You are on page 1of 10

Journal of Chromatography A, 1195 (2008) 117126

Contents lists available at ScienceDirect

Journal of Chromatography A
journal homepage: www.elsevier.com/locate/chroma

Enantiomer identication in the avour and fragrance elds by interactive


combination of linear retention indices from enantioselective gas
chromatography and mass spectrometry
Erica Liberto a , Cecilia Cagliero a , Barbara Sgorbini a , Carlo Bicchi a , Danilo Sciarrone b ,
Barbara DAcampora Zellner b , Luigi Mondello c , Patrizia Rubiolo a,
a

Dipartimento di Scienza e Tecnologia del Farmaco, Facolt`


a di Farmacia, Universit`
a degli Studi di Torino, Via Pietro Giuria 9, Turin 10125, Italy
Dipartimento Farmaco-Chimico, Facolt`
a di Farmacia, Universit`
a degli Studi di Messina, Viale Annunziata, Messina 98168, Italy
c
Campus-Biomedico, Via E. Longoni 47, Rome 00155, Italy
b

a r t i c l e

i n f o

Article history:
Received 19 February 2008
Received in revised form 16 April 2008
Accepted 18 April 2008
Available online 24 April 2008
Keywords:
Enantiomer identication
Mass spectral library
Enantioselective GCMS
Cyclodextrin
Linear retention indices
Mass spectra
Flavours
Fragrances

a b s t r a c t
This study describes the development of a gas chromatographymass spectrometry (GCMS) library to
identify optically active compounds in the avour and fragrance eld using enantioselective GC with
cyclodextrin derivatives (CDs) as chiral selectors in combination with MS. The library operates on the
interactive combination of linear retention indices (IT values) in parallel to MS spectra, so as to identify enantiomers reliably and to measure EE and/or ER unequivocally. Since MS is not a selective probe
to discriminate optical isomers, mass spectra (or diagnostic ions in SIM mode) are used to locate the
enantiomer(s) in the chromatogram, and IT values to identify it(them) safely and reliably in particular in
complex mixtures. The library has been built up through the following steps:
(a) Selection of CD derivatives able to cover a wide range of racemate separations. Four
cyclodextrin derivatives were used: 2,6-di-O-methyl-3-O-pentyl--CD, 2,3-di-O-methyl-6-O-tertbutyldimethylsilyl--CD, 2,3-di-O-ethyl-6-O-tert-butyldimethylsilyl--CD, and 2,3-di-O-acetyl-6-Otert-butyldimethylsilyl--CD.
(b) Determination of the analytes IT values and evaluation of their stability and reliability at both intraand inter-laboratory level.
(c) Determination of the range within which the IT of an enantiomer has to fall to be correctly identied,
i.e. determination of a common retention index allowance (RIA).
(d) Construction of the library, at the moment comprising the enantiomers of 134 racemates. A record has
been attributed to each enantiomer including IT values determined on the four CD coated columns,
mass spectrum, IUPAC chemical name, CAS number, molecular weight, and, when separated, racemate
enantiomer resolution on the CD investigated.
Some applications of the library are also reported.
2008 Elsevier B.V. All rights reserved.

1. Introduction
The interaction of a compound with a biological system has
long been shown to be stereoselective. Enantiomer recognition and
enantiomeric excess (EE) and/or ratio (ER) determination are a very
important task in avour and fragrance elds, so as (i) to dene
the correlation between chemical composition and organoleptic
properties; (ii) to implement quality control and detect fraud or
adulteration of natural samples; (iii) to determine the biosynthetic pathway when the formation of a compound is studied or

Corresponding author. Fax: +39 011 2367661.


E-mail address: patrizia.rubiolo@unito.it (P. Rubiolo).
0021-9673/$ see front matter 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.chroma.2008.04.045

to classify a sample; (iv) to determine the geographic origin of a


natural sample.
Cyclodextrin derivatives (CDs) have truly represented a milestone in enantioselective gas chromatography (GC); being rst
introduced by Sybilska and Koscielski at the University of Warsaw
in 1983 for packed columns [1] and applied to capillary columns
in the almost contemporary works of Juvancz et al. [2] and Schurig
and Nowotny [3]. Moreover, Schurig and co-workers rst proposed diluting CD derivatives in moderately polar polysiloxane
(OV-1701) to provide them with good chromatographic properties
and a wider range of operative temperatures [4]. Since then,
several groups have investigated CD derivatives as chiral selectors
for enantioselective GC applications, and several hundreds of
articles have been published dealing with the theory of chiral

118

E. Liberto et al. / J. Chromatogr. A 1195 (2008) 117126

GC recognition with CDs, synthesis of new CD derivatives, their


enantioselectivity and applications, many of them concerning the
avour and fragrance eld [58].
Chiral recognition of marker compounds in complex real-world
samples, as those in the avour and fragrance eld often are,
generally requires a two-dimensional approach, because enantioselective GC may double the number of peaks of optically active
analytes. This makes some parts of the total chromatograms even
more complex and, as a consequence, increases the probability
of interferences with a correct EE and/or ER determination. Two
complementary but distinct approaches can therefore be adopted;
the rst and most popular one is to introduce a second dimension in separation. Chiral recognition is here generally carried out
by conventional heart-cut GCGC [912], where the rst column,
coated with a conventional phase, serves to locate the peak(s) of the
optically active racemate(s), and the second column, coated with
a CD stationary phase, separates its(their) enantiomers after online transfer through the heart-cut interface. When the number of
components to be investigated is very high, comprehensive twodimensional gas chromatography (GC GC) can also be used but,
in this case, column geometry must be inverted because of the
high efciency required by the columns coated with CDs in order
to give reliable separations [13,14], the rst column has to be coated
with the enantioselective CD stationary phase while the second one
distributes the peaks over the chromatographic plane [15,16]. The
second approach involves the use of a second dimension in identication. In this case, the enantiomer is located and identied by MS
detection (or very rarely FT-IR). Single- or multiple-ion monitoringMS (SIM-MS) carried out after a careful choice of suitable diagnostic
ions of the optically active marker(s) of the sample under investigation can be applied to clean the part of the chromatogram
where the enantiomers elute, thus making correct EE and/or ER
determination possible.
In general, most GCMS software takes insufcient account of
the identication potential of GC, because the identication power
of mass spectrometry when used as detector for GC is considered to
be, and very often is, exhaustive. Retention indices (Is ) are the most
reliable and effective tool for analyte identication by GC data. They
were rst introduced by Kovats for isothermal analysis [17] and
then by Van den Dool and Kratz for temperature programmed analysis [18], the latter being better known as linear retention indices
(IT values). Most GCMS software packages do not include IT values
as identication criterion, and only some of them report IT values in
the library as blind or inactive data appearing in the legend of each
proposed identication record, making them only useful for further
or additional conrmation. On the contrary, the interactive use of
IT values (i.e. their use as an active identication parameter) can
be highly effective since it provides a second independent tool
to identify a compound, operating actively and simultaneously in
parallel to MS spectra. Moreover, IT values are based on a chemical
property of an analyte of a completely different nature compared
to MS, which can orthogonally and synergically be combined with
its MS fragmentation pattern, i.e. its chromatographic interaction
with a given chromatographic separation system or better with a
given stationary phase.
Mass spectrometry is well known to be unable to discriminate
between optical isomers, not being a selective chiral probe in this
sense, and therefore giving indistinguishable spectra. As a consequence it cannot be used alone to determine which enantiomer is
present in a sample, or to establish the predominant one or to measure its EE and/or ER. In enantioselective GCMS, a given optical
isomer can only unequivocally be identied through its IT obtained
with a column coated with a chiral selector suitable to separate
it from its enantiomer. In the chiral recognition of optically active
isomers in a complex mixture, the two identication parameters

(i.e. IT values and MS spectrum) must therefore be combined but,


unlike conventional GCMS analysis, mass spectra (or diagnostic
ion monitoring) are used to locate the two enantiomers in the
chromatogram, and IT values for their identication.
This study deals with the development of a MS library specic
for the identication of optically active compounds in the avour
and fragrance eld using interactive IT values in parallel to MS
spectra, so as to identify enantiomers reliably and, when necessary,
to enable the measurement of EE and/or ER unequivocally.
2. Experimental
2.1. Racemate standards and essential oils
Analyses of 134 racemate standards and pure enantiomers solubilised in cyclohexane at a concentration of 100 ppm were carried
out. The enantiomeric recognition of the marker compounds characteristic of commercially available essential oils (e.o.) and extracts
was also carried out, in particular balm lemon, bergamot, boronia, cornmint, lavender, lemon, peppermint, and rosemary e.o.s.;
apple avour and apricot, peach and coconut headspace sampled
by solid-phase microextraction (SPME) were analysed.
2.2. Enantioselective GC columns
The library has been built on the basis of IT values obtained on
four columns (25 m 0.25 mm I.D., df : 0.25 m) coated with four
cyclodextrin derivatives as chiral selectors diluted at 30% in PS-086:
2,6-di-O-methyl-3-O-pentyl--CD
(2,6DM3PEN--CD) [19,20]
2,3-di-O-methyl-6-O-tert-butyldimethylsilyl--CD
(2,3DM6TBDMS--CD) [21]
2,3-di-O-ethyl-6-O-tert-butyldimethylsilyl--CD
(2,3DE6TBDMS--CD) [22]
2,3-di-O-acetyl-6-O-tert-butyldimethylsilyl--CD
(2,3DA6TBDMS--CD) [21]
All columns were from MEGA (Legnano, Italy). Their performance were periodically tested through the Grob test [23,24]
and a laboratory-made chiral test [25] consisting of limonene,
2-octanol, camphor, isobornyl acetate, linalyl acetate, 2-methyl(3Z)-hexenyl butyrate, menthol, hydroxycitronellal, -decalactone
and -decalactone racemates.
2.3. Enantioselective GCMS conditions
A Shimadzu QP2010 GCMS system was used and results were
elaborated with the Shimadzu GCMS Solution 2.51 software (Shimadzu, Milan, Italy).
GC conditions: injection mode, split; split ratio, 1:20 for standard solutions, 1:50 for essential oils; injection volume, 1 l.
Temperatures: injection, 220 C; transfer line, 230 C; ion source:
200 C; temperature programme: from 50 to 220 C at 2 C/min (if
not specied otherwise). Carrier gas: He; ow rate: 1.0 ml/min.
2.4. Library setting-up
The library was created on the basis of the analysis of the
134 racemate standards and pure enantiomers, on each column,
recording their mass spectra, calculated IT values and enantiomer
resolution (R). The IT determination was carried out by injecting
an homologous series of n-alkanes containing 17 n-hydrocarbons
(C9 C25 ) purchased from Supelco (Bellefonte, PA, USA), each at

E. Liberto et al. / J. Chromatogr. A 1195 (2008) 117126

100 ppm in hexane. The enantiomer stereochemistry was conrmed either through authentic samples or by combining literature
data and analysis of essential oils or fruit avour headspaces or
extracts.
3. Results and discussion
3.1. Basic approach
As already mentioned, the safest approach to identify unequivocally a given optical isomer by enantioselective GCMS, in
particular in complex mixtures, is to combine two identication
parameters (i.e. IT values and MS spectrum), one of them suitable to
distinguish between enantiomers (IT values). Unlike conventional

119

GCMS analysis, mass spectra (or monitoring by diagnostic ions


(SIM)) are used to locate the two enantiomers in the chromatogram,
and IT values for their identication.
The present study was divided into two main steps: the rst
one concerns the building up of the library using the approach
described above and the evaluation of its reliability, and the second one involves its application to its everyday use in a routine
laboratory.
3.2. Development of the library
The library has been created through the following steps
(a) choice of a set of chiral selectors (CD derivatives) able
to cover a wide range of racemate separations, (b) analyte IT

Fig. 1. Chiral test proles carried out on the four columns investigated. (1) Limonene, (2) 2-octanol, (3) camphor, (4) isobornyl acetate, (5) linalyl acetate, (6) 2-methyl-(3Z)hexenyl butyrate, (7) menthol, (8) hydroxycitronellal, (9) -decalactone, (10) -dodecalactone; (a) (R) enantiomer, (b) (S) enantiomer, x and y: enantiomer conguration not
assigned.

120

E. Liberto et al. / J. Chromatogr. A 1195 (2008) 117126

Table 1
IT variation of three marker analytes as a consequence of different injection time intervals of the hydrocarbon standard mixture
Hydrocarbon Injection frequency: every week

Hydrocarbon Injection frequency:


every 3 days

Hydrocarbon Injection frequency: every ve


analyses

Anal. A1 Anal. A2 Anal. A3 Anal. A4 Anal. A5 IT

Anal. B1 Anal. B2 Anal. B3 IT

Anal. C1 Anal. C2 Anal. C3 Anal. C4 Anal. C5 IT

2,6DM3PEN--CD
(S)-()-Limonene
(R)-(+)-Limonene
(1R, 2S, 5R)-()-Menthol
(1S, 2R, 5S)-(+)-Menthol
(R)--Decalactone
(S)--Decalactone

1061
1068
1350
1352
1610
1616

1061
1068
1350
1352
1610
1615

1060
1067
1348
1351
1608
1614

1060
1067
1348
1350
1608
1613

1059
1067
1348
1350
1607
1613

2
1
2
2
3
3

1061
1068
1350
1352
1610
1616

1061
1068
1350
1352
1610
1615

1060
1067
1348
1351
1608
1614

1
1
2
1
2
2

1061
1068
1350
1352
1610
1616

1061
1068
1350
1352
1610
1616

1061
1068
1350
1352
1610
1616

1061
1068
1349
1352
1610
1615

1061
1068
1349
1352
1610
1615

0
0
1
0
0
1

2, 3DM6TBDMS--CD
(S)-()-Limonene
(R)-(+)-Limonene
Menthol
(R)--Decalactone
(S)--Decalactone

1082
1097
1371
1635
1645

1081
1096
1370
1633
1643

1081
1096
1370
1633
1644

1081
1096
1370
1633
1644

1081
1096
1370
1633
1643

1
1
1
2
2

1082
1097
1371
1635
1645

1081
1096
1370
1633
1643

1081
1096
1370
1633
1644

1
1
1
2
2

1082
1097
1371
1635
1645

1082
1097
1371
1634
1645

1082
1097
1371
1635
1645

1082
1097
1371
1634
1645

1082
1097
1371
1635
1645

0
0
0
1
0

2,3DE6TBDMS--CD
(S)-()-Limonene
(R)-(+)-Limonene
(1R, 2S, 5R)-()-Menthol
(1S, 2R, 5S)-(+)-Menthol
(R)--Decalactone
(S)--Decalactone

1056
1072
1282
1285
1573
1588

1056
1072
1282
1285
1573
1588

1057
1073
1283
1285
1574
1588

1056
1072
1282
1285
1573
1588

1056
1072
1282
1284
1573
1587

1
1
0
1
1
1

1056
1072
1282
1285
1573
1588

1056
1072
1282
1285
1573
1588

1057
1073
1283
1285
1574
1588

1
1
0
0
1
0

1056
1072
1282
1285
1573
1588

1056
1073
1283
1285
1574
1588

1056
1073
1282
1285
1573
1588

1056
1073
1282
1285
1573
1588

1056
1072
1282
1284
1573
1588

0
1
1
1
1
0

2,3DA6TBDMS--CD
Limonene
(1R, 2S, 5R)-()-Menthol
(1S, 2R, 5S)-(+)-Menthol
(R)--Decalactone
(S)--Decalactone

1053
1383
1393
1809
1819

1053
1384
1394
1809
1819

1051
1382
1392
1807
1817

1051
1081
1091
1807
1816

1051
1081
1091
1806
1816

2
3
3
3
3

1053
1383
1393
1809
1819

1053
1384
1394
1809
1819

1051
1382
1392
1807
1817

2
2
2
2
2

1053
1383
1393
1809
1819

1053
1384
1394
1809
1819

1053
1384
1394
1810
1820

1053
1383
1393
1810
1819

1053
1083
1093
1810
1820

0
1
1
1
1

Section A and B: IsT calculated every consecutive days on hydrocarbons injected on the day A1/B1. Section C: IsT calculated every consecutive days on hydrocarbons injected
each day every ve analysis.

determination and evaluation of their reliability, and (c) denition of a correct procedure to select a suitable retention index
allowance (RIA) that included determination of the optimal injection amount and measurement of the average total analyte tailing
factor.
All results and considerations reported in the present article are
based on data resulting from enantioselective GCMS analyses of
134 racemates whose IT values were determined on four columns
coated with different CD chiral selectors (see below).

3.2.1. Selection of the chiral selector set


Four cyclodextrin derivatives were chosen as chiral selectors to
build up this library (see Section 2.2). The choice of a relatively large
number of chiral selectors (columns) was mainly due to the fact that
a derivatised CD with a universal enantioselectivity has not yet been
found, therefore a number of derivatives suitable to cover most of
the usual separation in the avour and fragrance eld had to be
used. This lack is due to the intrinsic mechanism of chiral recognition with CD derivatives in gas chromatography that is based on a

Table 2
Average IT s and IT s of chiral test components carried out over three weeks, on a period of 2 months randomly repeated for three couples of 2 consecutive days for each CD
column
2,6DM3PEN--CD

2,3DM6TBDMS--CD

2,3DE6TBDMS--CD

2,3DA6TBDMS--CD

Week 1 Week 2 Week 3 IT Week 1 Week 2 Week 3 IT Week 1 Week 2 Week 3 IT Week 1 Week 2 Week 3 IT
(S)-()-Limonene
(R)-(+)-Limonene
(S)-2-Octanol
(R)-2-Octanol
(S)-()-Camphor
(R)-(+)-Camphor
(X)-Isobornyl acetate
(Y)-Isobornyl acetate
(R)-()-Linalyl acetate
(S)-(+)-Linalyl acetate
(X)-2-Methyl-(3Z)-hexenyl butyrate
(Y)-2-Methyl-(3Z)-hexenyl butyrate
(1R, 2S, 5R)-()-Menthol
(1S, 2R, 5S)-(+)-Menthol
(X)-Hydroxycitronellal
(Y)-Hydroxycitronellal
(R)--Decalactone
(S)--Decalactone
(S)--Decalactone
(R)--Decalactone

1061
1068

1061
1068

1061
1067

0
1

1147

1146

1146

1
0

1184

1184

1184

1269
1273

1269
1273

1269
1273

0
0

1240

1240

1240

1243
1245
1350
1352

1243
1245
1349
1352

1243
1245
1349
1352

0
0
1
0

1438

1438

1438

1610
1616

1610
1616

1610
1616

0
0

1637

1636

1636

1082
1097
1128
1130
1193
1199
1257
1261
1254
1256
1246
1249

1081
1096
1128
1130
1193
1198
1256
1260
1254
1255
1245
1249

1082
1095
1128
1130
1193
1199
1256
1260
1254
1256
1245
1249

1
2
0
0
0
1
1
1
0
0
1
0

1371

1370

1371

1447
1450
1635
1646
1641
1646

1446
1449
1634
1645
1640
1645

1447
1450
1634
1645
1641
1645

1
1
1
1
1
1

Analyses were repeated three times each day for a total of 18 determinations for each column.

1056
1072
1111
1112
1133
1141
1220
1222
1231
1237
1240
1244
1282
1285
1373
1374
1573
1588
1586
1588

1056
1073
1111
1113
1133
1141
1220
1222
1231
1237
1240
1244
1282
1285
1373
1374
1573
1588
1586
1588

1056
1071
1110
1112
1133
1141
1220
1223
1231
1237
1240
1244
1282
1285
1373
1374
1573
1588
1586
1588

0
1
1
1
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0

1052

1053

1053

1236

1237

1235

1241
1258

1242
1259

1242
1258

1
1

1321

1322

1321

1301
1303
1296
1298
1383
1393
1646
1651
1809
1819
1866
1875

1303
1304
1297
1299
1383
1394
1647
1652
1810
1820
1867
1876

1301
1303
1296
1298
1382
1392
1645
1651
1810
1820
1866
1875

2
1
1
1
1
2
2
1
1
1
1
1

E. Liberto et al. / J. Chromatogr. A 1195 (2008) 117126

121

Table 3
RIA window values expressed as average start-apex and apex-stop IT variation of the groups of racemate analysed with each CD column
Class

2,6DM3PEN--CD

2,6DM6TBDMS--CD

2,6DE6TBDMS--CD

2,6DA6TBDMS--CD

Start-apex

Apex-stop

Start-apex

Apex-stop

Start-apex

Apex-stop

Start-apex

Apex-stop

Esters
Alcohols
Ketones
Acids
Hydrocarbons
Aldehydes
Lactones

1.6
1.6
1.6
1.5
1.1
2.0
1.8

1.4
1.8
1.7
2.4
1.8
2.0
2.0

1.3
1.3
1.5
1.1
1.5
1.0
1.5

1.8
1.8
2.1
2.6
1.5
1.5
2.2

1.3
1.2
1.4
1.1
1.3
1.0
1.6

1.7
1.5
1.5
4.3
2.0
1.0
2.5

1.4
1.4
1.5
1.3
1.3
1.0
1.8

Mean

1.6

1.9

1.3

1.9

1.3

2.1

1.4

hostguest interaction of each enantiomer of a racemate with the


CD selector, depending the separation on the rather low difference
in the energy of interaction of each enantiomer with the CD chiral
selector [13,14]. On the basis of the authors experience, laboratories involved with enantioselective GC in the avour and fragrance
eld should be provided with at least two columns coated with different CD derivatives, which enables to separate at least 80% of the
most common racemates in this eld.
The selected columns were periodically tested through the Grob
test [23,24] to evaluate their chromatographic performance over
time and a homemade chiral test [25] containing racemates with
different volatility, structure and polarity to control their enantioselectivity. Its composition is reported in Section 2.2. Fig. 1
reports the chiral test proles carried out on the four columns
investigated.
3.2.2. Analyte IT determination and reliability
As mentioned above, IT values are the identifying parameter and
can actively and successfully be used only if they are highly stable
over time; the repeatability of IT values over time is fundamental for
a reliable identication. The development of this library required
some fundamental investigations, not necessary in routine work, to
evaluate the reliability of the system. The frequency of injection of
the hydrocarbon reference mixture to obtain the highest IT repeatability was rst studied. The intra-laboratory IT stability versus the
frequency of injection of the hydrocarbons was determined by a
series of periodical experiments carried out on some of the components of the chiral test on all columns investigated. Table 1 reports
the IT variation of the marker analytes by injecting hydrocarbon

Average start-apex RIA

Average apex-stop RIA

1.3
1.9
1.6
4.8
1.6
1.5
1.8

1.4
1.4
1.5
1.3
1.3
1.2
1.7

1.6
1.8
1.7
3.5
1.7
1.5
2.1

2.1

1.4

2.0

standard mixture at different intervals. These results clearly show


that the highest IT stability is obtained when the hydrocarbons
are injected every ve analyte injections. This frequency was then
adopted for the creation of the library. Less frequent hydrocarbon
injections are required for routine analysis.
Control analyses were carried out over a period of 2 months, they
were randomly repeated for three couples of 2 consecutive days for
each CD column. Analyses were repeated three times each day for
a total of 18 determinations for each column. Table 2 reports the
average IT values and IT of each enantiomer calculated over the
whole period investigated. These results show that IT values are
highly repeatable and their IT values never exceed two units.
3.2.3. Determination of the retention index allowance, RIA
The RIA window determination, i.e. the range within which the
IT of an analyte has to fall to be correctly identied, is a key point
for an univocal identication. A friendly operating library should
be based on a unique RIA window to be automatically applied to
all enantiomers when analysed on all columns investigated. The
ideal RIA should be narrow enough to include only one of the
two enantiomers; RIAs choice is therefore strongly conditioned by
the enantiomer resolution obtained with a given chiral selector and
it is critical, in particular, for those enantiomers whose resolution
on a given column is below 1.5, i.e. where partial peak overlapping
occurs. On the other hand, RIA cannot be too narrow to avoid that
the IT of a given analyte falls outside the range because of retention
variation. A reliable RIA therefore requires not only highly stable
IT values, as mentioned in Sections 3.2.2 and 3.2.4, but also highly
inert columns to avoid peak tailing that can affect enantiomer IT

Table 4
Inter-laboratory average IT s and IT s of the chiral test components
2,6DM3PEN--CD

2, 3DM6TBDMS--CD

(S)-()-Limonene
(R)-(+)-Limonene
(S)-2-Octanol
(R)-2-Octanol
(S)-()-Camphor
(R)-(+)-Camphor
(X)-Isobornyl acetate
(Y)-Isobornyl acetate
(R)-()-Linalyl acetate
(S)-(+)-Linalyl acetate
(X)-2-Methyl-(3Z)-hexenyl butyrate
(Y)-2-Methyl-(3Z)-hexenyl butyrate
(1R, 2S, 5R)-()-Menthol
(1S, 2R, 5S)-(+)-Menthol
(X)-Hydroxycitronellal
(Y)-Hydroxycitronellal
(R)--Decalactone
(S)--Decalactone
(S)--Decalactone
(R)--Decalactone

2, 3DE6TBDMS--CD

Laboratory 1 Laboratory 2 I

Laboratory 1 Laboratory 2 I

1061
1068

1062
1069

1
1

1147

1147

1082
1096
1128
1130
1193
1199
1257
1260
1254
1256
1245
1249

1082
1097
1129
1131
1193
1199
1256
1261
1254
1256
1246
1249

0
1
1
1
0
0
1
1
0
0
1
0

1371

1371

1447
1450
1634
1645
1641
1645

1447
1450
1634
1645
1641
1645

0
0
0
0
0
0

1184

1185

1269
1273

1270
1274

1
1

1240

1241

1243
1245
1350
1352

1244
1245
1350
1352

1
0
0
0

1438

1438

1610
1616

1611
1616

1
0

1637

1637

2, 3DA6TBDMS--CD

Laboratory 1 Laboratory 2 I
1056
1072
1111
1112
1133
1141
1220
1222
1231
1237
1240
1244
1282
1285
1373
1374
1573
1588
1586
1588

1056
1072
1110
1112
1134
1142
1221
1224
1233
1239
1241
1245
1283
1285
1374
1375
1573
1587
1587
1589

0
0
1
0
1
1
1
2
2
2
1
1
1
0
1
1
0
1
1
1

Laboratory 1 Laboratory 2 IT


1053

1053

1237

1237

1242
1258

1244
1260

2
2

1321

1322

1302
1303
1296
1298
1383
1393
1646
1652
1810
1820
1866
1876

1302
1304
1297
1298
1385
1394
1648
1654
1813
1823
1869
1878

0
1
1
0
2
1
2
2
3
3
3
2

122
Table 5
List of compounds included in the library
Hydrocarbons
-Phellandrene
-Pinene
-Citronellene
-Phellandrene
-Pinene
Camphene
Caryophyllene
Limonene
Sabinene
Heterocyles
Ambroxide
Menthofuran
Rose oxide
Esters
-Terpinyl acetate
Bornyl acetate
Bornyl isovalerate
Butyl butyrolactate
cis-2-Methyl-3-hexenylbutyrate
cis-Carvyl acetate
Dihydrocarvyl acetate
Dimethyl methylsuccinate
Ethyl 2-methylbutyrate
Ethyl 2-phenylbutyrate
Ethyl 3-hydroxybutyrate
Ethyl 3-hydroxyhexanoate
Ethyl 3-methyl-3-phenylglicidate
Isobornyl acetate
Isobornyl isobutyrate
Lavandulyl acetate
Linalyl acetate
Linalyl cinnamate
Linalyl propionate
Menthyl acetate
Methyl 3-hydroxyhexanoate
Methyl dihydrofarnesoate
Neomenthyl acetate
Nopyl acetate
Propylene glycolbutyrate
Styrallyl acetate
Lactones
Aerangis lactone
3-Methyl--decalactone
-Decalactone
-Dodecalactone
-Heptalactone
-Hexalactone
-Nonalactone
-Octalactone
-Undecalactone
-Decalactone
-Dodecalactone
-Decalactone
-Dodecalactone
-Heptalactone
-Hexalactone
-Nonalactone
-Octalactone
-Pentadecalactone
-Pentalactone
-Tetradecalactone
-Undecalactone
Massoia decalactone
Massoia dodecalactone
Whyskey lactone
Ketones
1,8-Epoxy p-menthan-3-one
3,6-Dimethylocta 2-en-6-one
3-Methylcyclohexanone
3-Oxocineole
-Damascone
-Ionone
-Irone

E. Liberto et al. / J. Chromatogr. A 1195 (2008) 117126


Table 5 (Continued )
Camphor
Camphorquinone
Carvone
Fenchone
Isomenthone
Menthone
Methylcyclopentenolone
Nootkatone
Piperitone
Pulegone
Verbenone
Aldehydes
Citronellal
Cyclamen aldehyde
Hydroxycitronellal
Myrtenal
Perillyl aldehyde
Alcohol
-Bisabolol
1-Octen-3-ol
1-Phenyl ethanol
1-Phenyl-1-propanol
1-Phenyl-2-pentanol
2-Butanol
2-Heptanol
2-Hexanol
2-Methylbutanol
2-Octanol
2-Pentanol
2-Phenyl-1-propanol
3-Hexanol
3-Octanol
4-Methyl-1-phenylpentanol
6-Methyl-5-hepten-2-ol
-Terpineol
Borneol
cis-Myrtanol
Citronellol
Fenchyl alcohol
Geosmin
Isoborneol
Isomenthol
Isopinocampheol
Isopulegol
Lavandulol
Linalool
Linalool oxide
Menthol
Neoisomenthol
Neomenthol
Nerolidol
Octan-1,3-diol
Patchouli alcohol
Perillyl alcohol
Terpinen-4-ol
Tetrahydrolinalool
trans-Myrtanol
Viridiorol
Acids
Citronellic acid
2-Methylbutyric acid
2-Phenylpropionic acid
Chrysanthemic acid

E. Liberto et al. / J. Chromatogr. A 1195 (2008) 117126

123

Fig. 2. Extract ion proles (77, 99, 114 m/z) of (): Cocus nucifera L. fruit head space sampled by SPME, and (- - -) -octalactone standard solution analysed on a 2,3DA6TBDMS-CD column.

values and/or resolution, in particular with polar analytes. RIA is


therefore directly related to the tailing factor of a peak and is determined by measuring the IT values at the start, apex and stop points
of the peak; the window is dened through the following expressions:
T
T
T
T
Istart
) < IxT < IxT + (Istop
Iap
)
IxT (Iap
T is the IT at the apex
where IxT is the IT of the analyte considered, Iap
T
T
is the IT at the peak starting point and Istop
is the IT
of the peak, Istart
at the peak nal point. This approach gives two sub-window values
(the rst one before the peak apex, the second one after the peak
apex) limiting identication mistakes with asymmetrical peaks. A
correct RIA is also conditioned by the injected analyte concentration that should be such to avoid column overloading in particular
with CD stationary phases because of the peak defocusing phenomenon [26].
The unique RIA for the present library was obtained from the
average RIA of each class of compounds, in its turn determined from
the individual RIA of each of the enantiomers of 134 racemates with
different structures, polarities and volatilities analysed three times
on the four enantioselective columns adopted. Table 3 reports the
average RIA window values of the different groups of compounds
analysed with each column. The good RIA homogeneity between
the different classes of compounds allowed us to adopt a unique
average RIA value of 1 and +2 for all analytes analysed on the
investigated columns.

3.2.4. Inter-laboratory IT s reliability


The IT reliability was also tested by an inter-laboratory round
robin test carried out in both the authors laboratories. The chiral
test was simultaneously analysed with two sets of four CD columns
under rigorously controlled conditions (see experimental). Each CD
column was rst submitted to a reconditioning cycle (2 h) and to a

Grob test to evaluate their performance. IT s of the chiral test components were then determined by injecting both the hydrocarbon
standard mixture and three times the chiral test. Table 4 reports IT
values of the chiral test components and IT values between the
two laboratories. These results show that the IT reproducibility is
very high and that the system proposed can be used at an interlaboratory level provided that rigorously standardised conditions
are applied.
3.2.5. Creation of the library
The library was then created by attributing a record to
each enantiomer. Each record includes retention indices on four
columns, mass spectrum, IUPAC chemical name, CAS number,
molecular weight, and, when separated, racemate enantiomer resolution on the chiral selector investigated and, in the included data
base, relative retention and tailing factor together with the original
source of the each enantiomer (see below). Table 5 reports the list
of the compounds at present included in the library.
The library with interactive IT values described here is dedicated
to GC/MS systems adopted here, and operates together with the
related data collecting software (see Section 2.3) [27]. This software
is provided with an automated retention index calculation option.
IT calculation rst requires the creation of the reference index table
obtained by injecting an homologous series of standards (in this
case the C9 C25 n-alkane series). IT values of enantiomers or of
real-world sample components are then determined relatively to
the reference index table. Moreover, up to ve distinct retention
index values (i.e. determined on ve stationary phases) for each
compound whose mass spectrum is appended to the mass spectral
library can be introduced in each record, and a column selection tool
is available in order to display in the similarity search result window only the IT values determined on the investigated stationary
phase.

Fig. 3. Enantioselective GCMS prole of Lavandula angustifolia P. Mill. essential oil analysed on a 2,6DM3PEN--CD column.

124

E. Liberto et al. / J. Chromatogr. A 1195 (2008) 117126

Fig. 4. Enantioselective GCMS prole of linalool in Lavandula angustifolia essential oil with a 2,6DM3PEN--CD column. Library search with inactive (1) and active (2)
RIA.

The correct elution order of the enantiomers of a racemate


was determined by analysing pure standards either commercially
available or supplied by other laboratories, and natural sources
containing identied enantiomer conrming the results with lit-

erature data. The determination of the correct elution order of


-octalactone enantiomers on the four columns is here reported as
an example of the latter approach. Coconut fresh fruits were analysed by HS-SPME-enantioselective GCMS on the four CD columns

Fig. 5. Identication of linalyl acetate in Lavandula angustifolia essential oil.

E. Liberto et al. / J. Chromatogr. A 1195 (2008) 117126

125

and -octalactone located in the four chromatograms through its


MS spectrum. (Fig. 2). The absolute R conguration was attributed
to the identied peak on the basis of literature data reporting the
elution order on the same stationary phases [28].
3.3. Application of the library to real-world samples
The reliability of the library was then evaluated through the
chiral recognition of some components of Lavandula angustifolia
P. Mill (lavender) e.o. These examples are also reported to show
how the long and careful work required for the library development resulted in a very easy and quick operation in the everyday
work. Since the MS spectra of optical isomers are not distinguishable, their EE and/or ER can directly be determined from the TIC
integration, provided that: (a) when complex matrices are analysed, their peaks are perfectly separated and/or diagnostic ions
suitable to discriminate them in extract ion mode from other coeluting peaks are present in their mass spectra or that a suitable
spectral deconvolution programme is available, and (b) a dramatically high difference of enantiomer abundances do not inuence
the mass spectrum relative ion abundances. Moreover, the fact that
each record also includes the enantiomer resolution of a racemate
on the four columns enables the operator to nd the column(s)
(within the four investigated) able to separate from an unresolved
racemate, provided that it has been identied through its mass
spectrum.
Three examples involving the identication and ER determination of well resolved racemates, unresolved racemates, and poorly
resolved compounds are discussed here.
3.3.1. Well resolved racemates
Two different strategies for library searches can be applied
depending if RIA option is operated or not. Enantiomers can be
distinguished through interactive IT values only when the RIA is
operative. This parameter indicates the IT window where to run
the search; only the spectra of analytes with IT values falling in
the selected window are considered, thus making possible their

Fig. 6. Enantioselective GCMS proles of linalyl acetate in Lavandula angustifolia


essential oil ( ) and of its racemate standard (- - -) analysed on a 2,3DE6TBDMS-CD column.

unequivocal identication. If RIA is not operative, IT values are displayed but they do not actively operate for identication: the
enantiomers are therefore located by MS but not distinguished.
These considerations are clearly illustrated by the enantioselective
GCMS analysis of linalool in lavender e.o with a 2,6DM3PEN--CD
column (Figs. 3 and 4) where R enantiomer was found to elute as
rst and to be the most abundant (ER >95%) [29].
3.3.2. Unresolved racemates
Fig. 5 shows the reported results after library search when
linalyl acetate is analysed with the identical column used above
for the same e.o. IT values and RIA are active but specic enantiomers are not identied because they are not separated. However,
library search displays the resolution obtained with the four chiral columns enabling us to nd the one separating them. Fig. 6

Fig. 7. Enantioselective GCMS prole of -pinene in Lavandula angustifolia essential oil analysed on a 2,3DE6TBDMS--CD column. Library search with narrow (1) and wide
(2) RIA.

126

E. Liberto et al. / J. Chromatogr. A 1195 (2008) 117126

reports the enantioselective GC prole of the same e.o. analysed on a 2,3DE6TBDMS--CD where linalyl acetate is base-line
separated, R-isomer eluted as rst and was prevailing (ER > 99%)
[29].
3.3.3. Poorly separated enantiomers
The choice of a correct RIA is fundamental for those racemates
whose resolution on a given column is below 1.5, i.e. when they
partially overlap. In this case, too wide a RIA may include both
enantiomers making their identication problematic, even if their
IT values are different; on the other hand, if RIA is too narrow,
the risk that IT falls outside the window is concrete, in particular
if the chromatographic system is not perfectly stable. The rst
possibility is clearly illustrated in Fig. 7 for -pinene enantiomer
identication in the same lavender essential oil when it is analysed
with a 2,3DE6TBDMS--CD column and a RIA between 3 and +3
is applied. On the other hand Fig. 7, shows that the general RIA
adopted for this library, from 1 to + 2, provides a correct enantiomer identication showing that the R isomer elutes as rst with
an ER above 75%.
4. Conclusions
The results reported here demonstrated the reliability of the
adopted approach for an unequivocal enantiomer identication by
enantioselective GCMS in the avour and fragrance eld and the
fundamental importance of combining IT values and mass spectra
actively for the correct identication of an enantiomer. IT values
have also been demonstrated to be highly stable, thus affording the
adoption of a RIA value common to all class of racemates investigated between +1 and 2 IT units, provided that the reference
hydrocarbon standard mixture is injected every ve analyses when
the library is created. The library has also been shown to operate
effectively in the analysis of real-world samples and at present it
includes the enantiomers of 134 racemates analysed on columns
coated with four different CD derivatives. The number of records is
being constantly increased although its implementation takes time
because of the difculty to nd new racemates and the related pure
enantiomers.

Acknowledgements
This research was carried out within the project entitled:
Sviluppo di metodologie innovative per lanalisi di prodotti
agroalimentari (FIRB Cod.: RBIP06SXMR 002) of the Ministero
dellIstruzione, dellUniversita` e della Ricerca (MIUR) (Italy).
References
[1] D. Sybilska, T. Koscielski, J. Chromatogr. 261 (1983) 357.
[2] Z. Juvancz, G. Alexander, J. Szejtli, J. High Resolut. Chromatogr. Chromatogr.
Commun. 10 (1987) 105.
[3] V. Schurig, H.P. Nowotny, J. Chromatogr. 441 (1988) 155.
[4] H.P. Nowotny, D. Schmalzing, D. Wistuba, V. Schurig, J. High Resolut. Chromatogr. 12 (1989) 383.
[5] C. Bicchi, V. Manzin, A. DAmato, P. Rubiolo, Flavour Fragr. J. 10 (1995) 127.
[6] C. Bicchi, A. DAmato, V. Manzin, in: K. Swift (Ed.), Flavours and Fragrances,
Royal Society of Chemistry, Cambridge, UK, 1997, p. 57.
[7] C. Bicchi, A. DAmato, P. Rubiolo, J. Chromatogr. A 843 (1999) 99.
[8] W.A. Koenig, D.H. Hochmuth, J. Chromatogr. Sci. 42 (2004) 402.
[9] G. Schomburg, H. Husmann, E. Hubinger, W.A. Koenig, J. High Resolut. Chromatogr. Chromatogr. Commun. 7 (1984) 404.
[10] A. Bernreuther, N. Christoph, P. Schreier, J. Chromatogr. 481 (1989) 363.
[11] A. Mosandl, U. Hener, U. Hagenauer-Hener, A. Kustermann, J. High Resolut.
Chromatogr. 12 (1989) 532.
[12] L. Mondello, M. Catalfamo, G. Dugo, P. Dugo, J. Chromatogr. Sci. 36 (1998) 201.
[13] M. Jung, D. Schamalzing, V. Schurig, J. Chromatogr. 552 (1991) 43.
[14] V. Schurig, H.P. Nowotny, Angew. Chem., Int. Ed. Engl. 29 (1990) 939.
[15] R. Shellie, P. Marriott, C. Cornwell, Anal. Chem. 74 (2002) 5426.
[16] R. Shellie, L. Mondello, G. Dugo, P. Marriot, Flavour Fragr. J. 19 (2004) 582.
Helv. Chim. Acta 41 (1958) 1915.
[17] E. Kovats,
[18] H. Van den Dool, P.D. Kratz, J. Chromatogr. 11 (1963) 463.
[19] C. Bicchi, G. Artuffo, A. DAmato, V. Manzin, J. High Resolut. Chromatogr. 15
(1992) 710.

[20] W.A. Koenig, B. Gehrcke, D. Ichelin, P. Evers, J. Donnecke,


W. Wang, J. High
Resolut. Chromatogr. 15 (1992) 367.
[21] B. Maas, A. Dietrich, D. Bartschat, A. Mosandl, J. Chromatogr. Sci. 33 (1995) 223.
[22] C. Bicchi, A. DAmato, V. Manzin, A. Galli, M. Galli, J. Chromatogr. A 742 (1996)
161.
[23] K. Grob, G. Grob, K. Grob, J. Chromatogr. 156 (1978) 1.
[24] K. Grob, G. Grob, K. Grob, J. Chromatogr. 219 (1981) 13.
[25] C. Bicchi, G. Artuffo, A. DAmato, V. Manzin, A. Galli, M. Galli, J. High Resolut.
Chromatogr. 16 (1991) 209.
[26] H.G. Schmarr, B. Maas, A. Mosandl, S. Bihler, H.P. Neukom, K. Grob, J. High
Resolut. Chromatogr. 14 (1991) 317.
[27] R. Costa, M.R. De Fina, M.R. Valentino, P. Dugo, L. Mondello, Nat. Product Commun. 2 (2007) 413.
[28] D. Lehmann, B. Maas, A. Mosandl, Z. Lebensm, Unters. Forsch. 201 (1995) 55.
[29] P. Kreiss, A. Mosandl, Flavour Fragr. J. 7 (1992) 187.

You might also like