Evaluation of stereo and chemical stability of chiral compounds

Article abstract of DOI:10.1002/chir.20941

A stopped-flow bidimensional recycle HPLC (sf-BD-rHPLC) configuration has been used to investigate simultaneously the stereo and chemical stability of labile chiral compounds. The single enantiomers of a racemate can be separated on chiral column (first dimension) and each one can be trapped in the achiral column (second dimension) that works as reactor.By filling the achiral column with the appropriate aqueous buffers it is possible to evaluate the stability of the trapped enantiomer toward aqueous buffer itself. It was possible to recycle the reaction products formed in the chiral column (first dimension) where they are separated by a second six valve port. The reaction rate constants were calculated for the different processes occurred in the achiral column by means of corresponding peak areas. The method was applied to a pharmacological active compound: (±)7-chloro-5-ethyl-3-methyl-3,4-dihydro-2H-benzo[1,2,4] thiadiazine 1,1-dioxide ((±)-1) to evaluate enantiostability and hydrolysis in conditions similar to those of biological fluid. A classical batchwise kinetic method was used to calculate rate constants of hydrolysis and enantiomerization at the same temperature and in the same solvents used in sf-BD-rHPLC. The good agreement of the results obtained validate the novel procedure developed. Furthermore, the results generated off-line were used to determine the influence of solvents on the racemization of (±)-1. Copyright

Full text of DOI:10.1002/chir.20941

CHIRALITY 23:851–859 (2011)
Evaluation of Stereo and Chemical Stability of Chiral Compounds
*
GIUSEPPE CANNAZZA, UMBERTO BATTISTI, MARINA M. CARROZZO, LIVIO BRASILI,
DANIELA BRAGHIROLI, AND CARLO PARENTI
Dipartimento di Scienze Farmaceutiche, Universita` degli Studi di Modena e Reggio Emilia, Via Campi 183, Modena, Italy
ABSTRACT
A stopped-flow bidimensional recycle HPLC (sf-BD-rHPLC) configuration has
been used to investigate simultaneously the stereo and chemical stability of labile chiral com-
pounds. The single enantiomers of a racemate can be separated on chiral column (first dimen-
sion) and each one can be trapped in the achiral column (second dimension) that works as
reactor. By filling the achiral column with the appropriate aqueous buffers it is possible to evalu-
ate the stability of the trapped enantiomer toward aqueous buffer itself. It was possible to
recycle the reaction products formed in the chiral column (first dimension) where they are sep-
arated by a second six valve port. The reaction rate constants were calculated for the different
processes occurred in the achiral column by means of corresponding peak areas. The method
was applied to a pharmacological active compound: (6)7-chloro-5-ethyl-3-methyl-3,4-dihydro-
2H-benzo[1,2,4]thiadiazine 1,1-dioxide ((6)-1) to evaluate enantiostability and hydrolysis in
conditions similar to those of biological fluid. A classical batchwise kinetic method was used to
calculate rate constants of hydrolysis and enantiomerization at the same temperature and in the
same solvents used in sf-BD-rHPLC. The good agreement of the results obtained validate the
novel procedure developed. Furthermore, the results generated off-line were used to determine
C
VV
the influence of solvents on the racemization of (6)-1. Chirality 23:851–859, 2011.
2011
Wiley-Liss, Inc.
KEY WORDS: stopped-flow bidimensional recycle HPLC; chiral; hydrolysis; benzothiadiazines
INTRODUCTION
If enantiomerization process does not occur during chro-
matographic run, it is possible to apply stopped-flow methods
in which separated enantiomers are interconverted for a cer-
tain time at the desired temperature by stopping the mobile
phase flow in the chiral column. Original and interconverted
enantiomers are then separated and enantiomerization rate
constants are calculated from the corresponding peak
areas.11
In both dynamic and stopped-flow methods, enantiomeriza-
tion occurs in the chiral environment of CSP that could
exerts stereoselective perturbing effects on the interconver-
sion process. It is important to note that rate constants calcu-
lated from dynamic and stopped-flow chromatographic inves-
tigations are apparent rate constants (ka1pp = kꢀap1p), which are
weighted means of the different enantiomerization rates in
the mobile phase and stationary phase.12,19,20 Efforts have
been done to combine the advantages of the on-column chro-
matographic methods with the option to perform enantiome-
rization in an achiral environment that does not influence
interconversion process rate.
The evidence is that frequently only one enantiomer is
pharmacologically active, whereas the other may be inactive
or toxic has lead, over the past two decades, to an even more
stringent demand for enantiopure drugs. Since recently
national and international organizations such as food and
drug administration require extensive stereochemical infor-
mation on chiral drugs like their enantiomeric stability and
enantiomeric purity, different methods have been developed
to evaluate configurational and conformational lability of chi-
ral compounds.1,2
Literature reports different approaches to evaluate stereo-
chemical integrity like spectroscopic, chirooptical, and chro-
matographic methods.3–11 Since several chiral stationary
phases (CSPs) are recently commercially available, chromato-
graphic techniques are currently widely used. Three experi-
mental approaches have been exploited to study interconver-
sion process by chromatography: classical batchwise methods
combined with enantioselective separations, stopped-flow
methods, and dynamic chromatography methods. The first
approach is time consuming and cumbersome because it
requires isolation of single pure enantiomer followed by an ex-
amination of batchwise racemization kinetics. Because
dynamic and stopped-flow chromatographic methods do not
require single pure enantiomers, they are widely used. If enan-
tiomerization process occurs at the same time-scale of enantio-
separation, it is possible to observe a characteristic plateau
between the two peaks corresponding to the enantiomers. The
height of plateau is proportionally to enantiomerization rate,
and by dynamic HPLC (DHPLC), it is possible to evaluate the
kinetic parameters of enantiomerization by several methods
like the theoretical plate model, the stochastic model, the con-
tinuous flow model, peak deconvolution method, approxima-
tion function, and unified equation of chromatography.3–18
Multidimensional liquid chromatography (MDLC), that
uses different column in series, is a high performance sepa-
ration system that recently is gaining popularity for the sepa-
ration of complex samples.
Recently, we have developed a stopped-flow multidimen-
sional chromatographic system (sf-MDHPLC) that permits to
Contract grant sponsor: MIUR (Ministero dell’Istruzione, dell’Universita` e
della Ricerca, Italy).
*Correspondence to: Giuseppe Cannazza, Dipartimento di Scienze Farmaceu-
tiche, Universita` degli Studi di Modena e Reggio Emilia, Via Campi 183,
41100 Modena, Italy. E-mail: giuseppe.cannazza@unimore.it
Received for publication 28 January 2010; Accepted 9 December 2010
DOI: 10.1002/chir.20941
Published online 20 September 2011 in Wiley Online Library
(wileyonlinelibrary.com).
C
VV
2011 Wiley-Liss, Inc.

852
CANNAZZA ET AL.
hydrolysis in acidic medium, it is important to evaluate ster-
eointegrity and chemical stability of (6)-1.27–31
Preliminary studies conducted by dynamic chromatogra-
phy have been demonstrated that (6)-1 undergoes rapid
enantiomerization in aqueous solvent and hydrolysis in
acidic mobile phase. sf-BD-rHPLC method was applied to
compound (6)-1 to evaluate enantiomerization and hydroly-
sis rate constants in conditions similar to physiological
fluids.
Moreover, the enantiomerization and hydrolysis kinetic
parameters were calculated by off-column method on single
pure enantiomer to evaluate matrix effect exerted by achiral
stationary phases on both enantiomerization and hydrolysis
processes.
Fig. 1. Compound investigated.
perform enantiomerization in an achiral environment by
using three columns in series: the first and the third column
are chiral, and the second column is achiral.19 After complete
separation of the enantiomers in the first chiral column,
at conditions where enantiomerization is suppressed, one
enantiomer is introduced into the second column using a
heart-cut technique. Then, the flow is stopped and enantio-
merization is performed at a chosen temperature for selected
time. Afterward, the resumed flow of the mobile phase intro-
duces the enantiomers to the third column where they are
separated. By this way, it is possible to perform enantiomeri-
zation in an achiral environment that does not exert stereose-
lective perturbing effects. Anyway, sf-MDHPLC method
constrains to evaluate enantiomerization rate constants in
mobile phase that could exert some influence on process
rate, whereas from a pharmaceutical point of view it is impor-
tant to evaluate stereointegrity of drugs in conditions similar
to that of physiological fluids.14,21,22
More recently, we have developed an on-column stopped-
flow bidimensional recycling HPLC (sf-BD-rHPLC) proce-
dure to obtain an enantiomeric enrichment starting from a
racemic mixture.23,24 The method developed was applied to
two chiral compounds of pharmaceutical interest, (6)(R,S)-
2,3,3a,4-tetrahydro-1H-pyrrolo[2,1-c][1,2,4]benzothiadiazine
5,5-dioxide and (6)-7-chloro-3-methyl-3,4-dihydro-2H-1,2,4-
benzothiadiazine 1,1-dioxide ((6)IDRA21), because the phar-
macological activity of the two benzothiadiazine derivatives
investigated has been ascribed to only one enantiomer. Start-
ing from a racemic mixture, it was possible to obtain about
95% of pure enantiomer.
EXPERIMENTAL
Instrumentation
The chromatographic apparatus was a Shimadzu LC-10AD Pump (Shi-
madzu Italia, Milano), a Merck Hitachi L-6200A Pump (Merck KGaA,
Darmstadt, Germany), a Rheodyne 7725 manual injector equipped with a
50 ll sample loop. As detector was used a Merck Hitachi L-7400UV
(Merck KGaA, Darmstadt, Germany). Chromatograms were recorded
with a Jasco J-700 program (Jasco Europe, Italy, Milan). Two Rheodyne
7000 valves were used to switch the mobile phase flow. Column tempera-
ture regulation was obtained with a Jasco CO-2067 column oven (Jasco
Europe, Italy, Milan). The columns used were Chiralcel OD-RH [cellu-
lose tris(3,5-dimethylphenylcarbamate); 150 mm 3 4.6 mm I.D.; 5 lm]
purchased from Daicel chemical industries, IllKirch, France, Chiraspher
NT [poly(N-acryloyl-S-phenylalanineethylester); 250 3 10 mm I.D.; 5
lm] purchased from Merck, German, Supelcosil LC-18 (250 mm 3 4.6
mm I.D.; 5 lm) purchased from Supelco Italy, Milan, and Supelcosil LC-
8 (250 mm 3 4.6 mm I.D.; 5 lm) purchased from Supelco Italy, Milan.
Melting points were determined with an electrothermal apparatus, and
they are uncorrected. IR spectra were recorded on a PerkinElmer Model
1600 FT-IR spectrometer. 1H NMR spectra were recorded with a
Brucker DPX 200 spectrometer with DMSO-d6 as solvent and tetrame-
thylsilane as external standard. Chemical shifts (d) are in part per million
and coupling constant (J) in hertz. Multiplicities are abbreviated as fol-
lows: s, singlet; d, doublet; dd, double doublet; t, triplet; m, multiplet.
The IR spectra were recorded with an FT-IR spectrophotometer Digilab
Scimitar Series FTS 2000. All pH measurements were made using Orion
Research EA940 pH-meter.
Synthesis
(6)7-Chloro-5-ethyl-3-methyl-3,4-dihydro-2H-benzo[1,2,4]thia-
diazine 1,1-dioxide ((6)-1). The compound was synthesized as
previously described by Philips et al.25
In this work, we have applied configuration system of
on-column sf-BD-rHPLC procedure to the aim to conduct
enantiomerization not only in achiral environment but also in
aqueous solution at different pH similar to physiological con-
ditions using a chiral and an achiral columns in series.
Recently, (6)7-chloro-5-ethyl-3-methyl-3,4-dihydro-2H-ben-
zo[1,2,4]thiadiazine 1,1-dioxide ((6)-1) (Figure 1) has been
attracted particular attention for its activity as positive alloste-
ric modulators of 2-amino-2,3-dihydro-5-methyl-3-oxo-4-isoxa-
zolepropanoic acid (AMPA) receptor indicating that it could
be useful in the treatment of neurodegenerative disorders
such as cognitive disorders, schizophrenia, depression, and
Parkinson’s disease.25 Moreover, because it has been
recently suggested that only S-enantiomer of (6)-1 is the
active one on AMPA receptor, it is relevant to develop chro-
matographic methods to separate the two enantiomers to
investigate their enantiomeric stability.26 Since previous stud-
ies have indicated that benzothiadiazine type compounds
undergo to rapid enantiomerization in aqueous solutions and
Yield 70.6%, mp 180–1838C, FTIR 3399, 3223, 2952, 1491, 1437, 1304,
1282 cm21 1H NMR (DMSO-d6) d 5 1.13 (t, 3H, J 5 7.4 Hz), 1.48 (d,
.
3H, J 5 7.1 Hz), 2.52 (m, 2H), 4.80 (m, 1H), 6.24 (s, 1H), 7.21 (d, 1H, J
5 2.2 Hz), 7.35 (d, 1H, J 5 2.2 Hz), 7.70 (d, 1H, J 5 11.8 Hz).
Chromatography
Analytical enantioseparation. Separation of enantiomers of (6)-1
was carried out isocratically at 08C on Chiralcel OD-RH column using
water:acetonitrile (60:40, v/v) as mobile phase. The compound was dis-
solved in ethanol and subsequently diluted 1:10 (v/v) with mobile phase
at final concentration of 100 lg/ml. The injection volume was 50 ll. The
detector was set at 254 nm. HPLC-grade acetonitrile and ethanol were
obtained from Baker.
Preparative enantioseparation. Pure (1) and (2) enantiomers
were obtained by semipreparative HPLC on Chiraspher NT column with
fraction collection of the respective peaks. The mobile phase consisted
of n-hexane and tetrahydrofuran (THF) 70:30 (v/v). The compound was
dissolved in THF at final concentration of 2 mg/ml. The injection volume
was 500 ll. The detectors was set at 254 nm. The collected fractions cor-
Chirality DOI 10.1002/chir

STEREO AND CHEMICAL STABILITY OF CHIRAL COMPOUNDS
853
Fig. 2. Schematic represententation of the stopped-flow bidimensional HPLC system. M, mobile phase; P, pump; V1, valve 1; V2, valve 2; I, injector; C1, chiral
column; C2, achiral column; W, waste; D, detector. Step 1: Racemic mixture is injected and enantioseparated on the chiral column C1. Step 2: One of the two
eluted enantiomers was trapped into the achiral C18 column by switching valve 2. Step 3: C2 is filled with different pH buffers by pump 2 and enantiomerization
and hydrolysis were performed at different temperatures for a set period of time. Step 4: By switching valve 1, the original mobile phase is reinforced in the achiral
column and the enantiomers and hydrolysis product were separated in the first chiral column.
responding to the enantiomers were analyzed by injection on the same
column and in the same chromatographic conditions. HPLC-grade n-hex-
ane and THF were obtained from Baker.
ture originated, consisting in both the two enantiomers and in the hydro-
lysis product was separated in the chiral column 1 (Step 4).
Off-Column Method
Chromatographic parameters. The separation factor (a) was cal-
culated as k2/k1 and retention factors (k1 and k2) as k1 5 (t1 2 t0)/t0,
where t1 and t2 refer to the retention times of the first and second eluted
enantiomers. The resolution factor (Rs) was calculated by the formula Rs
5 2(t2 2 t1)/(w1 1 w2), where w1 and w2 are the peak widths at base for
the first and second eluted enantiomer.
Single enantiomers of (6)-1, obtained by semipreparative chromatog-
raphy, were used to calculate enantiomerization and hydrolysis rates in
different solvents. Enriched enantiomers were dissolved in ethanol and
subsequently diluted 1:100 (v/v) with buffer (1 ml) at different pH (pH
5 1.2, 2.2, 4.2, and 7.4) at final concentration of 10 lg/ml and warmed at
378C.
Racemization was monitored by chromatography on the Chiralcel OD-
RH column by using water:acetonitrile 60:40 (v/v) as mobile phase. Hy-
drolysis was monitored by chromatography on a supelcosil LC-18 col-
umn by using water:acetonitrile (50:50, v/v) as mobile phase. Four
repeats of each experiment were made at different times. The peak cor-
responding to hydrolysis product of (6)-1 was identified by injecting
authentic 3-ethyl-2-aminobenzensulfonamide.
Buffer solution at pH 1.20 was prepared by mixing 407 ml of 0.2 M hy-
drochloric acid with 93 ml of 0.2 M potassium chloride and diluting with
water to one liter. Chloroacetate buffer solution at pH 2.20 was prepared
by mixing 50.99 ml of 0.1 M chloroacetic acid with 2.76 ml of 0.1 M
KOH and diluting with water to 100 ml. Acetate buffer solution at
Stopped-Flow Recycling Bidimensional HPLC
The HPLC procedure developed is described in Figure 2. The racemic
mixture was injected, and the individual enantiomers were separated on
the first chiral column in conditions where hydrolysis and enantiomeri-
zation do not take place (Step 1). In this first step, one of the two enan-
tiomers is trapped in the achiral column by switching valve at appropri-
ate time (Step 2). By pump 2, it is possible to fill the achiral column with
a selected solvent in which hydrolysis and racemization occur in stopped
flow conditions (Step 3). The achiral column and solvent were previously
warmed to the desired temperature. After an appropriate time, the origi-
nal mobile phase was reinforced in column 2 by using pump 2. The mix-
Chirality DOI 10.1002/chir

854
CANNAZZA ET AL.
pH 4.20 was prepared by mixing 43.03 ml of 0.1 M acetic acid with 9.93
ml of 0.1 M KOH and diluting with water to 100 ml. Phosphate buffer so-
lution at pH 7.40 was prepared by mixing 6.54 ml of 0.02 M KH2PO4
with 28.97 ml of 0.01 Na2HPO4 and diluting with water to 100 ml.
Evaluation of Kinetic Rate Constants and Free Energy
Barriers of Enantiomerization
The kinetic rate constants can be calculated by fitting the data
obtained to Eq. 1
ꢀ
ꢁ
ao
ln
¼ 2ktenant
ð1Þ
2at ꢀ ao
where k is the rate constant of forward or backward enantiomerization
(sec21), ao the peak of the initial enantiomer before enantiomerization
(5100%), at the relative peak area of the enantiomer remaining after
enantiomerization time t (%) (i.e., at 5 100-conversion), and tenant the
enantiomerization time(s).
From the kinetic rate constants, the corresponding activation energies
of enantiomerization (rotational energy barriers) DG#(T) can be
calculated by the Eyring equation (Eq. 2):
ꢀ
ꢁ
kh
DG#ðTÞ ¼ ꢀRT ln
ð2Þ
jkBT
where k is kinetic rate constant, kB the Boltzmann constant (kB
5
1.380662 3 10223 J K21), h Planck’s constant (h 5 6.626176 3 10234
J sec), R the universal gas constant (R 5 8.31441 K J mol21), j the trans-
mission coefficient (j 5 0.5 for the reversible microscopic interconver-
sion), and T the temperature (K).
Calculation of Kinetic Rate Constants of Hydrolysis
The kinetic rate constants can be calculated by fitting the data to Eq. 3:
At ¼ A0 expðꢀkitiÞ
ð3Þ
where ki is the rate constant of hydrolysis (sec21), Ao the peak of the
(6)-1 before hydrolysis (5100%), At the relative peak area of the injected
compound remaining after hydrolyzation time ti (%) (i.e., at 5 100-conver-
sion) and ti the hydrolysis time(s).
RESULTS AND DISCUSSION
Chromatography
Enantioseparation of (6)-1 was conducted with Chiralcel
OD-RH column in reversed phase condition using water:ace-
tonitrile (60:40, v/v) as mobile phase. Baseline enantiomeric
resolution was obtained performing enantioseparation at 08C
(k1 5 5.21, k2 5 6.35, a 5 1.22, Rs 5 2.23) (Fig. 3). The
observed plateau between the two peaks corresponding to
the two enantiomers of (6)-1 at temperatures above 258C
in reversed phase conditions indicates a rapid chiral intercon-
version during the chromatographic analysis (Fig. 3).
By changing the pH of the mobile phase (buffer pH
2.20:ACN 60:40 (v/v)), a third peak showed up in the chro-
matogram before the first eluted enantiomer (Fig. 3c).
Fig. 3. On-column enantiomerization and hydrolysis of (6)1; flow rate:
0.5 ml/min, Column: Chiracel OD-RH (150 mm 3 4.6 mm I.D.; 5 lm);
(a) water:acetonitrile 60:40 (v/v); T 5 08C, (b) water:acetonitrile 60:40 (v/v);
T 5 378C, (c) buffer pH 2.20 (0.05 M NaClO4/HClO4): acetonitrile 60:40
(v/v); T 5 378C.
Because we have recently reported that (6)-7-chloro-3-
methyl-3,4-dihydro-2H-1,2,4-benzothiadiazine
1,1-dioxide
((6)IDRA21), a benzothiadiazine structurally related to (6)-
1, undergoes to rapid enantiomerization and hydrolysis to
2-amino-5-chlorobenzensulfonamide and acetaldehyde in
aqueous acidic solvents, it is plausible that (6)-1 hydrolyzed
in 2-amino-7chloro-5-ethylbenzensulfonamide and acetalde-
hyde in acidic aqueous mobile phase.30
Authentic 7-chloro-2-amino-5-ethylbenzensulfonamide was
prepared and analyzed under identical experimental condi-
tions used for enantioseparation of (6)-1. The retention
times of unknown peak and pure 2-amino-7-chloro-5-ethylben-
zensulfonamide were identical confirming the identity of
hydrolysis product (Scheme 1). Recently, we have proposed
a mechanism to explain enantiomerization and hydrolysis of
chiral benzothiadiazine type compounds that involves a
stereochemical inversion at chiral carbon atom via an imine
achiral intermediate that could evolve toward hydrolysis
acid catalyzed of imine bond to give the corresponding ben-
zensulfonamide and aldehyde (Scheme 1) (Cannazza et al.,
submitted).
Chirality DOI 10.1002/chir

STEREO AND CHEMICAL STABILITY OF CHIRAL COMPOUNDS
855
Scheme 1. Proposed enantiomerization and hydrolysis mechanism.
whereas the forward (k1m) and backward (kꢀm1) rate constants
of enantiomerization should be the same in achiral environ-
ment of the mobile phase. If km is known from independent
measurements, ks1 and kꢀs 1 can be calculated according to the
following equations:
Dynamic Chromatography
By dynamic chromatography, it is possible to investigate
kinetics of enantiomerization and hydrolysis of chiral labile
compounds like benzothiadiazine type compounds.
Recently, Trapp et al. have been developed the useful soft-
ware DCXplorer to calculate kinetic constants of first-order
processes that occur during chromatographic run by unified
equation of chromatography.16–18,32–38
The software was herein used to calculate enantiomeriza-
tion rate constants of (6)-1 on column Chiralcel OD-RH and
water:acetonitrile 60:40 (v/v) or buffer pH 2.20:ACN 60:40
(v/v) as mobile phases at temperature of 158C and 378C
(Tables 1 and 2).
1
kenantðꢀÞ
1 þ kenantðꢀÞ
kenantðþÞ
1 þ kenantðþÞ
ka1pp
¼
km
þ
k1s
ð4Þ
ð5Þ
1 þ kenantðꢀÞ
1
ka1pp
¼
km
þ
ksꢀ1
1 þ kenantðþÞ
enant
(2)
enant
(1)
where k
and k
are the retention factors of the (2)
and (1) enantiomers, respectively.14,20
Moreover, the same software has permitted to determine
hydrolysis rate constants of (6)-1 on C18 column. By using
water:acetonitrile 60:40 (v/v) or 80:20 (v/v) as mobile phases
at temperatures of 158C and 378C, it was observed that no hy-
drolysis of (6)-1 takes place during chromatographic run.
By using buffer pH 2.20:ACN 60:40 (v/v) as mobile phase,
the plateau between the hydrolysis peak and (6)-1 peak is
too low to be determine by DCXplorer, suggesting that the
presence of organic modifier and/or C18 stationary phase
influence hydrolysis process rate. Accordingly, hydrolysis
rate constant was studied on C18 column and buffer pH
2.20:ACN 80:20 (v/v) as mobile phase at temperature of
378C. In these conditions, it was possible to determine by
DCXplorer the kinetic rate constant of hydrolysis (ki 5 5.5 3
1024 sec21) indicating that hydrolysis process rate has been
influenced by the presence of organic modifier, decreasing
with the increasing of percentage of acetonitrile (Table 3).
Because dynamic chromatographic methods constrain cal-
culation of kinetic constants in the mobile phase that usually
contains organic modifier in different percentage, the reac-
tion rates obtained could be different from those in pure sol-
vent calculated by off-column methods.
The rate of (6)-1 enantiomerization in buffer pH 2.20:ace-
tonitrile 60:40 (v/v) and at temperature of 378C was deter-
mined independently by off-column method (km 5 2.88 6
0.69 3 1024 sec21) using pure enantiomers collected by
semipreparative chromatography on Chiraspher NT.
The km value obtained have been used to calculate, by
eqs. 4 and 5, ks1 and kꢀs 1 (k1s 5 2.36 6 0.23 3 1024 sec21 and
ksꢀ1 5 2.43 6 0.15 3 1024 sec21).
The results confirm that the CSP stabilized the initial state
of enantiomers being km greater than ks.
Stopped-Flow Bidimensional HPLC
MDLC is an high performance separation system that is
gaining popularity for the separation of complex samples.
Two-dimensional liquid chromatography are currently
widely using to separate enantiomers of complexed biologi-
cal samples by coupling achiral-chiral columns. Previously,
we have developed an on-column sf-BD-rHPLC procedure
to obtain an enantiomeric enrichment starting from a race-
mic mixture.24 The configuration system has been used to
calculate enantiomerization and hydrolysis rate constants
of (6)-1.
Figure 2 shows a schematic representation of the sf-BD-
rHPLC, in which column C1 is the column containing the
CSP and C2 is the achiral column that works as reactor. At
the beginning of the experiment, the racemic mixture was
injected and the individual enantiomers were separated on
the first chiral column in conditions where hydrolysis and
racemization do not take place. The switching valve 2 located
Moreover, the rate constants of enantiomerization
obtained by dynamic chromatography investigations are
apparent rate constants, which are weighted means of the dif-
ferent enantiomerization rates in mobile phase and stationary
phase. If enantioseparation of a labile compound can be
accomplished, the forward (ks1) and backward (ksꢀ1) rate
constants of enantiomerization may be different in a chiral
environment or in the presence of a CSP, respectively,
Chirality DOI 10.1002/chir

856
CANNAZZA ET AL.
Chirality DOI 10.1002/chir

STEREO AND CHEMICAL STABILITY OF CHIRAL COMPOUNDS
857
TABLE 3. Hydrolysis of (6)-1 performed with sf-BD-rHPLC, off-column, and DCXplorer methods at 378C
Hydrolysis
method
aapp
E1i
aapp
E2i
Solvent
k
(sec21
)
k
(sec21
)
kmi (sec21
)
kci app (sec21
)
Sf-BD-rHPLC
Off-column
Buffer pH 2.20
Buffer pH 7.40
Buffer pH 1.20
Buffer pH 2.20
Buffer pH 4.20
Buffer pH 7.40
Buffer pH 2.20:ACN
(80:20 (v/v))
2.92 6 0.67 3 1024a
8.01 6 0.81 3 1026a
2.80 6 0.44 3 1024a
7.76 6 0.96 3 1026a
2.29 6 0.11 3 1023a
5.15 6 0.06 3 1024a
1.78 6 0.04 3 1025a
8.79 6 0.07 3 1026a
DCXplorer
5.5 6 0.01 3 1024a,
*
**
H2O:ACN (80:20 (v/v))
Columns Chiralcel OD-RH, Supelcosil LC-18. Column operation temperature at 08C. n 5 4. Time intervals for hydrolysis at 37.58C 5 50, 100, 150, 200, 300.
a,Rate constants in C18 column.
*, mobile phase buffer pH 2.20:ACN 80:20 (v/v).
**, any hydrolysis occurred during chromatographic run.
between the two columns is initially set to direct the flow
from column 1 through a detector to waste. At the appropri-
ate time by switching the valve two, one of the two enantiom-
ers was trapped into the achiral column that was filled with
the selected aqueous buffer by pump 2 (Step 2). Because the
aqueous buffer selected as racemization medium has a low
eluitropic force, it was possible to retain on C18 column the
trapped enantiomer. The enantiomerization and hydrolysis
were affected by heating at selected temperature column 2 in
conditions of stopped-flow. Afterward, the original mobile
phase was resumed in column 2 by pump 2 and the valve 1
was switched in position (b), allowing to the mobile phase to
run through the chiral column 1. By this way, the enantiom-
ers and hydrolyzed product were introduced into the chiral
column where they are separated.
lated from the corresponding peak areas, from the enantio-
merization time and from the enantiomerization temperature
as described in the experimental part by fitting the data to
eqs. 1 and 2. For both enantiomers, the corresponding plots
gave straight lines with reasonable correlation coefficients.
It is assumed that the precision of DG# is not influenced
within the experimental standard deviation by the warm-up
times on the column 2 for stopped-flow multidimensional
experiments, in which separation is carried out at low
temperature (08C) and the enantiomerization at higher
temperature (378C). However, in case of very short enantio-
merization times, this systematic error increases and it can-
not be ignored.39–41
By sf-BD-rHPLC, it was possible to study enantiomeriza-
tion and hydrolysis processes, because the two reactions
occur simultaneously in the achiral column under the same
stopped-flow conditions.
The chromatogram reported in Figure 4 refers to a repre-
sentative example of sf-BD-rHPLC of compound (6)-1: peak
18 arises from the hydrolysis process and peaks 28 and 38
corresponding to the enantiomers of analytes injected, one
arise from enantiomer trapped in the achiral column and the
other one corresponding to its interconverted product. The
elution order was validate by calculating retention times of
single enantiomers and hydrolysis product on both column.
The kinetic parameters for both enantiomers (apparent
rate constants of pseudo first-order of enantiomerization
The hydrolysis apparent rate constants of pseudo first-
aapp
E1i
order for both enantiomers (1k
and 1kaEa2pip) were calculated
from the residual concentration of the parent racemate and
by fitting the time-residual concentration data to a first-order
exponential decay (Eq. 3), as reported in the experimental
part. For both enantiomers, the corresponding plots gave
straight lines with reasonable correlation coefficients.
The enantiomerization and hydrolysis of (6)-1 were inves-
tigated at pH 2.20 and 7.40 and at temperatures of 158C and
378C (Tables 1–4).
1ka1app and kaapp and apparent free energy barrier of forward
1
ꢀ1
DG1]aapp and backward DG]aapp enantiomerization) were calcu-
ꢀ1
Fig. 4. Chromatogram of enantiomerization and hydrolysis experiment of (6)-1 performed by sf-BD-rHPLC procedure: peak 18 hydrolysis product, peak 28and
38 corresponding to enantiomers of (6)-1. Columns: chiracel OD-RH (150 mm 3 4.6 mm I.D.; 5 lm); Supelcosil LC-18 (250 mm 3 4.6 mm ID; 5 lm). Mobile
phases: water:acetonitrile 60:40 (v/v); chloroacetate buffer at pH 2.20. Time interval for enantiomerization/hydrolysis at 37.58C 5 20 min. Flow: 0.5 ml/min.
Chirality DOI 10.1002/chir

858
CANNAZZA ET AL.
TABLE 4. Hydrolysis of (6)-1 performed with sf-BD-rHPLC and DCXplorer methods at 158C
aapp
ꢀ1
Hydrolysis method
Sf-BD-rHPLC
Solvent
Buffer pH 2.20
ka1app (sec21
)
k
(sec21
)
k
c app (sec21
)
2.85 6 0.31 3 1025a
2.82 6 0.20 3 1025a
Buffer pH 7.40
Buffer pH 2.20:ACN (80:20 (v/v))
H2O:ACN (80:20 (v/v))
*
*
**
**
DCXplorer
Columns: Chiralcel OD-RH, Supelcosil LC-18. Column operation temperature 08C. n 5 4. Time intervals for hydrolysis at 37.58C 5 50, 100, 150, 200, 300, 600, and
1200.
aRate constants in C18 column.
*, any hydrolysis occurred in 6 h.
**, any hydrolysis occurred during chromatographic run.
The sf-BD-rHPLC method allowed to determine kinetic
rate constants of enantiomerization and hydrolysis in an
achiral environment, permitting to suppress the perturbing
effect of CSP originated from the diastereomeric nature of
the transient adduct formed by interaction of the enantiom-
ers with the chiral selector.
As shown in Tables 1 and 2, the apparent kinetic rate con-
stants of pseudo first-order of the forward and backward
enantiomerization are quite similar as expected since any
stereospecific interactions occurs with stationary phase.
By using a C18 column as enantiomerization and hydroly-
sis reactor, it was possible to change the mobile phase sol-
vent with the desired buffers because aqueous solvent are
unable to elute the trapped enantiomer. For pharmaceutical
compounds became important to evaluate stereo and chemi-
cal-stability in conditions similar to those they will meet in bi-
ological fluids when administered in vivo.
culate enantiomerization and hydrolysis rate constants in
achiral environment and in selected solvents. The method
was applied to the pharmacological active compound (6)-1
to evaluate its enantio- and chemical-stability in conditions
similar to that of body fluid. The data obtained by off-column
method were in good agreement with those obtained by
sf-BD-rHPLC validating the accuracy and applicability of the
novel method developed.
The sf-BD-rHPLC method offers different advantages
respect previously developed stopped-flow HPLC system. It
allows to determine kinetic rate constants of enantiomeriza-
tion and hydrolysis in an achiral environment, permitting to
suppress the perturbing effect of CSP originated from the
diastereomeric nature of the transient adduct formed by
interaction of the enantiomers with the chiral selector. It is
possible to apply sf-BD-rHPLC method to racemic mixture
that is only partial enantioresolved by cutting the peaks at
appropriate times. sf-BD-rHPLC system requires only one
chiral column.
Anyway, enantiomerization and hydrolysis occur in C18 sta-
tionary phase that might not be inert in principle, it can act
as promoter or inhibitor agent, increasing or decreasing the
enantiomerization and hydrolysis rate constants of the stud-
ied compound.
sf-BD-rHPLC method together with dynamic methods
could be a valuable tool in the investigations of stereo and
chemical integrity of chiral compounds particularly in the
pharmaceutical field where the knowledge of chemical stabil-
ity is an important prerequisite for the further drug develop-
ment process.
To the aim to evaluate this matrix effect, enantiomerization
and hydrolysis rate constants were calculated by classical
batchwise off-column method in the same buffers and at the
same temperatures used for sf-BD-rHPLC (Tables 1–4).
Comparisons of the kinetic rate constants of enantiomeri-
zation and hydrolysis obtained by sf-BD-rHPLC and off-col-
umn methods indicate that the processes were significantly
faster in mobile phase than in C18 column, indicating that
stationary phase could exerts an inhibitory effects. The enan-
tiomerization and hydrolysis rate constants determined by
sf-BD-rHPLC were performed in the hydrophobic conditions
of C18 stationary phase that could be decrease the rates of
both processes. The influence of polarity on enantiomeriza-
tion and hydrolysis rate constants have been evaluated by
using the less apolar C8 stationary phase instead of C18 sta-
tionary phase as achiral column in sf-BD-rHPLC. The rate
constants obtained on C8 column (pH 2.20 and T 5 158C)
were 5.16 6 0.19 3 1024 sec21 and 3.89 6 0.20 3 1025 sec21
for enantiomerization and hydrolysis, respectively, that are
slight higher than that calculated in the same conditions on
C18 column indicating that polarity of stationary phase could
exert some effects on enantiomerization and hydrolysis rates.
LITERATURE CITED
1. Announcement. FDA’s policy statement for the development of new ster-
eoisomeric drugs. Chirality 1992;4:338–340.
2. De Camp WH. The FDA perspective on the development of stereoiso-
mers. Chirality 1989;1:2–6.
3. Schurig V, Keller F, Reich S, Fluck M. Dynamic phenomena involving
chiral dimethyl-2,3-pentadienedioate in enantioselective gas chromatogra-
phy and NMR spectroscopy. Tetrahedron Asymmetry 1997; 8:3475–3480.
4. Bringmann G, Heubes M, Breuning M, Gobel L, Ochse M, Schoner B,
Schupp O. Atropisomerization barrier of configurationally unstable biaryl
compounds, useful substrates for atroposelective conversionsa to axially
chiral byiaryls. J Org Chem 2000;65:722–728.
5. Mannschreck A, Andert D, Eiglsperg E, Gmahl E, Buchner H. Chiropti-
cal detection. Novel possibilities of its application to enantiomers. Chro-
matographia 1988;25:182–188.
6. Mannschreck A, Kiessel A. Enantiomerization during HPLC on an opti-
cally active sorbent. Deconvolution of experimental chromatograms.
Chromatographia 1989;28:263–266.
7. Trapp O, Schoetz G, Schurig V. Determination of enantiomerization bar-
riers by dynamic and stopped-flow chromatographic methods. Chirality
2001;13:403–414.
8. Wolf C. Stereolabile chiral compounds: analysis by dynamic chromatog-
raphy and stopped-flow methods. Chem Soc Rev 2005;34:595–608.
CONCLUSIONS
9. Schurig V, Reich S. Determination of the rotational barriers of atropi-
sf-BD-rHPLC has been successfully applied to calculate
enantiomerization and hydrolysis rate constants of stereo-
and chemical-labile chiral compounds. It was possible to cal-
someric polychlorinated biphenyls (PCBs) by
a novel stopped-flow
multidimensional gas chromatographic technique. Chirality 1998;10:
316–320.
Chirality DOI 10.1002/chir

STEREO AND CHEMICAL STABILITY OF CHIRAL COMPOUNDS
859
10. Schurig V. Contributions to the theory and practice of the chromato-
graphic separation of enantiomers. Chirality 2005;17:205–226.
zides: a new subgroup of AMPA Receptor modulators with improved
affinity. Bioorg Med Chem 2002;10:1229–1248.
11. Mannschreck A, Zimmer H, Pustet N. The significance of the HPLC time
scale: an example of interconvertible enantiomers. Chimia 1989;43:
165–166.
26. Harpsøe K, Varming T, Gouliaev AT, Peters D, Liljefors T. Identification
of a putative binding site for 5-alkyl-benzothiadiazines in the AMPA
receptor dimer interface. J Mol Graph Model 2007;26:213–225.
12. Veciana J, Crespo MI. Dynamic HPLC: a method for determination rate
constants, energy barriers, and equilibration constants of molecular
dynamic processes. Angew Chem Int Ed Engl 1991;30:74–77.
27. Cannazza G, Braghiroli D, Baraldi M, Parenti C. Chiral resolution of
the enantiomers of 7-chloro-3-methyl-3,4-dihydro-2H-1,2,4-benzothia-
diazine 1,1-dioxide using high performance liquid chromatography on
cellulose-based chiral stationary phases. J Pharm Biomed Anal 2000;23:
117–125.
13. Jung M, Schurig V. Determination of enantiomerization barriers by
computer simulation of interconversion profiles: enantiomerization of
diaziridines during chiral inclusion gas chromatography. J Am Chem Soc
1992;114:529–534.
28. Cannazza G, Braghiroli D, Tait A, Baraldi M, Parenti C, Lindner W.
Studies of enantiomerization of chiral 3,4-dihydro-1,2,4-benzothiadiazine
1,1-dioxide type compounds. Chirality 2001;13:94–101.
14. D’Acquarica I, Gasparrini F, Pierini M, Villani C, Zappia G.
Dynamic HPLC on chiral stationary phases: a powerful tool for the
investigation of stereosimulation processes. J Sep Sci 2006;29:1508–
1516.
29. Cannazza G, Braghiroli D, Iuliani P, Parenti C. Energy barrier determina-
tionof enantiomerization of chiral 3,4-dihydro-1,2,4-benzothiadiazine
1,1-dioxide type compounds by enantioselective stopped-flow HPLC.
Tetrahedron Asymmetry 2006;17:3158–3162.
15. Bu¨rkle W, Karfunkel H, Schurig V. Dynamic phenomena during enan-
tiomer resolution by complexation gas chromatography. A kinetic study
of enantiomerization. J Chromatogr 1984;288:1–14.
30. Cannazza G, Carrozzo MM, Braghiroli D, Parenti C. Enantiomerization
and hydrolysis of (6)-7-chloro-3-methyl-3,4-dihydro-2H-1,2,4-benzothiadia-
zine 1,1-dioxide by stopped-flow multidimensional high-performance liq-
uid chromatography. J Chromatogr A 2008;1212:41–47.
16. Trapp O, Schurig V. Stereointegrity of Troger’s base: gas chromato-
graphic determination of the enantiomerization barrier. J Am Chem Soc
2000;122:1424–1430.
31. Carrozzo MM, Cannazza G, Battisti U, Braghiroli D, Parenti C. Simulta-
neously determination of enantiomerization and hydrolysis kinetic
parameters of chiral N-alkylbenzothiadiazine derivatives. Chirality
2010;22:389–397.
17. Trapp O, Schurig V. ChromWin-a computer program for the determina-
tion of enantiomerization barriers in dynamic chromatography. Comput
Chem 2001;25:187–195.
18. Trapp O, Schurig V. Approximation funtion for the direct calculation of
rate constants and Gibbs activation energies of enantiomerization of race-
mic mixture from chromatographic parameters in dynamic chromatogra-
phy. J Chromatogr A 2001;911:167–175.
32. Trapp O. A unified equation for fast and precise access to rate constants
of first order reactions in dynamic and on-column reaction chromatogra-
phy. Anal Chem 2006;78:189–198.
33. Trapp O. Fast and precise access to enantiomerization rate constants in
dynamic chromatography. Chirality 2006;18:489–497.
19. Weseloh G, Wolf C, Koenig WA. A new application of capillary zone
electrophoresis: determination of energy barriers of configurationally
labile chiral compounds. Angew Chem Int Ed Engl 1995;34:1635–
1636.
34. Trapp O. The unified equation for the evaluation of first order reactions
in dynamic electrophoresis. Electrophoresis 2006;27:534–541.
35. Trapp O. A dynamic molecular probe to investigate catalytic effects and
Joule heating in enantioselective micellar electrokinetic chromatography.
Electrophoresis 2007;28:691–696.
20. Cannazza G, Carrozzo MM, Braghiroli D, Parenti C. Enantiomerization
of chiral 2,3,3a,4-tetrahydro-1H-pyrrolo[2,1-c][1,2,4]benzothiadiazine 5,5-
dioxide by stopped-flow multidimensional HPLC.
2008;875:192–199.
J Chromatogr B
36. Trapp O, Weber SK, Bauch S, Hofstadt W. High-throughput screening of
catalysts by combining reactions and analysis. Angew Chem Int Ed Engl
2007;46:7307–7310.
21. Zimmermann T, Pustet N, Mannschreck A. Separation and thermal race-
mization of the enantiomers of spiro[cyclohexadiene-dihydroacridines].
Monats Chem 2000;131:1083–1090.
37. Trapp O. Gas-chromatographic high-throughput screening techniques in
catalysis (review article). J Chromatogr A 2008;1184:160–190.
22. Gasparrini F, Lunazzi L, Mazzanti A, Pierini M, Pietrusiewicz KM,
Villani C. Comparison of dynamic HPLC and dynamic NMR in the
study of conformational stereodynamics: case of the enantiomers of
38. Trapp O. A novel software tool for high-throughput measurements of
interconversion barriers: DCXplorer. J Chromatogr B 2008;875:42–47.
a
hindered secondary phosphine oxide.
J Am Chem Soc 2000;
39. Schurig V, Reich S. Determination of the rotational barriers of atropiso-
meric polychlorinated biphenyls (PCBs) by a novel stopped-flow multidi-
mensional gas chromatographic technique. Chirality 1998;10: 316–320.
122:4776–4780.
23. Lorenz K, Yashima E, Okamoto Y. Enantiomeric enrichment of stereola-
bile chiral spiro compounds by dynamic HPLC on chiral stationary
phases. Angew Chem Int Ed Engl 1998;37:1922–1925.
40. Reich S, Schurig V. Stopped-flow multidimensional gas chromatography:
a
new method for the determination of enantiomerization barriers.
J Microcolumn Sep 1999,11:475–479.
24. Cannazza G, Carrozzo MM, Battisti U, Braghiroli D, Parenti C. On-line
racemization by high-performance liquid chromatography. J J Chroma-
togr A 2009;1216:5655–5659.
41. Reich S, Trapp O, Schurig V. Enantioselective stopped-flow multidi-
mensional gas chromatography: determination of interconversion bar-
25. Philips D, Sonnenberg J, Arai AC, Vaswani R, Krutzik PO, Kleisli T,
Kessler M, Granger R, Lynch G, Chamberline AR. 50-Alkyl-benzothiadia-
rier of 1-chloro-2,2-dimethylaziridine.
487–498.
J
Chromatogr
A
2000;892:
Chirality DOI 10.1002/chir