H. Frank, W. Woiwode, G. I. Nicholson, E. Bayer, Liebigs Ann. Chem.
1981, 354.
[I811 W. Woiwode, H. Frank, G. J. Nicholson, E. Bayer, Chem. Ber. 111
(1978) 371 I.
(I821 H. Frank, A. Eimiller, H. H. Kornhuber, J. Chromatogr. 224 (1981)
177.
11831 H. Frank, A. Rettenmeier, H. Weicker, G. J. Nicholson, E. Bayer, Clin.
Chim. Acra 105 (1980) 201.
[184] J. H. Liu, W. W. Ku, J. Chromatogr. 271 (1983) 309.
[185] J. H. Liu, W. W. Ku, J. T. Tsay, M. P. Fitzgerald, S. Kim, J. Forensic.
Sci. 27 (1982) 39.
11861 H. Bretting, G. Jakob, I. Benecke, W. A. Konig, J. Thiem, Carbohydr.
Res.. in press.
[I871 E. Kiisters, H. Allgaier, G . Jung, E. Bayer, Chromafographia 18 (1984)
287.
(1881 E. Koch, G . J. Nicholson, E. Bayer, J. High Resolut. Chromarogr. Chromalogr. Commun. 7 (1984) 398.
765
Документ
Категория
Без категории
Просмотров
1
Размер файла
2 009 Кб
Теги
analytical, complex, optically, method, gas, phase, new, enantiomers, separating, free, metali, activ, stationary, chromatography
1/--страниц
Пожаловаться на содержимое документа