the assay - university of toronto · 2020. 4. 6. · the development of a stereospecific assay for...
TRANSCRIPT
THE DETELOPMENT OF A STEREOSPECIFIC ASSAY FOR SELEGILINE
HM>ROCHLORIDE
Maureen A. McLaughlin
A thesis submitted in conformity with the requirements for the degree of
Master of Science in Pharmacy
Graduate Department of the Faculty of Pharmacy
University of Toronto
O Copyright by Maureen A. McLaughiin (1997)
National Library 1*1 of Canada Bibliothèque nationale du Canada
Acquisitions and Acquisitions et 8ibliographic Services services bibliographiques
395 Wellington Street 395. rue Wellington Ottawa ON K1A O N 4 ûttawaON KIAOiU4 Canada -da
Your itk votre reterenm
Our iUe Notre reterence
The author has granted a non- L'auteur a accordé une Licence non exclusive licence ailowing the exclusive permettant à la National Library of Canada to Bibliothèque nationale du Canada de reproduce, loan, distribute or sen reproduire, prêter, distribuer ou copies of this thesis in microform, vendre des copies de cette thèse sous paper or electronic formats. la forme de microfiche/^, de
reproduction sur papier ou sur format électronique.
The author retains ownership of the L'auteur conserve la propriété du copyright in this thesis. Neither the droit d'auteur qui protège cette thèse. thesis nor substantial extracts from it Ni la thèse ni des extraits substantiels may be printed or othemise de celle-ci ne doivent être imprimés reproduced without the author's ou autrement reproduits sans son permission. autorisation.
T h e Development of a Stereospecific Assay for Selegiline hydrochloride
Maureen A. McLaughlin
Master of Science 1997
University of Toronto, Faculty of Pharmacy
Seiegihe ([RI [LI - (-) - N-, a -dirnethyl - N-(prop-2-ynyl) phenylethylarnine
hydrochloride) is a potent type B monoamine oxidase inhibitor used in the heatment of
Parkinson's disease. The major metabolites ((-) amphetamine and (-) methamphetamine]
of Çeleghe are devoid of cenbal nervous system stimulation, however a small amount
of (+) Seleghe wodd produce the cenbal nervous system stimulants (+) amphetamine
and (+) methamphetamine. Therefore, a chiral assay for Selegiline drug substance and
tablets has been developed by derivatization with (-) menthyl diloroformate followed
by separa tion of the dealkylated carbamates on a silica column.
The reaction of Selegiline with (-) menthyl diloroformate consistently resulted in
the formation of one product Kising from the deavage of the propargyl group. The
developed assay procedure has been applied to the determination of (-) SelegiLine in
commercial tablet preparations. Analysis of ten tablets showed a recovery of 99.70h
(+ 1.4) of the label daim. The assay is not suffiuently sensitive to alIow determination of
Selegiline in biological samples. In atternpts to potentiaiiy increase sensitivity additional
experiments with two achiral chloroformates, ethyl chloroformate and 9-fluorenyl
methyl chloroformate were performed. Experiments with these two adiiral reagents
resulted in the formation of a mixture of produa carbamates.
Acknowledgments
1 would like to express my sincere gratitude to my advisor Dr. J. Barry
Robinson for bis patience and advice throughout my research endeavors.
In addition 1 wodd like to thank the other members of my thesis cornmittee;
Dr. J. 1. Thiessen, Dr. M. Spino and Dr. J. R. Ballinger for their time and
interest in my work. The financial support (Grant in Aid) provided by Apotex
Inc. was greatly appreciated. The support and friendship of my fellow
graduate students: Susan, Gay, Mike and Rosa will always be remembered.
Lastly I would like to thank the anonymous Parkinson patient who
generously provided me with the Selegiline tablets which were used in this
research.
Table of Contents
.......................... Introduction .................................................. .,,............ 1
Stereochemical Concepts ................................................................................................ 6
S tereochemical Techniques ............................................................................... ....... 10 Chromatograp hy ............................................................................................................ 14
Ma thematical Principles of Chromatography ......................................................... 21 Chiral Separations-A Direct Approach using Chird Stationary Phases ............ 26
Chiral Separation-Direct Separations using Chiral Mobile Phase Additives ... 35
Chiral Separation via the Indirect Method Using Chiral Derivatizing
Agents ............................................................................................................................... 37
Chiral Separations of Amines with Particular Reference to Amphetamine
.......................................................................................................... and its Analogues 45
.................................................................................... Discussion of Experimental 56
.......................... Physical. Chemical Characteris tics of Selegiline hydrochloride 56
Preliminary Studies of Po tential Assay Procedures ................... ...... ............... 56
........................... ................................... Chiral Columns and Chiral Additives ..... 57
................................................................. ................... Reverse Phase Separa tions .. 59
5 mg Preparation . Methamphetamine and Selegiline carbama te ..................... 64
Reverse Phase Chromatography . Lntemal Standard Selection .......................... 65
................................. Normal Phase Separations ..... 3
Normal Phase Chromatography- Interna1 Standard Selection ........................... 77
............... Preparation of Calibration Cumes . Methamphetamine .... .. ..... ... 81
. . Preparation of Calibration Cumes . Seleplme ........................................................ 82
............................. Application of the Assay to Seelegiline hydrochloride Tablets 86
Studies Employing other Chloroformate Reagents .......................................... eJ37
S tudies wi th 9-Fluorenylmethyl Chloroformate (FMOC) .................................... 89
Experimental ................................................................................................................... 97 ................... Achiral Separation . Reverse Phase Chromatographie Conditions 98
...................... Chiral Separation . Normal Phase Chromatographie Conditions 98
Preparation of [RI and [SI Methamphetamine Calibration Curves/ Separa tion
by Normal Phase Chroma tography ........................................................................... 99
Preparation of Initial [RI and [SI Selegiline Calibration Cumes
........... Normal Phase ................. .... ........... ..... 100 ............................ Preparation of [RI Selegiline Standard Calibration C u ~ e #3 101
............................................................. Extraction Procedure of Selegdine Table ts 101
................... E thyl Chloroformate Deriva tiza tion Reactions .. ............................ 102 . .
FMOC Derivatrzation Reactions ................................ .................... ............. -102 Resul ts of Tablet Assays ......................................................................................... 114
Discussion ..................................................................................................................... 118
References ...... ........... ...................... ............................................................................ 123
List of Tables. Charts and G r a ~ h s
Table I- Types of Chiral Stationary Phases ................................................................ 28
Table II a Method Development using various solvent systems ........................ 61 Table III O Reverse Phase = Interna1 Standards ......................................................... 70
Table IV- Normal Phase = Internal Standards ......................................................... 79
[RI Methamp he tamine Calibration Cume Data .................................................... 104 ............................................................ [RI Methamphe tamine Calibraiion Graph 105
[SI Methamp he tamine Calibration Curve Da ta ................... .......................... .... 106
[S 1 Methamp hetamine Calibration Graph ............... .... ..............=...................... 107 .......................................... [SI Selegiline Calibration Curve Data ................... ... 108
. [SI Selegiline Calibration Curve Graph ................... ........................................... 109
.................... [RI Selegiline Calibration C u ~ e (#1) Data ..................................... 110
[RI Selegiline Calibration Curve (#1) Grap h ...................... ............................... 111
............................................................ [RI Selegiline Calibration Curve (#3) Data 112
[RI Selegiline Calibration Cunre (#3) Graph .......................................................... 113 Chinoin Tablet Data- Lot# 0600289 ................................................... -1 15
Chiesi Jurnex Tablets O First Batch . Lot# 118 ...................... .............................. 116
Chiesi Jumex Tablets O Second Batch O Lot# 118 .................................................... 117
List of Figures
Figure 1 O Selegiline and Metabolites ........................................................................... 2
Figure 2 O 2. 3- Dihydroxybutanoic acid ....................................................................... 8
Figure 3 . Types of Chromatography ......................................................................... 15
Figure 4 = Chernical Structure of Silica ...................................................................... 18
Figure 5 O Silica "'T'ails" ............................................................................................... 18
Figure 6 O Chromatogram Depicting the Separation of two Compounds ......... 22
Figure 7 = Resolution of Neighbonng Peaks .......................................................... -22 Figure 8 . Generalized Chiral Recognition Mode1 Between a Chiral Stationary
................................................ Phase and a chiral Dinitrobenzamide Enantiorner 30
Figure 9 . Structure of B- Cyclodextrin Units ................... ..................... ............. 32
................................................... . Figure 10 Some Chiral Denvatizing Reagents 39
................. ................... . Figure 11 2-Aryl Propionic Acid Derivatives .................... 40
........................ . Figure 12 Potential Chiral Derivatizing Reactions for Amines 42
............................ . Figure 13 Stereochemical Precursors of S (4 Amphetamine 48
Figure 14 . Derivatization Reaction of an enantiomer of Methamphetamine
............ with 2r 3,4,6.tetra.O.acetyl.B. D.glucopyranosy1 isothiovanate (GITC) 48
..... . Figure 15 Reaction of 2 -Naphthyl Chloroformate with a Tertiary amine 50
..... . . Figure 16 Tertiary amines Derivatized with 2 Naphthyl Chloroformate 51
............ . Figure 17 Promethazine derivatized with (-1 Menthyl chloroformate 53
Figure 18 . Possible Products of the denvatization of Selegiline with (-1
.............................................................................................. Menthyl Chlorof ormate -54
Figure 19 - Chromatogram - (k) Methamphetamine derivatized with (-)
Menthyl Chloroformate ............................................................................................. 62
.................. Figure 20 - Relationship Between Peak Separation and Resolution 63
Figure 21 - Chromatogram - (0) Menthyl carbarnate derivatives of (k)
Methamp hetamine, 70130 , methanollwater ......................................................... -66
Figure 22 - Chromatogram - (0) Menthyl carbamate derivatives of (k)
Methamphetamine - Effect of 2% isopropanol ................... ...........................-.......67 Figure 23 - Chromatograrn - (-1 Menthyl carbamate derivatives of (f)
...................................................... Methamp hetamine - Eff ect of 5% isopropanol 68
Figure 24 - Chromatogram - (-) Menthyl carbamate denvatives of (k)
. Methamphetamine Peaks starting to tail ................... ..................................... .... 71 Figure 25 - (-1 Menthyl carbamate derivatives of [RI Amphetamine Intemal
Standard, Peaks are tailing ........................................................................................... 72
Figure 26 - Chromatograrn - (-) Menthyl carbarnate derivatives of (_+)
Selegiline, Nonnal Phase Cluomatography ............................................................ 74 Figure 27 - Chromatogram - (-1 Menthyl carbamate derivatives of (k)
Methamp hetamine, Normal Phase Cluomatograp hy, 96/4 ............................ *75
Figure 28 - Chromatogram - (-1 Menthyl carbamate derivatives of (&)
Methamphetamine, Normal Phase Chrornatograp hy, 95/5 .......................... .... ..76 Figure 29 - Chromatogram - (k) Selegiline carbamates with a, a-
Dimethylbenzylamine ................................... ............................................................ 80
Figure 30 - Chrornatogram - [RI Selegiline and a, a- Dimethylbenzylamine
derivatized with (0) menthyl chloroformate ........................ ............................. 85
Figure 31 - Chromatogram - (f) Methamphetamine derivatized with Ethyl
chloroformate ................... .. ..... .. ................................... -91
Figure 32 - Chromatogram - (k) Desmethyl selegiline derivatized with Ethyl
chloroformate ................ ......,, ................................................................................. 92
Figure 33 - Chromatogram - (c) Selegiline derivatized with Ethyl
................................................................................................................. chloroformate 93
Figure 34 - Chromatogram - (2) Methamphetamine derivatized with
.................................................................................... FMOC ............. .............*.....*.... 94
A~pendix
Preparation of a - Phenylacetoacetonitrile ................... ........... ....O................... A
Preparation of Benzy 1 Methyl Ketone ........................................................................ B
Preparation of (f) -N, a -Dimethy l-13-pheny le thylamine ................... ... .......... C
Preparation of (t) - N-propargyl-a-methy 1-B-phenylethylamine ....................... D
Prepara tion of (I) -N-a -Dimethyl-N-propargyl-B-p henylethylamine ............... E
Preparation of the Menthyl carbamate of (+) Methamphetamine ................ F
Preparation of the Menthyl carbarnate of (t) Desmethyl Selegiline ................ ..G
Preparation of N-Acetyl-2-Phenylisopropylamine ............................................... H
Preparation of 2-Amino-3 phenylpropane .......... ....................................................... I
Preparation of N-Formyl-2-phenylisopropylamine ................................................. J
............................................................. Preparation of a-a-Dimethylbenzylamine K
Introduction
The work described in this thesis concerns studies aimed at the
development of a stereospecific assay for the drug Selegiline ([LI [RI- (-)-N, a-
dimethyl-~-(prop-2-~~yl) phenylethylamine hydrochloride) (figure l) which is
one of the agents currently employed in the treatment of Parkinson's disease.
Parkinson's syndrome or Parkinson's disease, first described by James Parkinson
in 1817, is an age related progressive disorder involving the loss of dopaminergic
nigrostriatal neurons and is clinically characterized by akinesia and tremor.
Biochemically patients with Parkinson's disease show a pronounced deficiency
of dopamine in the striatum (caudate nucleus and putamen) (Cobias, G. C.,
1971).
Selegiline is a potent monoamine oxidase type B inhibitor which has been
shown to delay the progression of the disease when taken in its early stages, Le.,
shortly after diagnosis. Additionally, when used in combination with L-Dopa,
[LI-(-)-3-(3,1-dihydroxy phenyl) alanine, it allows for a reduction in the levels of
L-Dopa, and therefore a reduction in toxic side effects normally associated with
L-Dopa therapy (Birkmayer, W., Riederer, P., Ambrozi, L. and Youdim, M. B. H.,
1977)
ïhe discovery of Selegiline was prompted by a series of pharmacological
studies with racemic methamphetamine, racemic amphetamine and
dextroamphetamine which indicated that, depending on the dose, one could
achieve an increase in catecholamine or serotonin release. In 1963, the N-
propargyl derivative of N-methyl amphetarnine1 was synthesized and found to
l Although the compound employed was the racemic form of Methamphetamine it should be
noted that the name Methamphetamine as employed in the USP %XII, designates the (+)
enantiomer.
Selegiline and Metabolites
Amp he tamine
N-Methyl amphetamine
be a potent irreversible inhibitor of monoamine oxidase. in order to possibly
diminish the amphetamine like stimulus, the individual enantiomers were
prepared and it was found that the (-) enantiomer was approximately 500 times
more potent than the (+) enantiomer as a MAO inhibitor in vitro (Knoll, J., 1983)
Monoamine oxidase (EC 1.4.3.4.; MAO) is a flavin-adenosine-dinucleo tide
(FAD) containing enzyme located on the outer mitochondria membrane. The
enzyme is thought to be responsible for the regulation of the levels of certain
biogenic amines within the brain and peripheral tissues. Specifically the enzyme
catalyzes the oxidative deamination of biogenic amines such as dopamine (DA),
noradrenaline (NA) and serotonin (5-HT) to their corresponding aldehydes
(equation l), and is therefore involved in their biological inactivation in vivo
(Kinemuchi, H., Fowler, C.J. and Tipton, K.F., 19û4)
Prior to the advent of Selegiline, certain monoamine oxidase inhibitors
had been employed briefly in the treatment of hypertension and depression.
However, due to the inhibition of monamine oxidase, patients eating foods rich
in certain amines e.g. tyramine found in cheese, red wine, pickled herring etc.,
were unable to metabolize the amines which resulted in increased adrenergic
activity. Studies showed that, although Selegiline inhibited monoamine oxidase,
it also inhibited the release of biogenic amines from the nerve terminais and was
free of the "cheese effect" in vitro and in vivo (Knoll, J., Vizi, E.S. and Somogyi,
G., 1968 and Knoli, J., 1978)
In 1968 Johnston differentiated monoamine oxidase into two forms,
designated A and B. The differences in the two forms are based on their substrate
and inhibitor selectivity. MAO-A preferentially deaminates serotonin and
noradrenaline and iç selectively inhibited by clorgyline (approximately 10-8 M)
(Johnson, J. P., 1968), harmaline, and LY51641 (Fuller, R. W., 1968), whereas
MAO-B deaminates B-phenylethylamine (and the synthetic substrate
benzyiamine) and is selectively inhibited by L-Selegiline (approximately 10-8 M)
(Knoll, J., and Magyar, K., 1972)
Selegiline belongs to a class of inhibi tors known as "suicide" or irreversib le
inhibitors. Such inhibitors forrn a covalent bond with the enzyme, resulting in
prolonged or irreversible inhibition. Selegiline reacts with MAO-B initiallv in a
reversible reaction forming a non covalent complex, followed by a tirne
dependent irreversible reaction leading to the reduction of the enzyme bound
FAD and simultaneous oxidation of the Selegiline. Oxidized Selegiline reacts
covalently with FAD at the N-5 position of the isoalloxazine ring. The following
equation describes the Lnhibition mechanism ;
where [EI] represents the initial enzyme inhibitor complex, [E14] the enzyme
activated inhibitor complex and [EI"] the irreversible inhibitor complex
(Geriach, M., Riederer, P. and Youdim, M. B. H., 1992)
Selegiline's major metabolites are (-) amphetamine, (-) methamphetamine
and (-) desmethyl selegiline. Presently there are both chiral and achiral methods
available to assay methamphetamine and amp hetamine, either in combination or
separatelv. In addition there is also an HPLC method that separates racemic
amphetamine, methamphetamine, desmethyl selegiline and Selegiline (Beaulieu,
N., Cyr, T.D., Graham, S. J. and Lovering, E. G., 1991). Currently, the proposed
USP method for Selegiline (in-process revision) is an achiral separation
(Convention, U. S. P., 1991) . However, Pharmeuropa has suggested an
alternative chiral method to detect (S)-(+)-Selegihe, suggesting the possibility of
(+) Selegiline in (-) Selegiline raw material and tablets (Pharmeuropa, 1996). The
presence of (+) Selegiline in tablets would result in the formation of (+)
amphetamine and (+) methamphetamine as metabolites. Clearly, additional
central nervous system stimulation would not be beneficial for a Parkinson's
patient. Taking into account the above, together with the fact that Selegiline is
marketed solely as the (-) enantiomer, it seems prudent that a stereospecific assay
for Selegiline be developed.
Therefore, the main objective of this research has been to develop a
chiral assay for Selegiline using high pressure liquid chromatography. The
developed assay is based upon the reaction of a tertiary amine (Selegiline)
with a chloroformate derivative [(-) menthyl chloroformate] followed by
dealkylation to yield the carbarnate ester of a secondary amine.
While the assay is shown to be capable of application to both Selegiline
drug substance and Selegiline in its dosage form, it is not of sufficient
sensitivity to allow detection of Selegiline in biological fluids. Accordingly,
additional studies have been initiated involving the reaction of Selegiline,
desmethyl selegiline and methamphetamine with other chloroformate
derivatives in an attempt to identify other potential derivatizing reagents
which could provide the basis of a more sensitive assay.
Stereochemical Concepts
There has been an increased interest in the area of stereochemistry in
recent years due to recognition of the importance of chiral molecules in both
biological and chernical systems. Often biologically active compounds show a
high degree of stereospecificity with regard to their affinity toward receptor
proteins and their resultant pharmacological activity. Many dmgs in the past
were marketed as racemates, with the assumption that there was no difference
between enantiomers. Further investigation revealed that there are frequently
significant differences between enantiomers, not only in the magnitude of the
pharmacological response but also in their distribution, metabolism, elimination
and side effects. One example is isoprenaline, (R)-(-)-isopropyl norepinephrine;
the (-) isomer is about 800 times more effective as a bronchodilator than the (+)
içorner (Allenmark, S. G., 1988). Such differences have required the
pharmaceutical indus t~ to provide the regula tory agencies with detailed studies
demonstrating either that both enantiomers are equally efficacious and safe, or
that one enantiomer is lacking in the required pharmacological activity and
devoid of any other activity or toxicity.
Stereochemistry is an area of chemistry that looks at molecules in three
dimensions. When Louis Pasteur examined the Faces of tartaric acid he noticed
some of the crystals had "hemihedral" faces. These hemihedral or asymmetric
crystals were different from his racemic acid crystals. The asymmetric crystals
were mirror images of each other, some were right handed and others were left
handed. These right and left handed crystals were called "chiralM(derived from
the Greek word "cheiros" meaning hand). A substance is chiral if its mirror
images are not supenmposable (Ramsey, O. B., 1981).
Stereoisomers are compounds that contain
but differ in the spatial arrangement of the atoms.
the same structural formula
Classical stereochemistry, as
defined by Eliel, divides stereoisomers into optical isomers, diastereomers and
geometric (cis -trans) isomers. Mislow has suggested a new system of
classification of stereoisomers which is based on symmetry and energv criteria
and therefore includes only enantiomers and diastereorners. (Eliel, E. L., 1962)
Two stereoisomers that are mirror images of each other and are not
superimposable are called enantiomers. Enantiomers have identical physical
properties such as melting points, boiling points, molecular weight, solubility
and density. The only difference between a pair of enantiomers is the direction in
which they rotate plane polarized light and in the rate of reaction with other
enantiomenc cornpounds. The magnitude of the rotation is the same for a pair of
enantiomers, but opposite in direction.
Stereoisomers that are not mirror images of each other, Le., are not
enantiomers are called diastereorners. Unlike enantiomers, diastereomers do not
have the same physical properties. They have different solubilities, melting
points, boiling points and can have both different magnitudes and directions
with respect to optical rotation. Diastereomeric molecules contain minimally two
chiral centers. Figure (2) shows the various stereoisorners of 2, 3 -
dihydroxybutanoic acid. Compounds (1) and (2) and compounds (3) and (4) are
enantiomers of each other. Stereoisomer (1) is not a mirror image of (3) or (4) so it
is a diastereomer of both (3) and (4). Similarly, stereoisomer (2) is not a mirror
image of either (3) or (4) and is also a diastereomer of both. To convert a
molecule with two chiral centers to its enantiomer, both chiral centers must be
changed. Reversing the configuration at a single chiral carbon results in a
diastereomer (Carey, F. A., 1987).
2,3, - Dihydroxybutanoic acid
dias tereomers I
-
enantiomers
diastereomers
<
enan tio mers
I dias tereomers
Figure 2
Aside from the classification of stereoisomers there are also a few tenns
used to describe the various ratios of enantiomers such as racemic mixtures,
enantiomeric excess and enantiomeric composition. Pasteur discovered that
paratartaric acid or racemic acid was optically inactive because 5096 of the
crystals were of the (+) configuration and 50% were of the (-) configuration, in
other words a 1:I enantiomeric mixture. This definition has withstood the test of
time and a racemic mixture is still defined as a 50:50 mixture of the (+) and (-)
enantiorners. The enantiomeric composition of a sample may be described by a
simple dimensionless mole ratio or mole percent of the major enantiomer.
Enantiomeric excess is another term often used to describe the excess of
one enantiomer over another. The enantiomeric excess (or ee) is described by
equation (3) , where [RI and [SI represent the mole fractions of the R and S
enantiomers.
S tereochemical Techniques
There are many techniques available to determine the enantiomeric
composition or purity of a compound such as polarimetry, various enzyme
based techniques, nuclear magnetic resonance (NMR), and chrornatographic
methods. Polarimetry, NMR, calorimetry and enzyme techniques do not require
the separation of the enantiomers, whereas chromatography does. Other
techniques such as x-ray crystallography, circular dichroism and optical rotary
dispersion may be used also to determine the absolute configuration (Allemark,
S. G., 1988).
One of the oldest techniques used to study optically active compounds is
polarimetry. Historically polarirnetry owes its beginnings to the investigations
done by physicists on the behavior of light through various crystals. As early as
1669 Erasmus Bartholinus noted the double refraction of light by calcite. Later, in
1809, Malus noted that light which was reflected from a transparent surface sudi
as water or glass has the same characteristics as one of the beams produced by
the double refraction of calcite. He called this characteristic "polarization".
Subsequent to this, the reflection plane itself, Le., plane passing through the
incident ray which was normal to the reflecting surface was called the "plane of
polarization" (Pasteur, L., 1901).
The magnitude of the rotation is dependent on solute concentration,
optical path length, solvent, temperature, and wavelength used. In order to
compare individual measurements to each other? specific conditions m u t be
specihed. Thus, the specific rotation [a] is defined by equation (4),
100a [a]: = - Ic
where a= measured optical rotation, T= temperature (C), A , the wavelength ,
[usually the D 1ine of sodium, A = 529 nm], l= path length of the cell (dm), c =
concentration of the compound (g/lOOml). If the specific rotation of an optically
pure compound is denoted by [a],, then the optical purity, P (%) of a given
sample of a specific optical rotation [a] can be calculated from equation (5).
This equation defines the optical purity which is based on experimentally
derived values which in themselves may have errors associated with them .
Optical purity is linearly related to enantiomeric purity only when there is no
rnolecular association between the enantiomers in solution. Therefore methods
that allow for a separation of enantiomers are generally more reliable. Since the
precision associated with polarimeter measurements is about 1-2% the optical
purity will give only a rough estimate of enantiomeric composition.
Additionally, the use of polarimetry for purity determinations requires that the
specific rotation of the optically pure compound is known.
The optical purity of amino acids can be determined using enzyme
techniques that exploit the fact that enzymes are stereoselective (Greenstein, J. P.
and Winitz, M., 1961). Using specific enzyme catalyzed reactions high
enantiomeric purity can be determined in the presence of a small amount of its
antipode. For example it is possible to detect as little as 0.1% of one enantiomer
in the presence of 99.9% of its antipode. There are two types of reactions that
have been used extensively in this technique, one is an oxidation reaction for
which both L- and D- amino acid oxidases are available. The other reaction, a
decarboxylation reaction can only be catalyzed by L-amino acid decarboxylase
and therefore only the optical purity of D-arnino acids cm be determined.
1 / 2 O2 A4 - oxidase oxidation : H,N-CHR-CO,H - R-CO-C02H +Mi3
AA - decarboxylase decarboxylation : H2N-CHR-C02H 2 R-CH2-NH2 + CO,
An NMR spectrum will no' differentiate between enantiomers and
racemates but is capable of distinguishing between the resonance of specific
atorns within diastereomers or diastereomeric complexes. Enantiomers are
potentially differentiated by reaction with a chiral derivatizing agent or
complexation with a chiral solvent. Mislow and Mosher developed one of the
first chiral derivatizing reagents used for this purpose known as Mosher's acid ;
MITA ((R)- (+)- a-methoxy- cc-trifluromethyl-phenylacetic acid), used to convert
chiral alcohols and amines to diastereomeric esters and amides.
In addition to converting enantiomers into diastereomers using a chiral
derivatizing agent, two direct methods have been also been developed. Pirkle
used an optically active solvent, (R)- (-)-2,2,2-trifluro-1-phenylethanol, to induce
a chemical shift difference between the enantiomers. The optical purity of the
solvent does not influence the integration results, but only the peak separation.
The peak splitting resulting kom the solvent induced chemical shift difference is
a consequence of the differential interaction of one of the enantiomers with the
chiral solvent. This is the same type of recognition that occurs within
enantiorneric compounds and chiral stationary phases within chrornatographic
processes. These NMR studies actually assisted Pirkle and CO-workers in the
design of chiral stationary phases. A more powerful NMR method used to
differentiate enantiomers uses the optically active lanthdde shift reagents. This
technique combines the high resolution obtained with pseudo contact downfield
shifts with splitting of resonance Iines by enantioselective interaction with the
chiral lanthanide reagent. Results obtained from NMR using peak integration
methods give the concentration ratio of the enantiomers and then enantiomenc
excess can be calculated.
Although the above rnethods are often used to determine the
enantiomeric purity of a compound they all (except for NMR) rely on a
measurement from an optically pure compound. Chrornatographic methods
which result in the complete separation of enantiomers represent one of the most
powerhl techniques available to determine enantiomeric composition and do
not necessarily require an optically pure standard, a partiaily purified standard
will suffice to allow identification of the individual chromatographic peaks
(Allenmark, S. G., 1988).
Chrornatoma~hy
A Russian botanist named Mikhail Tswett performed the first separation
by column chromatography. In 1900 Tswett, working with plant pigments,
separated chlorophylls and xanthophylls by passing solutions of the compounds
in organic solvents through a column packed with calcium carbonate. As the
separation took place, colored bands appeared on the column. The appearance of
these colored bands aided Tswett in naming the separation technique
"chromatography" from the Greek chroma meaning "color" and graphein
meaning "to write" (Skoog, D. A. and Leary, J. J., 1992). Although Tswett's work
was dune in the early 1900's, further development in the area of chromatography
was not pursued until 1941 when Martin and Synge produced the first
mathematical treatment of chrornatographic theory (Martin, A. J. P. and Synge,
R.L.M., 1941).
Chromatography employs two major components, a stationary and a
mobile phase. The classification of the vanous types of chromatography depends
on the nature of these two phases. The mobile phase can be either a gas, liquid or
supercritical fluid. The stationary phase may be either a column or a plate that
has a chemical phase either bonded or adsorbed to it. In al1 chrornatographic
separations the sample is dissolved in the mobile phase which is forced through
an immiscible stationary phase. The two phases are chosen so that the sample
can distribute itself between the mobile and stationary phases. Samples that are
strongly retained by the column move slowly with the flow of the mobile phase.
Conversely, samples that are weakly retained by the stationary phase, have1
much quicker. ïhese differences in retention are responsïble for the separation of
the various sample components and allow for both qualitative and/or
quantitative analysis. Figure (3) is a chart of the various types of
chroma tography.
Liquid chromatography (LC) Liquid-Iiquid, or partition (mobiIe phase: liquid)
Liquid-bondcd phase
Liquid-solid, or adsorp- tion
Ion exchange Sizc exclusion
Gas chromatography (GC) Gas-liquid (mobile phase: gas)
Gas-bonded phase
Supercritical-fluid chroma- tography (SFC) (mobile phase: supercritical fluid)
Liquid adsorbcd on a solid
Organic spccics bonded to a solid surfacc
Ioncxchange rcsin Liquid in interstices of a polymeric solid
Liquid adsorbed on a solid
Organic specics bondcd to a solid surface
Organic species bondtd to a solid surface
Partition bctwecn immis- cible liquids
Partition bctween liquid and bunded surface
Adsorption
Ion exchange Partitionlsicving
Partition betwcen gas and liquid
Partition ktwecn liquid and bonded surface
Adsorption
Partition betwttn super- critical fluid and bondcd surface
Figure 3
(Skoog, D. A. and Leary, J. J., 1992)
In addition to the column and mobile phase, a typical high pressure liquid
chromatography (HPLC) system also consists of a pump to deliver the mobile
phase through the system , a detector, an injection system (either a manual or
automated injection system) and a data collection system. Today most
Laboratories use integrators that are run by computer software, but a simple strip
chart recorder will also suffice to record the da ta.
Chromatography has corne a long way since the days of Tswett, and today
perhaps the most common type of chromatography used is high pressure liquid
chromatography- High pressure liquid chromatography (HPLC) offers two main
modes of separation, termed normal phase or reverse phase chromatography.
There are many other types of chromatography such as ion-exchange, size
exclusion and affinity chromatography; however for the purpose of considering
chiral separations, the discussion will be limited to reverse phase and normal
phase systems and how they apply to chiral separatioffi.
Since its inception chromatography has gone through tremendous
development and irnprovements. Improvements often bring new terminology,
leaving the scientist confused over the meaning of the original and/or revised
terminology. Thus several different names may be used to describe the same
type of chromatography. Normal phase chromatography is also called
adsorption and straight phase chromatography, while reverse phase has also
been referred to as liquid/liquid or partition chromatography. The term
"liquid/liquidW refers to the type of column that has a liquid film of stationary
phase covering the column packing. Due to the inconsistencies associated with
liquid/liquid columns, they are rarely used today and have been replaced with
bonded phase columns (Wainer, 1. W., 1994).
In normal phase chromatography the column is composed of a polar
stationary phase such as silica , and a mobile phase containing non polar solvents
such as hexane and ciidilorornethane. It will be noted that it was a normal phase
separation which was originally performed by Tswett on the plant pigments.
Silica is an absorbent with many uses; it can be used alone as a stationary phase
or can be chemically modified to prepare many other types of chrornatographic
phases. Silica contains silicon atoms bridged three dimensionaily by oxygen
atoms. The silicon-oxygen Iattice is then saturated by OH groups (silanol groups,
figure 4). The silano1 groups are the active sites in the stationary phase and can
form weak bonds with nearby molecules that are capable of hvdrogen bonding,
dipole-induced dipole interactions, dipole-dipole interactions or rc-cornplex
bonding. The silica gel in the chrornatographic bed is surrounded on al1 sides by
mobile phase molecules. A sample molecule is adsorbed only if it interacts more
strongly than the solvent with the adsorbent. Separation is achieved by the
differences in adsorption between various componens within a sample.
In normal phase separations, both sample and solvent molecules are
arranged on the silica surface so that their polar functional groups or double
bonds are closer to the silanol groups. Therefore, any hydrocarbon "tails" are
diverted away from the silica (figure 5) . A silica column cannot distinguish
between compounds that are identical apart from their aliphatic moiety. The
strength of the interaction, and therefore the separation, depends not only on the
functional groups contained on the sample molecule, but also on steric factors.
Therefore, molecules with different stenc structures, such as geometncal isomers
are suitable for separation
compounds are generally
(Meyer, V. R., 1994).
by normal phase chromatography. Typically polar
best separated by normal phase chromatography
Chernical Structure of S ilica
Figure 4
Silica "Tails"
Figure 5
In reverse phase chromatography, the stationary phase is less polar than
the mobile phase. The most frequently used reverse phase packing is composed
of a silica backbone to which n-octadecyl chains (ODS, C-18) have been
chemically bonded. The long chain hydrocarbon groups are aligned parallel to
one another and perpendicular to the particle surface, resulting in a brush or
bristle Like structure. The exact mechanisrn by which these surfaces retaui solute
molecules is a t present not entirely clear. Some scientists feel that the bmsh like
surface acts as a liquid hydrocarbon layer and separation occurs due to the
solubility differences of the various solutes for either the mobile phase or the so
called "hydrocarbon liquid layer". This theory follows the separation mechanism
of a typical liquid/liquid stationa. phase. Other scientists feel that the brush
coating acts as a modified surface with which the solute molecules compete
along with the mobile phase for adsorption. The mobile phase is usually a
combination of polar solvents, such as methanol, water or buffer solutions. In
normal phase chromatography, the most non-polar compounds are eluted first,
whereas in reverse phase diromatography, a more non-polar compound would
interact strongly with the stationary phase and would be eluted last. In addition
to the much used C-18 column, several other types of reverse phase packing are
available e.g., C-8, phenyl and cyano columns. The decision as to which type of
reverse phase column to use is generally dependent on the samples you wish to
separate. A more non-polar sample would be best separated by a C-18 or a C-8
column, whereas more polar samples may be more suitable for a phenyl or cyano
column. The longer the alkyl group (C-18 vs. C-8) the more non-polar the
column, and therefore longer retention times would be expected.
Eluted compounds are trans~orted bv the mobile hase to the detector * L J L
and the response is recorded using either an integrator or strip
signals that are recorded are Gaussian shaped peaks. Each peak
recorder. The
can give both
quantitative and qualitative information about the solute. The time required for
a substance to elute from a column is known as its retention tirne. ï h i s retention
time will remain constant under identical chrornatographic conditions.
chromatographic conditions are specified by the type and dimensions of a
column , mobile phase composition, detection wavelength (for UV or
fluorescence detection) and temperature. Therefore, if the retention time of a
known substance is determined under certain chromatographic conditions it c m
be used as a comparison to aid in the identification of an unknown analyte.
Quantitative measurements are possible by comparing the area under the curves.
The area of a particular peak is proportional to its concentration. Therefore, by
consmicting a calibration curve, the concentration of an unknown sample may
be detennined.
Ln addition to quantitative and qualitative information about an unknown
solute, chromatograms can also give information regarding the efficiency of a
separation. In the following paragraphs chromatographic terms such as the
capacity factor, resolution and the separation factor will be presented. During
method development these expressions allow for a quantitative comparison
among different chromatographic separations.
Mathematical Principles of Chromatoma~hv
The efficiency of a particular separation can be described mathematically
by two separate terms, the capacity factor and the separation factor. Different
components of a sarnple are identified on a chromatogram by their retention
times, tR in minutes. A compound that is not retained by the column has a
retention time of t, and is the mean time for the solvent to flow through the
column and is the same for al1 non-retained compounds within a particular
chromatogram (figure 6 ) . Retention time is a function of mobile phase flow
velocity and column length. If the mobile phase flow rate is low or the column is
long, both the t and tR would be large and by themselves not suitable for
characterizing a separation. Therefore, the capacity factor or k' is used to describe
the migration rates of retained analytes on columns and is described by the
following equa tion;
The capacity factor, k' is independent of both col umn length and mobile phase
flow rate and represents the molar ratio of the compound in the stationarv and
mobile phase. Capacity factors between 1 and 5 are generally preferred. As can
be seen from the equation, if the k' values are very low then the separation may
be inadequate ( compounds pass too quickly through the column, little column
interaction). However, high k' values are accompanied by long analysis times.
The capacity factor, k', can also be described by the following equation:
Signai
Somple injection
Chromatogram Depicting the Separation of two Compounds (Meyer, V. R., 1994)
Figure 6
Pcak- heig ht ratio 1:l
Resolutions of two Neighboring Peaks (Meyer, V. R., 1994)
Figure 7
where K is the distribution coefficient of the solute between the stationary and
mobile phase, Vs is the volume of stationary phase and Vhf the volume of mobile
phase in the column. In other words the capacity factor is directly proportional to
the volume occupied by the stationary phase and particularly to the specific
surface area of the adsorbents. A column packed with porous layer beads
produces lower k' values and therefore shorter retention times. A silica column
packed with narrow porous material produces larger k' values than a wide
porous material. In order for two components to be separated they must have
different k' values (Meyer, V. R., 1994)
The extent by which two species, 1 and 2 are separated from one another
on a particular column is described by a parameter known as a or the separation
factor, a quantity which compares the relative retention times of the hvo peaks
and is described by;
In this equation kt2 > kVl , therefore tR2 > tR, and a should be greater than 1. If a
=1, then no separation occurs.
h o t h e r important quantity to be determined, especially with respect to
chiral separations is the resolution. The resolution or R5 of a column is a
quantitative measure of its ability to separate two analytes. Figure (7) correlates
the various R, values with individual separations. The Resolution or R, of two
analytes is defined by the following equations;
In equation (9), W is the peak width, which can be measured either by the use of
an integrator or manually with a mler and A 2 is the difference between the two
retention times. A cornplete separation between two peaks is obtained when R, =
1.5, however an R, = 1.25 would be sufficient for quantitative analysis with an
integrator.
Chromatograms can also be used to determine column efficiency. Column
efficiency is usually described by two related terms, the plate height, H and the
number of theoretical plates, N. The foollowing expression relates the two
quanti ties;
where L is the length in centimeters of the column packmg. The efficiency of a
column increases as the number of theoretical plates increases and the plate
height decreases. An increase in the number of theoretical plates is a function of a
better packing, longer column length and an optimum mobile phase flow rate. A
column with a high number of theoretical plates c m separate a mixture of closely
related compounds, Le. similar a values. The term "plate height" and "number of
theoretical plates" cornes from the mathematical treamient of Martin and Synge
(Martin, A. J. P. and Synge, R. L. hd. 1941). They viewed a chrornatographic
column as though it were made up of nurnerous nanow layers called theoretical
plates. They proposed that at each plate, equilibration of the analyte between the
mobile and stationary phase was established. Movement of a solute through the
column was then treated as a series of stepwise transfers of equilibrated mobile
phase from one plate to the next. Although this theory accounts for the Gaussian
shape of chrornatographic peaks and the rate of movement through a column, it
does not account for band broadening. ï h i s plate theory was later dismissed and
replaced with the rate theory. Although the terms theoretical plates and plate
height are still being used as applied to the rate ~ e o r y , it should be noted that
the concept that a column contains plates where an equilibrium exists is not
accepted. Ln fact, an equilibrium can never be realized with a mobile phase in
constant motion2(Skoog, D. A. and Leary, J. J., 1992). Both the plate height H,
and the number of theoretical plates N, also can be calculated using the
following equa tions;
Another method for calculating N, which some scientists feel is more reliable is
to determine the WlI2, the width of the peak at half its maximum height. Using
W,, , equation (12) becomes;
The above equations represent a series of mathematical expressions that are often
used to either describe a particular separation, the efficiency of a column or used
in method development to improve a separation. With respect to the separation
of enantiomers or diastereomers the most important parameters are the
separation factor, a and the resolution, R, .
However, plate height or height equivalent to a theoretical plate can be considered as that
length of column over which the equivalent of an equilibrium is established.
Chiral Seuarations - A Direct Approach usine Chiral Stationam Phases
In the 1960's Gil-Av used chiral stationary phases to separate enantiomers
by gas liquid chromatogaphy (GLC) (Gil-Av, E., Feibush, B. and Charles-Sigler,
R., 1966). Also at this time, enantiomers were converted to diastereomers using
chiral derivatizing agents followed by separation by achiral GLC or thui laver
chromatography. As the knowledge of chiral recognition interactions increased,
chiral derivatizing agents for liquid chromatography (LC) were developed.
Eventually by the 1980's chiral stationary phases for liquid chromatography
were developed and direct separation of enantiomers was possible. Today, many
more chiral stationary phases and chiral derivatizing agents have been
developed and studied. Although analysis by gas chromatography is still a
useful method, the requirement that the analyte is both volatile and thermally
stable often necessitates prior derivatization. Presently the most popular method
for analyzing a chiral compound is by high pressure liquid chromatography.
High pressure liquid chromatography affords two basic methods to
separate chiral compounds, a direct and an indirect method. A direct separation
includes the use of a chiral column or of a chiral mobile phase. An indirect
separation involves derivatization of the compound using a chiral reagent
resulting in the formation of diastereomers followed by separation on an achiral
column. In sorne cases a combination of the two methods is used e.g., a solute
may be derivatized with a chiral or adiiral reagent prior to separation on a chiral
column.
Direct separations using chiral stationary phases (CSPs) are the result of
the formation of a weak diastereomeric complex between the enantiomeric solute
and the chiral selector of the column. The ability to separate a solute on a
particular chiral column is dependent on the correct "interaction sites" on the
solute. Therefore, the presence of specific functional groups in the correct
orientation is required for a solute to be separated on a chiral stationas. phase.
In some cases the correct "interaction site" is not part of the solute and needs to
be added, by prior derivatization.
Dalgliesh has proposed a "three point interaction" model to explain the
chiral recognition process for the solute-CSP cornplex. According to this model,
there are required minimally three interaction sites between the solute and the
chiral selector, with one of these interactions being dependent on the
stereochemical structure of the solute (Dalgliesh, C. E., 1952). ï h e differences in
free energy between the individual enantiomers/chiral selector will determine
the magnitude of the separation. For example, if the (+) isomer forms a more
stable (lower energy) bond with the chiral column, this isomer will be retained
longer. The actual efficiency of the separation is dependent on the specific
interactions between the chiral column and the enantiomeric solutes.
The best way to appreciate the different types of chiral stationary phases
is to group them according to the type of solute/CSP interaction that takes place.
Table 1 (Wainer, 1. W., 1993) lists some of the more popular columns along with
their ~referred substrate interactions. Interaction sites can be classified as x
-basic, rr -acidic, aromatic rings, acidic sites, basic sites , steric sites or sites for
electrostatic interactions. Aromatic rings represent a source of rc -n. interactions,
whereas acidic sites rnay allow electrostatic interactions or hydrogen bonding to
take place. Typically the hydrogen bonding involves such functional groups as
amide, carbarnate, urea, amine or alcohols, although ether, sulfinyl or
phosphinyl oxygens may also be involved in hydrogen bond formation.
Electrostatic interactions may occur at charged groups or with permanent or
induced dipoles. Steric interactions occur between large groups.
Type I I -cellulose
Cyclobond 1, 11 ,111
Mixed Mode CSP
WE, Wl1, W M
AGP OVM H S A BSA
TVpe I
I I
1 I
11 1
I + III
IV
v Proteins
Interaction hydrogen bonding, dipnle slacking, iZ - K
Combina tion of h y drogen hmding, dipolc stacking, A - n fitting in cavity part of all the solutc rnust f i t ir the cavlty, a h hydrogen bonding and Ir-lt
hydrogen bond ing dipole stacking Ir- lt and cavi ty inclusion complex with transition metnl (M), chiral moleculc (A A) (L) and enantiomer o f racemic solute hydrogen bonding and hydrophobie interactions bctwcen protein and çolute. Alsri electrostatic
salute a lkyl carbinols, aryl substitutcd liydan toins lactanis, sulfoxides Contain phenyl t;roups, amides, iniides, keloncwj, esters
phcnytoin and mctab»lites, mpod crmtaining rnultiplc pheny 1 groups such as alcohols, amide, ester, ketone aromatic amines, alcohols, multiple phcnyl group aimpounds
a-amino acids mino and dicarboxylic acids containing a-OH
cationic and anioni mmpaunh, ephedrine, a, B- amino acids 1-1SA/BSA-anionic and neutral mmpounds. Necd aromatic and polai moicty
chiral selector (R)-N-(3,5 Dinitro benzoyl) phenyl glycine (S)-N-(3,5 DNBPG)
Microcrystalline cëllulose triacetate
Cellulose triaceta te Cellulose Mbenzoatc B-cyclodex trin y- cyclodextrin a - cyclodex trin Poly (triphenylmet hyl methacrylate) Chiral 18-crown-6-ethcr
Type 1- naphthylethyl carbamate Type 111- cyclodextrin
(WE) N-carboxymethyl(lR, 2s)-diphenylamino ethanol (W 11)- L-Praline (WM) - L-tert-Leucinc
AGP-ai-Acid protein OVM - Ovomucoid HSA-tiuman serurn albumin BSA- Bovine serum albumin
- Derivatizing :onvert alcohols, amines 07 :arboxylic acids to arornatic amide, urea or carliamate
h h o l s to csters, espccially para-nitrobcnzoic esters if Ot- near chiral ccnter, benzyl amides (or benzyl esters
lntroduce phcnyl groups. Chiral aliphatic amines hav w n converted to benazmide derivalives clicarl>«xylic acids have lieen converted to benzyl esters or amides Type 1 -Non aromalic amines and alcohols must be convertec Io dinitrophenyl derivative
derivatizd amino acids including N-acetyl, amide, N carbamoy 1, N-carbobenzoxy
Por HSA/DSA -most amino acids need ptecalurnn derivalizalion wilh N-acety benzenesulphonyl, DANSY L Often used in achiral/chiral systerns
One of the most popular CSP's used today is the type 1 or " bmçh type"
phase (see Table 1) . Bmsh type phases are based on amino acid derivatives,
which contain additional polar groups which can facilitate hydrogen bonding.
The most widely used r-electron accep tor columns have a 3,5-dinitrobenzoy l
derivatized phenylglycine or leucine group linked to a silica backbone. Type 1
CSP's rely on rc -r interactions, hydrogen bonding and dipole stacking for the
recognition process to take place. Figure (8) depicts the chiral recognition mode1
as suggested by Pirkle for the longer retained enantiomer of the N-(3,5-
dinitrobenzoyl) derivative of an amine separated on a N-(2-naphthy1)alanine
CSP. The three interactions are the R. -n interactions between the arornatic groups,
hydrogen bonding between the acid N-H proton of the amine and the carbonyl
group of the CSP and finally hydrogen bonding between the basic site (8) of the
amine to the N-H group on the CSP (Pirkle, W. H., 1986). Chiral recognition is
usuallv successful when the three chiral recognition sites are adjacent to the
stereogenic center. The presence of an aromatic group is required for chiral
separation on this type of CSP's , and on most other CSP's as well. Typical solutes
suitable for this type of CSP include aryl-substituted hydantoins, lactams,
sulfoxides and amino alcohols. While most solutes contain an aromatic n -basic
group not al1 solutes contain a x -acidic group, and must be derivatized. In many
cases derivatization with 3, 5 dinitrobenzoyl is necessary. This type of CSP
would use a mobile phase similar to that used in normai phase diromatography,
e.g. hexane, heptane and a more polar modifier such as tetrahydrafuran or
methanol (Pasutto, F. M., 1992).
Several types of CSP's rely on the formation of an inclusion complex. In
such instances, the solute must be able to fit inside the cavity created by the
CSP's structure. Both Cellulose type 1 and II and cyclodextrin CSP's depend on
the formation of an inclusion complex.
Generalized Chiral Recognition Model Between a Chiral Stationary
Phase and a chiral Dinitobenzamide Enantiomer
(Pirkk, W. H., 1986)
Figure 8
30
Cellulose 1 CSPs are composed of D-O-glucose units arranged in chahs in either
parallel (CTA type 1) or anti parallel directions (CTA type II) . The arrangement
of the D-B-glucose rnoieties allow for the formation of cavities between the units
and for spaces between the polysaccharide chahs. One of the requirements for
this CSP is a phenyl group which must enter the cavity of the CSP to form the
solute/CSP complex. Additional interactions include hydrogen bonding and
dipole stacking. Typically enantiomeric cyclic amides, imides, esters, ketones,
and alkyl substituted indenes are separated on such CSP's. Chiral alcohols are
best resolved when they are converted to esters. Cellulose II type CSP's also
contain a cavity which appears to be important in the recognition process.
Pharmaceutical compounds such as warfarin, verapamil and propranolol have
been separated on this type of CSP. Typical solutes contain one or more aromatic
ring and a carbonyl, sulfinyl or nitro group. Amines and carboxylic acids are
often derivatized to improve the chrornatographic efficiency of the column.
Mobile phases for these type of CSP's are usually composed of hexane and
isopropanol or a similar modifier.
The cyclodextrin type CSP's are cyclic oligosaccharides composed of a
-D-(+) glucose units linked at the 1,4 position. The most cornmon forms of this
molecule are the a -, B- and y -cyclodextrins, which contain six, seven and eight
glucose uni& respectively. The mouth of the cyclodextrin molecule has a larger
circumference than its base, giving it the overall shape of a cup. The mouth of the
cup is lined with the secondary hydroxyl groups of the C-2 and C-3 atoms of
each glucose unit, while the C-6 primary hydroxyl groups line the base of the
cup (Figure 9) (Ahuja, S., 1991). The separation mechanism is based on the
formation of an inclusion complex within the hydrophobic cavity. The size,
shape and polarity of the solute are the most critical factors affecting the stability
of the inclusion complex. Similar to the cellulose CSP's, solutes for cyclodextrin
Smicture of B- Cyclodextrin Units
Figure 9
CSP's must also contain an aromatic moiety at or adjacent to the stereogenic
center. The aromatic portion of the compound must be inserted into the
cyclodextrin cavity. The size of the aromatic portion of the molecule, thus
determines which cyclodextrin CSP is appropriate. Cyclodextrin CSP's have been
used in both reversed and normal phase modes, although the most common
mobile phases are composed of water, or buffers modified with methanol.
Cyclodextrin CSP's have been used to separate enantiomers of phenytoin along
with several aromatic amino acids such as D,L-pheny lalanine, D , L - ~ N ~ tophan
and D,L-tvrosine (Wainer, 1. W., 1993).
The last CSP to be discussed is the type V or protein based chiral
stationary phase. Proteins are polymers of hgh molecular weight that are
composed of L- amino acids. Since proteins are hvolved in the uptake and
transport of dmg substances, it seems logical that such proteins would display
stereospecific binding. Experiments using human al - acid glycoproteui (AGP)
indicated that Spropranolol is bound io a greater extent than the R-enantiomer.
The ability of proteins to stereoselectively bind small molecules fostered the
development of a series of protein based CSPs made by immobilizing the
particular protein ont0 a s i k a backbone. In addition to AGP, ovomucoid (OVM),
human serum albumin (HSA) and bovine serum albumin (BSA) have al1 been
immobilized ont0 silica. Several pharmaceutical compounds have been separated
on AGP columns , e.g., atropine, ephedrine, and verapamil. BSA type phases
have separated amino acids, warfarin and sulfur containing cornpounds.
Separations with protein phases are normally used with an aqueous mobile
phases containing a phosphate buffer at neutral pH. Under these conditions,
hydrogen bonding, electrostatic and hydrophobic interactions c m take place.
Although the exact cnteria for separation on a protein type CSP are not fully
understood, it has been shown that adjustments in the mobile phase, flow rate or
pH tend to affect the s tereoselectivity sugges ting multiple types of interactions
may be necessary for a successhl separation (Ahuja, S., 1991).
Chiral Seuaration - Direct Se~arations us in^ Chiral Mobile Phase Additives
Aside from the use of a chiral column, a direct separation may be obtained
by the addition of a chiral additive to the mobile phase. When a chiral selector is
added to the mobile phase conventional non-chiral columns cm be used. The
stereoselective separation can be due to the formation of a stereoselective
cornplex in the mobile phase, the adsorption of the chiral selector to the
stationarv phase creating an in situ chiral column or the formation of
diastereomeric complexes with different distribution properties between the
mobile and stationary phase. In ail cases, except the formation of diastereomeric
complexes (ion pairs), it is necessary to have the typical "three point interaction"
for separation to occur. A two point interaction consisting of an electrostatic
attraction and hydrogen bonding may be sufficient for diastereomeric
complexes. (Pettersson, C., 1988).
There are several different types of chiral mobile phase additives
(CMPAs) such as L-proline, cyclodextrins, albumin, (R, R)-tartaric acid mono-n-
octylamide, quinine and u I acid glycoprotein. Many of these CMPAs have
borrowed principles that are used in other types of chromatographv. For
example ligand exchange chromatography was originally developed for the
separation of enantiomenc amino acids. It involved the use of a stationary phase
containing L-proline (ligand) bonded to a resin on which copper was loaded.
Subsequently, ligand exchange with the chiral metal complexing agent in the
eluent was developed. An injection of an racemic amino acid into a
chromatographie system with an optically active ligand (L-Lig) and copper ions
u.i the mobile phase results in the formation of two ternary diastereomeric metal
complexes (LePage, L., Linder, W., Davies, G., Seitz, D. and Karger, B., 1979). The
the two diastereomeric
a chiral mobile phase
separation is due to the differences in the structure of
complexes. Ligand exchange chromatography with
additive is used with conventional reverse phase columns and is used to separate
racemic amino acids (Pettersson, C., 1988).
The use of crown ethers and cyclodextrins in the mobile phase has been
used successfully for chiral separations. Separations with mandelic acid and
derivatives indicated that the asymmetric carbon atom rnust have functional
groups capable of bonding to the hydroxyl groups of the B-cyclodextrin
(Debowski, J., Sybilska, D. and Jurczak, J., 1982). Unfortuna tely, relatively low
steroselectivity is found when cyclodextrins are used alone as the mobile phase
additive. improved selectivity has been shown with the addition of a counter ion
such as d-camp hor-10-sulfonic acid.
The addition of an ion pairing reagent has often been used in reverse
phase chromatography. Similarly, the use of chiral counter ions has been shown
to be able to separate chiral acids, amines and alcohols. The separation is based
on the formation of diastereomeric ion pairs in the mobile phase. It appears that
a two point interaction between the ions is adequate for separation. For a
reasonable separation, the counter ion should contain the ionized h c t i o n and
have a hydrogen bonding group near the chiral center. Quinine and the related
cinchona alkaloids have been used for the separation of chiral carboxylic acids
(Pettersson, C. and No, K., 1983). Enantiomers of propranolol and similar amino
alcohols have been separated using d-camphor-10-sulfonic acid (Pettersson, C.
and Schill, G., 1981) or N-benzoxycarbonyl-glycyi-L-proline (ZGP) (Pettersson,
C. and Josefesson, M., 1986). The chiral separation is optimized by varying the
concentrations of the counter ion in the mobile phase and a competing
compound with the same ionized h c t i o n as the enantiomers. In many cases low
polarity solvents are used to obtain a high degree of ion pair formation. This
method is usually used for separation of enantiomeric amines, acids and
alcohols.
Chiral Separation via the Indirect Method: Usine Chiral Derivatizine Agents
Chiral separation by the indirect method involves the use of a chiral
derivatizing agent (CDA) to convert enantiomers into diastereomers. In some
cases derivatization is done to provide greater sensitivity by the detector, but
may also be used to introduce needed interaction sites for subsequent separation
on a dural column. CDA's have also found use in asymmetric synthesis (Olofson,
R. A., Schnur, R.C., Bunes, L. and Pepe, J.P., 1977). Since diastereomers have
different physical chernical characteristics, separation on conventional achiral
columns is possible. Despite the expanding availability of chiral stationary
phases a wide variety of chiral separations are still performed using CDA's since
chiral columns generally are not as rugged and reproducible as haditionai
normal and reverse phase columns, and they are also very mu& more expensive.
The main consideration in the use of a CDA is its purity and stability
under derivatization conditions. If a racemic compound is derivatized with a
pure R-reagent the following diastereomers would be formed; S-dmg-R-reag (A)
+ R-drug-R-reag (B). However if the CDA was contaminated with a small
arnount of Sreagent then there would also be formed S-drug-S-reag (C) + R-
dmg-S-reag (D) in addition to compounds (A) and (B). Compounds (A) and (B)
are diastereomers as are compounds (C) and (D) which can be separated on an
achiral column. However, compounds (A) and (D) are enantiomers as are (B) and
(C). The use of an achiral column would not allow for the separation of (A) from
(D) or (B) from (C) and therefore would result in art inaccurate enantiomeric
determination. The precision of the analytical results as well as the optical purity
of the collected peaks is thus limited by the purity of the CDA. Further the
reaction conditions must not be so harsh as to result in racemization at the chiral
centers of either the solute or the CDA.
The chiral substances rnost often resolved as diastereomers are amines,
alcohols and carboxylic acids. Chiral alcohols are usually derivatized with chiral
acids, chloroformates or isocyanates, giving diastereomeric esters, carbonates
and carbamates respectively. Esters are usually prepared from an alcohol and an
"activated" chiral carboxylic acid, usuaily in the form of a diloride, anhydride or
imidazole derivative. Acyl chlorides are highly reactive and are used to
derivatize hindered secondary alcohols and amines. Unfortunately acyl
chlorides, particularly those used containing a proton attached to the chiral
carbon are susceptible to racemization under extreme acidic or basic conditions
or at elevated temperatures. The most commonly used reagents are the dilorides
of 313-acetoxy-A3 - etienic acid, (-)menthoxyacetic acid and S(-)-N-
(triAuoroacetyl)proline. Figure (10) shows the chernical structures of some of the
more common CDA's. The formation of carbamates by the reaction of alcohols
with chiral isocyanates has become quite popular for the resolution of chiral
alcohols and prirnary and secondary amines. Typical reagents include R(+)/S(-)
phenylethyl isocyana te and R(+) /S(-)-b(1- naphthy1)ethyl isocyana te. The
diastereorneric carbamates are generally stable without racemization. Alcohols
react slowly and may require heating for several hours, whereas phenols react
rapidly, even at room temperature (Wainer, 1. W., 1993).
Many non steroidal anti inflammatory drugs (NSAIDs) such as ibuprofen,
fenoprofen, and flurbiprofen contain a 2-arylpropionic acid group in their
structure and are chiral due to the asymmetric center a to the carboxyl group
(Figure 11). These drugs are marketed in their racemic form , with the full
knowledge that most of the desired activity resides in the S - isomer. The R -
isomer has little or no activity. In addition, it has been shorvn that many of these
NSAIDs undergo in vivo metabolic biotransformation of the inactive R
R (+)/S a(-)-a-Methoxy-a- (tnflurome thy1)pheny lacetic acid (Mosher's acid)
i C-C-OH
3f3- AC^ toxy-~--e tienic acid
2,3,4-tri-O-acetyl-a-D- arbinopyranosyl isothiocyanate (AITC)
giucop yranosy l iso thiocyana te
Some Chiral Derivatiting Reagents
Figure 10
2-Aryi Propionic Acid Derivatives
CHCOOH
Fenoprofen
- CHCOOH
Flurb ip rofen
Figure 11
enantiomer to the active S enantiomer. This interesting finding ( o d y possible
from use of chiral chrornatographic separations) has prompted most of the
publications dealing with derivatizing carboxylic acids to be focused on
arylpropionic acid derivatives (Hutt, A. J. and Caldwell, J., 19%). Chiral alcohols
typically used in such derivatizations are 2-octanol (Johnson, D. M., Rerter, A.,
Collins, M. and Thompson, G.E., 1979), (-)-2-butanol (Kamerling, J. P., Duran, M.,
Gerwig, G.I., Ketting, D., Bruinvis, L., Vliegenthart, J.F.G. and Wadman, S.K.,
1981) and (-)-menthol (Hasegawa, M. and Matsubara, I., 1975). In addition to
carboxylic acids the use of chiral aIcohoIs to Çom diastereomeric esters has also
been used for the resolution of amino acids (Hasegawa, M. and Matsubara, I.,
1975) hydroxy acids and 2-alkyl-substituted carboxylic acids (Konig, W. A. and
Benecke, 1-, 1980).
Amines represent one of the easiest functional groups to derivatize, a Çact
which is reflected in the number of different reagents available. Figure (12)
depicts the various derivatization routes possible for amines. The formation of
amides from chiral acylating reagents is one of the most popular derivatizing
techniques for prirnary and secondary arnino groups. N-heptafluorobutyryl -L-
prolyl chloride and CNitrophenylsulphonyl-L-prolyl chloride have both been
used to derivatize amphetamine. Primary and secondary amino groups react
with isocyanates and isothiocyanates to give urea and thiourea derivatives
respectively. Miller et al compared the use of four chiral reagents; (R)-(+)-1-
phenylethyl isocyanate (PEIC), (-) - a -methoxy-a (trifluoromethy1)phenylacetyl
diloride (MPTA Cl) ,2,3,4,6-tetra-O-acetyl-E-D glucopyranosyl
isothiocyanate (GITC), and 2,3 ,4-tri-o-acetyl - a -D-arabinopyranosyl
isothocyanate (AITC) for the derivatization of a series of chiral ring substituted
1-phenyl-2-aminopropanes (amphetamines) and 1-phenylethylamine. Al1 four
CDA's formed diastereomeric products with each solute using mild reaction
Alkylation Amine
Chloroformate Carbornate
Isothiocyonate Thiourea
Isocyonate Ureo
RlNH2 + R2SH + acHo \ - RJ" ~ CHO
OPT/ chiral th101 Isoindole
Acyla tion Amide
Potential Chiral Derivatizing Reactions for Amines
Figure 12
conditions. However, better resolutions (R-values) were seen wi th AITC, GITC
and MPTA 4 Cl than with PEIC. Separations were performed using a reversed
phase C-18 column with methanol - water as the mobile phase (Miller, K. R., Gal,
J. and Ames, M. M., 1984).
Primary amines when reacted with O-phthalaldehyde (OPA) in the
presence of a thiol form isoindoles (Simons, S. S. and Johnson, D.F., 1976). Ln
addition thiols may be assayed by derivatization with OPA in the presence of an
excess of a suitable primary amine. This procedure is often used for the
determination of amino acids (Deyl, Z., Hyanek, J-, and Horakova, M., 1986). The
use of a chiral thiol for the derivatization with OPA results in the formation of
diastereomers in the presence of primary amines- Amino alcohols, amino acids
and primary amines have been resolved using OPA and N-acetyl-L-cysteine,
Boc-L-cysteine and N-acetyl-D-penicillamine as the chiral thiol (Buck, R. H. and
Knunmen, K., 1984, Buck, R. H. and Krurnmen, K., 1987).
Thus far the reactions discussed for amines are suitable for either prirnary
or secondary amines. However, chloroformates are suitable reagents for primary,
secondary and tertiary amines. The reaction of a chloroformate with a prirnary or
secondary amine results in the formation of the corresponding carbamate and
chiral chloroformates are readily synthesized from chiral alcohols and phosgene.
One of the unique properties of chloroformates however is their ability to N-
dealkylate tertiary amines to produce the carbamate of the correspondine
secondary amine ( equation 14).
Witte et al derivatized promethazine with (-)menthyl chloroformate and
separated the diastereomers by reverse phase HPLC (Witte, D. K., de Zeeuw,
R.A. and Drenth, B. F. H., 1990) - A series of antihistamine derivatives were
reacted with 2-Naphthyi chloroformate forming fluorescence carbarnates that
could be separated using RP-HPLC (Gubitz, G-, Wintersteiger, R. and Hartinger,
A., 1981). The potential use of this reaction as the basis for an assay process for
Selegiline will be discussed following consideration of chiral assay procedures
for amphetamine and its analogues.
Chiral Separations of Amines with Particular Reference to
Amphetamine and its Analogues
Many chiral pharmaceutical compounds contain pnmary or secondary
amine functions and have provided much of the impetus for the development of
enantiomer separations of basic substances using chromatographie methods. The
foollowing discussion will center predominantly upon the separation of
enantiomers of amphetamine (LW), methamphetamine (MAI') and Selegiline
as examples of primary, secondary and tertiary amines respectively and will
consider separations utilizing high pressure liquid diromatography specifically.
It should be noted however, that (although not discussed here) enantiomer
separations of many of these compounds have also been reported using
gas/liquid chromatography and capillary electrophoresis.
Amphe tamine Methamp he tamine
CH-
While derivatization prior to enantiomer or diastereomer separation on a
chiral or achiral column is undoubtedly the more reliable method for
development of a chiral assay for these compounds, some direct enantiomeric
separations have been reported. Thus Makino (Makino, Y., Ohta, S. and Hirobe,
M., 1996), using a chiral crown ether column ( 15 x 0.4 an Daicel Crovmpack (+),
Daicel Chemicals, Japan) reported separation of racemic amphetamine but not
racemic methamphetamine using aqueous perchloric acid (pH 1.8) mobile phase.
Similarly, using what is a poorly defined B-cyclodextrin coated support, (the
colurnn is reported to be awaiting patent approval) the enantiomers of Selegiline
were separated (a = 1.095 from reported k' (+)-Selegiline = 1.47 and k' (-)-
Seledine = 1.61) using a methanol/50 mM potassium phosphate buffer (60/40)
mobile phase (Cserhati, T. and Forgacs, E., 1994). This separation however
appears to be very solvent sensitive since replacement of the methanol in the
mobile phase with ethanol yields identical values for the capacity factor (kt) of
the enantiorners (Forgacs, E., 1995).
While the above are examples of direct enantiomer separations using
chiral supports, prior derivatization of amphetamine and its analogues would
appear to greatly increase the potential for a successful separation of the
enantiomers. Thus derivatization of a series of amines (including
methamphetamine) by 2-naphthylchloroformate (NCF) has been followed by
separation of the product carbarnates on a Pirkle type covalent chiral stationary
phase ((Il)-N-(3, 5-dinitrobenzoy1)phenyl glycine bonded to silica). The
derivatization introduces the R-basic naphthyl function whch can interact with
the IF acidic 3, 5 -dinitrobenzoyl group of the column leading to a successful
"th.ree point attadiment" to the stationary phase (see previous discussion) (Doyle,
T. D., Adams, W.A., Fry, F.S. Jr. and Wainer, LW., 1986). Similarly, the N-benzoyl
amide derivatives of (2) AMP and (k) MAP have been separated on chiral
cellulose based columns (Chiralcel OB (tri-O-benzoylcellulose) and Chiralcel OJ
(tri-O-tolylcellulose)). Resolutions for both MAI? and AMF was superior on the
Chiralcel OB column when using a mobile phase of hexane/2-propanol (90/10)
with a flow rate of 1.0 ml/min (Nagai, T. and Kamiyama, S., 1990).
More frequently however, derivatization is accomplished using chiral
reagents followed by separation of the product diastereomers upon an achiral
column. Several examples exist of this procedure in which chiral acvlating
agents, isothiocyanates or chloroformates have been employed.
N-acyl-L-prolyl chlorides have been shown to be useful chiral reagents
when used prior to separation by gas chromatography (Liu, J. H. and Ku, W. W.,
1981). The similar use of such reagents in high pressure liquid chromatography is
typified by the reaction of 4-nitrophenylsulfonyl-(S)-prolyl chloride with AMP
and MAP. Separations were carried out on a silica column in the case of AMP
derivatives using a mobile phase of chloroform/heptane (80/20) whereas the
diastereomers of MAP were separated on a Zorbax C-18 column with a mobile
phase of methanol/ water (60/40) (Barksdale, J. M. and Clark, C R . , 1985).
Among the chiral isothiocyanates employed as derivatizing agents,
2,3,4,6-tetra-0-acetyi-E-D-glucopyranosl isothiocyanate (GITC) is frequently
used. Noggle et al (Noggle, F. T. J., DeRuiter, J. and Clark, CR., 1986) describe
the separation of the diastereomers of (+)-methamphetamine and the use of such
a separation in studies of forensic samples containhg (S)-methamphetamine and
significant quantities of (IR, 2s)-ephedrine (figure 13). The amines are
derivatized in chloroform solution in the presence of a 10 molar excess of reagent
followed by separation on a C-18 column using a mobile phase of
water/THF/acetic acid (70/35/1). The isothiocyanate group in GITC is very
reactive and undergoes nucleophilic attack by the amino group of primary and
secondary amines to form thiourea products (figure 14 ). (N. B. The presence of
small quantities of (IR, 2s)-ephedrine or (lS, 2s)-pseudoephedrine in samples of
(5)-methamphetamine would probably indicate a poorly purified sample derived
(IR, 2S) Ephednne S (+) Methamphetamine
Stereochemical precursors of S (+) Amphetam
Figure 13
ine
CH; I
X = H for Methamphetamine/ X = OH for Ephedrine
Derivatization Reaction of an enantiomer of Methamphetamine with
2,3,4,6-tetra-O-acetyI-i3-D-glucopyranosyl isothiocyanate (GITC)
Figure 14
48
from illici t synthesis; s imilarly , the presence of solel y (R)-(-)-methamp he tamine
in urine samples would be indicative of therapeutic use of Selegiline rather than
use of illicit methamphetamine. Such distinctions require the use of chiral assav
procedures). The chiral derivatizing agents 2,3,1,-tri-O-acetyl-a-D-
arabinopyranosyl isothiocyanate (AITC) and (R)-(+)-1-phenylethyl isocvanate
(PEIC) have also been used for the separation of amphetamines (Miller, K. R.,
Gal, J. and Arnes, M.M., 1984).
Mention has already been made of the use of 2-naphthyl chloroformate
(NCF) as an acylating agent for primary and secondary amines prior to
separation of the enantiomeric carbamates on a chiral colurnn (see separation of
amphetamine enantiomers above). Many of the carbamates derived from 2-NCF
are fluorescent making detection of the products extremelv sensitive. In addition
however, chloroformates react with tertiary amines but the product quatemary
carbamate is unstable, dealkylating at slightly elevated temperatures to the
carbamate of the secondary amines. Thus, a number of tertiary amines of
therapeutic importance have been analyzed as the carbamate of the
demethylated secondary amine by this method e-g. diphenhydramine,
tenalidine, diphenylpyraline, clemastine (see figures 15 and 16) (Gubitz, G.,
Wintersteiger, R. and Hartinger, A., 1981).
Reaction of 2-Naphthyl Chloroformate with a tertiary amine
Figure 15
Diphenhydramine
Diphenyip yraline
T e r t i q Amines derivatized with 2-Naphthyl chloroformate
Figure 16
Other chloroformates commercially available a re ( - menthyl
chloroformate and the strongly fluorescent compounds 9-fluorenylmethyl
chloroformate (FMOC) and (+)-1-(9-F1uorenyl)ethyl Chloroformate (FLEC). This
latter reagent has been used to separate amino acid enantiomers and various
chiral amines. The derivatization of amino acids with FLEC occurs at room
temperature under basic conditions in an aqueous environment. At the
completion of the reaction, excess reagent and its hydrolysis product is removed
by pentane extraction and the required diastereomeric products separated on a
C-8 column. Enantiomers of Metoprolol, a secondary amine, have been separated
using the FLEC reagent (Einarsson, S., Josefsson, B., Moller, P. and Sanchez, D.,
1987). The secondary amine was mixed with borate buffer (lM, pH 7.85) and
reacted with FLEC (1 mM in acetone) at room temperature for 30 minutes. Excess
reagent was removed by reaction with an excess of hydroxyproline and the
product diastereoisomeric carbarnate derivatives separated on a C-8 column
using a mobile phase of acetonitrile/ water (60 /XI).
Recently, a sensitive HPLC rnethod for the determination of the three
main metabolites of Selegiline in human plasma has been developed using
FMOC as a fluorescent derivatking agent (La Croix, R., Pianezzola, E., Benedetti
and Strolin Benedetti, M., 1994). Using FMOC with fluorescence detection
considerably enhances the detection of desmethyl selegiline, N-methyl
amphetamine and amphetamine. Samples (in 0.1 M HC1) are mixed in borate
buffer (pH Il) , FMOC (4 mM in acetonihile) is added, the samples allowed to
react at 500 C for five minutes. Excess reagent was reacted with proline and the
mixture injected directly ont0 a Nova-Pak Phenyl colurnn using a mobile phase
of acetonitrile/50 mM phosphate buffer (50/50) (pH 6.0). Detection was by
fluorescence using an excitation wavelength of 260nm and an emission
wavelength of 315 nm. Linearity of response was obtained over the concentration
range of 0.5 - 80.0 ng/ml plasma for both amphetamine and desmethyl-
selegiline. While this study was aimed at the detection of the metabolites of
Selegiline i-e. primary and secondam amines, n o mention is made of studies of
the reaction with the tertiary amine Selegiline- Thus, while the reaction
conditions are mild, it is uncertain whether all the metabolites detected have
corne from the metabolic processes or whether some may have arisen from the
reaction of the small amount of unchanged Selegiline in the plasma with FMOC
and subsequent chernical dealkylation.
(-) - Menthyl chloroformate is a further chemicallv stable chiral
chloroformate which has been employed in chiral assays involving prior
derivatization to form diastereorners. The matenal is capable of reacting with
primary, secondary or tertiary amines. As previously mentioned, the product
from reaction with tertiary amines dealkylates on heating to yield the carbarnate
of the secondary amine. An example of the use of this reagent for the
determination of tertiary amines is provided by studies on the chiral amine
Promethazine (figure 17 ) (Witte, D. K., de Zeeuw, R. A. and Drenth, B. F. H.,
1990).
The product formed from dealkylation of Promethazine is unambiguous
in structure. However, the application of this dealkylation reaction to the
development of a chiral assay for Selegiline could result in either or both of two
possible products i.e. that resultuig from dealkylation at the N-methvl function
and that resulting fxom dealkylation at the N-propargyl function (figure 18 ).
Thus,
formed will
Figure 18
if the propargyl group is cleaved from Selegiline, the product
be the menthyl carbamate of methamphetamine whereas, if the
methyl group is cleaved preferentially, the product will be menthyl carbamate of
desmethyl selegiline.
Kapnang and Charles (Kapnang, H. and Charles, G., 1983) studied the
reactions between various amines and chloroformates in order to examine the
ease of cleavage of different substituents on nitrogen. When one of the
substituents was a benzyl group, its cleavage was favored over the loss of
methyl, cyclohexyl or n-pentyl. Through a series of related experiments they
reported the following preference with respect to cleavage at the nitrogen; N-
debenzylation > N-deallylation > deamination of cyclohexylamines >N-
demethylation. This information would seem to suggest with respect to
Selegiline that the N-propargyl group wouid be cleaved more readily than the N-
methyl function. However, the nature of the products formed following reaction
with menthyl chloroformate with Selegiline and subsequent dealkylation can
readily be determined by cornparison with the products of reaction with
methamphetamine and with desmethyl-Selegiline.
Provided the dealkylation reaction results in only one carbarnate product,
the use of (-)-menthyl diloroformate as the derivatking agent should provide a
useful starting point to the developrnent of a chiral assay for Selegiline. The work
presented in this thesis will report upon this reaction and the studies to develop
a chiral assay for Selegiline alone and within its dosage form (tablets), along with
quantitative data of Selegiline tablets from two different lots. In addition, some
qualitative results will be reported on studies of the reaction of Selegiline,
desmethyl selegiline and methamphetamine with other chloroformates (ethyl
chloroformates and 9-fluorenyhethyl chloroformate) since there is only limited
information in the literature pertaining to the reaction of tertiary amines with
chloroformates. Any additional information could be valuable to scientists
interested in chiral separations of iertiary amines or dealkylation reactions.
Discussion of Expenmental
Phvsical. Chemical Characteristics of Seleeiline hvdrochloride
Selegiline hydrochloride is a near white powder, with a calculated
molecular weight (Cifii7N HCL) of 223.75 and a reported melting point
between 141 to 144 O C (Ecsery, Z., Kosa, L, Knoll, J., Somfai, E., 1967, Fowler, J.
S., 1977, and Robinson, J. B., 1985). X- ray diffraction studies on (-) Selegiline
hydrochloride have determined that the crystals are orthorombic and belong to
the P212121 space group (Simon, K., Podanyi, B., Ecsery, 2. and Torok, Z., 1986,
Simon, K., Bocskei, Z. and Torok, Z., 1992) . Chafetz et al (Chafetz, D., 1980, L.,
Desai, M.P. and Sukonik, L., 1994) determined by titration a pKa value of 7.4 at
250 C for Selegiline hydrochloride, a value which agrees with predictions
(Robinson, J. B., 1985,, Ullrich, K. J., Rumrich, G., Neiteler, K., and Fritzsch, G.,
1992) derived from the known pKa of amphetamine (pKa = 9.92 at 250 C)
(Girault-Vexlearschi, G., 1956) and the known base-strengthening and base
-weakening effects following successive alkylation of the amine function with
methyl and propargyl functions (Perrin, D. D., 1980) Selegiline hydrochloride, as
expected is freely soluble in water and methanol but does display an unusually
high solubility in chloroform [Chafetz et al report a water/chlorofonn partition
coefficient of close to 1 without however reporting the pH of the aqueous phase
(Chafetz, L., Desai, M. P. and Sukonik, L., 1994). As will be shown in a
subsequent section, this unusual solubility of Selegiline hydrochloride in
chloroform proved advantageous in the extraction of the drug frorn its dosage
form.
Preliminaw Studies of Potential Assav Procedures
While the main objective of ths study was to develop a stereospecific
assay procedure for Selegiline Hydrochloride, the availability of achiral assay
procedures were employed initially, particularly for confirmatory purposes in
the syntheses of (5)-Selegiline, (+)-desmethyl-selegiline and (f) -
methamphetamine. The method reported by Beaulieu et al used employing a
cyan0 column (Waters pBondapak CN; 150 x 3.9 mm) with a mobile phase of 0.1
M ammonium phosphate dibasic (adjusted to pH = 3. 1 with 85% phosphoric
acid) and acetonitrile at a ratio of 8515 with a flow rate of 1.0 ml/min (Beaulieu,
N., Cyr, T.D., Graham, S. J. and Lovering, E. G., 1991). Detection was by means
of a variable UV detector set at 254 nm and the diluent used for bolfi the samples
and standards was 15% acetonitrile in distilled water. For the purposes of
confirming the identity of newly synthesized material, both samples and
standards were prepared at a concentration of 0.1 pg/ml. The method is capable
of separating amphetamine, methamphetamine, desmethyl selegiline and
Selegiline.
Chiral columns and Chiral Additives
As part of the initial investigation of potential chiral separatory methods
two chiral columns were also studied, namely a Bakerbond Chiral DNBPG
column [chiral discriminator of (R) -N-(3, 5,-dinitrobenzoyl) phenyl glycine
covalently bonded to silica; 250 x 4.6 mm; 5 pm ] and a Chiralcel OC column N-
phenylcarbamate ester of cellulose which is covalently bonded to silica; 230 x 4.6
mm; 10 pm]. Solutions of (f) Selegiline, (f) desmethyl Selegiline and (2)
methamphetamine failed to show enantiomer separation when employing a
mobile phase of hexane/isopropyl alcohol (97:3) at a flow rate of 1.0 ml/min.
Further, brief studies employing the diastereoisomeric mixtures derived frorn the
reaction of (-) menthyl chloroformate with each of the above racemates similarly
failed to show separa tion of the dias tereoisomeric compounds.
An alternative means of potentially separating racemates is to ernploy a
chiral mobile phase with the hope that the product diastereoisomeric complexes
(ion pairs) will show separation. Accordingly, using the cyano column
previously employed for identification of the various synthetic racemates (see
above), the original mobile phase was modified by replacing the ammonium
phosphate with either (+) tartaric acid (concentration 0.05 or 0.025 M) or with d-
carnphor-10-sulfonic acid. Not unexpectedly, these systems failed to yield a
çepara tion of the enan tiomeric forms of Selegiline, desmethyl-çelegiline or
methamp hetamine-
Although, several authors have demonstrated the use of a low polarity
solvent such as methylene chloride (separation on a normal phase column)
favors the formation of "ion pairs" it should be noted that amphetamine was
separated (although not baseline separation) using the above mentioned aqueous
mobile phase (Pettersson, C., 1988, Pettersson, C. and Sdiill, G., 1981, Pettersson,
C. and Josefçson, M., 1986).
As previously rnentioned rnost chiral recognition processes require the
usual "three point" interaction , whereas the separation of diastereomeric
complexes (ion pairs) can sometimes be achieved by a two point interaction. It
appears that the underlying mechanisms responsible for a chiral separation
based on selectors in the mobile phase are mudi more cornplex. In addition to
the interaction behveen the chiral selector and solute, the entire equlibriurn of the
diromatographic system must be taken into account (Pettersson, C. and Schill,
G., 1981).
Chiral mobile phases were also produced using solutions of chemically
modified B-cyclodextrin. Specifically, three forms of hydroxy1propyl-B-
cyclodextrin (containing varying degrees of hydroxypropyl substitution) were
added at either 1 or 5 mM concentration to a mobile phase of phosphate buffer
(M/15; pH 8.2) containing 30% acetonitrile. Chromatography was studied using
a silica column ( Brownlee; Spheri-5; 100 x 4.6 mm) at a flow rate of 0.5ml/min
with detection at 254 nm. While the retention time of Selegiline hydrochloride
was modified by the presence of the hydroxypropyl-D-cyclodextrin, the system
failed to show separation of the enantiomers of Selegiline hydrodiloride.
Reverse Phase Separations
Having established that it was not possible to separate Selegiline without
prior derivatization using either a chiral column or a chiral mobile phase
additive, separation of the diastereomers formed from reaction with (-) menthyl
chioroforma te using a reverse phase system was inves tigated. Separations were
started using various C-8 columns with mobile phases containing rnethanol,
water and modifiers such as tetrahydrofuran, diethyl ether, or isopropanol.
Experiments with a 150 x 4.6 mm Zorbax C-8 column (5 pM) showed inferior
separations as compared to other C-8 columns. Such Zorbax columns are end
capped and often used to separate cornpounds that exhibit tailing. The inferior
separation seems to indicate that the presence of some untreated silanol groups
are needed for the separation. It was also noticed that the addition of either
tetrahydrofuran or diethyl ether improved the peak shape. Although, the peaks
were not adequately resolved with these mobile phases, thinner peaks were
obtained. Table II h t s some of the different coiumns and mobile phases studied.
Eventually, an Altex C-8 (130 x 4.6 mm) was employed with a mobile
phase consisting of 70/30, methanol/water at a 80w rate of 1.3 rnl/rnin. The
separation had a resolution l& of 1.12, and a separation factor, a of 1.07. The
retention times of the two diastereomers were 27.6 and 29.4 minutes. As can be
seen from the chromatogram (figure 19) there are clearly two peaks; however
baseline resolution is not achieved (Rs = 1.5). An S value of 1.12 for two peaks
with a Peak height ratio of 111 represents between a 98.0 O/O to 99.4 Oh resolution
(see figure 20). Since it was assumed that adjustments to the chromatography
might have to be made for the tablet assay, this separation was considered
adequate. Separate injections of individual enantiomers showed that the [SI
carbarnate of methamphetamine eluted fïrst followed by the [RI denvative. Since
one of the objectives of this work will be to apply the chiral separation to assay
Selegiline tablets (expecting only to see the [RI enantiomer) this is the preferred
order, as it is always desirable to'have the minor antipode eluted before the
major due to the possibility of tailing (Lough, W. J., 1989).
Table II Method Development us in^ M e h o I Water
Method Develo~rnent Usine MethanoWater / Diethvl ether
MethanoüH,-O/Diethyl ether
Method DeveIo~ment Using MethanoVWater/TKF
* *Zohau CoIumn ***Flow rate in d m i n
(-) Menthyl carbamate derivative of (f) Methamphetamine RG = 1.12 COI-: ~ l t e x C-8 (150 x 4.6-1, Mobile 70/30 m e t h L i / ~ ~ o Flow rate: 1.3 ml/min, Detector set at 254 nm, 20 pl manual injection
Figure 19
Fig. 20.4 Resolution of neighbouring peaks, peak-height ratio 1 :1
Resolution
Relationship Between Peak Separation and Resolution (Meyer, V. R., 1994)
Figure 20
5 me Prevarations - Metharn~hetamine and SeIedine carbarnate
Having conducted al1 of the preliminary separations on the bulk
derivatives of either methamphetamine or desmethyl-selegiline, conditions for
the derivatization reaction at a concentration sirnilar to that of Selegiline tablets
(5 mg) needed to be determined. In addition, the reaction of (-1 menthyl
chioroformate and Selegiline had not been performed on a microscale and it was
suspected that harsher reaction conditions may be necessary for the tertiary
amine. The derivatization conditions used were a modification of those used by
Gubitz et al (Gubitz, G-, Wintersteiger, R. and Hartinger, A., 1981). The reaction
was performed in benzene with a 30 fold molar excess of potassium carbonate
and a 10 fold molar excess of (-) menthyl chloroformate. Specifically, 5mg of the
amine hydrochloride, lOOmg of potassium carbonate, 500 pl of (-) menthyl
diloroformate (10% v /v solution in benzene) and 500 pl of benzene were added
to a Silli-vap vial. The vials were sealed with appropriate septa and heated.
Studies showed that the derivatization reaction with Selegiline had to be
conducted at 1000 C, whereas the carbamates of methamphetamine and
amphetamine couid be formed in higher vields at 600 C. Therefore, samples
were heated for one hour at 1000 C and allowed to cool to room temperature.
The liquid layer was poured into a centrifuge tube, allowing the solids to remain
in the vial. In order to remove any excess menthyl chloroformate 3 mls of
methanolic potassium hydroxide was added and the mixture vortexed. To aid in
phase separation 5 mls of distilled water was added and the tube was then
centrihged. The benzene layer was removed, dried (molecular sieve), dilrited as
necessary and injected onto the HPLC.
As expected the yields of the Selegiline carbamates were much lower than
those formed from either methamphetamine or desmethyl selegiline. Calculating
peak areas of Selegiline against those of methamphetamine, it was estimated that
there was only an 8% yield from the Selegiline hydrochloride derivatization
reaction. Extending the time for the Selegiline hydrochloride reaction did not
seem to improve the yields significantly [It was difficult to tell at this stage in the
absence of an interna1 standard if an increase in yield was due to evaporation or
the additional reaction tirne]. Although the yields were low it was determined
that the N-propargyl group was cleaved from the Selegiline molecule, since the
carbarnate retention times of both methamphetamine and Selegiline were
identical and different from those of desmethyl Selegiline. Although the reaction
worked, with continued column use the separation deteriorated, the peaks
widened resulting in a decreased separation, R, = 0.91 (figure 21). To improve
the chromatography the mobile phase was adjusted to include 5% isopropanol.
Figures 22 and 23 demonstrate the effect of isopropanol on the chromatography.
Reverse Phase Chromatoera~hy - Interna1 Standard Selection
For the present tirne the separation was adequate and an intemal standard
had to be chosen. Generally the use of an intemal standard ensures a high degree
of analytical precision, specifically the uncertainties introduced by manual
injection can be avoided. Since this separation requires a derivatization reaction,
the use of an intemal standard overcomes any slight incowistency in the reaction
conditions and extraction process. An ideal internal standard would be either an
achiral or optically pure primary or secondary amine. Table III lists the internal
standards evaluated along with their retention times. Based on a reasonable
retention tirne (15 minutes) and ease of derivatization, R-amphetamine used in
the free base forrn was chosen.
The extremely low yields of the Selegiline reaction were a concem and it
was considered that changing the solvent from benzene to dichloroethane might
increase the yield. Dichloroethane was chosen because, it had been used by other
(-) Menthyl carbarnate derivatives of (k) Methamphetamine Chromatography deteriorating, Rs = 0.91 Column: Altex C-8 (150 x 4.6mm), Mobile phase: 70/30 methanol/HzO Flow rate: 1.3 m l h i n , Detector set at 254 nm, 20 pl rnanual injection
Figure 21
(-) Menthyl carbarnate derivatives of (k) Methamphetamine Effect of 2% isopropanol ( P A ) on the separation Column: Altex C-8 (150 x 4.6mm), Det'ector set at 254 nm, 20 pl injection Mobile phase: 68/30/2 methanol/H20/IPA, Flow rate: 1.3 ml/min
I Figure 22
(-) Menthyl carbamate derivatives of (k) Methamphetamine Effect of 5% isopropanol (PA) on the separation Column: Altex C-8 (150 x 4.6mm), Detector set at 254 nm, 20 pl injection Mobile phase: 65/30/5 methanol/H20/IPA, Flow rate: 1.3 ml/min
Figure 23
investigators in similar reactions and has a high boiling point (830 C) with a W
cut off of 230 m. Chloroform was considered but it has boiling point of 610 C
and a UV cut off of 245 nrn. Experiments indicated that replacing benzene with
dichloroethane increased the yields of the Selegiline carbamates as well as those
of methamphetamine and desmethyl selegiline. Although there was a substantial
improvement (-35% for Selegiline) using dichloroethane, yields from both
methamphetamine and desmethyl selegiline were still greater than those of
Selegiline. At this point the amount of isopropanol in the mobile phase was
increased to IO%, for a final mobile phase composition of 60:30:10, methanol:
water: isopropanol with a flow rate of 1.3 ml/min.
Having corne to the point where the conditions of the reaction and
chromatography were
amphetamine were
set, two calibration curves of R-methamphetamine and R-
prepared. In one curve the interna1 standard, R-
amphetamine was derivatized separately from methamphetamine and then
mixed together prior to injection. Ln the second calibration curve, both R-
amphetamine and R-methamphetamine were derivatized in the same vial. The
slopes and intercepts of the two curves were very similar, indicating that there
was enough (-) menthyl chloroformate in the reaction vial to derivatize both
compounds. Although the results from the calibration curves were encouraging,
by the end of the second calibration curve the chromatograms showed tailing
and poor peak shapes (figures 24 and 25).
Table LU
Reverse Phase - Interna1 Standards
Compound Retention time
R- Amphetamine 14.24 minutes
Pargyline (two peaks) 20.61/22 minutes
a , a - Dimethy 1 phenylethylamine 24.20 minutes
R-a-Methyl pargyline (peak tails badly) 25.80 minutes
S-N-GDimethy 1 N-propargy 1 l3-phenyle thy lamine 36.61 minutes
N- a , a Trimethylpropargyl B-phenylethylamine 50.50 minutes
Deriva tized Metham~hetamine
25.80 minutes
(-) Menthyl carbarnate derivatives of [RI Methamphetamine and [RI Amphetamine / Peaks starting to tail Column: Altex C-8 (150 x 4.6mm), Detector set at 254 MI, 20 pl injection Mobile phase: 60/30/10 rnethanol/H20/IFA, Flow rate: 1.3 ml/min
Figure 24
(-) Menthyl carbamate derivative of [RI Amphetamine Interna1 Standard Peaks are T a i h g
Column: Altex C-8 (150 x 4.6mm), Detector set at 254 nm, 20 pl injection Mobile phase: 60 /3O/lO methanol/H20/IFA, Flow rate: 1.3 ml/min
Figure 25
Ofien when peaks are not gaussian in shape it can be due to either the column
deteriorating, too much sarnple being injected (sample loading) , solvent
incompatibility or an inappropriate pH (Meyer, V. R., 1994). Each standard via1
contained 100 pl of dichloroethane and 200 pl of methanol. Perhaps the 2:l ratio
of methanol : dichioroethane was a problem. Not really sure if it was a solvent
incompatibility problem or just column deterioration, normal phase separations
were studied. Dichloroethane would be more compatible with a normal phase
separation.
Normal Phase Separation
The C-8 column was replaced with a silica column [Brownlee Spheri- 5,
220 x 4.6 mm 1 and various concentrations of hexane and diethyl ether or
isopropanol were evaluated for the mobile phase. Using isopropanol in the
mobile phase did not result in any separation. Isopropanol is probably too polar
and therefore prevents the derivatized products from interacting with the
stationary phase to allow for a separation. Experiments with diethyl ether
proved successful and showed a superior separation as compared with that of
the reverse phase system (figure 26). While optimizing the amount of diethyl
ethyl ether to be used in the mobile phase it was noticed that a slight
modification in the amount of diethyl ether (1%) resulted in a significant
difference in the retention times of the resulting carbamate derivatives. An
example of th& is seen in figures 27 and 28. With a flow rate of 1.5 ml/min and a
mobile phase of 96:4 (hexane: diethyl ether) the retention times of the
methamphetamine carbamate diastereomers are 25.42 and 26.45 respectively,
whereas using a mobile phase of 95:5 (hexane: diethyl ether, 1.5 ml/min)
resulted in retention times of 21.03 and 21.84 for the same derivatives.
(-) Menthyl carbamate derivatives of (k) Seleghe (-) Selegiline - 18.93 min / (+) Selegdine - 19.71 min Normal Phase Cluomatography Column: Silica Brownlee Spherî 5 (220 x 4.6mm) Mobile phase: 95 /5, hexane/diethyl ether, Flow rate 1 Jml/ min Detector set at 254 nm, 20 pl manual injection
Figure 26
(-) Menthyl carbarnate derivatives of (k) Methamphetamine (-) Methamphetamine - 21.03 min / (+) Methamphetamine 21.84 min Normal Phase Chromatography 96/4 Hexane/diethyl ether Column: Silica Brownlee Spheri 5 (220 x 4.6mm) Mobile phase: 96 /4, hexane/diethyl ether, Flow rate 1 Sml/ rnin Detector set at 254 nm, 20 yl manual injection
Figure 27
(-) Menthyl carbarnate derivatives of (k) Methamphetamine (-) Methamphetamine - 25.42 min /(+) Methamphetamine - 26.46 min Normal Phase Chromatography, 95/5 Hexane/diethyl ether Column: Silica Brownlee Sphen 5 (220 x 4.6mm) Mobile phase: 95 /5, hexane/diethyl ether, Flow rate 1 Sml/ min Detector set at 254 nm, 20 pl manual injection
Figure 28
Typically when working in reverse phase duomatography a larger percentage of
the more non-polar component in the mobile phase is needed for such a change
in retention time. Having noticed the pronounced effect that the concentration of
diethyl ether had on the retention tirnes, the mobile phase container was tightly
sealed preventing evaporation of the ether. However it should be noted that
changes in retention times were not totally preventable, factors such as
environmental conditions (room temperature) and slight differences in the
hexane ( due to different manufacturers) resulted in minor variations of retention
times. At this time a mobile phase of 9 6 4 (hexane: diethylether) at a fiow rate of
1.5 ml/min was chosen and such conditions were employed to select an
appropriate intemal standard. In addition to improving the chromatography,
changing from a reverse phase system to a normal phase system also reversed
the elution order the diastereomers. In the normal phase systern the order is R (-)
followed by S (+).
Normal Phase Chrornatoera~hv - Intemal Standard Selection
Using the same criteria as previously mentioned in the reverse phase
separation various interna1 standards were evaluated. Although [RI
amphetamine was used in the reverse phase separation, its retention time of 24
minutes was too close to that of [RI methamphetamine. Table IV lists the
compounds that were evaluated using normal phase chromatographic
conditions, along with their retention times. Since a , a -Dirnethylbenzylamine
did not interfere with the quantitation of the Selegiline carbarnates it was chosen
as the intemal standard. Since the separation of the derivatives and the interna1
standard was suffïcient it was decided to adjust the mobile phase to 95:5 (hexane:
diethylether) in order to shorten the total run time of the separation. The flow
rate remained the same at 1.5 ml/min. Figure (29) shows a chromatogram of the
separation of derivatized Selegiline with a , a -Dimethylbenzylamine as the
interna1 standard.
Table TV
Normal Phase Intemal Standards
Com~ound
a, a- Dimethylp heny lethy lamine
Aniline
[RI Arnp hetamine
[SI - a-Methylbenzylamine
a, a- Dimethy lbenzylamine
[RI - a-Methyl benzylamine
Derivatized Methamrhetamine
[RI Methamphetamine
[SI Me thamp he taamine
Retention T h e '
6.18 minutes
6.22 minutes
24.00 minutes
26.92 minutes
34.78 minutes
39.43 minutes
25.42 rninu tes
26.46 minutes
--
The retention tirnes are based on a mobile phase composition of 96:4 (hexane: diethyl ether)
with a flow rate of 1.5 ml/rnin.
(+) Selegiline Carbarnates with a , a -Dimethylbenzylamine Retention tirne of Interna1 Standard 24.63 minutes Normal Phase Chromatography, 95/5 Hexane/diethyl ether Column: Silica Brownlee Spheri 5 (220 x 4.6mm) Mobile phase: 95 /5, hexane/diethyl e h , Flow rate 1.5ml/min Detector set at 254 nm, 20 p1 manual injection
Figure 29
Preparation of Calibration Curves - Metham~hetamine
Using normal phase chromatography individual calibration curves for the
(-) menthyl ~Moroformate deriva tives of [SI and [RI methamphetamine and [SI
and [RI Selegiline using u , a -Dimethylbenzylamine as an intemal standard
were prepared. Having realized the benefit of dichloroethane, both the intemal
standard and the (-) menthy 1 chloroformate were dissolved in dichloroethane.
Calibration curves for methamphetamine were prepared by weighing
approximately 2.5 to 5.5 mg of an individual enantiomer into a silli-vap vial. To
each vial 500~1 of (-) menthyl chloroformate (10% v/v in dichloroethane), 500pl
a , a -dimethylbenzylamine (0.5% v / v in dichloroethane) and 100 mg of
potassium carbonate was added. The vials were sealed with the appropriate
septa and heated for one hour at 1000 C. After allowing the vials to cool, the
liquid layer was poured into a l5ml centrifuge tube, leaving the solid behind and
3 ml of methanolic potassium hydroxide was added. The mixture was vortexed
for two minutes and 5 ml of distilled water added. The tubes were then
centrifuged for ten minutes. The top aqueous layer was removed by a pasteur
pipet and discarded. The dichloroethane layer was poured into a clean dry silli-
vap vial containing molecular sieves ( type 4A to remove any residual water).
The standard solution (100 pl) was diluted to 1.0 ml with hexane. Solutions were
then injected until three injections yielded area ratios within 2% of each other.
The range of 2.5 to 5.5 mg was diosen because it brackets the expected 5mg dose
contained in a tablet. Results of the first calibration curve of [RI
methamphetamine showed inconsistent internal standard peak areas.
Specifically, as the concentration of methamphetamine was increased the peak
areas of a , a -dimethylbenzylamine decreased. This suggested that there was
not enough (-) menthyl chloroformate to derivatize both methamphetamine and
the internal standard. To alleviate this problem the stock solution of (-) menthyl
chloroformate was increased from 10% v/v to 20% v/v. This solved the problem,
the intemal standard peak areas were consistent again, and in al1 subsequent
reactions 500pl of (-) menthyl chloroformate (20.0 v/v in dichloroethane) was
added to the derivatization reactions. The above procedure was conducted for
both enantiomers of methamphetamine. Linear regression analysis using the
least squares method showed straight lines for both curves with R2 values of
0.99490 and 0.98924 for the [RI and [SI calibration curves respectively. Although
the y-intercept (0.15413) of [SI methamphetamine was not optimal the data does
indicate that the derivatiza tion process is quantitative. Statistical data and
corresponding graphs are presented in the Experimental section.
Preparation of Calibration Curves - Seledine
Realizing that Selegiline hydrodiloride would have to be extracted from
the tablets it was decided that a procedure be developed and used to extract
Selegiline hydrodiloride from the tablets as weil as to prepare calibra tion curves.
Since Selegiline hydrochloride is freely soluble in chloroform and not as soluble
in dichloroethane it was decided to extract the hydrochloride salt into
chloroform and then derivatize as usual with b o t - the interna1 standard and (-)
menthyl chloroformate in dichloroethane. Chloroform with a boiling point of
61.20 C and a UV cut off of 245nm would not be a suitable derivatization
solvent, and therefore would only be used as the extraction solvent. Experiments
showed that extracting in situ (in the silli-vap vial) resulted in the greatest yields.
Attempts at dissolving 5mg of Selegiline hydrochloride in water, making the
solution basic and then extracting into chloroform resulted in low yields and
large variations possibly due to the known volatility of Selegiline base. The first
calibration curves of [RI and [SI Selegiline hydrochloride were prepared by
weighing a specific amount of Selegiline hydrodiloride (2.5 to 5.5 mg) into a silli-
vap vial, recording the weight and adding Iml of chloroform to each vial. The
chloroform layer was evaporated to dryness under a gentle stream of air. To eadi
vial 100 mg of potassium carbonate, 500pl(-) menthyl chloroformate (20% V/V in
dichloroethane) and 500pI a , a -dimethylbenzylamine (0.25% v / v in
didiloroethane) was added. The silli-vap vial was heated for one hour at 1000 C ,
then allowed to cool. As with the methamphe tamine standard preparation the
liquid layer was poured into a centrifuge tube, 3ml of methanolic potassium
hydroxide solution added and the contents vortexed. Distilled water (5ml) was
added and the mixture centnfuged for ten minutes. Using a disposable pasteur
pipet the top aqueous layer was removed and discarded. The remaining
dichloroethane Iayer was placed into a clean silli-vap vial containing molecular
sieve (type 4A) for the removal of residual water. Due to the differences in yields
relative to pure methamphetamine, the final standard solution was prepared by
adding 100~1 of standard solution to 200~1 of hexane. Each standard solution was
injected until three injections pelded area ratios within 2% of each other.
Cornparisons of the two curves showed similar dopes of 0.02329 and
0.02070 for the [RI and [SI enantiomers indicatuig consistent extraction and
derivatization of the two enantiomers. Regression analysis showed correlation
coefficients of 0.993 and 0.994 for the [RI and [SI calibration curves respectively.
Although these results showed reasonable correlation coefficients, one final
modification was made. To eliminate errors associated with weighing srnall
amounts (2.5 - 5.5 mg) of Selegiline hydrochloride into individual vials, a stock
solution of Selegiline hydrochoride in chloroform was prepared. Using an
Eppendorf pipet specific amounts ranging from 100 ~1 to 500 pl of Selegiline
hydrochloride stock solution were added to individual silli-vap vials, which
were then evaporated to dryness under a stream of air. The remainder of the
derivatization procedure was carried out exactly as had been done previously.
Regression analysis of the [RI curve showed improvernent with a correlation
coefficient of 0.999. Figure 30 shows a chromatogram of [RI Selegiline and a , a
-dirnethylbenzylamine standard solution derivatized with (-) menthyl
chloroformate. Statistical information and the calibration curves are presented in
the Experimental section.
-0.0050 Aboorbancm
O.OQOQ O-OQSO I l ,
0.OIOQ 1 1 I 1 t 1
1 1 I I
[RI Selegiline and a , a -dimethylbenzylamine standard solution derivatized with (-) menthyl chloroformate, Colurnn: Silica Brownlee Spheri 5 (220 x 4.6mrn) Mobile phase: 95 /5, hexane/diethyl ether, Flow rate 1 Sml /min Detector set at 254 m, 20 pl manual injection
Figure 30
Avvlication of the Assav to Seleeiline hvdrochloride Tablets
The final portion of this work was to apply the assay to analyze Selegiline
hydrochloride tablets. Once again using the favorable solubility of Selegiline
hydrochloride in chioroform, experiments were conducted to develop a suitable
extraction technique. The major difference between assaying the tablets and the
raw material is the presence of the excipients. Each Selegiline hydrochloride
tablet contains 5mg of Selegiline hydrochloride, 84 mg of Lactose, 9 mg of
Polyvinyl povidone (PVP), 3 mg of talc, 3 mg of magnesium stearate and 16 mg
of starch for a total tablet weight of 150 mg'. Experiments showed that a tablet
would not dissolve by simply placing a tablet in a solution of chloroform and
shaking by means of a mechanical shaker. Therefore it was determined that a
tablet (weight recorded) would have to be cmshed uçing a mortar and pestle and
then extracted. Once the Selegiline was extracted the derivatization could be
carried out as had been done previously.
Borrowhg from the technique used to prepare the calibration curves, an
individual tablet was cmshed and a known amount (approximately 100 mg) of
the powder was placed into a test tube to which 1.0 ml of chiorofom was added.
The tube was allowed to shake for 15 minutes on a mechanical shaker. At this
time experiments were conducted to separate the excipients from the Selegiline
by means of filtering. Filter paper, scintered glass filters and a nail filter were
employed for this purpose. Results from these experiments showed very low
yields. Realizing that the Selegiline was being absorbed ont0 the various filters it
was decided to centrifuge the crushed tablet and pour off the remaining liquid. A
single tablet was therefore weighed, crushed using a mortar and pestle and an
aliquot added (100 mg) to a centrifuge tube. Chloroform was added (1 ml), the
** T h information was listed on package of Selegiline tablets
86
tube was centrihiged for 5 minutes and the liquid layer poured into a silli-vap
vial. The chloroform was evaporated to ddryness under a gentle stream of air and
100 mg of potassium carbonate, 500 pl of (-) menthyl chloroformate, and 500 pl of
internal standard solution was added. The derivatization procedure was
continued as previous. Results showed a mean recovery of 96.5O/0 with a
standard deviation of 2.01%-
in order to increase the recovery the crushed tablet was extracted twice
with two 1 ml fractions of chloroform. Results indicated 102.0Y0 recovery, this
procedure was then used to analyze the two different lots of Selegiline
hydrochloride tablets. To demonstrate the lack of interference from the
excipients, an excipient mix was prepared using the above quantities and
assayed using the same procedure (including the internal standard). Results
showed no interference.
Studies employin~ other chloroformate reagents
As previously mentioned, a further objective of this study was to
investigate the reactions of Selegiline, methamphetamine and desmethyl
selegiline with other chloroformate reagents. While the reaction of menthyl
chloroformate with the tertiary amine Selegiline yielded the carbarnate
derivative of methamphetamine as the sole identified product i.e. was the result
of loss of the propargyl function from the quatemary intermediate, i t was
unknown whether reaction of Selegiline with other chloroformates would
similarly yield a single product from the dealkylation reaction. The
chloroformate reagents employed in this study were ethyl chloroformate (EC)
and 9-fiuorenylmethyl chloroformate (FMOC). Although neither of these
reagents are chiral, they are both structurally different from (-) menthyl
chloroformate, the one (ethyl chloroformate) being a simple aliphatic reagent
while FMOC is the derivative of a highly bulky (planar) and lipophilic aromatic
alcohol. The possibility exists that these reagents may yield products which:
a) dealkylate more readily than the corresponding menthyl chloroformate
derivatives giving rise to the carbamate derivative of a secondas. amine in
higher yield
b) deaikylate by competing routes giving Ne to a mixture of products
C) dealkylate by an unknown mechanism yielding the carbamate
derivative of desrnethyl selegiline as the major or only product.
In addition , the results obtained from investigating FMOC as a
derivatizing reagent would possibly indicate the potential for employing the
commercially available (but extremely expensive) (+) -1-(9-fluorenyl) ethyl
chloroformate (FLEC) as a derivatizing agent for Selegiline. The latter
derivatizing agent (and other fiuorenyl derivatives) is highly chromophoric and
also strongly fluorescent and would thus provide a far more sensitive assay
procedure than when using enantiomers of menthyl chloroformate.
The studies with ethyl chloroformate were carried out using the
conditions. previously established for reaction of Selegiline with menthyl
chloroformate with dichloroethane as the derivatizing solvent and
chromatographie separations performed using the normal phase system. Figures
31, 32 and 33 show chromatograms of racemic methamphetamine, racemic
desmethyl selegiline and racernic Selegiline. Looking at the chromatogram of
racemic Selegiline derivatized with ethyl chloroformate (figure 33) there are
three peaks present (13.76, 16.24 and 31.98 minutes). The peak at 31.98 minutes
corresponds to the peak obtained from the derivatization of racemic
methamphetamine with ethyl chloroformate (figure 32). The peak at 16.24
minutes appears to be the same peak obtained from the derivatization of ethyl
chloroformate with desmethyl selegiline which has a retention time of 16.35
minutes (figure 32), while the peak at 13.76 minutes is only seen on this
chroma togram and does no t correspond to any ethyl carbama te p roduct.
Although further experiments would be needed it does appear that the
derivatization of Selegiline with ethyl chloroformate does result in a mixture of
products. However the predominant carbamate produced is the
methamphetamine carbamate.
Interestingly the peak area corresponding to the desmethyl carbamate is
slightly higher than that of the methamphetamine carbamate (5mg of both
methamphetamine and desmethyl selegiline were derivatized with ethyl
chloroformate). As with (-1 menthyl chloroformate the yields are higher
(approximately 32.0%) for the methamphetamine ethyl carbamate than the
Selegiline ethyl carbamate. This is not surprising since the derivatization reactior.
of methamphetamine with (-) menthyl chloroformate (in dichloroethane, normal
phase separation) has a 65% greater yield than that of Selegiline derivatized with
(-) menthyl chloroformate.
Studies with 9-fluorenvlmethvl chloroformate (FMOC)
Experiments were performed employing both benzene and
dichloroethane as the derivatizing solvent. Although the experiments performed
in benzene gave confusing results, reactions performed in dichloroethane gave
results similar to those of ethyl chloroformate. Derivatization reactions of
Selegiline, desmethyl selegiline and methamphetamine with FMOC in
dichloroethane showed that the derivatization of Selegiline resulted in the
formation of carbamates corresponding to both desrnethyl selegiline and
methamphetamine 9-fluorenylmethyl carbama tes. (Figures 34,35 and 36). Once
again the major product of the reaction of Selegiline with FMOC was the
formation of the methamphetamine ethyl carbamate. In addition to the peaks
corresponding to amine carbamates was an additional peak which elutes
immediately after the methamphetamine carbamate peak. Derivatization of a
blank solution indicate that this additional peak cornes from unreacted 9-
fluorenylmethyl chloroformate and is most likely the fluorenyl alcohol. In a
similar readion reported by La Croix et al(La Croix, R., Pianezzola, E., Benedetti
and Strolin Benedetti, M., 1994) FMOC was used to derivatize and
amphetamine, methamphetamine and desmethyl selegiline from plasma. Ln this
reaction proline was used to remove any unreacted 9-flurorenyl methyl
chioroformate. The use of either a suitable amine as done by La Croix or an
extraction could be used to remove any unreacted 9-fluorenyl chloroformate.
As seen in ail other derivatization reactions, the carbamate of
methamphetamine is produced in much higher quantities than the Selegiline
carbamate. As expected due to the highly chromophoric nature of the FMOC
reagent, the reactions of FMOC with al1 three amines is a much more sensitive
assay than reactions with either (-1 menthyl chloroformate or ethyl
chloroformate.
(f) Methamphetamine Derivatized with Ethyl Chloroformate Deriva tiza tion in Dichloroethane, Column: Silica Brownlee Spheri 5 (220 x 4 . 6 m ) Mobile phase: 95 /5, hexane/diethyl ether, Flow rate 1.5ml/ min Detector set at 254 nm, 20 pl manual injection Peak at 30.39 - Methamphetamine ethyl carbarnate
Figure 31
(+_)Desmethyl selegiline Deriva tized with Ethyl Chloroforma te Derivatization in Dichloroethane, Separation on Silica Column Column: Silica Brownlee Spheri 5 (220 x 4.6mm) Mobile phase: 95/5, hexane/diethyl ether, Flow rate 15ml/min Detector set at 254 run, 20 pl manual injection Peak at 16.35 - Desmethyl seleghe ethyl carbarnate
Figure 32
(f) Selegiline Derivatized with Ethyl Chloroformate Derivatization in Dichioroethane, Separation on Silica Column Co lum: SiLica Brownlee Spheri 5 (220 x 4.6mm) Mobile phase: 95/5, hexane/diethyl ether, Flow rate l.Sml/rnin Detector set at 254 nm, 20 pl manual injection Peak at 16.24 - Desmethyl selegiline ethyl carbamate Peak at 31.98 - Methamphetamine ethyl carbarnate
Abeorbanca - a m s a a. 3 0 0 ~ a. 0050 o.0100
0.00 l 1 1 I i 1 1 1 1 1 I l l I I I
1 J
4 --5 -. - !:Il 1.5'1
10.00 -
Figure 33
*
1o.00 - ,
10.00 - i - 31.96 t
* f '
39.99 1 I I i i l K 8 1 1 " 1 i 1 1 -0.0050 0.0000 0.0050 0.0100
(+) Methamp hetamine deriva tized with FMOC Derivatization in Dichloroethane, Separation on Silica Column Column: Silica Brownlee Spheri 5 (220 x 4.6mm) Mobile phase: 95 /5, hexane/diethyl ether, Flow rate lSml/ min Detector set at 254 nm, 20 pl manual injection Peak at 48.66 - Methamphetamine FMOC derivative Peak at 52.15 - Excess FMOC reagent
Figure 34
(i) Desmethyl selegiline derivatized with FMOC Derivatization in Didiloroethane, Separation on Silica Column Column: Silica Brownlee Spheri 5 (220 x 4.6mm) Mobile phase: 95/5, hexane/diethyl ether, Flow rate 1 Sml/ min Detector set at 254 nm, 20 pl manual injection Peak at 25.66 - Desmethyl selegiline FMOC derivative Peak at 51.55 - Excess FMOC reagent
Figure 35
Msorbauca -o.itaso 0.0000 o.oa!o O.QldO
L 1 1 I I I 1 1 1 1
1 I fi t I I I 1
1 . 1 1 1 1 I I I I 1 ' I I I I
1
-Q .O050 O.OOQ0 O.0OSO Q.0100
(f) Selegiline Derivatized with FMOC Derivatization in Didiloroethane, Separation on Silica Colurnn Column: Silica Brownlee Spheri 5 (220 x 4.6mm) Mobile phase: 95/5, hexane/diethyl ether, Flow rate 1 . h l / min Detector set at 254 nm, 20 pl manual injection Peak at 25.89 - Desmethyl selegiline FMOC derivative Peak at 48.46 - Methamphetamine FMOC derivative Peak at 52.02 - Excess FMOC reagent
Figure 36
E x ~ e r h e n ta1
Al1 weighing were done on a five place Mettler balance, model number
86. (-) Menthyl chloroformate ( 99% ee by GLC), ethyl chloroformate (97%
purity) and 9-fluorenyl methyl chloroformate were al1 purchased frorn Aldrich.
(-) Selegdine hydrochloride used to prepare al1 (-) Selegiline standard curves was
generously donated by Apotex Inc. Individual enantiomers of amphetamine
hydrochloride, methamphetamine hydrochloride and the (+) enantiomer of
Selegiline hydrochloride were previously synthesized by J. Barry Robinson.
Potassium carbonate (anhydrous), ammonium phosphate dibasic and potassium
hydroxide were al1 reagent grade. Synthesis of racemic methamphetarnine,
desmethyl selegiline, Selegiline, a, a-dimethylbenzylamine and al1 precursors
are listed in the appendix. Al1 solvents used were either reagent or HPLC grade.
Samples were derivatized in Silli-vapB vials fi tted with appropriate septa
and heated using a heating block equipped with sample wells designed to hold
the vials. The heating unit contains a thermometer to monitor the temperature.
Infrared Spectra of synthetic materials was performed using a Perkh
Elmer Infrared Spectrophotometer model 1330 and were run as liquid films or
Nujol mulls between NaCl plates. Melting points determinations were done
using a Thomas Hoover Capillary melting point apparatus and are uncorrected.
Initial reverse phase chiral and achiral separations were performed using
a Waters M-45 solvent delivery system , Waters 490E programmable
multiwavelength detector with either a Honeytvell chart recorder or Shimadzu
C-RIA Chromatopac integrator. In both cases manual injection was by a
Rheodyne valve rnodel7725i fitted with a 20 pl loop .
The HPLC used for the later chiral reverse phase separations and al1 chiral
normal phase separations was a Beckman system consisting of a programmable
pump model 116 and a variable wavelength detector model 166. The HPLC
system is controlled via the System Gold software@ package. Injections were
made manually using a Rheodyne manual injector mode1 7725i fitted with a 20p1
loop. Chromatograms were printed on a Canon inkjet printer .
Chiral reverse phase separations used a variety of columns and mobile
phases. The C-8 columns used include an Altex Ultrasphere 5 ym [150 x 4.6
mm], Brownfee 5 ~III cartridge [IO0 x 4.6 mm ] and a Zorbax 5 pm [IO0 x 4.6
mm]. A variety of mobile phases and flow rates were employed. However, in al1
cases the mobile phase was filtered using a 45 pm nyIon filter and degassed by
sonicacing for 20 minutes. Distilled water and KPLC grade solvents were used to
prepare al1 mobile phases.
Achiral Se~aration - Reverse Phase Chromatoaravhic Conditions
The achiral separation was done using a Waters pBondapak CN column
[150 x 3.9 mm ] with a mobile phase of 82/15, 0.1M ammonium phosphate
dibasic (pH = 3.1, adjusted with phosphoric acid, 85%)/acetonitrile at a flow rate
of 1.0 ml/min. The detector, a Waters 490E was set at a wavelength of 254 m n
and a range of 0.1 AUFS. The Shimadzu C-RIA Chromatopac ùitegrator was nin
at a chart speed of 10 mrn/min. Racemic amphetamine hydrochloride,
me thamphe tamine hydrochoride, desmethyl seiegiline and Selegiline
hydrochloride were prepared at a concentration of 0.1 &ml in 15% acetonitrile
in \va ter.
Chiral Separation -Normal Phase Chrornatoaraphic Conditions
Normal phase separations were achieved using a Brownlee Spheri-5 Silica
cartridge [220 x 4.6 mm ] with a mobile phase of 95:5, hexane : diethylether at a
flow rate of 1.5 ml/min. The mobile phase was filtered through a 45 p m nylon
filter and degassed by sonicating for 20 minutes. The detector was set to 254 nm
at a range of 0.01 AUFS.
Preparation of IR1 and rS1 Methamuhetamine Calibration Curves/ Separation
bv Normal Phase Chromatoera~hv
Specific amounts ranging from 2.5 to 5.5 mg of an individual enantiomer
was weighed into a Silli-vap vial. To each via1 500 pl of (-) menthyl chlorofomate
solution (20% v / v in dichloroethane), 5 0 0 ~ 1 of u , a -dimethylbenzylamine
solution (0.5% v /v in dichloroethane) and 100 mg of potassium carbonate was
added. The vials were sealed with the appropriate septa and heated for one hour
at 100 0 C. After coolinq the liquid contents of the via1 was poured into a 13 ml
centrifuge tube and 3 ml of a saturated solution of methanolic potassium
hydroxide was added. The mixture was vortexed for two minutes and 5 ml of
distilled water added. The tubes were centrifuged for ten minutes and the top
aqueous layer removed by pasteur pipet. The remaining dichloroethane layer
was poured into a clean dry Silli-vap vial containing molecular sieves (type 4A)
to remove any residual water. Each derivatized standard (100 pl) was diluted to
1.0 ml with hexane. Solutions were injected until three injections (20 pl) yielded
area ratios within 2% of each other. Chromatographie conditions as listed above
for chiral normal phase separations were employed. Graphs and statistical data
are presented at the end of this section.
Pre~aration of Initial IR1 and [SI Seledine Calibration Curves - Normal Phase
Çpecific amounts ranging from 2.5 to 5.5 mg of an individual enantiomer
was weighed into a Silli-vap vial. To each vial 1 ml of chloroform was added, the
mixture vortexed and the chloroform layer was then evaporated to drvness
under a stream of air. Potassium carbonate (100 mg), 500 pl (-) menthyl
chloroformate solution (20% v/v in dichloroethane) and 500 pl a, a-
dimethylbenzylamine solution (0.250/0 v /v in dichloroethane) was added to each
vial. The vials were seaied and heated for one hour at 1000 C. After cooling, 3 ml
of methanolic potassium hydroxide solution was added and the contents
vortexed. Distilled water (5 ml) was added and the mixture centrifuged for ten
minutes. The top aqueous layer was removed by pasteur pipet and the
remaining dichloroethane layer was placed into a clean Silli-vap vial containing
molecular sieves. Solutions for injection were prepared by adding 100 pl of
derivatized standard to 200 pl of hexane. Each standard was injected (20 pl) until
three injections yielded area ratios within 2 % of each other. Chroriiatographic
conditions as listed above for chiral normal phase separations were employed.
Graphs and statistical data are presented at the end of this section.
Prevaration of TRI Seledine Calibration Cunre #3
%s calibration curve was prepared exactly as the above curves except
instead of weighing specific amounts of [RI Selegiline into each vial an aliquot of
stock solution of [RI Selegiline is added into each vial. A stock solution of [RI
Selegiline at a concentration of 9.92 mg/ml in chloroform was prepared. Using
an Eppendorf pipet 230 pl to 500 pl of stock solution was added to separate Silli-
vap vials and the chloroform layer evaporated to dryness under a current of air.
The remainder of the derivatization reaction was exactly as described for the
initial [RI and [SI calibration curves. The graph and statistical data are presented
at the end of this section.
Extraction Procedure of Seleeiline Tablets
An individual tablet was weighed, crushed using a mortar and pestle and
an aliquot ( approximately 100 mg) added to a centrifuge tube. One ml of
diloroform was added to the tube and the tube shaken and then centrifuged for
five minutes. The chloroform layer was poured into a silli-vap vial and
evaporated to dryness under a Stream of air. To the original centrifuge tube a
further 1 ml of chloroform was added and the tube was centrifuged for five
minutes. The chloroform layer was added to the original silli-vap via1 and
evaporated to dryness. To each vial, 500 p1 of (-) menthyl chloroformate solution
(20 % v/v in dichloroethane), 500 pl of a, a-dimethylbenzylamine solution
(0.25% v/v in didiloroethane) and 100 mg of potassium carbonate was added.
The vials were sealed and heated for one hour at 1000 C. After cooling, 3 ml of
methanolic potassium hydroxide solution was added and the contents vortexed.
Distilled water (5 ml) was added and the mixture centrifuged for ten minutes.
The top aqueous layer was removed by pasteur pipet and the remaining
dichloroethane layer was placed into a clean Silli-vap vial containing molecular
sieves. Solutions for injection were prepared by adding 100 p1 of derivatized
standard to 200 pl of hexane. Each standard was injected (20 pl) until three
injections yielded area ratios within 2 '10 of each other. Chrornatographic
conditions as listed above for chiral normal phase separations were employed.
Ethvi Chloroformate Derivatization Reactions
To individual Silli-vap vials 5.0 mg of racemic methamphetamine (free
base), 5.0 mg desmethyl selegiline (free base) and 5.0 mg of racemic selegiline
(free base) was added. To each vial 100 mg potassium carbonate and 500 pl ethyl
chloroformate solution (10% v/v in dichloroethane) was added. The vials were
sealed and heated for one hour at 1000 C. After cooling, 3 ml of methanolic
potassium hydroxide solution was added and the contents vortexed. Distilled
water (5 ml) was added and the mixture centrifuged for ten minutes. The top
aqueous layer was removed by pasteur pipet and the remaining dichloroethane
layer was placed into a clean silli-vap vial containhg molecular sieve. The
derivatized methamphetamine and desmethyl selegiline derivatives were
diluted (100 pl) to 1.0 ml with hexane, whereas the derivatized Selegiline was
diluted (100 pl) to 0.5 ml with hexane. In al1 cases 20 pl was injected and
diromatographic conditions as listed above for chiral normal phase separations
were employed.
FMOC Derivatization Reactions
Derivatization reactions in dichloroethane were perfonned by adding 5.0
mg of racemic methamphetamine (Free base), 5.0 mg desmethyl selegiline (free
base) and 5.0 mg racemic Selegiline (free base) into individual silli-vap vials. To
each vial 100 mg of potassium carbonate and 500 p1 of FMOC solution (4 mM in
dichloroethane) was added. The vials were heated for one hour at 1000 C and
allowed to cool. The liquid contents of each vial was decanted into individual
centrifuge tubes to which 3 ml methanolic potassium hydroxide was added. The
tubes were vortexed for two minutes and 5 ml of distilled water was added. The
tubes were centrifuged for ten minutes and the top aqueous layer removed by
pasteur pipet. The dichloroethane layer was poured into a clean dry silli-vap via1
containing molecular sieves (type 4A). Samples (100 pl) were diluted to 1.0 ml
with hexane and 20 pl manual injections were made. Chrornatographic
conditions as iisted above for chiral normal phase separations were employed.
IR1 Metham~hetamine Calibration Curve
*R-ICILAMC - R-Methamphetamine denvatized with (-) rnenthyl chloroformate "DMBA- Dimethylbenzylamine internai standard derivatized with (-) rnenthyl chloroforma te
Standard # Concentration
pgs înjected
R-bWMC* Peak Area
-DMBA** Peak Area
Ratio Mean &
Standard Dev 1
R - Methamphetamine Calibration Curve
6 9
ugs in 20ul injection
[SI- Metham~hetamine - Calibration Curve Data
ta tistical Da ta r2 = 0.98924 m = 0.05925 b = 0.15413
* S-MAhfC-SMethamphetamine derivatized with (-) menthyl chioroformate
1 Standard #
** DMBA-Dimethÿlbenzylamine intemal standard derivatized with (-) menthyl chloroformate
Ratio Mean &
Standard Dev Concentration ugs injecteci
S - W I C * Peak Area
DMBA" Peak Area
S - Methamphetamine Calibration Curve
6 9
ugs in 20ul injection
[SI Seleeiline Calibration Curve Data
Statistical Data rz = 0.99464 m= 0.02070 b= - 0.07630
' SSelegiluie MC-ESelegdme derivatized with (-) menthyl chloroformate (selegiline weighted out)
**DMBA-Dirneth y lbenzylamine intemal standard derivatized rvith (-) menthyl chioroformate
Ratio DPVIBA** Peak Area
Standard # Mean &
Standard Dev a
Concentration ugs injecteci
S-Selegihe MC* Peak Area
S Selegiline Calibration Curve
-O 4 8 12 16 20 24 28 32 36 40
uns in 20ul injection
IR1 Selegiline Calibration Curve #1
ta tis tical Da ta r2 = 0.99343 m = 0.02329 b = - 0.15079
Standard #
* R -Sele@ne MC- R-Selegiline derivatized with (-) rnenthyl chloroformate (selegilrne solid used)
** DMBA-Dimethylbenzylamine intemal standard derivatize with (-) menthyl chloroformate
Concenha tion ugs injecteci
R-Seleghe MC* 1 DMBA" ~a t i o Peak Area
-
T G ~ & Standard Dev Peak Area
1
R Selegiline Calibration Curve #1
O 4 8 12 16 20 24 28 32 36 40
uas in 20 ul iniection
IR1 Selegiline Calibration Curve #3
Concentration R-SelegLLine MC* DMBA" ugs injected Peak Area Peak Area
1 Ratio Standard Dev
Statistical Data rz = 0.99952 m = 0.03042 b = - 0.20560
*R-ÇelegJltie MC-R-Selegiluie derivatized with (-) menthyL chloroformate (selgiline in diloroform stock solution)
**DMBA-DLmethylbenzy Lamhe interna1 standard derivatized with (-) menthyl chloroformate
R - Selegiline Calibration Curve #3
ugs in 20ul iniection
Results of Tablet Assays
Two separate lots of Selegiline hydrochloride tablets were assayed using
(-) menthyl chloroformate as the derivatizing agent and separated 5y the normal
phase chromatographie conditions as described previously. Result of the analysis
of ten Chinoin tablets (lot 06000289) show a mean label claim of 99.776 with a
standard deviation of 4.4. Tnese results demonstrate that the method developed
is accurate and applicable to the analysis of Çelegiline tablets.
Analysis of another lot of tablets produced by Chiesi (lot 118) have a
mu& lower mean label claim of 90.5% with a standard deviation of 12.5. m e n
obtained these tablets had already reached their expiry date and therefore lower
resulh are not totally unexpected. Additional analysis of these tablets did not
show an improvement in the data, rather the results were lower with a mean
label daim of 71.0 '/O and a standard deviation of 6.0 (n = 4). Calibration curves
using a , a -dimethylbenzyIamine as the interna1 standard were used to
calculate al1 tablet data.
Chinoin Tablet Data
Lot # 0600289
PA - Peak Area
Sta tistical Data
Mean Label daim = 99.7%
Standard deviation + 4.4
Chiesi lumex Tablets
First Batch
Lot #Il8
PA - Peak Area
!
StatisticaI Data
Mean label daim = 90.5%
Standard deviation + 12.5
103.5 1 Tab 2 150.0
146.9
F
100.6
?.
Tab 1A
1.92702 1.W78
154397
1
24.28
19.42 1 -56861
- -
4.62951 )0.41625 )0.41471 23.00
23.55
102.0
4.56337
5.05733 5.26389
101.4
0.29799
1
105.6 1 1
0.38203
0.41317 [
334767 Tab 3 100.2 33.87
82.5 0.30529
0 3 4 3 7
Tab 1A
1
I
2 - 8 2
1
I
1 0.30161
4.00476 149.0
0.37968
r 1.47894 1.53053
100.2
I 1
Tab 4
0.296971 19.22 Z . 18-42
149.8
143.0
0.35639 1.47-185 1.43625
1.38092 1.34201
0.28960 102.9
104.2
2.28086 2.23991
1.17677 1.21501
4.69831 ' 0.29392
97.6 21.79 4.133% 10.35677
Tab 1C
Tab 2B
4.47311
1 2230
1 82.6 ' 1
4.03435
7.87387 7.73639
6.2035
149.7
148.8
18.90
101.0
0.30002
0.35601
0.28967 0.28953
0.2S970 6.33015
5.90527
22.91
0.140821 14.70
152.5 Tab 5
0.19194 23.34 62.5
1.96728 2.01206
94.7
-- - 5.92277 10.33972
1
1 1
22.08 20.92 0.33315 0.33644
Chiesi ïumex Tablets Second Batch
Lot #il8
PA - Peak Area
Statistical Data
Mean Iabel daim = 71.0
Standard deviation + 6.0
Aliquot in1 Tab via1 (mg) 1 #
i
109.0 16
103.2 7
Ave (PA)
0.3749
0.30312
Tablet / Sel wt (mg) ( (PA)
!
DMBN Ratio
151.0
102.4
(PA)
17.40
15.96
1.40533 1.45126
0.30016
(PA)
22-47
22.S2
ugs ( ug's
0.30607
77.4
202.6
1
0.325-M 0.32216
J
Assayed
15.22
16.70
3
151.4 12.15474 7.14283
8.49438 8.55248
ii
8 0.32378
'10 label daim Expected
24.02
22-72
5.46350 5.63052
7.03992 1 2.14397 I
151.9 12.76408 2.75529
Std # '
0.25622 0.25775
9 69.9 3
63.2
73.3
0.27983 149.9 6.11123 5.2û457
3
3
1.60356 0.26240 0.2975 / 1.57083
Discussion
While there are several derivatizing reagents available for the chiral
separation of primas. and secondary amines, there are few that are suitable for
tertiary amines- (-) Menthyl chloroformate has been demonstrated to be a
suitable derivatizing agent for the chiral separation of (5) Selegiline
hydrochloride. Ushg the favorable solubility of Selegiline in chloroform a
successful method (99.7 % label clairn) has been developed for a chiral HPLC
assay of Selegiline hydrochloride tablets.
As previously mentioned the reaction of Selegiline with (-) rnenthyl
chloroformate could result in the formation of three possible carbarnates due to
the deavage of either the propargyl group, the methyl group or a combination of
both. Studies clearly show that the reaction of (-) menthyl chloroformate with
Selegiline produce the same carbamate as the reaction of methamphetamine with
(-) menthyl chloroformate. Therefore, the reaction of Selegiline with (-) menthyl
chloroformate results in the dealkylation of the propargyl group. These results
are consistent with results previously reported by Kapnang et al where the
preferred cleavage order for a series of related tertiary amines is N-benzvl > N-
allyl> N-ethyl > N-methyl ' (Kapnang, H. and Charles, G., 1983)
Although the main objective of this thesis was to develop a chiral assay
for Selegiline hydrochloride and Selegiline hydrochloride tablets, the chiral
derivatization and assay procedure developed using (-) rnenthyl chloroformate
can ako be used for the separation of racemic amphetamine, racemic
methamphetamine and racemic desmethyl selegiline. Another benefit of the
reaction of (-) menthyl diloroforrnate with Selegiline hydrochloride is that it
consistently affords only one product. h al1 the reactions performed, the
' This cleavage order is reported for both ethyl and vinyl diloroformates
118
propargyl group was consistently cleaved. Finally, the separation is performed
on a silica c o l u m which is more rugged and less expensive than a chiral
colunin*
The main disadvantage of the derivatization procedure developed is the
lengthy reaction time ( 60 minutes) needed for the reaction of (-) menthyl
chloroformate with Selegiline hydrochloride. Although amphetamine,
methamphetamine and desmethyl selegiline are formed using shorter reaction
times, Seleglute needs both the extended time and high temperature. The
retention time of the intemal standard a, cc-dimethylbenzylamine at
approximately 25 minutes results in a chromatographie run time of thirty
minutes. Studies could be done to identify another suitable amine as an interna1
standard that would elute prior to the Selegiline carbamate to reduce the run
tirne of the chromatogram. The low yields of the reaction are also a concern. As
previously mentioned the yields of the reaction of (-) menthyl chloroformate
with S e l e g h e are substantially lower than the reaction of (-) menthyl
diloroformate with either methamphetamine or desmethyl selegrluie. Trying to
increase the yields by extending the reaction time from one hour to two and
three hours resulted in an increase in yield of Selegrline carbamate. However it
was also noticed that the dichloroethane had evaporated. Therefore, the
increased yields could have been due to a more concentrated solution. The
yields may be increased if the reaction could be camed out in a sample via1
whi& could withstand the heat and not allow for evaporation . Another way to
increase the yield would be to use a more concentrated solution of (-) menthyl
chloroformate. Perhaps a greater excess of (-) menthyl diloroformate reagent
would increase the reaction rate and the overall yield.
In addition to the development of a chiral HPLC assay of Selegiline
hydrochloride, the reactions of methamphetamine, desmethyl selegiline and
Selegiline with two additional achiral chloroformates, ethyl and 9-
fluorenylmethyl chloroformate (FMOC) have also been evaluated. Studies have
shown that the reaction of Selegiline with both ethyl chloroformate and FMOC
resuit in formation of both desmethyl selegiline and methamphetamine
carbamates.
The reactions of the chloroformates studied with methamphetamine,
desmethyl selegiline and Selegiline appear to be affected by the size of the
groups attached to the nitrogen, the temperature of the reaction and the polarity
of the derivatizing solvent. Reactions of al1 three chloroformates with
methamphetamine showed much higher yields than those obtained with similar
reactions with Selegiline. These results are consistent with those of Kometani et
al (Kometani, T., Shunsaku, S. and Mitsuhashi, K., 1976) who reported that the
reactivity of an amine to ethyl chloroformate decreased with an increase in the
size of the group around the nitrogen atom. Even in the reaction of FMOC with
Selegiline, the peak corresponding to the methamphetamine carbamate is larger
than the peak corresponding to the desmethyl selegiline carbamate.
Studies of al1 three chloroformates with methamphetamine, desmethyl
selegiline and Selegiline showed much higher yields when the reactions were
carried out in dichloroethane as compared to those employing benzene as the
derivatizing solvent. Although the exact mechanism of the reaction is not fully
understood, the reaction is believed to proceed by way of a quaternary
intemediate followed by nucleophilic attack by chloride on one of the nitrogen
substituents (Cooley, J. H. and Evain, E. J., 1989). The increased yields seen in
dichloroethane may be due to the increased stability of the quaternary
ammonium intermediate in a more polar solvent.
Studies in benzene showed that the reaction of (-) menthyl chloroformate
with amphetamine and methamphetamine can form the resulting carbamates at
600 C, whereas a temperature of 1000 C is needed to form Selegiline menthyl
carbamate. Similarly, when reacting Selegiline with FMOC in benzene only a
very small single product (correspondhg to the methamphetamine carbamate)
was formed after two hours of heating, while the yields of the resulting
carbarnates of methamphetamine , amphetamine and desmethyl selegiline were
higher after only one h o u at 1000 C. Once again the reactivity may be related to
the size of the substituents on the nitrogen atom as suggested by Kometani et al
which would require harsher conditions for more complicated amines, such as
Selegiline (Kometani, T., Shunsaku, S. and Mitsuhashi, K., 1976) .
As expected due to the highly fluorescent nature of the compound, the
reaction of FMOC with methamphetamine, desmethyl selegiline and Selegiline
resulted in a more sensitive assay than the reactions with either ethyl
chloroformate or (-) menthyl chloroformate.
Although the results of the above experiments do provide further
information pertaining to the reactions of chloroformates with primary,
secondary and tertiary amines, it would be interesting to investigate the
reactions of the above mentioned chiorofornates with tertiary amines bearing
different substituents on the nitrogen. As previously mentioned Kapnang et al
report a preferred cleavage order of N-benzyl > N-allyl > N-ethyl > N-methyl
using vinyl and ethyl chloroformate. However the order rnay be different when
derivatizing with either (-) menthyl chloroformate or FMOC. Ln addition the
reaction conditions should be investigated, specifically the reactions of ethyl
chloroformate and FMOC with Selegiline. The reaction after one hour in
dichloroethane at 1000 C resulted in the formation of two carbarnates. However
if the reaction was heated for a shorter period, at a lower temperature or at a
reduced temperature for a shorter period of time, perhaps only one product
would be formed. If one product could be formed then the reaction of Selegiline
with the chiral (+) -1-(9-fluorenyl) ethyl chloroformate (FLEC) should be
evaluated.
Although 1 chose to use a chiral derivatking agent followed by separation
on a convention adùral column the discussions presented throughout this thesis
demonstrate there are many alternative methods that could employed.
Specifically the use of 2-naphthyl chloroformate followed by separation
on either a cellulose or cyclodextrin chiral column may also result in a chiral
separation. The introduction of the 2-naphthyl group would allow for increased
sensitivity as well as introduce the needed naphthyl group for separation on
either of these chiral colurnns. As with the FMOC derivatization, excess reagent
would have to be removed either by extraction or by reaction with a suitable
amine.
The work presented here illustrates the development of a chiral HPLC
separation for Selegihe hydrochloride. Although alternative methods for the
chiral separation of Selegiline hydrochloride may be possible, derivatization with
(-) menthyl chloroformate followed by separation on a conventional silica
column has been shown to be fairly simple and applicable to the analysis of
Çelegiline tablets. The information obtained regarding the reactions of primary,
secondary and tertiary amines with (-) menthyl chloroformate, ethyl
chloroformate and 9-fluorenylmethyl chloroformate could be valuable to both
the synthetic and analytical chemist.
References
Ahuja, S., Chiral Sepmations by Iiqzl id chronintogrnphy, Amencan Chemical Society,
Washington, KI 1991, pp. 43-100.
Menmark, S. G., Chrornatographic Enantiosepnrntion: Methods and Applicntiorzs, John Wiley & Sons, Toronto, 1988, pp. 15-16,27- 41,90-140.
Barksdale, J. M. and Clark., C.R., Liquid chrornatographic determination of the enantiomeric composition of amphetamine and related dmgs by diastereomenc derivatization \. Cliromntogr. Sci., 23 (1985) 176-180.
Beaulieu, N., Cyr, T.D., Graham, S. J. and Lovering, E. G., Liquid diromatographic method for Selegiline Hydrochloride and related compounds in raw matenals and tablets 1. Assoc. Off. Anal. Chem, 74 (1991) 453-455.
Birkmayer, W., Riederer, P.. Arnbrozi, L. and Youdim, M. B. H., Implications of combined treatment with Madopar and 1-deprenyl in Parkinson's disease Lmicet, II (1977) 43943-
Buck, R. H. and Knunmen, K., Resolution of amino acid enantiomers by high performance iiquid chromatography using automated pre-colurnn derivatization with a chiral reagent [. Chromntogr., 315 (1984) 279-285.
Buck, R. H. and Krummen, K., High-performance liquid chrornatographic determination of enantiomeric acids and alcohoIs after derivatization with a
chiral reagent 1. Chromntogr.. 387 (1987) 255-265.
Carey, F. A., Organic Chemistnj, McGraw-Hill Book Company, New York, 1987, p p 252-286.
Chafetz, L., Desai, M.P. and Sukonik, L., Trace decomposition of Selegiline. Use of worse-case kinetics for a stable drug 1. Phnrm. Sci., 83 (1994) 1250-1252.
Cooley, J. H. and Evain., E. J., Amine dealkylations with acyl dilorides Synthesis, 1 (1989) 1-7.
Cotzias, G. C., Deuelopments In Trentment For Parkinson's Disease , hledcom Press, 1971, pp. 25.
Cçerhati, T. and Forgacs, E., Retention of some monoamine oxidase inhibitory drugs on a Bcyclodextrin polymer-coated silica column 1. Chrornatogr. A,, 660 (1994) 313-318.
Dalgliesh, C. E., The optical resolution of arcmatic amino acids on paper
diromatograms 1. Chem. Soc., (1952) 3490-394.
Debowski, J., Sybilska, D. and Jurczak, J., The resolution of some chiral
compounds in reversed-phase high-performance liquid chromatography by
means of B-cydodextrin inclusion complexes Chrornatographin, 16 (1982) 198-200.
Deyl, Z., Hyanek, J. and Horakova, M., Profihg of amino acids in body fluids
and tissues by means of liquid diromatography [. Chromatogr., 379 (1986) 177-
250.
Doyle, T. D., Adams, W. A., Fry, F. S. Jr. and Wainer, LW., The application of HPLC chiral stationary phases to stereochernical problems of phamtaceutical interest: a general method for the resolution of enantiomenc amines as B- naphthyl-carbarnate derivatives \. Liqirid Chrornatogr., 9 (1986) 455471.
Ecsery, Z., Kosa, I., Knoll, J. and Somfai, E., Neth. Pat. App. 6,605,956 Chem. Abs.,
67 (1967) 21611~.
Einarsson, S., Selective determination of secondas. amino acids using precolumn derivatization with 9-fluorenylmethyl chloroformate and reversed-phase high- performance liquid chrornatography 1. Chromatogr., 348 (1985) 213-220.
Einarsson, S., Josefsson, B., Moller, P. and Sanchez, D., Separation of amino acid enantiomers and chiral amines using precolumn derivatization with (+)-b(9- Fluoreny1)ethyl chloroformate and reversed-phase liquid chromatography And.
Chem., 59 (1987) 1191-1195.
Eliel, E. L., Stereochemisty of Carbon Compouncis, McGraw-Hill, New York, 1962.
Forgacs, E., Use of principal cornponent analysis for the evaluation of the retention behavior of monoamine oxidase infubitory drugs on B-cyclodextrin
column 1. Phnrm. Biomed. Anal., 13 (1995) 525-532.
Fowler. J. S., 2-Methyl-3-butyn-2-01 as an acetylenic precursor in the Mannich reaction. A new synthesis of suicide inactivators of monoamine oxidase 1. Org. Chem, 42 (1977) 2637-2639.
Fowler, C. J., Callingham, B. A., Mantle, T.J. and Tipton, K.F., Monoamine
oxidase A and 8: A usefd concept? Biochemical Phnnnacology, 27 (1978) 97.
Fuller, R. W., Inhibition of monoamine oxidase by N-(phenoxyethyl) cydopropylamine. Correlation of inhibition with Hammett constants and
partition coefficients [. Med. Chem., 11 (1968) 397-398.
Gerlach, M., Riederer, P. and Youdim, M. B. H., The molecular pharmacologv of L-Deprenyl Eilr. J. Phannacol., 226 (1992) 97-108.
Gil-Av, E., Feibush, B. and Charles-Sigler, R., Separation of enantiomers by gas
liquid chromatography with an optically active stationary phase Tetrahedron Letf.,
(1 966) 1009-1015.
Girault-Vexlearschi, G., Influence de la ramifications des chaines hydrocarbonees sur la baside des amines III, Etude de l'equilibre d'ionisation des amines B d l .
Soc. Chim. Fr., 23 (1956) 589-606.
Greenstein, J. P. and Winitz, M., n e Chemistry ofilmino Acids, Wiiey, New York,
1961, pp. 1738.
Gubitz, G., Wintersteiger, R. and Hartinger, A., Fluorescence derivatization of
tertiary amines with 2-naphthyl chloroformate Chromatugr., 218 (1981) 51-56.
Hasegawa. M. and Matsubara, I., Gas chromatographie determination of optical
purities of amino acids using N-Trifluoracetyl menthyl esters Anal. Biochem., 63 (1975) 308-320.
Hutt, A. J. and Caldwell, J., The importance of Stereochemistry in the clinical
pharmacokinetics of the 2-aryl propionic acid non-steroidal anti-uillammatory
drugs Clin. Pharmacokinet., 9 (19%) 371-373.
Johnson, J. P., Some observations upon a new inhibitor of monoamine oxidase in
brain tissue Biochem. Phannacol., 17 (1968) 1285.
Johnson, D. M., Rerter, A., Collins, M. and Thompson, G.E., Enantiomenc purity of Naproxen by liquid chromatographie analysis of its diastereomeric octyl esters
J. Pham. Sci., 68 (1979) 112-111.
Julian, P. L., Oliver, J.J., Kimball, R.H., Pike, A.B. and Jefferson, GD., Methyl
benzyl ketone Org. Syn., 18 (1938) 66-69.
Julian, P. L. and Oliver, J.J., a-Phenylacetoacetonitrile (Acetobenzyl cyanide) Org. Syn., 18 (1938) 66-69.
Kamerling, J. P., Duran, M., Gerwig, G.J., Ketting, D., Bruinvis, L., VLiegenthart,
J.F.G. and Wadman, S.K., Determination of the absolute configuration of some
biologically important 2-hydroxydicarboxylic acids by capillary gas-liquid chromatography \. Chromatogr., 222 (1981) 276-283.
Kapnang, H. and Charles, G., Reaction des chloroformates sur les amines tertaires: cornpetition entre N-demethylation, desamination, N-debenzylation de N-dcsallylation Tetraliedron Lett., 24 (1983) 3233-3236.
Kinemuchi, H., Fowler, C.J. and Tipton, K.F., Snbstrate specz$ïcicities of the two f o m s of monoamine oxidase, Academic Press, London, 1984, pp. 253.
Knoll, J., Vizi, E.S. and Somogyi, G., Phenyiisopropyhnethylpropinylamine (E- 250), a monoamine oxidase inhibitor antagonizing effects of tyramine. Arzneim-
Forsch., 17 (1968) 1285-1297.
Knoll, J. and Magyar, K., Monoamine Oxiriase-New Vistas : Advances in
Biochemical Psycholpharmacology, Vol 5, Raven Press, New York, 1972, pp. 393
Knoll, J., The possible mechankm of action of (-)deprenyl in Parkinson's disease.
J. Neurnl TTansrn., 57 (1978) 33-38.
Knoll, J., Deprenyl (Selegiline): the history of its development and
pharmacological action Acta Neirrol. Sciznd. Suppl. , 95 (1983) 57-80.
Kometani, T., Shunsaku, S. and Mitsuhashi, K., On the cleavage of tertiary
amines with ethyl chloroformate Chem. Pham. Biil l . , 24 (1976) 342-349.
Konig, W. A. and Benecke, L, Gas chromatographic separation of chiral 2- hydroxy acids and Zalkyl substituted carboxylic acids J. Chromatogr., 195 (1980)
292-296.
Krimen, L. 1. and Cota, D.J., The Ritter Reaction Organic Renctions, 17 (1969) 213-
325.
La Croix, R., Pianezzola, E., Benedetti, and Strolin Benedetti, M., Sensitive high- performance liquid chrornatographic method for the determination of the three
main metabolites of Selegiline (L-deprenyl) in human plasma 1. Chromatugr. B, 656 (1994) 251-258.
LePage, L., Linder, W., Davies, G., Seitz, D. and Karger, B., Resolution of the
optical isomers of dansvl amino acids by reversed phase Liquid chromatography
with optically active metal chelate additives Annl. Chem., 51 (1979) 433-435.
Liu, J. H. and Ku,W. W., Detemination of enantiomeric N-Trifluoroacetyl-L-
prolyl diloride amphetamine derivatives by capillary gas chromatographylmass
spectrometry with chiral and achiral stationa. phases Annl. Cliem., 53 (1981)
21 80-2184.
Lough, W. J., Chiral Liqnid Chrornntography, Blackie and Son Ltd., Glasgow, 1989,
pp. 39-80.
Makino, Y., Ohta, S. and Hirobe, M., Enantiomeric separation of amphetamine by
high performance liquid chromatogcaphy using chiral crown ether-coated
reversed-phase packing: application to forensic analysis Forensic Sci. M., 78
(2996) 65-70.
Martin, A. J. P. and Synge, R. L. M., A new form of chromatogram employing
two liquid phases. 1. A theory of chromatography 2. Application to the micro-
determination of the higher monoamino-acids in proteins Biochem [., 35 (1941)
1358-1368.
Meyer, V. R., Prncticnl High-Prrfonnnnce Litpiri Chromatogrnphy, Second Edn, John
Wiley & Sons, Chichester, 1994, pp. 17-32.123-157-
Miller, K. R., Gal, J. and Ames, M. M., High Performance liquid diromatographic
resolution of enantiomers of 1-phenyl-2-aminopropanes (amphetamines) with
four dural reagents 1. Chromatogr., 307 (1984) 335-342.
Nagai, T. and Kamiyama, S., Assav of the optical isomers of methamphetamine
and amphetamine in rat urine using high-performance Liquid chromatography
with chiral cellulose-based columns 1. Chromatogr., 525 (1990) 203-209.
Noggle, F. T. J., DeRuiter, J. and Clark, C.R., Liquid chromatographie determination of the enantiomeric composition of methamphetamine prepared
from ephedrine and pseudoephedrine And. Cliem., 58 (1986) 1643-1648.
OLofson, R. A., Sdinur, R.C., Bunes, L. and Pepe, J.P., Selective N-dealkylation of
tertiary amines with vinyl chloroformate: An improved synthesis of naloxone Tetrnhedron Lett., (1977) 1567-1570.
Pasteur, L., The Fortncilrtions of Stereo Chernistry: Memoirs by Pasteur, Van 't Hoff, Le Bel and Wislicenw, Amencan Book Company, New York, 1901, pp. 1-33.
Pasutto, F. M., Mirror Images: The analysis of pharmaceutical enantiomers 1. Clin. Pharmacol., 32 (1992) 927-924.
Perrin, D. D., Physical Chernical Properties of Dnigs, Marcel Dekker, New York.
1980, pp. 1-48.
Pettersson, C. and Schill, G., Separation of enantiomeric amines by Ion- pair chrornatography 1. Chromatogr., 204 (1981) 179-183.
Pettersson, C. and No, K., Chiral resolution of carboxyiic and sulphonic acids by ion pair chromatography 1. Chromntogr., 282 (1983) 671-684.
Pettersson, C. and Josefsson, M., Chiral separation of amino alcohols by ion-pair chromatography Chromntopphin, 21 (1986) 321-326.
Pettersson, C., Liquid diromatographic separation of enantiomers using chiral
additives in the mobile phase Trends A d . Chem., 7 (1988) 209-217.
Pharmacopeial Convention, U. S., Sekgdine Hydrochloride- In process revision Phnmacopeial Forum, 20 (1994) 714.1-7150.
Pharmeuropa, Alternative method proposed for the defection of IS) Selegdine, 1996, pp. 24.
Pirkle, W. H., Pochapsky, T. C., Mahler, G. S., Corey, D. E., Reno, D. S. and
Alessi, D. M., Useful and easily prepared &al stationary phases for the direct
chrornatographic separation of the enantiomers of a variety of derivatized amines, arnino acids, alcohols and related compounds 1. Org. Chem., 51 (1986)
4991-5000.
Ramey, O. B., SLereochemist y, Heyden & Son Ltd., Philadelphia, 1981, pp. 69-97.
Ritter, R. J. and Minieri, P.P., A new reaction of nitriles 1; Amides from alkenes
and mononitdes J. Amer. Chem. Soc., 70 (1948) 4045-4048.
Ritter, J. J. and Kalish, J., a,a - Dirnethyl-P-phenylamine (Phenethylamine a, a -dimethyl) Org. Syn., 44 (1964) 44.
Robinson, J. B., Stereoselectivity and isoenzyme selectivity of monoamine oxidase inhibitors Biochern. Phamzacol., 34 (1985) 4105-4108.
Salonen, J. S., Determination of the amine metabolites of Selegiline in biological
fluids by capillary gas diromatographv J. Chromatogr.-,527 (1990) 163-168.
Simon, K., Podanyi, B., Ecsery, 2. and Torok, Z., Absolute configuration and
conformational analpis of (-)-(R)-Deprenyl and its homologues 1. Chem. Soc. Perkin Trtrns. 11, (1986) 11 1-1 15.
Simon, K., Bocskei, Z. and Torok, Z., X-ray diffraction study of Selegiline and
related compounds Acta. Pham. Hung., 62 (1992) 225-230.
Simons, S. S. and Johnson, D.F., The structure of the fluorescent adduct formed
in the reaction of O-phthalaldehyde and thiols with amines 1. Am. Chem. Soc., 98
(1976) 7098-7099.
Skidmore, M. W ., Hondbook of Derimtives for Chrornntography, Second Edn, John Wiley & Sons, Chichester, 1993, pp. 215-249.
Skoog, D. A. and Leary, J. J., Principles of lnstnimental Annlysis, Fourth Edition
Edn, Saunders CoLIege Publishing, Philadelphia, 1992, pp. 579-680.
Tipton, K. F., Fowler, C.J. and Houslay, M. D., Monoamine oxidase: Basic and
clinicalfrontiers, Excerpta Medica, Amsterdam, 1982, pp. 87.
Ullrich, K. J., Rurnrich, G., Neiteler, K., and Fritzsch, G., Contraluminal transport
of organic cation in the proximal tubule of the rat kidney Pflugers Arch., 420 (1992) 29-38.
Wainer, 1. W., Drug S tereochem ist y, Analyt ical Methods and Pharmacology, Second
Edn, Marcel Dekker, New York, 1993, pp. 65-106,139-182.
Wainer, 1. W., High Performance Liqitid Chrornatogtaphy: Fundarnerztal Pnnciples nnd
Pracf ice, John Wiley & Som, Toronto, 1994.
Witte, D. K., de Zeeuw, R-A. and Drenth, B. F. H., Chiral derivatization of
Promethazine with (-)menthyl chloroformate for enantiomeric separation by RP- HPLC 1. High Resolut. Chrornntogr.: Chromatogr. Commun., 13 (1990) 569-571.
Pre~aration of a- Phenylacetoacetonitrile - <Julian, 1938 #58>
h t o a 2 liter round bottomed flask fitted with a separatory funne1 and
reflux condenser, each protected with a CaC12 drying tube, was placed absolute
ethanol (700 ml) and clean sodium metal (70g, 3 moles) added in srnail pieces
at such a rate as to conhol the refuing of the solution. To the warm solution
of sodium ethoxide was then added slowly with stirring a solution of benzyl
cyanide (234g ; 2 moles) in dry ethyl acetate (264g : 293 mis ; 3 moles). The
mixture was refluxed on a steam bath for two hours and then allowed to
stand at room temperature ovemight. The mixture was diluted with water (2
L) and stirred to dissolve the solid. Crushed ice (IL) was added, the mixture
extracted with ether (3 x 500 ml) and the ether extracts discarded. The
separated aqueous phase was freed of any residual ether by bubbling air
through the solution during 30 minutes and the solution was then acidified
with a solution of glacial acetic acid (150m.l) in water (400ml). The precipitate
was filtered, washed with water and the crude product (a-
phenylacetoacetonitrile) used without hr ther drying or purification for the
preparation of benzyl methyl ketone.
Pre~aration of Benzyl Meth 1 Ketone cJulian, 1938 #59>
Concentrated sulfuric acid (350 ml) was added to a 3 liter round
bottorned flask and cooled to -100 C. The cmde, moist a-
phenylacetoacetonitrile was added slow ly to the cooled sulfuric acid with
shaking and maintaining the reaction mixture temperature below 200 C.
When all the a-phenylacetoacetonitrile was added, the flask was warmed on
a steam bath until the solution was complete and the then the flask was
cooled to 00 C. Water (1750 ml) was added rapidly and the flask heated on a
steam bath for 48 hours during which time the ketone layer separated (ketone
layer contains much unreac ted a-phenylace toacetonitrile , which is difficult
to decarboxylate without adequate stirdtg of the reaction mixture). The
ketone layer was separated and the acidic aqueous phase extracted with ether
(500 ml). The combined oil and ether extract was dried over sodium sulfate,
filtered and the solvent evaporated. [The residue tends to deposit crystals of
a-phenylacetoacetonitrile which were filtered out and recycled through the
decarboxylation reaction]. The residue was distilled the fraction b.p. 100-105°
C at 12 mm being collected. The product was redistilled at atmospheric
pressure, the material b.p. 214-2160 C being collected (literature b.p. 2160 C).
KOH
h t o a round bottomed flask was placed methylamine hydrochloride
(12.58 g , 0.1863 moles), potassium hydroxide (6.7 g, 0.1 194 moles) and
benzylrnethyl ketone (5 g, 0.0373 moles) dissolved in methanol (50ml). The
mixture waç stirred for one hour at room temperature. Sodium
cyanoborohydride (1.5 g, 0.0227 moles) was added and the mixture stirred for
twenty-four hours. The reaction mixture was evaporated under reduced
pressure , water (30 ml) added, the solution acidified with hydrodoric acid,
extracted with diloroform (3 x 30ml) and the chloroform extracts discarded.
The aqueous solution was made alkaline with ammonium hydroxide and
extracted with chloroform (3 x 30ml). The chloroform extracts were collected,
dried (Molecular Sieve; Type 4A) filtered and evaporated under reduced
pressure. The residue was distilled using a bulb distillation apparatus, the
fraction b.p. 1000 C at 10 mm being collected.
i.r. (liquid film) 3320 cm-' (NH); 1600 un-' (aromatic); 1345 cm-', 1370 cm-'
(C-N); 700 cm-1 ,750 cm-' (mono substituted benzene)
(f) Methamphetamine hydrochloride (recrystallized from acetone)
m.p. 134.5 C [Lit m.p. 135-137 Cl
In a round bottom flask, methanol (25 ml), propargylamine
hydrochloride (5.5 g , 0.06 moles), potassium hydroxide (2.5 g , 0.0445 moles)
and benzyl methyl ketone (2 g, 0.015 moles) were added and auowed to stir at
room temperature for one hour. Sodium cyanoborohydride ( 0.6 g, 0.0095
moles) was added and the mixture stirred for twenty-four hours. The mixture
was evaporated under reduced pressure and water (30 ml) added. The
solution was acidified with hydrochloric acid and extracted with ether ( 3 x 30
ml). The extracts were discarded and the aqueous solution made alkaline
with ammonium hydroxide and extracted with chloroform (3 x 30 ml). The
chloroform extracts were collected and dried (Moleculas Sieve; Type 4A),
filtered and evaporated under reduced pressure. The residue was fractionally
distilled, the fraction b.p. 48-500 C and 0.1 mm being collected. Yield 1.3 g
i.r. (liquid film) 3300 cm- (c eC-H), 1600 an-' (aromatic); 700 cm-1 ,753 cm-1
(mono subs tituted benzene)
To a solution of N-methylpropargylamine hydrochioride (4.8 g, 0.045
moles) in anhydrous methanol (12 ml) was added potassium hydroxide
(1.683 g. 0.03 moles). After dissolving the potassium hydroxide, benzyl rnethyl
ketone (3 g, 0.022 moles) was added and the mixture stirred for one hour at
room temperature. Sodium cyanoborohydride (0.44 g, 0.007moles) was added
and the mixture stirred for twenty-four hours. The solution was evaporated
to dryness under reduced pressure. Water (60 mls)was added and the solution
acidified with hydrochloric acid. The mixture was extracied with ether (3 x 30
ml), and the extracts discarded. The aqueous solution was made alkaline with
ammonium hydroxide and extracted with chloroform (3 x 30 ml). The
extracts were dried (Molecular Sieve,Type M) filtered and evaporated. The
residue was fractionally distilled (bulb distillation), the fraction b.p. 58-600 C
at O.lmm being collected. Yield 0.6 g
i.r. (liquid film) 3300 cm- l (C C- H), 2100 cm- 1 ( C C), 1600 cm- 1,1595 cm- 1
(aromatic); 705 cm-' ,745 cm-' (mono substituted benzene)
Preoaration of the Menthyl carbarnate of ( f l Metharn~hetarnine
[Diastereorneric Methodl
A CH, CH3
Into a round bottomed flask was placed potassium carbonate (1.5 g), (k)
methamphetamine (1 g, 0.0067 moles) in benzene (10 ml) and (-) menthyl
chloroformate (3 ml: 3.06g: 0.01399 moles). The flask was stoppered and the
reaction mixture stirred at room temperature for 4 days. Methanolic
potassium hydroxide (25 ml, saturated solution) was added, the reaction
mixture stirred ( 2 hours) and water (30 ml) added. The benzene layer was
separated, washed with water and evaporated under reduced pressure. The
residual liquid was placed in a sealed container. The infra-red spectum (liquid
film) showed absorption peaks indicative of a carbamate, 1750 cm-' (C=O),
1260 cm-' (C-O) .
Pre~aration of the Menthyl carbamate of (f) Desmethvl Seleeiline
Into a round bottomed fiask was placed potassium carbonate (1.5 g), (+)
methamphetamine (1 g, 0.00578 moles) in benzene (10 ml) and (-) menthyl
chloroformate (2.5 ml: 2.55 g: 0.01166 moles). The flask was stoppered and the
reaction mixture stirred at room temperature for 4 days. Methanolic
potassium hydroxide (25 ml, saturated solution) was added, the reaction
mixture stirred ( 2 hours) and water (30 ml) added. The benzene layer was
separated, washed with water and evaporated under reduced pressure. The
residual liquid was placed in a sealed container. The mfra-red spectum (liquid
film) showed absorption peaks indicative of a carbamate and an acetylenic
group.
i.r. (liquid film), 3300 cm-' (CmCH), 1745 un-' (GO), and 1260 cm-1 ( C-O).
Pre~aration of N-Acetvl-2-Phenvlisooro~vlamin~
(slight modification of the method of Ritter et al <Ritter, 1948 #6b)
(0.2 mole; 23.64g ; 26ml) and acetonitrile (0.3 mole;
12.2g ; 15.61ml) were mixed and added to a solution of p-toluene sulphonic
acid monohydrate (0.2 mole; 38g) dissolved in glacial acetic acid (100ml). The
mixture was allowed to stand at room temperature with stirring for 72 hours
and then poured onto cmshed ice (400g). The mixture was made alkaline
with ammonia and extracted with ether (3 x 100ml), the extracts dried over
potassium carbonate, filtered and the solvent evaporated. The residue was
fractionally distilled and the following fraction collected.
i) b.p. 113 -118 at 0.4 mm 4.88g
ii) b-p. 125 -131 at 0.4 mm 2.82g
iii) b-p. 131 -140 at 0.4 mm
Only fractions i) and ii) showed infrared spectra characteristic of amides.
Preparation of 2-Amino-2-ohenvl~ro~ane ( a-a-Dimethvlbenzvlamine)
N-Acetyl-2-phenylisopropy1amine (7.6g) was dissolved in 95% ethanol
(IOml) and added to a solution to alcoholic KOH (50rnl; 20%) and the solution
reflwed during 96 hours. Water (2001111) was added and the solution steam
distiiled, the distillate being collected in aqueous hydrochloric acid (volume
of distillate approxima tely 3OOml). The distillate was evapora ted to dryness
under reduced pressure, the residue dissolved in the minimum amount of
water, made alkaline with ammonia and extracted with ether
(3 x 50ml). The extracts were dried (molecular sieve), filtered and the solvent
evaporated. The residue was distilled (Kugelrohr), the material b.p. 750 C at
8mrn being collected. Yield 85 mgm.
Pre~aration of N-Formyl-2-~henvliso~ropvlamine
(Modification of Ritter method <Ritter, 1964 #62>
Into a 500ml three necked round bottomed flask fitted with a stirrer,
thermometer, dropping funnel and reflux condenser comected to a trap
containing 20% sodium hydroxide solution was added glacial acetic acid (125
mls). The material was cooled in an ice bath to 20 C and potassium cyanide
(0.5 mole; 32.6g) added slowly while maintainhg the reaction mixture below
200 C. From the separatory funnel was added a previously cooled mixture of
concentrated sulhric acid (68ml ; 1.25 mole) in glacial acetic acid (62.5m1). The
addition was made slowly with stirring to keep the mixture below 200 C.
The ice bath was removed and a-methylstyrene (0.5 mole; 59g; 65ml)
added during 20 minutes while the temperature slowly rose to approximately
450 C. The mixture, which was quite viscous, was stirred at room
temperature overnight. Air was bubbled through the mixture until free of
hydrogen cyanide and the reaction mixture then poured into ice-water (750 g),
the mixture neutralized with sodium carbonate and extracted with ether (3 x
100ml). The extracts were dried over sodium sulfate, filtered, the solvent
evaporated and the residue distiiled. The material had a b.p. of 100-1100 C at
0.25mm and was a colorless viscous liquid.
The infra-red spectrum showed absorption peaks indicative of an
amide and was almost completely superimposable upon a spectrum of N-
acetyl-2-phenylisopropylamine. Yield 31g
Pre~aration of a-a-Dimethvlbenzvlamine
N-Fomyl-2-phenylisopropylarnine (30g) was dissolved in absolute
ethanol (150ml), potassium hydroxide pellets (30g) added and the solution
refluxed during 40 hours. The mixture was diluted with water (400ml) and
steam distilled until approximately 750ml of distillate had been collected. The
acidified distillate was evaporated under reduced pressure and the solid
residue dissolved in water (30 ml), the solution made alkaline with ammonia
and extracted with ether (3 x 50 ml). The extracts were dried (molecular sieve),
filtered and evaporated to leave an oily residue which was distilled, the
fraction b.p. 67-790 C at 2.5mrn being collected. Further distillations of the
material showed b-p. 77-800 C at 8.5 mm. Yield 6.lg (57% based on formyl
derivative consumed. [Lit b.p. 196-1970 C at 980 C at llmrn; 1000 C at 22mm,
Dictionary of Organic Compounds, Heilbron and Brunbury]
[N.B. Unchanged amide (17g) was recovered by ether extraction of the steam
distillation residues]
IMAW LVALUATION TEST TARGET (QA-3)
k g 12 II- = 1111pL . 12 - L ,,A IlIll&
APPLIEO - IMAGE. lnc 1653 East Main Street - -. - - Rochester. NY 14609 USA -- -- - - Phone: i l 6/482-O3ûû -- -- - - FU: i l 6/288-5989
O 1993. Applied Image. Inc.. Ail Rights Resewed