S

Gt

LI

a

ARRAA

KMSTGMC

1

pCepeaa

eosrtmpeEastt

0d

Journal of Chromatography A, 1218 (2011) 7061– 7064

Contents lists available at ScienceDirect

Journal of Chromatography A

jou rn al h om epage: www.elsev ier .com/ locat e/chroma

hort communication

as chromatographic retention indices of trimethylsilyl derivatives oferpene alcohols

ech Szczepaniak ∗, Valery A. Isidorovnstitute of Chemistry, Białystok University, 15-399 Białystok, Poland

r t i c l e i n f o

rticle history:eceived 2 June 2011eceived in revised form 27 July 2011ccepted 31 July 2011vailable online 6 August 2011

a b s t r a c t

This paper presents the experimentally determined retention indices (RITMS) for a set of 75 silylated ter-penols (33 monoterpenols and 42 sesquiterpenols). The attempt was made to assess the dependence ofRITMS on RI (for non-silylated terpenols) and on RIAc (for acetylated terpenols). Satisfactory linear regres-sion parameters (RITMS = b0 + b1RI) were observed for tertiary substituted monoterpenols and primary orsecondary substituted sesquiterpenols. The mass spectra of silylated terpenols that were not found inthe available literature are in Supplementary information.

eywords:onoterpene alcohols

esquiterpene alcoholsrimethylsilyl derivativesas chromatographic retention indicesass spectra

© 2011 Elsevier B.V. All rights reserved.

alculated retention indices

. Introduction

Terpene alcohols are widespread in nature, particularly inlants. Together with monoterpenes C10H16 and sesquiterpenes15H24, they make the major part of essential oils and plant tissuexudates. As constituents of plant material, they appear in differentroducts, both food and others, such as raw and processed veg-tables and fruits, beverages, cosmetics and medicines. Their widepplication comes also from the well-documented antimicrobialction.

Usually, the GC/MS identification of terpenols in samples likessential oils does not show any major difficulties, since for manyf them literature provides important analytical characteristics,uch as electron impact mass spectra and gas chromatographicetention indices for phases of different polarities. However,erpene alcohols may also constitute a part of more complex

ixtures containing multifunctional substances, such as polyols,henol carboxylic acids, flavonoids and chalcones. Plant tissuextracts, plant exudates and propolis belong to these mixtures.ssential oil from birch buds is a rich source of caryophyllene-typelcohols (betulenols) [1–3]. During analysis of such mixtures, the

ingle injection method is preferred [4]. This is possible only underhe condition of derivatization of polar compounds, for instanceheir silylation. However, available literature provides mass

∗ Corresponding author. Tel.: +48 85 7457800; fax: +48 85 7470103.E-mail address: lech@uwb.edu.pl (L. Szczepaniak).

021-9673/$ – see front matter © 2011 Elsevier B.V. All rights reserved.oi:10.1016/j.chroma.2011.07.106

spectra of TMS derivatives of terpene alcohols and their chromato-graphic retention indices only for a few substances of this class. Itsignificantly limits the possibility of terpenols identification.

The purpose of this work is to present both gas chromatographicretention indices and mass spectra for a collection of silylated ter-penols with different types of substitution (primary, secondary,tertiary, endo-, exo-cyclic –OH groups). There were used both com-mercially available terpene alcohols, as well as ones separated forthis purpose from plants and commercial essential oils.

2. Experimental

2.1. Chemicals

Commercial sabinene hydrate, nerol, geraniol, linalool,(Z)-verbenol, (E)-verbenol, menthol, isomenthol, isopulegol,(E)-pinocarveol, myrtanol, terpineols (mixture of �-, �- and�-isomers, 1- and 4-terpineols), (E)-nerolidol, (Z,E)-farnesol, (E,E)-farnesol, �-bisabolol, �-eudesmol, (−)-globulol and (+)-cedrolwere purchased from Sigma–Aldrich, Fluka and Roth (Poznan,Poland). Part of commercial preparation of monoterpene alco-hols was obtained from Chemistry of Natural Products Divisionof Institute of Chemistry, University of Białystok (Białystok,Poland). Commercial essential oils from Amyris balsamifera,

Aniba roseadora, Malaleuca alternifolia (tea tree oil), Juniperusvirginia (cedar oil), sandal-wood oil and ylang-ylang oil were pur-chased from AVICENA-OIL (Wroclaw, Poland). Pyridine, n-hexaneand bis(trimethylsilyl)trifluoroacetamide (BSTFA) containing

7 hrom

1(f

2

feanatcupgtrhcocctdt

tpDtwteaotwit

2

htothwSsosT27Ti2

uta

062 L. Szczepaniak, V.A. Isidorov / J. C

% trimethylchlorosilane were purchased from Sigma–AldrichPoznan, Poland). Ethyl acetate and diethyl ether were purchasedrom POCH SA (Gliwice, Poland).

.2. Isolation of terpenols from plant materials

Some of not commercially available terpenols were isolatedrom essential oils or ether extracts from plant material of differ-nt origins. Essential oils from leaves of Scots pine (Pinus sylvestris)nd thuya (Thuja occidentalis), as well as from black poplar (Populusigra) buds, were extracted by hydrodistillation in a Clevenger-typepparatus [5] and dried over anhydrous Na2SO4. These prepara-ions as well as commercial essential oils were fractionated byolumn chromatography on a silica column. Pure n-hexane wassed to elute the mixtures of non-polar and slightly polar com-ounds (terpene hydrocarbons, carbonyls, ethers and esters). Next,radient elution (n-hexane/ethyl acetate) was used to isolate frac-ions containing terpene alcohols. Some of these fractions wereechromatographed to provide a fraction enriched in terpene alco-ols of interests. There was no attempt to isolate a pure singleompound and the only task of column chromatography was tobtain fractions containing a few (two to four) compounds whichould be well separated (without overlapping of peaks) on the GCapillary column. All these fractions were analyzed by GC/MS andhe terpenol identification was performed using both mass spectralata and calculated retention indices which were compared withhe ones reported by Adams [6].

Because there were not enough birch buds to isolate the essen-ial oil by hydrodistillation, the buds of two species of birch (Betulaubescens and Betula litwinowii) were extracted by diethyl ether.ried at 40 ◦C, buds (5 g) were crushed and while constantly stirred,

hree portions of ether 25 mL each were applied. The joint extractsere filtered through a paper filter and the solvent was removed

o a volume of ca. 1 mL on a rotor evaporator. These concentratedxtracts were subjected to column chromatography (as above) andll the fractions were analyzed by GC/MS. The automatic systemf processing data of GC/MS is supplied by NIST (US National Insti-ute of Standards and Technology) and Wiley mass spectra librarieshich do not contain the mass spectra of betulenols. Hence, the reg-

stered spectra of these alcohols were compared with published inhe literature [1,2].

.3. Sample preparation and analysis

TMS derivatives of commercial preparations and terpene alco-ols isolated from plant materials were synthesized according tohe unified procedure. 1–2 mg of the materials were put into a vialf 2 mL in volume and dissolved in 200–250 �L of pyridine, afterhat, 50–80 �L of BSTFA was added. The reaction mixtures wereeated at 70 ◦C for 0.5 h. The terpenols and their TMS derivativesere analyzed on an Agilent 6890 Gas Chromatograph with Mass

elective Detector 5973 and Autosampler 7683, electronic pres-ure control and split/splitless injector. Separation was performedn the HP-5ms (30 m × 0.25 mm I.D.; 0.25 �m film thickness) fusedilica column. Helium flow rate through the column was 1 mL/min.he injector worked in split (1:50) mode; injector temperature50 ◦C. The EIMS spectra were obtained for ionization energy of0 eV, at the source temperature 230 ◦C and quadrupol one 150 ◦C.he MSD was set to scan 41–600 a.m.u. Chromatograms were reg-stered in a linear temperature programmed regime from 50 ◦C to20 ◦C at the rate of 3 ◦C/min.

Hexane solution of C10–C22 n-alkanes was chromatographednder above conditions. Linear temperature programmed reten-ion indices (RI) were calculated from the results for this mixture,nd for solutions of terpene alcohols and their TMS derivatives.

atogr. A 1218 (2011) 7061– 7064

3. Results and discussion

The success of the GC/MS identification in the routine analysisdepends entirely on the availability of such analytical parameters asmass spectra and chromatographic retention indices of analytes [7].In case of TMS derivatives of terpene alcohols, an access to such datais very limited. In the most comprehensive of available databases[8], there are presented the mass spectra for only five TMS deriva-tives of C10H18O monoterpenols and four C15H26O sesquiterpenols.Because of this, in some papers, the authors are limited to the iden-tification on the group level (and not the particular substances) incase of finding terpenols in samples. For example, in the study [9],13 sesquiterpenols were registered in the form of TMS derivatives,but only two substances were positively identified ((E)-nerolidoland bisabolol) and one (guaiol) – presumably.

The identification of terpenol TMS derivatives gets complicatedalso by a small informative value of their mass spectra. In manycases, after silylation, mass spectra of alcohols get significantly sim-pler and contain only a few intensive peaks, what is the case for e.g.�-eudesmol and its TMS derivative, as shown in Fig. 1(A) and (B).However, even when a researcher possesses standard mass spectraof TMS derivatives, correct identification of terpenols is not alwaysguaranteed due to high similarity between the spectra of these sub-stances. In Fig. 1(C) and (D) there are presented the mass spectra ofderivatives of two sesquiterpenols with completely different struc-tures. It can be seen that the set of fragment ions is identical in bothcases, and the usual distortions of the mass spectra do not give achoice possibility in favor of one of the two alcohols on the basis ofpeaks intensity differences (for the m/z 161, 204, 105, and 189). Insuch a situation, the role of chromatographic retention indices asreference analytical parameters is rising. Besides, aforementioneddatabase [8] contains retention indices of TMS derivatives of onlyone C10H16O alcohol, five C10H18O monoterpenols and five C15H26Osesquiterpenols. There are no mass spectra, nor retention indices,for C15H24O TMS sesquiterpenols derivatives in the available liter-ature.

Table 1 presents experimental values of retention indicesfor 75 terpenols (33 mono- and 42 sesquiterpenols). Measuredretention indices for terpene alcohols are in good accordance(usually in range ±3 i.u.) with ones presented by Adams [6],and their mass spectra – with ones contained in the librariesof the used GC/MS apparatus software. Mass spectra of terpenolTMS derivatives (absent in [8]) are presented in Supplementaryinformation.

Since the limited number of commercial and plant-isolated ter-pene alcohols was available, it was interesting to calculate theretention indices of terpenol TMS derivatives not included in ourdata. One may assume that the retention indices of silylated ter-penols can be calculated from the relationship between RITMS andRI or between RITMS and RIAc (for this purpose the values of RIAc –the retention indices for acetylated terpenols from [6] were gath-ered in Table 1). After initial data analysis, it was found that thefact if a given terpenol is tertiary substituted (or not) has a signif-icant influence on relationships between RITMS and RI or betweenRITMS and RIAc. The regression equation (model) parameters fordifferent subgroups (all, tertiary, non-tertiary substituted) are pre-sented in Table 2. It was assumed that regression parameters shouldmeet the following conditions: (1) n > 10, (2) SE < 15 u.i., (3) regres-sion parameters statistically significant at the level of 5% (P < 0.05).Only two models (RITMS–RI relationships) fulfilled these conditions:for tertiary monoterpenols, and for non-tertiary sesquiterpenols.It is a surprise, that RITMS–RIAc relationships do not satisfy these

conditions. Perhaps larger set of data RIAc would change this sit-uation. It is believed that greater possibilities of RITMS predictionwill be achieved with the application of the chemometric approachwith the use of molecular descriptors. The better opportunities are

L. Szczepaniak, V.A. Isidorov / J. Chromatogr. A 1218 (2011) 7061– 7064 7063

Table 1Retention indices of monoterpenols and sesquiterpenols.

No. Terpenol Formula Sa RI RIAc RITMS

1 (Z)-sabinenehydrate C10H18O 3 1067 1219 12072 (E)-sabinenehydrate C10H18O 3 1092 1253 12323 Linalool C10H18O 3 1099 1256 12404 endo-Fenchol C10H18O 2 1104 1220 12025 exo-Fenchol C10H18O 2 1113 1234 12126 (Z)-p-menth-2-en-1-ol C10H18O 3 1120 12597 Terpin-1-ol C10H18O 3 1132 12608 (E)-pinocarveol C10H16O 2 1137 1297 12709 (E)-p-menth-2-en-1-ol C10H18O 3 1139 127610 (Z)-verbenol C10H16O 2 1140 1282 124811 �-Terpineol C10H18O 3 1142 1297 126812 (E)-verbenol C10H16O 2 1145 1292 125213 Camphene hydrate C10H18O 3 1147 128314 Isopulegol C10H18O 2 1149 1275 125115 Isoborneol C10H18O 2 1150 121616 �-Phellandren-8-ol C10H16O 3 1157 129317 Borneol C10H18O 2 1161 122618 �-Phellandren-8-ol C10H16O 3 1165 130519 Lavandulol C10H18O 1 1170 123420 Terpin-4-ol C10H18O 3 1177 1328 130321 Isomenthol C10H20O 2 1182 124322 Myrthenol C10H16O 1 1192 1327 130023 �-Terpineol C10H18O 3 1192 1350 132224 (Z)-piperitol C10H18O 2 1195 1330 128525 �-Terpineol C10H18O 3 1196 132726 Menthol C10H20O 2 1206 1294 126427 (E)-piperitol C10H18O 2 1208 1340 129828 (E)-carveol C10H16O 2 1217 1337 128029 (Z)-carveol C10H16O 2 1229 1362 129130 Nerol C10H18O 1 1229 1365 134331 Citronellol C10H20O 1 1229 1354 131932 Myrtanol C10H18O 1 1250 1381 134033 Geraniol C10H18O 1 1255 1383 137034 1-epi-Cubebol C15H26O 3 1494 161735 Cubebol C15H26O 3 1515 163036 Birkenol C14H24O 1 1520 156537 �-Copaen-11-ol C15H24O 3 1534 163438 (Z)-nerolidol C15H26O 3 1534 166139 Elemol C15H26O 3 1549 163840 (E)-nerolidol C15H26O 3 1564 1714 169141 Germacrene d-4-ol C15H26O 3 1575 167442 Spathulenol C15H24O 3 1576 167343 (−)-Globulol C15H26O 3 1584 168444 Gleenol C16H26O 2 1585 163045 Viridiflorol C15H26O 3 1590 168946 Carotol C15H26O 3 1595 164247 Cedrol C15H26O 3 1596 1762 174548 Guaiol C15H26O 3 1597 1724 168749 1,10-di-epi-Cubenol C15H26O 3 1610 164650 Cubenol C15H26O 3 1627 166751 Agarospirol C15H26O 3 1628 173452 Caryophylla-2(12),6(13)-dien-5b-ol C15H24O 2 1628 167653 Acorenol C15H26O 3 1630 172354 �-Eudesmol C15H26O 3 1633 1778 173955 Hinesol C15H26O 3 1633 1780 173656 Caryophylla-2(12),6(13)-dien-5a-ol C15H24O 2 1634 168257 6-Hydroxy-caryophyllene C15H24O 2 1636 168458 tau-Cadinol C15H26O 3 1640 170159 tau-Muurolol C15H26O 3 1642 170760 �-Eudesmol C15H26O 3 1649 1788 175161 �-Cadinol C15H26O 3 1651 174862 7-epi-�-Eudesmol C15H26O 3 1654 174863 �-Eudesmol C15H26O 3 1656 1790 175664 Patchouli alcohol C15H26O 3 1659 171965 14-Hydroxy-9-epi(E)-caryophyllene C15H24O 1 1662 172866 14-Hydroxy-caryophyllene C15H24O 1 1664 172567 Bulnesol C15H26O 3 1667 176768 �-Bisabolol C15H26O 3 1679 175269 14-Hydroxy-4,5-dihydro-isocaryophyllene C15H26O 1 1681 173770 epi-�-Bisabolol C15H26O 3 1683 1801 175671 2,3-Dihydro-(6E)-farnesol C15H28O 1 1688 176272 (2Z,6E)-Farnesol C15H26O 1 1696 1818 178873 14-Hydroxy-4,5-dihydro-�-caryophyllene C15H26O 1 1704 176274 14-Hydroxy-�-humulene C15H24O 1 1709 176575 (2E,6E)-Farnesol C15H26O 1 1720 1843 1812

Typical errors for RITMS are in the range 1–8 i.u., according to Supplementary material.a Type of substitution –OH group (1, primary; 2, secondary; 3, tertiary).

7064 L. Szczepaniak, V.A. Isidorov / J. Chromatogr. A 1218 (2011) 7061– 7064

Fig. 1. Mass spectra of �-eudesmol (A) and some of isomeric C15H26O sesquiterpene alcohols after silylation: (B) �-eudesmol, TMS derivative; (C) guaiol, TMS derivative(bicyclic 5-azulenemethanol); (D) �-acorenol, TMS derivative (�,�,4,8-tetramethylspiro[4.5]dec-7-ene-1-methanol).

Table 2Regression data for monoterpenols and sesquiterpenols.

No. Dataset Model n b0 SD (b0) b1 SD (b1) R2 SE

1 Monoterpenols RITMS = b0 + b1RI 33 440 113 0.71 0.10 0.637 262 Monoterpenols only tertiary RITMS = b0 + b1RI 13 234 40 0.91 0.03 0.984 53 Monoterpenols non-tertiary RITMS = b0 + b1RI 20 190a 138 0.91 0.12 0.775 234 Sesquiterpenols RITMS = b0 + b1RI 42 359 119 0.83 0.07 0.761 265 Sesquiterpenols only tertiary RITMS = b0 + b1RI 29 454 147 0.78 0.09 0.727 246 Sesquiterpenols non-tertiary RITMS = b0 + b1RI 13 −262 108 1.19 0.07 0.968 137 Monoterpenols RITMS = b0 + b1RIAc 22 146a 73 0.86 0.05 0.924 138 Monoterpenols only tertiary RITMS = b0 + b1RIAc 6 130 44 0.88 0.03 0.994 49 Monoterpenols non-tertiary RITMS = b0 + b1RIAc 16 110a 100 0.89 0.08 0.907 14

10 Sesquiterpenols RITMS = b0 + b1RIAc 10 65a 137 0.94 0.08 0.950 911 Sesquiterpenols only tertiary RITMS = b0 + b1RIAc 8 254a 188 0.84 0.11 0.911 912 Sesquiterpenols non-tertiary RI = b + b RI 2 – – – – – –

R RITMS

fe

A

t

R

[[

[

[[[

[

[Reference Database Number 69, National Institute of Standards and Tech-

TMS 0 1 Ac

2, coefficient of determination; SE, standard error; RI, retention index of terpenol;a b0 value is statistically not significant at the 5% level.

or RITMS–RI relationships, due to the big number of collected (RI)xperimental data.

ppendix A. Supplementary data

Supplementary data associated with this article can be found, inhe online version, at doi:10.1016/j.chroma.2011.07.106.

eferences

1] B. Demirci, K.H. Bas er, T. Ozek, F. Demirci, Planta Med. 66 (2000) 490.2] K.D. Klika, B. Demirci, J.-P. Salminen, V.V. Ovcharenko, S. Vuorela, K.H. Can Bas er,

K. Pihlaja, Eur. J. Org. Chem. (2004) 2627.

[

, retention index of silylated terpenol; RIAc, retention index of acetylated terpenol.

3] V.A. Isidorov, U. Krajewska, V.T. Vinogorova, L.V. Vetchinnikova, I.L. Fuksman, K.Bal, Biochem. Syst. Ecol. 32 (2004) 1.

4] I. Molnár-Perl, J. Chromatogr. A 845 (1999) 181.5] J.F. Clevenger, Am. Perfumer Essent. Oil Rev. (1928) 467.6] R.P. Adams, Identification of Essential Oil Components by Gas Chromatogra-

phy/Mass Spectroscopy, 2nd ed., Allured Pub. Corp., Carol Stream, IL, USA,1995.

7] V.I. Babushok, P.J. Linstrom, J.J. Reed, I.G. Zenkevich, R.L. Brown, W.G. Mallard,S.E. Stein, J. Chromatogr. A 1157 (2007) 414.

8] P.J. Linstrom, W.G. Mallard, NIST Chemistry WebBook, NIST Standard

nology, Gaithersburg, MD 20899, 2011, http://webbook.nist.gov/chemistry/(retrieved 01.04.11).

9] W. Greenway, J. May, T. Scaysbrook, F.R. Whatley, Z. Naturforsch. 42c (1991)111.