Switchable Enantioselective Three- and Four-Dimensional Dynamic Gas Chromatography-Mass Spectrometry: Example Study of On-Column Molecular Interconversion

Article abstract of DOI:10.1021/acs.analchem.7b00853

A novel hybrid online enantioselective four-dimensional dynamic GC (e4D-DGC) approach to study reversible molecular interconversion through specific isolation of a diastereo and enantiopure oxime, 2-phenylpropanaldehyde oxime, from prior multidimensional separation, is described. It incorporates a pre-enantioseparation step that applies comprehensive two-dimensional GC (GC × GC), prior to multiple microfluidic (Deans) switching for selection of components of a diastereomeric (E,Z) and enantiomeric (R,S) oxime into a third reactor column where isomerization occurs. This is followed by E/Z separation in a fourth analytical column. The enantioselective first dimension (¹D_enant) yields enantioseparation of E(R), Z(R), E(S), and Z(S) isomers, with a characteristic interconversion zone between the E and Z isomers. However, these are contaminated with underlying stereoisomers. Selected separation regions were then modulated and separated using a second dimension (²D) column via GC × GC, resolving the interfering stereoisomers. Individual pure enantiomers were then selectively heart-cut from within the 2D separation space, cryofocused, then eluted on a ³D reactor column for E ? Z isomerization under controlled oven temperature and flow. Heart-cuts taken over the resulting interconversion distribution were cryotrapped at the inlet of a ⁴D column, on which achiral separation allows precise quantification of each E and Z isomer of the enantiomer. From peak areas and isomerization time, the forward and backward rate constants (k_E→Z and k_Z→E) were determined. The described methodology is suited to other configurationally labile molecules (for instance, hydrazones and imines), which exhibit isomerization, and can be used to isolate individual compounds from multicomponent samples, without requiring pure compound synthesis, or complex mathematical models or in-silico simulations.

Full text of DOI:10.1021/acs.analchem.7b00853

Subscriber access provided by ORTA DOGU TEKNIK UNIVERSITESI KUTUPHANESI
Article
Switchable Enantioselective Three- and Four-Dimensional
Dynamic Gas Chromatography-Mass Spectrometry:
Example Study of On-column Molecular Interconversion
Yong Foo Wong, Chadin Kulsing, and Philip John Marriott
Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.7b00853 • Publication Date (Web): 11 Apr 2017
Downloaded from http://pubs.acs.org on April 19, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth
Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 10
Analytical Chemistry
1
2
3
4
5
6
Switchable Enantioselective Three- and Four-Dimensional
Dynamic Gas Chromatography-Mass Spectrometry: Example
Study of On-column Molecular Interconversion
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Yong Foo Wong, Chadin Kulsing and Philip J. Marriott*
Australian Centre for Research on Separation Science, School of Chemistry, Monash University, Wellington Road, Clayton,
VIC 3800, Australia.
ABSTRACT: A novel hybrid onꢀline enantioselective fourꢀdimensional dynamic GC (e4DꢀDGC) approach to study reversible moꢀ
lecular interconversion through specific isolation of a diastereo, and enantiopure oxime, 2ꢀphenylpropanaldehyde oxime from prior
multidimensional separation, is described. It incorporates a preꢀenantioseparation step that applies comprehensive twoꢀdimensional
GC (GC×GC), prior to multiple microfluidic (Deans) switching for selection of components of a diastereomeric (E,Z) and enantioꢀ
meric (R,S) oxime into a third reactor column where isomerization occurs. This is followed by E/Z separation in a fourth analytical
1
column. The enantioselective first dimension ( Denant) yields enantioseparation of E(R), Z(R), E(S) and Z(S) isomers, with a characꢀ
teristic interconversion zone between the E and Z isomers. However these are contaminated with underlying stereoisomers. Selectꢀ
2
ed separation regions were then modulated and separated using a second dimension ( D) column via GC×GC, resolving the interferꢀ
ing stereoisomers. Individual pure enantiomers were then selectively heartꢀcut from within the 2D separation space, cryofocussed,
3
then eluted on a D reactor column for E ⇌ Z isomerization under controlled oven temperature and flow. Heartꢀcuts taken over the
4
resulting interconversion distribution were cryotrapped at the inlet of a D column, on which achiral separation allows precise quanꢀ
tification of each E and Z isomer of the enantiomer. From peak areas and isomerization time, the forward and backward rate conꢀ
stants (kE→Z and kZ→E) were determined. The described methodology is suited to other configurationally labile molecules (for inꢀ
stance, hydrazones and imines) which exhibit isomerization, and can be used to isolate individual compounds from multiꢀ
component samples, without requiring pure compound synthesis, or complex mathematical models or inꢀsilico simulations.
Investigation of the molecular dynamics of configurationally
labile stereoisomers has become increasingly important across
isomers constituting interconversion profiles confounds direct
estimation of epimeric, isomeric or enantiomeric ratios at any
given point. This normally requires iterative comparison of
experimental and simulated elution profiles to assess interconꢀ
version rate constants and Gibbs activation. The Schurig group
investigated many such molecular processes of chiral moleꢀ
1
a broad range of research areas (e.g. drug development, proꢀ
2
3
4
tein folding, biochemistry, supramolecular chemistry, etc.).
Transformation or interconversion processes of labile isomers
can be treated as a change toward racemic or nonꢀracemic
mixtures of stereoisomers at thermodynamic equilibrium. Two
different processes can be identified, (i) irreversible structural
9,11,14,15,21,22
cules using dynamic chromatographic approaches.
In DGC experiments, an interconversion or plateau region
will arise from original injected peaks if the prevailing thermal
energy does not greatly exceed the interconversion barrier enꢀ
ergy, otherwise ‘peak coalescence of the thirdꢀtype’ may
5
transformation (e.g. and decomposition ), and (ii) reversible
6
molecular interconversion (e.g. racemization , enantiomerizaꢀ
7
4
tion and diastereomerization ). Dynamic chromatography has
been used for the study of reversible configurational changes,
23
arise. Determination of kinetic and activation data for interꢀ
conversion is based on the total envelope shape of the dynamic
chromatogram, accompanied by mathematical modelling or
simulation. Trapp and Schurig developed a Windowsꢀbased
computer program (ChromWin and ChromWin 2D) for interꢀ
pretation of the interconversion process based on simulation,
employing the theoretical plate model, stochastic model, and a
modified stochastic model to describe chromatographic eluꢀ
8ꢀ10
including gas chromatography (GC), high performance liqꢀ
1,2,11ꢀ13
14ꢀ16
uid chromatography,
capillary electrophoresis,
thin
17
layer chromatography,
supercritical fluid chromatogꢀ
18,19
raphy,
and in conjunction with nuclear magnetic resoꢀ
2
0
nance. Interconversion, phase partitioning, and distribution
of the E and Z isomers of oximes in a ‘reactor column’ can be
represented by Supporting Information Figure S1.
9,24,25
For ‘dynamic reaction’ in single dimensional (1D) chromaꢀ
tography, nonꢀGaussian elution profiles may be observed, inꢀ
dicating onꢀcolumn interconversion or decomposition during
separation. The shapes of the peaks (peakꢀbroadening; plateauꢀ
formation between two terminal peaks; peak coalescence) deꢀ
pend on the stereochemical stability or lability of the investiꢀ
gated molecules, and the rate of ‘reaction’ of the isomers unꢀ
der prevailing conditions on the chromatographic separation
timescale. Chromatographically, distorted distributions of
tion.
Krupčik et al. developed computer assisted methods
8,26
to deconvolute interconversion profiles. In 2006, Trapp deꢀ
veloped a unified equation to calculate rate constants of a firstꢀ
order reaction directly from chromatographic parameters (peak
width; retention time), applying it to calculate kinetic data for
27,28
alkylated diaziridines.
The program DCXplorer was further
introduced by Trapp, utilizing the unified equation of chromaꢀ
2
9,30
tography for evaluation of interconversion barriers.
ACS Paragon Plus Environment

Analytical Chemistry
Page 2 of 10
Multidimensional GC techniques (e.g. stoppedꢀflow MDGC transfer line to the FID. H/C switching of effluent flow from
1
3
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
and comprehensive twoꢀdimensional GC; GC×GC), have been
introduced to physically deconvolute isomeric forms underꢀ
neath the interconversion plateau. Trapp and Schurig develꢀ
oped stoppedꢀflow MDGC for the determination of epimerizaꢀ
tion barriers of stereolabile compounds (e.g. Tröger's base,
Denant to either D
or the FID was controlled through the
react
events option in MassHunter software via a three channel auxꢀ
iliary electronic pressure control (EPC) module (G1570A; Agꢀ
ilent). The modulator was held at a constant modulation temꢀ
perature (T ) of 20 °C and modulation period (P ) of 8 s, with
M
M
#
−1
chalcogran, etc.), with ꢁG values ranging from 70 kJ mol
carbon dioxide as cryogen coolant. FID (250 ˚C) was operated
at a sampling frequency of 100 Hz to monitor the very narrow
GC peaks eluting from D . A schematic of the system used
here is shown in Figure 1A, for GC×GC mode (without heartꢀ
gas
−1 9,10,31
to 200 kJ mol .
Marriott and coꢀworkers applied high
2
resolution GC×GC to study oxime interconversions, with novꢀ
el presentation of the isomerization process over the 2D chroꢀ
IL
2
matographic space where the second dimension ( D) resolves
cutting, indicated by orange arrow) and for transfer of heartꢀ
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
25,32ꢀ34
3
E and Z isomers across the distribution.
Recently, Kröger
cuts to the D
column (indicated by red arrow), respectiveꢀ
react
1
2
et al. employed enantioselective columns in both D and D to
investigate reversible molecular interconversion behavior of a
chiral aromatic oxime, illustrating features of molecular reꢀ
ly. A workflow describing the different steps in the operation
is shown in Figure 1B.
An SGE liquid CO cryogenic trapping device (CT; Trajan
2
3
4
versible process in the 2D patterns.
Scientific, Ringwood, Australia) was positioned at the beginꢀ
3
4
The present work demonstrates a novel enantioselective fourꢀ
dimensional DGC (e4DꢀDGC) system with accurate mass
timeꢀofꢀflight mass spectrometry (accTOFMS) for the microꢀ
scopic study of molecular interconversion of a chiral aromatic
oxime. It operates as a sequential enantioselective GC×GCꢀ
MDGCꢀGC system, to allow unique isolation of the modulated
peaks of diastereoꢀ and enantiopure isomers in 2D space (i.e.
by heartꢀcutting (H/C) as described by Mitrevski and Marꢀ
ning of D
and D , to trap and reꢀfocus the H/C solutes
react
np
2
3
from D to D
and/or trap and reꢀinject the oxime interꢀ
IL
react
3
4
conversion zones eluting from D
PressꢀTight connector (Restek Corp) was used to connect the
two columns. Isothermal operation (140 ˚C) was used for
to D . A deactivated
react
np
3
4
GC×GC experiments, whilst D
and D were operated at
react
np
various isothermal oven temperatures (140 ‒ 160 ˚C). Helium
was used as carrier gas (99.999% purity) with constant presꢀ
35
1
3
riott ), prior to study of the dynamic behavior of each indiꢀ
vidual compound. Here we report details of performance, reliꢀ
ability and applicability of e4DꢀDGC‒accTOFMS for direct
determination of isomeric ratios and rate constants of reversiꢀ
ble interconversion of 2ꢀphenylpropanaldehyde oxime. Data
are used to discuss reversible molecular processes of the interꢀ
converting molecules on the timeꢀscale of chromatographic
elution.
sure (46 psi in D ; 30.5 psi in Dreact). For stoppedꢀflow
enant
analysis, the EPC for the DS was reduced to 2 psi (typically
the lowest pressure setting that can be employed for the sysꢀ
tem, assumed to approximate stoppedꢀflow experiments). The
injector T was 230 ˚C, and injection volume 1 ꢂL. Since
QTOFMS was operated in total transfer of ion (TTI) mode
through the quadrupole sector, this is referred to hereafter as
accTOFMS. The ion source T, emission current and electron
ionization voltage were set at 280 ˚C, 4.6 ꢂA and 70 eV, reꢀ
spectively, with a mass range of 45–300 u. This system repreꢀ
sents a 4D system, because the cryogenic modulation process
mimics conventional GC×GC, the DS at the outlet of the seꢀ
cond column operates as it would in MDGC, and a CT device
EXPERIMENTAL SECTION
Reagents and chemicals. 2ꢀphenylpropanaldehyde (≥ 97%),
and 1ꢀtridecanol (≥ 97%) were purchased from SigmaꢀAldrich
(St. Louis, MO). HPLC grade nꢀhexane and dichloromethane
were purchased from Merck (Darmstadt, Germany).
at the inlet of the fourth column function to trap, refocus then
reinject the resulting oxime distribution eluting from D
3
4
D analysis: GCenant×GC ꢀDGCꢀGC ‒accTOFMS. Enantiꢀ
to
IL
np
react
4
oselective dynamic multidimensional experiments were conꢀ
ducted on an Agilent 7890A GC coupled to an Agilent 7200
accurate mass QTOFMS (Agilent Technologies, Mulgrave,
Australia), equipped with a flame ionization detector (FID)
and PAL3 Auto Sampler (CTC Analytics AG, Zwingen, Switꢀ
zerland), retrofitted with an Everest model longitudinally
modulated cryogenic system (LMCS; Chromatography Conꢀ
D . 3D experiments (eGCꢀDGCꢀGC‒accTOFMS, approxiꢀ
np
mating a GCenant ‒ GCreact ‒ GC arrangement) were conducted
using the same column configuration and conditions, but
without GC×GC modulation.
Supporting Information (Section 1ꢀ2) provides further details
of additional experimental methods (synthesis of 2ꢀ
phenylpropanaldehyde oxime, GCꢀFID system, and data hanꢀ
dling).
np
cepts, Doncaster, Australia). The GC×GC experiments were
1
performed using a MEGAꢀDEX DETꢀBeta column as D
enant
(
diethyl tertbutylsilylꢀβꢀcyclodextrin; 25 m × 0.25 mm × 0.25
RESULTS AND DISCUSSION
ꢂm d ; MEGA s.n.c, Milan, Italy), and a SLBꢀIL111 column
Interconversion of 2ꢀphenylpropanaldehyde oxime in
DGC. The 1D chromatogram obtained using a polyethylene
f
2
(1.8 m × 0.1 mm × 0.1 ꢂm d ; Supelco, Bellefonte, PA) as D
f
2
column ( D ). A deactivated PressꢀTight connector (Restek
glycol
(PEG)
column
demonstrated
that
2ꢀ
IL
Corp, Bellefonte, PA) was used to connect the two columns.
phenylpropanaldehyde oxime isomerizes on a time scale
commensurate with the chromatographic separation process.
Isomerization gives rise to a plateau between the terminal
peaks of E and Z isomers, and at a higher T of 180 ˚C, only a
single broad band is obtained, although much wider than an
inert solute of the same retention (Supporting Information
Figure S2A).
®
SUPELCOWAX 10 column (15 m × 0.25 mm × 0.25 ꢂm d )
f
3
3
was used as D reactor column ( Dreact), and a DBꢀ5ms Ultra
Inert column (30 m × 0.25 mm × 0.25 ꢂm d ; Agilent) as
f
4
fourth dimension column ( D ). A microfluidic Deans switch
np
(
DS; Agilent) for heartꢀcut (H/C) effluent switching was used
2
3
to interface the end of D to the start of D
with a deacꢀ
react
IL
tivated fusedꢀsilica tubing (DFS; 1.8 m × 0.1 mm I.D.) as
ACS Paragon Plus Environment

Page 3 of 10
Analytical Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
2
3
4
Figure 1. (A) Instrument schematic for e4DꢀDGC‒accTOFMS. DS, Deans switch; D: first, D: second, D: third, and D: fourth dimenꢀ
sion columns; DFS, deactivated fusedꢀsilica transfer line; CT, cryotrap; C and C , pressꢀfit union; C , capillary union; FID, flame ionizaꢀ
tion detector; LMCS, longitudinal modulation cryogenic system; EPC, electronic pressure control; SV, switching valve; QTOFMS, quadꢀ
rupoleꢀtimeꢀofꢀflight mass spectrometer. (B) Description of the process over the four column arrangement.
1
2
3
Separation of solutes into respective stereoisomers on the enꢀ
antioselective column was observed without evidence characꢀ
teristic of isomerism (Supporting Information Figure S3), inꢀ
dicating negligible interconversion. This is presumably beꢀ
cause the chiral selector is dispersed in a low polarity
methylpolysiloxane based matrix unlikely to promote interꢀ
by reduced retention), the distribution of the isomers can colꢀ
lapse into a single, broad peak, which should progress into a
narrower symmetric peak (i.e. t (E) ~ t (Z)) as the rate further
increases.
In Figure 2C (w for first eluted peak > 60 s; w for second
R
R
b
b
eluted peak > 80 s), both antipodes of the Z isomer were found
to coꢀelute with the later E isomer when higher T is used, as
verified by determination of their respective normalized peak
area ratios. By using a longer PEG column (i.e. increased soꢀ
lute residence time, due to reduced flow), a larger proportion
of the molecules undergo at least one interconversion event,
with an elution profile characterized by a broad band (Figure
2D, band width > 10 min). This illustrates the inability of the
separation process to adequately provide interpretation of the
different overlapping species, with peak distortion indicative
of extensive E ⇌ Z isomerization. Note that using DGC×GC
will assist in assigning all the relevant components in Figure 2
through physical separation of the total dynamic GC envelope.
It is clear that onꢀcolumn interconversions of several moleꢀ
cules at the same time can lead to quite complex peak profiles,
with confounding competing reactions resulting in misinterꢀ
preted reaction kinetics. Based on these observations, e3Dꢀ
DGC and e4DꢀDGC approaches were devised to investigate
36
conversion; it is assumed that interconversion of 2ꢀ
phenylpropanaldehyde oxime proceeds primarily in the PEG
stationary phase. As expected, an increase in T for the PEG
phase increases the rate of reaction (i.e. faster E/Z isomerizaꢀ
tion), and eventually, the interconverting isomers appeared as
a single broad peak (w ~ 39 s) without evidence of terminal
b
peaks (Supporting Information Figure S2B).
An ‘enantioselective+PEG’ coupledꢀcolumn ensemble with
different lengths of the PEG phase was tested to obtain enantiꢀ
oseparation as well as to promote onꢀcolumn isomerization.
On the enantioselective column, the chiral oxime is both reꢀ
solved into its (E)ꢀ and (Z)ꢀ isomers, with each isomer also
enantioseparated into (R) and (S) antipodes. The PEG column
then promotes interconversion of the two E/Z pairs of isomers
(i.e. R and S enantiomers), which enter the PEG column at difꢀ
ferent times. This distribution envelope differs from the distriꢀ
bution when the isomers are directly injected into a single
PEG column. If isomerization is slow compared to the chroꢀ
matographic timescale (e.g. by using a relatively short PEG
column) and the chiral stationary phase can distinguish beꢀ
tween the enantiomers, four baselineꢀseparated peaks will be
obtained.
onꢀcolumn
E/Z
isomerization
reactions
of
2ꢀ
phenylpropanaldehyde oxime.
e3DꢀDGC‒accTOFMS: D Enantioselective‒ D PEG‒ D
DBꢀ5ms. Initially, a three column strategy was evaluated, in
1
2
3
1
which D
separates the E/Z isomer maxima, and their corꢀ
enant
Figure 2A and 2B illustrate the enantioresolution, interconꢀ
version region and retentions of E(R)ꢀ, E(S)ꢀ, Z(R)ꢀ and Z(S)ꢀ
responding enantiomers. Individual physically resolved R/S
antipodes of the E/Z isomers in D
1
can then be selectively
enant
2
2
ꢀphenylpropanaldehyde oxime (note that the order of elution
H/C into Dreact, to promote isomerization of the isomer at deꢀ
fined T and pressure. Thus it is not necessary to synthetically
prepare individual pure E and Z isomers for separate study.
PEG phases result in high interconversion rates for oxime
molecules, so this was chosen as the reactor column. A liquid
CO static CT device was installed at the front section of the
Dreact to retain H/C components after which the oven was
cooled/heated to an appropriate T and/or pressure was reduced
of the R and S enantiomers has not been elucidated). A shoulꢀ
der marked with the arrow in Figure 2A preceding the Z isoꢀ
mers, will arise from interconversion of Z to E isomer as
shown in Figure S2, and here is indicated to elute before the Z
isomers on the PEG phase. At slower flow (Figure 2B) peak
distortion becomes apparent due to increased isomerization.
As interconversion kinetics increase (at higher T accompanied
2
2
ACS Paragon Plus Environment

Analytical Chemistry
Page 4 of 10
to a preset setting, then closing the CO supply to the CT reꢀ leases them to elute under isothermal conditions.
2
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2. Chromatograms of 2ꢀphenylpropanaldehyde oxime on an ensemble column pair (in enantioselectiveꢀPEG order) with PEG colꢀ
umn lengths, oven T and column pressures of (A) 15 m, 130 C, 15 psi; (B) 15 m, 130 C, 5 psi; (C) 15 m, 150 C, 5 psi; and (D) 30 m,
30 C, 15 psi. IS, internal standard (1ꢀtridecanol); I, impurity. Normalized peak area % were used to assist interpretation and peak assignꢀ
̊
̊
̊
1
̊
ments in (A), (B) and (C). Chromatographic interpretation of the interconversion region becomes difficult with an increased reaction rate.
1
An “onꢀtheꢀfly” analysis of analyte peak shapes (e.g. based
separation in Denant. Hence, direct H/C of the E/Z isomers in
27
1
on a unified equation) in which effluent is cryofocused at the
Denant might confound the correct determination of the interꢀ
conversion isomeric ratios. For instance, DS operation of a
2
static CT and directly remobilized to Dreact at the set oven T or
pressure (with CO supply to CT switched off and without a
specific region of E(R/S) isomer which contained an indetermiꢀ
2
second dedicated oven and/or pressure program) is not approꢀ
priate for this study, because determination of rate constants
depend on the employed pressure settings. Hence, the stoppedꢀ
flow approach was applied here. Since rate constants depend
on T, eGC‒GCreact with cold trapping was evaluated for transꢀ
fer of solutes, prior to oven cooling/heating and/or pressure
change. Revolatilization of cryotrapped solute at the preset T
nate amount of Z
isomer (i.e. the compound is not pure) to
(R/S)
2
Dreact for isomerization processes, might result in incorrect
estimation of the isomeric ratio, leading to erroneous rate conꢀ
stant values. Based on these observations, we devised an e4Dꢀ
DGC approach to investigate onꢀcolumn E/Z isomerization
reactions of 2ꢀphenylpropanaldehyde oxime, aiming to isolate
and study all individual pure isomers. If interconversion does
not proceed during enantiomeric separation, a modified e3Dꢀ
DGC system can be employed, (Supporting Information Figꢀ
allows isomer resolution into their corresponding E/Z isomers
3
on D , and reduces the broadening or distortion of peak
np
2
shapes arising from the D
column. System balance for H/C
ure S5) in which the isomerized solute (together with correꢀ
sponding interconversion plateau) eluting from D
cryotrapped at the inlet of D via the cryogenically cooled
react
2
operation without leakage to the other channel is confirmed
can be
react
3
(Supporting Information Figure S4).
np
Figure 3A illustrates an example of a selected (asterisked)
trapping device and the trap rapidly moved upstream to allow
isomer elution.
antipode corresponding to (Z)ꢀ2ꢀphenylpropanaldehyde oxime
successfully H/C, as shown in Figure 3B. Onꢀtheꢀfly analysis
1
2
e4DꢀDGC‒accTOFMS: D Enantioselective× D IL 111‒
2
3
4
without cold trapping of the solutes eluting from Dreact, results
D PEG‒ D DBꢀ5ms. Integrating chemical reactions (here,
in a broad band (> 4 min) for the Eꢀisomer which subsequently
overlapped with the Z isomer (Figure 3C). A much better
interconversion) and multidimensional separations (GC×GC
and H/C MDGC) in a single GC system relies on appropriate
manipulation of solutes within the columns, possible switching
of carrier flows, with precise control of pressure and DS balꢀ
peak profile was obtained by applying cryogenic trapping then
3
rapid delivery into the D separation stage (Figure 3D), alꢀ
np
lowing precise determination of the isomeric ratio. It was obꢀ
served that a small measure of interconversion of 2ꢀ
phenylpropanaldehyde oxime occurred during enantiomeric
ancing, and solute trapping and release as required. Of the five
1
discrete steps described here (Figure 1B), D
separates the
enant
E/Z isomers, and their corresponding enantiomers, according
ACS Paragon Plus Environment

Page 5 of 10
Analytical Chemistry
to their retention factors with a small extent of E/Z interconꢀ
version. By proper choice of the D column phase (here,
2
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
Figure 3. e3DꢀDGC‒accTOFMS analysis of 2ꢀphenylpropanaldehyde oxime: (A) D
FID response; the peak to be H/C is asterꢀ
3
enant
1
2
isked; (B) D
FID response, with (R/S)ꢀE H/C to D ; (C) Total ion chromatogram (TIC) of the H/C eluting from D without
react np
3 3 3
enant
cryogenic trapping at the inlet of D ; (D) TIC of the H/C eluting from D with cryogenic trapping at the inlet of D ; Imp, impuꢀ
np
np
np
rity. The static CT is cooled to ‒25 C
isomers to elute at about 6.8 min after release to D . Imp, impurity.
̊
and collects solute between 50 to 100 min, then the CO supply was switched off to allow
2
3
np
4
IL111 phase), it is possible to physically resolve (E/Z)ꢀ2ꢀ
phenylpropanaldehyde oxime over the chromatographic proꢀ
file. Individual resolved R/S antipodes of the E/Z isomers can
then be selectively H/C from the 2D separation space into
the inlet of D via the CT device. Revolatilization at the preꢀ
np
set T allows isomer resolution into corresponding E/Z isomers
on D , and reduces broadening/distortion of the peak shapes
arising from the transfer of plateau zones from Dreact. Both
4
np
3
3
Dreact, to promote isomerization of just that specific pure isoꢀ
column head pressure and EPC pressure have to be finely
tuned to obtain a balanced pressure setting for complete and
quantitative flow diversion, particularly for vacuum detection.
System balance for microfluidic switching without leakage to
the other channel is confirmed (Supporting Information Figꢀ
ure S8).
Figure 4 illustrates an example of the operation of e4Dꢀ
DGC‒accTOFMS. Fast cryogenic modulation and GC separaꢀ
tion at the end of D , gives discrete separation on D of all
2
mer at defined T and pressure. The selection of D column
phase (for GC×GC) requires careful consideration to balance
2
the degree of resolution and interconversion in the D fast colꢀ
2
umn; a D PEG and/or SLBꢀIL76 phase can promote rapid inꢀ
2
terconversion in D (Supporting Information Figure S6A and
S6B). This phenomenon can be visualized through a small inꢀ
crease in baseline (i.e. interconversion plateau) between the
2
1
2
two resolved E/Z isomers on D for a single modulation. A
enant IL
more substantial plateau will be found at higher T, accompaꢀ
nied by reduced resolution. Compromising between resolution,
interconversion and good modulated peak shape (Supporting
the interconverting isomers (Figure 4B), as shown. Figure 4B
represents the FID signal as the classical 3D plot typical of
GC×GC operation. By applying DS sampling, a selected (asꢀ
2
Information Figure S7), SLBꢀIL111 was chosen as the D
terisked)
antipode
corresponding
to
pure
in the next inꢀ
(E)ꢀ2ꢀ
3
phase.
phenylpropanaldehyde oxime is H/C to D
jection (Figure 4C). The targeted region (arrowed) has been
excised from the 2D space (i.e. it is absent on the FID signal).
Any band or region in the contour plot can thus be selected for
react
3
Without microfluidic DS transfer to Dreact, isomers are disꢀ
played as classical contour or 3D plots typical of GC×GC. A
liquid CO static CT device was installed at the front section
2
3
3
of the D
to retain components excised from the 2D separaꢀ
diversion to D
depending on how the DS ON/OFF operaꢀ
react
react
tion space, after which the oven was cooled or heated to an
appropriate T and/or with reduction of EPC pressure to apꢀ
tion is chosen. Onꢀtheꢀfly analysis without cold trapping of the
solutes eluting from Dreact, results in a broad band (> 4 min)
for the interconverted E isomer and overlapped with Z isomer
(Figure 4D). A much better peak profile is obtained by cryoꢀ
3
proximate stoppedꢀflow analysis; closing the CO supply to
2
the CT allows isothermal elution. Isomerized solute (and the
3
4
interconversion plateau) eluting from Dreact is cryotrapped at
genic trapping then rapid delivery into the D separation
np
ACS Paragon Plus Environment

Analytical Chemistry
Page 6 of 10
3
stage (negating the effect of dispersion on Dreact; Figure 4E),
allowing for precise determination of the isomeric ratio, giving
efficient trapping and transfer.
Figure
4.
e4DꢀDGC‒accTOFMS
analysis
FID response; (B)
of
2ꢀ
1
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
phenylpropanaldehyde oxime: (A) D
3D plot presentation without H/C, the peak to be H/C is asterꢀ
isked; (C) microfluidic switching of (R/S)ꢀEꢀoxime to D
enant
3
react
removes this component from the 2D FID separation space;
4
(D) TIC of the H/C E(R/S) component eluting from D without
np
4
cryogenic trapping at the inlet of D ; (E) Equivalent result to
np
4
(
D) but now with cryogenic trapping at the inlet of D ; the
np
static CT is cooled to ‒25 C̊ and collects solute between 50 to
1
00 min, then the CO supply was switched off to allow isoꢀ
2
4
mers to elute at about 6.8 min after release to D . Imp, impuꢀ
np
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
rity.
In the current set up, reaction T can be set independent of the
1
2
4
separation T in D × D and D , to control reaction rate,
enant
IL
np
3
or time of reaction (via the column flow) in Dreact. This can be
interpreted as a high resolution strategy to employ a wellꢀ
balanced microfluidic switching technique where one enantiꢀ
opure isomer (consisting of multiple modulated peaks) is H/C
3
from within the 2D separation space and transferred to Dreact,
which may be cooled/heated and/or with higher/lower pressure
settings e.g. to adjust the interconversion kinetics. Then the
4
isomerized fraction is separated in D (E/Z peaks) to deterꢀ
np
mine the de novo diastereoisomeric ratio for quantitative
measurement of interconverting isomers that lead to the obꢀ
served unresolved dynamic GC peak in 1D DGC. By repeat
injection, this can be accomplished over the whole distribuꢀ
tion, then conducted for different isothermal T or pressure setꢀ
tings.
Advantages of this design include: (i) it is unnecessary to
isolate individual antipodes from a product mixture prior to
GC analysis (i.e. analytes need not be preꢀpurified), since othꢀ
er components and impurities can be directly separated from
1
2
4
solutes in D × D and D separation stages (Figure 4),
enant
IL
np
which can be problematic for isomers which cannot be readily
synthesized; (ii) a spot, band, or region of interest in the 2D
3
space can be selected for diversion to the Dreact; (iii) enrichꢀ
ment of enantiomers (i.e. of low abundance) can be achieved
by multiple injections of the sample and directing target comꢀ
pounds into the cryogenically cooled static CT (at the inlet of
3
Dreact; Figure 1), to increase the mass of collected component.
This combines a number of contemporary GC methods (enꢀ
antioselective GC, GC×GC, inꢀoven CT, onꢀcolumn reaction
GC, and H/C MDGC) into a single GC system to study interꢀ
conversion of configurationally labile chiral molecules, or inꢀ
deed any other likeꢀprocesses, enabling separation and direct
visualization of the species under study, and determination of
rate constants. Massꢀbased deconvolution of the interconvertꢀ
ing peak envelope here, e.g. by using accTOFMS, is usually
not possible – the isomers often have the same mass spectra
(Supporting Information Figure S9). A more selective MS
dimension will enable extracted mass analysis ion product
scans, for selective detection/analysis of molecules that underꢀ
go other transformations (for instance, thermallyꢀinduced
3
7ꢀ39
transformations of terpenic molecules in essential oils ).
Reaction rate estimation. Both enantiomers of E (or Z) isoꢀ
mers have equal magnitude distribution coefficient (K) values
3
in D
(achiral phase). When forward and reverse rate conꢀ
react
stants are equal (k = k ), interconversion can be considered as
f
b
40
a reversible firstꢀorder reaction. The rate constants of isomerꢀ
ization (k) can then be calculated according to Eq S1 (Section
S3, Supporting Information)ꢀfrom the experimentally obtained
isomeric ratios of the major isomer just before elution (a ) and
i
ACS Paragon Plus Environment

Page 7 of 10
Analytical Chemistry
after elution (a ) from the reactor column with the assumption
that no isomerization takes place while the isomers were
trapped during the brief oven heat/cool step. In several cases,
ues obtained from the multidimensional experiments. Proceꢀ
dures for determination of tiso are shown in Section S4 (Supꢀ
porting Information). Note that for stereoisomers that display
different MS ionization probabilities (normally more proꢀ
nounced in electrospray ionization), the isomeric peak areas
have to be adjusted by appropriate correction factors. The calꢀ
t
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
the isomer ratios after interconversion (i.e. a ) can be < 50%
t
which are beyond the limit for calculation using Eq S1, e.g.
resulting in a negative natural logarithm. With the assumption
that kE→Z = 0 or kZ→E = 0, the pseudoꢀirreversible first order
culated isomeric ratio, t , and reaction rate constants of (E,
iso
4
0
reaction can be applied. k values can then be calculated acꢀ
R/S)ꢀ, and (Z, R/S)ꢀ2ꢀphenylpropanaldehyde oxime for differꢀ
ent pressure settings (including stoppedꢀflow analysis) are
summarized in Table 1. Isomeric ratios and rate constants deꢀ
rived using e3DꢀDGC and e4DꢀDGC show slight differences.
However, in the case of complex samples where solutes suffer
cording to Eq S2 (Section S3, Supporting Information).
In this study, it was observed that kE→Z < kZ→E where Z(R/S)
→
E(R/S) yielding > 50% E(R/S). Thus, Eq S1 cannot be applied.
However, it is also not effective to assume that Z(R/S) → E(R/S) is
a pseudoꢀirreversible first order reaction since E(R/S) → Z(R/S)
was also experimentally observed to be significant. This imꢀ
plies that Eq S2 cannot be effectively applied. Therefore, the
calculation of k values obtained using both e3DꢀDGC‒
accTOFMS and e4DꢀDGC‒accTOFMS experiments was perꢀ
formed, according to the direct solution of rate law for a reꢀ
versible firstꢀorder reaction where only one species is present
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
37ꢀ39
coꢀelution (particularly for plant derived extracts ) and/or
isomerization at high temperature (i.e. leading to fast interconꢀ
version time), an e4DꢀDGC method will be necessary. For inꢀ
stance, Figure 6 indicates elution profiles showing coalesꢀ
cence (E(S/R) overlaps both Z enantiomers) at elevated temperaꢀ
ture (160
vidual isomer is required to study its stereodynamics. The posꢀ
sible E/Z isomerization mechanisms of 2ꢀ
phenylpropanaldehyde oxime may be proposed as nitrogen
̊C ), where chromatographic isolation of pure indiꢀ
at the beginning of the reaction (a = 100%), as
i
4
1ꢀ
ꢉꢊꢄ ꢋꢄ ꢌꢂ
ꢁ
ꢁ
ꢂ
ꢄ
ꢅ
ꢆꢄ
ꢇ
ꢈ
ꢇ
ꢅ
ꢍꢎꢏ
inversion, or rotation of the OH group around the C=N bond
=
(1)
43
ꢃ
ꢄ
ꢇ
ꢆꢄ
ꢅ
, as shown in Figure S10 (Supporting Information). Whilst
the exact mechanism remains uncertain, the reported kinetic
data are still valid for the evaluation of E/Z isomerization of 2ꢀ
phenylpropanaldehyde oxime on the PEG phase.
Eq S1 is the specific case of Eq 1 when k = k = k, and Eq S2
f
b
is derived when k = 0. Since there are two parameters (k and
b
f
kb) in Eq 1, at least two independent experiments are required.
In this study, interconversion of individual (E, R/S)ꢀ, and (Z,
R/S)ꢀ2ꢀphenylpropanaldehyde oxime (ai,E = ai,Z = 100%) was
investigated and the calculation of rate constants can be obꢀ
tained according to
The precision of the e3DꢀDGC‒accTOFMS and e4DꢀDGC‒
accTOFMS was assessed by repeating interconversion experꢀ
iments five times on the same day (n = 5, intraꢀday precision),
at 140 ˚C. The relative standard deviations (RSD) obtained for
1
2
3
retention time in D (< 0.20%), D (< 2.27%), D (< 0.10%)
4
and D (< 0.10%), and isomeric ratios (≤ 1.02%) are summaꢀ
ꢉꢊꢄ
ꢋꢄ
ꢌꢂ
ꢁ
ꢁ
ꢂ,ꢐ
ꢄꢑ→ꢐꢀꢆꢀꢄꢐ→ꢑꢈ
ꢐ→ꢑ ꢑ→ꢐ ꢍꢎꢏ
=
=
(2)
(3)
rized in Supporting Information Table S1, demonstrating acꢀ
ceptable repeatability of the method. The first eluted isomer (E
isomer) of each enantiomeric pairs is enriched during the
chromatographic experiment (in 1D DGC with the PEG phasꢀ
es) because it is formed more rapidly than the first eluted isoꢀ
ꢃ,ꢐ
ꢄ
ꢐ→ꢑꢀꢆꢀꢄꢑ→ꢐ
ꢉꢊꢄ
ꢋꢄ
ꢌꢂ
ꢁ
ꢁ
ꢂ,ꢑ
ꢀꢄꢐ→ꢑꢆꢀꢄꢑ→ꢐ
ꢑ→ꢐꢀꢆꢀꢄꢐ→ꢑ
ꢈ
ꢑ→ꢐ ꢐ→ꢑ ꢍꢎꢏ
ꢃ,ꢑ
ꢄ
mer (k > k ), and the Z isomer is depleted to a greater extent
ꢀ1
1
kE→Z and kZ→E values can be obtained by fitting at,E and at,Z
calculated by using Eq 2 and Eq 3 with the corresponding valꢀ
due to its longer residence time in the reactor column.
Table 1. Forward and reverse rate constants measured by e3DꢀDGC‒accTOFMS and e4DꢀDGC‒accTOFMS methods (n=3).
E(R/S) → Z(R/S)
Z(R/S) → E(R/S)
Enantioselective multiꢀ
dimensional DGC
a
b
Peak area E
isomer (%)
73.9 ± 0.1
t
k
Peak area Z
t
k
Z→E
ꢀ3 ꢀ1
iso
E→Z
iso
ꢀ3 ꢀ1
(min)
29.6 ± 0.1
(× 10 s )
isomer (%)
(min)
(× 10 s )
e3DꢀDGC‒accTOFMS
3.7 ± 0.8
29.2 ± 0.1
34.0 ± 0.1
10.4 ± 2.2
e3DꢀDGC ‒accTOFMS
72.4 ± 0.2
37.2 ± 0.1
6.1 ± 1.0
27.6 ± 0.3
42.3 ± 0.1
15.9 ± 2.8
sf
e4DꢀDGC ‒accTOFMS
72.5 ± 0.4
73.6 ± 0.2
29.9 ± 0.1
37.4 ± 0.1
3.1 ± 0.1
4.6 ± 1.3
27.5 ± 0.4
25.5 ± 0.2
33.2 ± 0.2
42.2 ± 0.1
8.1 ± 0.2
e4DꢀDGC ‒accTOFMS
12.9 ± 3.6
sf
a
2
3
Peak area (%) of the E isomer peak, after isomerization. E(R/S) is H/C to D
(e3DꢀDGC) or D
(e4DꢀDGC) for interconverꢀ
react
react
sion.
b
2
3
Peak area (%) of the Z isomer peak, after isomerization. Z(R/S) is H/C to Dreact (e3DꢀDGC) or D
(e4DꢀDGC) for interconverꢀ
react
sion.
sf correspond to stoppedꢀflow experiments.
The described method should permit advanced chromatoꢀ
graphic chemical isolation in 2D space, excising any target
compound from the matrix, and permitting analysis of pure
material under dynamic conditions. This approach does not
specifically require a ‘dynamic’ reaction process to be studied,
but is applicable to any situation requiring isolation and analyꢀ
sis. Potentially, this configuration also offers opportunities for
investigation of interconversion processes of floral emissions
such as scent profiles from Darwin’s orchid and/or emissions
from Populus trichocarpa trees containing a series of aldoxꢀ
38,39
ime species , allowing realꢀtime online chromatographic
sample extraction, where target components, resolved from
matrix by the H/C process, are analyzed according to the
above strategies to study dynamic behavior in realꢀtime.
ACS Paragon Plus Environment

Analytical Chemistry
Page 8 of 10
mers) provided they take place on the chromatographic time
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
CONCLUSIONS
scale. It should be a valuable adjunct for improved DGC
methodology, offering high resolution and high throughput
analysis as well as straightforward investigation of oxime moꢀ
lecular interconversion processes, which could be readily exꢀ
tended to investigation of stereodynamic interconversion or
To the best of our knowledge, this is the first demonstration of
a novel automated functional online e4DꢀDGCꢀMS system
which incorporates a number of contemporary GC methods in
the one integrated instrument, to study onꢀcolumn isomerizaꢀ
tion reactions. The system provides initial assessment of enanꢀ
tiomeric composition of chiral molecules through D
ysis with FID. GCenant×GC separation was performed in order
to improve purity of solutes, then permitted interconversion of
individual oximes to be studied. Microfluidic switching selects
3
7ꢀ39
degradation processes of phytochemicals in plant extracts.
1
analꢀ
enant
ASSOCIATED CONTENT
Supporting Information
IL
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The Supporting Information is available free of charge on the
ACS Publications website.
Additional details, calculation, and figures (PDF)
AUTHOR INFORMATION
Corresponding Author
*Email: philip.marriott@monash.edu
Notes
The authors declare no competing financial interests.
ACKNOWLEDGMENT
YFW gratefully acknowledges the provision of a Monash Graduꢀ
ate Scholarship and Monash International Postgraduate Research
Scholarship. The authors would like to thank Assoc. Prof. Dr. Daꢀ
vid Lupton and Jacob Grant for synthesizing the 2ꢀ
phenylpropanaldehyde oxime used in this study, and Dr. Sungꢀ
Tong Chin for technical assistance. We are grateful to MEGA
s.n.c. for provision of the enantioselective column used in this
project. PJM acknowledges the award of a Discovery Outstanding
Researcher Award by the Australian Research Council, Grant
DP1310217.
REFERENCES
(1) Akapo, S.; McCrea, C.; Gupta, J.; Roach, M.; Skinner, W. J.
Pharm. Biomed. Anal. 2009, 49, 632ꢀ637.
(
2) Shibukawa, M.; Miyake, A.; Eda, S.; Saito, S. Anal. Chem. 2015,
7, 9280ꢀ9287.
8
(
(
3) Dugave, C.; Demange, L. Chem. Rev. 2003, 103, 2475ꢀ2532.
4) Wolf, C. Chem. Soc. Rev. 2005, 34, 595ꢀ608.
(5) Scheer, A. M.; Mukarakate, C.; Robichaud, D. J.; Nimlos, M. R.;
Carstensen, H. ꢀH.; Ellison, G. B. J. Chem. Phys. 2012, 136,
044309.
1
Figure 6. (A) D
FID response; (B) 2D plot presentation
enant
without H/C with the peak to be H/C denoted by the dotted
rectangle; (C) Microfluidic switching of compound Eꢀ2ꢀ
phenylpropanaldehyde oxime from the end of DIL to D
removes the compound from the 2D separation space; the H/C
component is absent on the FID signal whilst the remaining
signal is almost identical to (B).
(6) Welch, C. J.; Kress, M. H.; Beconi, M.; Mathre, D. J. Chirality
2
003, 15, 143ꢀ147.
2
3
(7) Trapp, O. In Differentiation of Enantiomers II; Schurig, V., Ed.;
Springer International Publishing: Cham, 2013; pp 231ꢀ269.
react
(
8) Krupčík, J.; Mydlová, J.; Májek, P.; Šimon, P.; Armstrong, D. W.
J. Chromatogr. A 2008, 1186, 144ꢀ160.
(
(
9) Trapp, O.; Schurig, V. J. Am. Chem. Soc. 2000, 122, 1424ꢀ1430.
10) Trapp, O.; Schurig, V. Chem. Eur. J. 2001, 7, 1495ꢀ1502.
regions of the chromatographic elution from 2D space of a
single pure stereoisomer for interconversion on Dreact. Reacted
products are quantified on D . This integrates a decision proꢀ
(11) Trapp, O.; Trapp, G.; Schurig, V. J. Biochem. Bioph. Meth. 2002,
54, 301ꢀ313.
(12) Wolf, C.; Xu, H. Tetrahedron Lett. 2007, 48, 6886ꢀ6889.
3
4
np
(
(
13) Maier, F.; Trapp, O. Angew. Chem. Int. Ed. 2012, 51, 2985ꢀ2988.
14) Wistuba, D.; Trapp, O.; GelꢀMoreto, N.; Galensa, R.; Schurig, V.
Anal. Chem. 2006, 78, 3424ꢀ3433.
cess based on pure component selection from either eGC
(e3DꢀDGC) or eGC×GC (e4DꢀDGC) separation. Determinaꢀ
tion of de novo diastereoisomeric ratios by monitoring the
whole time course of conversion of pure H/C isomers permits
exact determination of isomeric ratios and calculation of rate
constants. This is an advantage compared with usual 1D DGC
methods, which require iterative comparison of experimental
and simulated elution profiles to determine interconversion
barriers. The described methodology could be applied to inꢀ
vestigation of fast reactions (coalescing interconverting isoꢀ
(
15) Schoetz, G.; Trapp, O.; Schurig, V. Anal. Chem. 2000, 72, 2758ꢀ
2
764.
(16) Heegaard, N. H. H.; Jørgensen, T. J. D.; Cheng, L.; Schou, C.;
Nissen, M. H.; Trapp, O. Anal. Chem. 2006, 78, 3667ꢀ3673.
(17) Liberek, B.; Augustyniak, J.; Ciarkowski, J.; Plucinska, K.;
Stachowiak, K. J. Chromatogr. A 1974, 95, 223ꢀ228.
(
18) Oswald, P.; Desmet, K.; Sandra, P.; Krupcik, J.; Armstrong, D.
W. Chirality 2002, 14, 334ꢀ339.
ACS Paragon Plus Environment

Page 9 of 10
Analytical Chemistry
(
19) Wolf, C.; Pirkle, W. H.; Welch, C. J.; Hochmuth, D. H.; König,
W. A.; Chee, G. ꢀL.; Charlton, J. L. J. Org. Chem. 1997, 62, 5208ꢀ
210.
20) Mazzanti, A.; Chiarucci, M.; Bentley, K. W.; Wolf, C. J. Org.
Chem. 2014, 79, 3725ꢀ3730.
(38) Nielsen, L. J.; Møller, B. L. Phytochemistry 2015, 120, 3ꢀ18.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(39) Irmisch, S.; McCormick, A. C; Boeckler, G. A.; Schmidt, A.;
Reichelt, M.; Schneider, B.; Block, K.; Schnitzler, J. ꢀP.;
Gershenzon, J.; Unsicker, S. B.; Köllner, T. G. Plant Cell 2013, 25,
4737ꢀ4754.
(40) Krupčík, J.; Oswald, P.; Májek, P.; Sandra, P.; Armstrong, D. W.
J. Chromatogr. A 2003, 1000, 779ꢀ800.
5
(
(
21) Reich, S.; Trapp, O.; Schurig, V. J. Chromatogr. A 2000, 892,
4
87ꢀ498.
(22) Trapp, O.; Trapp, G.; Kong, J.; Hahn, U.; Vögtle, F.; Schurig, V.
Chem. Eur. J. 2002, 8, 3629ꢀ3634.
(41) Trapp, O.; Shellie, R.; Marriott, P.; Schurig, V. Anal. Chem.
2003, 75, 4452ꢀ4461.
(42) Andrzejewska, A.; Lapinski, L.; Reva, I.; Fausto, R. Phys. Chem.
Chem. Phys. 2002, 4, 3289ꢀ3296.
(43) Nsikabaka, S.; Harb, W.; RuizꢀLópez, M. F. J. Mol. Struc-
THEOCHEM 2006, 764, 161ꢀ166.
(
(
(
23) Schurig, V. Chirality 1998, 10, 140ꢀ146.
24) Trapp, O.; Schurig, V. Comput. Chem. 2001, 25, 187ꢀ195.
25) Trapp, O.; Shellie, R.; Marriott, P.; Schurig, V. Anal. Chem.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
003, 75, 4452ꢀ4461.
(26) Krupcík, J.; Oswald, P.; Spánik, I.; Májek, P.; Bajdichová, M.;
Sandra, P.; Armstrong, D. W. Analusis 2000, 28, 859ꢀ863.
(
(
(
(
27) Trapp, O. Anal. Chem. 2006, 78, 189ꢀ198.
28) Trapp, O. Chirality 2006, 18, 489ꢀ497.
29) Trapp, O. J. Chromatogr. B 2008, 875, 42ꢀ47.
30) Bihlmeier, A.; Rotzler, J.; Rickhaus, M.; Mayor, M.; Klopper,
W. Phys. Chem. Chem. Phys. 2015, 17, 11165ꢀ11173.
31) Reich, S.; Schurig, V. J. Micro. Sep. 1999, 11, 475–479.
32) Chifuntwe, C.; Zhu, F.; Huegel, H.; Marriott, P. J. J.
Chromatogr. A 2010, 1217, 1114ꢀ1125.
(
(
(
33) Marriott, P. J.; Trapp, O.; Shellie, R.; Schurig, V. J. Chromatogr.
A 2001, 919, 115ꢀ126.
(34) Kröger, S.; Wong, Y. F.; Chin, S. ꢀT.; Grant, J.; Lupton, D.;
Marriott, P. J. J. Chromatogr. A 2015, 1404, 104ꢀ114.
(
(
35) Mitrevski, B.; Marriott, P. J. Anal. Chem. 2012, 84, 4837ꢀ4843.
36) Marriott, P. J.; Aryusuk, K.; Shellie, R.; Ryan, D.; Krisnangkura,
K.; Schurig, V.; Trapp, O. J. Chromatogr. A 2004, 1033, 135ꢀ143.
37) Filippi, J. ꢀJ.; Cocolo, N.; Meierhenrich, U. J. J. Chromatogr. A
2015, 1383, 134ꢀ143.
(
ACS Paragon Plus Environment

Analytical Chemistry
Page 10 of 10
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table of Contents Graphic
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment