- Email: [email protected]
Fast enantiomeric analysis of a complex essential oil with an innovative multidimensional gas chromatographic system
Journal of Chromatography A, 1105 (2006) 11–16
Fast enantiomeric analysis of a complex essential oil with an innovative multidimensional gas chromatographic system Luigi Mondello a,∗ , Alessandro Casilli a , Peter Quinto Tranchida a , Masanao Furukawa c , Kyoichi Komori c , Kozo Miseki c , Paola Dugo b , Giovanni Dugo a b
a Dipartimento Farmaco-chimico, Facolt` a di Farmacia, Universit`a degli Studi di Messina, viale Annunziata, 98168 Messina, Italy Dipartimento di Chimica Organica e Biologica, Facolt`a di Scienze MM.FF.NN., Universit`a di Messina, Salita Sperone, 98165 Messina, Italy c Shimadzu Corporation, 1 Nishinokyo-Kuwabaracho, Nakagyo-ku, Kyoto 604-8511, Japan
Available online 24 August 2005
Abstract The present research is focussed on the evaluation of a recently developed high performance multidimensional gas chromatographic (MDGC) system employed in the fast analysis of a series of chiral compounds contained in rosemary essential oil. The heart of the MDGC system consists in a simple transfer device for the rapid sequential re-injection of analyte “heart-cuts” from the first to the second dimension. The transfer system has no temperature restrictions, presents very low dead volumes and achieves multidimensional analysis through a pressure-balance mechanism. The MDGC set-up is characterized by two GC ovens (enabling independent temperature programming) and the possibility of mass spectrometric (MS) and/or flame ionization detection (FID). Multiple-cut conventional and fast MDGC–FID methods were developed and the results obtained compared, in order to evaluate the effectiveness of the system. In this respect, the rapid method provided the same analytical result in a greatly reduced time (approximately five times less). Furthermore, quali/quantitative data reproducibilty was very good. Fast MDGC was achieved by using micro-bore (0.1 mm I.D.) columns in both dimensions. © 2005 Elsevier B.V. All rights reserved. Keywords: Multidimensional gas chromatography; Rosemary essential oil; Chiral analysis; Transfer system; Enantiomeric ratio
1. Introduction It is well-known that single column GC is often an insufficient tool when the total separation of a complex matrix is required. Component co-elution is a common event even for low-complexity samples, as the distribution of chromatographic peaks is, generally, random [1]. In particular, major difficulties are encountered when trace-level analytes overlap with more concentrated solutes. As a consequence, the development of multidimensional instrumentation and methods is one of the foremost goals of modern-day gas chromatography. With this respect, two directions are generally followed for the attainment of more extensive and reliable analytical information: the combination of two independent GC columns (with different selectivities) or the linkage of a mass spectrometer to a gas chromatographic system.
∗
Corresponding author. Tel: +39 090 676 6536 fax: +39 090 676 6532. E-mail address: [email protected] (L. Mondello).
0021-9673/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.chroma.2005.08.013
GC–MS is an effective procedure not only for single peak identification and quantitation [2] but also for the unravelling of multi-component peaks through the use of deconvolution methods [2,3]. It is obvious, though, that MS library peak assignment is more reliable (especially for the more complex samples) when pure mass spectra relative to entirely resolved compounds are obtained. Total component separation is, therefore, always desirable, but often difficult to achieve. Comprehensive GC (GC × GC) is a remarkable chromatographic method which enables the bidimensional analysis of the entire initial sample. As such, it is characterized by a greatly enhanced peak capacity with respect to conventional multidimensional gas chromatographic (MDGC) (and to any other type of GC method) [4]. A series of outstanding separations on complex samples have been reported in the literature [5–7]. The introduction of a third mass spectrometric dimension in a comprehensive GC system may be considered the most powerful option for volatile analysis today available. Nevertheless, it may also be affirmed that GC × GC method optimization is not an easy task, especially for analysts who do not possess a great experience in this field. Furthermore, the attainment of com-
12
L. Mondello et al. / J. Chromatogr. A 1105 (2006) 11–16
prehensive GC quantitative data is dependent on commercial softwares which are sold with industrially-produced GC × GC instrumentation. An additional shortcoming is represented by the fact that the separation power of a short secondary column (generally 1 m), in some cases, may be insufficient for the resolution of all components contained in each single primary column slice. In many cases, a high number of theoretical plates is required for the thorough separation of specific fractions in the second dimension. Conventional “heart-cutting” multidimensional gas chromatography has been on the chromatographic scene for more than 40 years [8]. The analytical approach is based, essentially, on the transfer of selected solute bands from a first to a secondary column. The aim is to subject unresolved 1D compounds to a further independent separation step [9]. The second chromatographic process must not, in any way, invalidate component resolution previously achieved on the primary column. The entire procedure is enabled by the presence, between the two columns, of a specific transfer device. Column flows are diverted, basically, by exploiting two different principles: valve [10] or Deans type switching [11,12]. A thorough account of the most commonly employed interfaces is outside the scope of the present paper. Accurate descriptions and advantages/disadvantages of the most important MDGC interfaces have been reported in the literature [13,14]. It must be noted that the employment of MDGC has undergone quite a substantial decline with the vast availability of bench-top GC–MS systems and since the introduction of GC × GC in 1991 [15]. Nevertheless, heart-cutting MDGC may be considered, for some applications, the most appropriate analytical choice. In general, MDGC method optimization is much easier if compared to comprehensive GC; the combined column efficiencies, for a specific fraction, are higher with respect to GC × GC; data elaboration is simple; pure solute bands are usually delivered to the detection system (deconvolution is not necessary). Obviously, with this type of approach continuous heart-cutting over the entire sample is not possible as this would cause the loss of primary column peak separation [16]. With respect to the maximum number of cuts achievable in a single MDGC run, this is dependent on the sample-type and on the analytical conditions. The present research is based on the analytical evaluation of a new high performance MDGC system. Both conventional and rapid MDGC methods were developed and applied to the analysis of hydro-distilled rosemary essential oil. The degree of peak retention time shift during multiple-cut analysis and the analytical quali/quantitative reproducibility were determined in the more drastic of the two applications. To the authors knowledge, no previous publications on fast MDGC analysis with microbore columns have been reported.
The following pure standard components were purchased from Sigma–Aldrich (Milan, Italy): (+)-camphene; (−)camphene; (+)-sabinene; (+)--pinene; (−)--pinene; (+)limonene; (−)-limonene; (+)-camphor; (−)-camphor; (+/−)isoborneol; (+)-borneol; (−)-borneol; (+)-terpinen-4-ol; (+)-␣terpineol; (−)-␣-terpineol. The chiral compounds were diluted 1:10 (v/v) in n-hexane prior to analysis in all applications. 2.2. Instrumentation and chromatographic conditions The MDGC system consists of two Shimadzu GC-2010 gas chromatographs [from now onwards denominated as GC 1 (location of the primary column) and GC 2 (location of the secondary column], a Shimadzu GCMS-QP2010 quadrupole mass spectrometer (not employed in the present research) and a Shimadzu AOC-20i autoinjector (Shimadzu Corporation, Kyoto, Japan). GC 1 and GC 2 present a split/splitless injection system (the injection system in GC 2 was not used) and a flame ionization detector [from now onwards denominated FID 1 (situated in GC 1) and FID 2 (situated in GC 1)]. GC 1 and GC 2 are connected by means of a heated transfer line (200 ◦ C). A schematic of the transfer device, which is partly located in GC 1, is shown in Fig. 1a (stand-by position) and b (cut position). The interface consists of an advanced pressure control (APC) unit [and a pressure sensor (PS)] which supplies carrier gas, at constant
2. Experimental 2.1. Samples and sample preparation Hydro-distilled rosemary essential oil was purchased from Sigma–Aldrich (Milano, Italy). The sample was diluted 1:10 (v/v) in n-hexane prior to analysis in all applications.
Fig. 1. (a) A schematic of the tranfer system in stand-by position; (b) a schematic of the tranfer system in cut position (for abbreviations refer to text).
L. Mondello et al. / J. Chromatogr. A 1105 (2006) 11–16
pressure (defined as P), to a three-way solenoid valve (defined as V). The valve is linked to two stainless steel tube branches that are bridged by a fused silica restrictor (R1). The latter, positioned in the flow pathway, causes a pressure drop indicated as P. A second restrictor (R2), which is located within GC 1, acts as a connection between the first and second column. R2, as can be seen in the figure, is also subjected to flow rates that derive from the two stainless tube branches (P and P − P). When V is in the normally opened position the transfer system is in the “stand-by” status. In this situation, the pressure (supplied by the APC unit) at the entrance of the secondary column (P) is higher than the pressure (supplied by the APC unit) at the end of the primary column (P − P). As a consequence, the access of the first dimension gas flow to the second column is prevented and is, thus, directed solely to FID 1. The FID 2 flame is maintained by a gas flow that derives from the APC unit pressure P. On the contrary, when V is in the normally-closed position the transfer system is in the “cut” status. In this case, the pressure (supplied by the APC unit) at the entrance of the primary column (P) is higher than the pressure (supplied by the APC unit) at the head of the secondary column (P − P). As a consequence, the first dimension mobile phase flows freely to the second column and to FID 2. The FID 1 flame is maintained by a gas flow that derives from the APC unit pressure P. MDGC method calculations were carried out with the support of a MDGC Calculation software (Shimadzu, Kyoto, Japan). The MDGC system was controlled by a MDGC software (Shimadzu, Kyoto, Japan). Data were acquired by a GC solution software (Shimadzu, Kyoto, Japan). The experimental conditions in the conventional MDGC application were as follows: column 1 was a SE 52 (5% diphenyl + 95% polydimethylsiloxane) 25 m × 0.25 mm I.D. × 0.25 m; column 2 was a DEtTBuSililBETA-086 (diethyltert-butyl-silil--cyclodextrin) 25 m × 0.25 mm I.D. × 0.25 m (MEGA, Legnano, Italy). Column 1 was passed through the transfer line and connected to column 2 by using a press-fit. The transfer device was connected to FID 1 by means of a 0.25 m × 0.15 mm I.D. retention gap (Restek, Bellefonte, PA, USA). An inlet pressure of 250 kPa (He) was applied to the transfer device by means of an APC unit. Injection mode GC 1 (250 ◦ C): split (25:1). Injection volume: 1 L. Carrier gas: He. Constant inlet pressure: 345 kPa. Initial linear velocity column 1: 36 cm/s. Initial linear velocity column 2: 86 cm/s. Tempera◦ ture program GC 1: 50 ◦ C to 280 ◦ C at 3 C/min. Temperature ◦ ◦ program GC 2: 45 C (12 min) to 180 C at 2 ◦ C/min. FID 1 (280 ◦ C): H2 flow: 50.0 mL/min; air flow: 400.0 mL/min; make up: 0.0 mL/min. Sampling rate: 80 ms. FID 2 (210 ◦ C): H2 flow: 50.0 mL/min; air flow: 400.0 mL/min; make up: 50.0 mL/min. Sampling rate: 40 ms. The experimental conditions in the fast MDGC application were as follows: column 1 was a SE 52 (5% diphenyl + 95% polydimethylsiloxane) 10 m × 0.1 mm I.D. × 0.1 m; column 2 was a DEtTBuSililBETA-086 (diethyl-tert-butyl-silil-cyclodextrin) 10 m × 0.1 mm I.D. × 0.1 m (MEGA, Legnano, Italy). Column 1 was passed through the transfer line and connected to column 2 by using a press-fit. The transfer device was connected to FID 1 by means of a 0.5 m × 0.1 mm I.D.
13
retention gap (Restek, Bellefonte, PA, USA). An inlet pressure of 450 kPa (He) was applied to the transfer device by means of an APC unit. Injection mode GC 1 (250 ◦ C): split (100:1). Injection volume: 0.5 L. Carrier gas: H2 . Constant inlet pressure: 600 kPa. Initial linear velocity column 1: 50 cm/s. Initial linear velocity column 2: 131 cm/s. Temperature program GC 1: 50 ◦ C to 280 ◦ C at 10 ◦ C/min. Temperature program ◦ GC 2: 45 ◦ C (2.5 min) to 180 C (8 min) at 10 ◦ C/min. FID 1 (280 ◦ C): H2 flow: 20.0 mL/min; air flow: 400.0 mL/min; make up (He): 40.0 mL/min. Sampling rate: 20 ms. FID 2 (210 ◦ C): H2 flow: 50.0 mL/min; air flow: 400.0 mL/min; make up (He): 30.0 mL/min. Sampling rate: 20 ms. 3. Results and discussion As seen previously, the MDGC set-up is characterized by the presence of two separate carrier gas inlets connected to the injection system and, via APC, to the transfer system. It is obvious that the influence of each inlet pressure, in terms of column flow rates and interface pressure variations, would require a considerable expenditure of time to define. As such, a dedicated user-friendly software was employed for these calculations during MDGC method optimization in all applications (see Section 2.2). A conventional MDGC method was developed and applied to the analysis of rosemary essential oil. This specific matrix is particularly suited for the testing of a bidimensional GC system as it contains several chiral components [17]. Helium was used as carrier gas (at constant pressure) with an initial linear velocity of 36 cm/s applied in the first dimension and 86 cm/s in the secondary. The primary column flow rate may be considered as optimum, while the chiral column value is rather high. In general, it may be affirmed that the MDGC system configuration enables rather good (but not optimum) flow rates if both columns are considered. Prior to “heart-cutting”, a “stand-by” analysis was carried out in order to define the retention time cutting windows of the chiral components to be re-analyzed in the second dimension. A 12.5 min chromatographic expansion (the entire analysis time was 34 min) relative to this application is illustrated in Fig. 2 (upper chromatogram). As it can be seen, the retention times windows of 8 peaks corresponding to 9 chiral pairs [(+/−)-sabinene and (+/−)--pinene co-elute] were defined. Peak identification (in all applications) was carried out by using pure standard components and GC–MS data previously obtained on a rosemary essential oil sample [17]. The eight peaks were then cut and then transferred to the second dimension. It must be added that, in some cases, only a portion of the entire peak (i.e. limonene) was diverted onto the second column. It is well known that in quality control-aimed chiral applications on essential oils the most important parameters to be defined are enantiomeric ratios [13]. Consequently, the quantitative transfer of all optically active compounds is not necessary (except obviously for low amount components). In the case of highly concentrated components, the transfer of only small peak portions avoids second dimension column overloading. The second dimension chiral result for the eight transferred peaks is reported in a 26 min chromatographic expansion illustrated in Fig. 2 (lower chromatogram). It is evident that all the
L. Mondello et al. / J. Chromatogr. A 1105 (2006) 11–16
14
Fig. 2. Upper chromatogram: a 12.5 min chromatographic expansion relative to the conventional MDGC rosemary oil application with the transfer system in the stand-by position; peak identification: (1S,4R)-(−)-camphene and (1R,4S)-(+)-camphene (peak 1), (1S,5S)-(−)--pinene and (1R,5R)-(+)-pinene (peak 2), (1S,5S)-(−)-sabinene and (1R,5R)-(+)-sabinene (peak 3), (S)-(−)-limonene and (R)-(+)-limonene (peak 4), (1S,4S)-(−)-camphor and (1R,4R)-(+)-camphor (peak 5), (1R,2R,4R)-(−)-isoborneol and (1S,2S,4S)(+)-isoborneol (peak 6), (1S,2R,4S)-(−)-borneol and (1R,2S,4R)-(+)-borneol (peak 7), (R)-(−)-terpinen-4-ol and (S)-(+)-terpinen-4-ol (peak 8), (S)-(−)-␣terpineol and (R)-(+)-␣-terpineol (peak 9). Lower chromatogram: a 26 min second dimension chromatographic expansion relative to the conventional MDGC rosemary oil application. Peak numbering is the same as above.
ously in both dimensions. As a consequence, a 6 min isothermal stand-time (approximately retention time of the first cut) was added to the original chiral column temperature program (a method previously optimized in single column chiral analysis). At this point a fast MDGC method was developed and applied to the same matrix. A twin set of 0.1 mm I.D. micro-columns was installed; the configuration of the transfer system remained unaltered. The effectiveness of the micro-bore column approach in achieving fast high resolution GC has been several times reported [18,19]. The objective in the present investigation was to reproduce the conventional analytical result in a much shorter time. Hydrogen was employed as mobile phase with an initial linear velocity of 50 cm/s applied in the pre-micro-bore column and 131 cm/s in the analytical micro-bore column. The first dimension flow rate may be considered as near-to-optimum, while the chiral column value is more than double the ideal value. It must be added, though, that the use of micro-bore columns enables the application of higher-than-optimum mobile phase velocities with little loss in terms of column efficiency [18]. A “stand-by” rapid MDGC application is illustrated in a 4.5 min chromatographic expansion reported in the upper chromatogram illustrated in Fig. 3 (the entire analysis time was 9.3 min). Although all peaks of interest are well-separated, it must be noted that, with respect to the conventional application, there is a slight loss in terms of peak resolution and sensitivity. The former is probably due to the better operational conditions (in terms of flow rates) in the primary conventional column,
18 enantiomers are well-resolved apart from (+)-sabinene which partially overlaps with an unidentified high quantity compound. Nevertheless, peak integration was possible for this enantiomer. The enantiomeric ratios relative to the 18 components are listed in Table 1. The overall analytical run-time requested (elution time of the last component of interest on the second column) was 43 min. It must be added that data acquisition started simultaneTable 1 Percentage amounts of the most abundant enantiomers (out of each pair) analyzed by using conventional and fast MDGC Enantiomer
%Amount (conventional MDGC)
%Amount (fast MDGC)
(−)-Camphene (−)--Pinene (−)-Sabinene (+)-Limonene (−)-Camphor (+)-Isoborneol (−)-Borneol (−)-Terpinen-4-ol (+)-␣-Terpineol
68.6 87.1 74.1 62.6 57.0 52.9 84.6 67.8 58.2
68.6 87.1 – 63.2 57.5 50.7 83.1 73.1 57.0
Fig. 3. Upper chromatogram: a 4.5 min chromatographic expansion relative to the fast MDGC rosemary oil application with the transfer system in the standby position; peak identification: tricyclene (peak A); a-phellandrene (peak B); unknown (peak C); a-terpinolene (peak D); bornyl acetate (peak E). Lower chromatogram: a 5 min second dimension chromatographic expansion relative to the fast MDGC rosemary oil application. For numbered peaks refer to Fig. 2 legend.
L. Mondello et al. / J. Chromatogr. A 1105 (2006) 11–16
15
Table 2 Fast MDGC first dimension chromatogram: mean retention times (min) and peak areas, standard deviations, coefficients of variation and retention time shifts for five components (not transferred in the second column) in five consecutive MDGC analyses on rosemary essential oil Component
Mean retention time
s
CV%
Retention time shift (s)
Mean areas
s
CV%
Tricyclene (A) ␣-Phellandrene (B) Unknown (C) ␣-Terpinolene (D) Bornyl acetate (E)
1.535 2.282 2.895 3.239 5.826
0.001 0.001 0.000 0.001 0.001
0.04 0.02 0.00 0.02 0.01
−0.06 −0.24 −0.72 −0.24 −0.18
4161 5053 10662 6697 28424
10.88 123.61 50.72 47.07 148.13
0.26 2.47 0.47 0.70 0.52
while the latter is due to the introduction onto the primary fast column of a lower amount of sample (micro-bore columns are characterized by a low sample capacity). Resolution (RS ) for several peak pairs across the chromatogram was measured and compared to values derived from the conventional application. Although the RS values generally decreased in the rapid MDGC application, the observed reductions were never over 30%. The peaks defined by letters are components that were not be subjected to chiral analysis; they were selected in order to evaluate the degree of retention shift during multiple heart-cutting. It is obvious that in any multidimensional gas chromatography system this type of occurrence must be, at the most, neglectable. After the “stand-by” application, the eight peaks were cut and then transferred to the second dimension. Except for peak 6, only internal peak slices were diverted onto the analytical column. This, in order to avoid column overloading during the rapid chiral separation. Five consecutive MDGC analysis were carried out in order to evaluate the fast MDGC analytical reproducibility. The retention time reproducibility and the degree of peak shift were evaluated for the five untransferred components (Table 2). As it can be seen from the CV% values, the retention time reproducibility was excellent. The degree of peak shift was quite acceptable with a maximum value of −0.72 s observed after the first three cuts. The observed peak shift after the subsequent five cuts (peaks 4–8) was very low (−0.18 s measured at peak E). It must be emphasized that, although this negative
aspect had little or no effect on the final analytical performance, future research will be dedicated to the elimination of this minor problem. Peak area reproducibility was also evaluated; the correspondent analytical data is also reported in Table 2. As it can be observed, the CV% values are very low (a maximum value of 2.47), demonstrating a satisfactory peak area reproducibility under the fast operational conditions. The MDGC chiral result is illustrated in Fig. 3 (lower chromatogram). The final peak elutes at approximately 8.7 min, with a speed gain of nearly a factor of five if compared to the conventional application. The greater presence of impurities (clearly visible in the chromatogram) in this chiral analysis must be emphasized. This negative feature is connected to the slightly reduced peak resolution attained in the first dimension. As such, a higher amount of undesired analytes were diverted together with the chiral solutes onto the secondary capillary. In general, peak resolution decreases, while peak sensitivity is quite satisfactory (it is essentially maintained if compared to the conventional analysis). Both of these aspects are related to the high linear velocity in the chiral column, which reduces both column efficiency and solute band broadening (higher and narrower peaks are delivered to FID 2). Resolution, for several peak couples, was again calculated and compared to values derived from the conventional chromatogram. RS losses were much higher in the first 5 min part of the chiral chromatogram, if compared to later parts. This, as a direct consequence of the gradual reduc-
Table 3 Fast MDGC second dimension chromatogram: mean retention times (min) and peak areas, standard deviations and coefficients of variation for chiral components in five consecutive MDGC analyses on rosemary essential oil Component
Mean retention time
s
CV%
Mean areas
s
CV%
(−)-Camphene (+)-Camphene (+)--Pinene (−)--Pinene (−)-Limonene (+)-Limonene (+)-Camphor (−)-Camphor (−)-Isoborneol (+)-Isoborneol (−)-Borneol (+)-Borneol (−)-Terpinen-4-ol (+)-Terpinen-4-ol (−)-␣-Terpineol (+)-␣-Terpineol
4.099 4.280 4.459 4.533 5.755 5.927 6.813 6.879 7.253 7.335 7.464 7.553 8.048 8.091 8.538 8.670
0.000 0.000 0.001 0.001 0.001 0.000 0.001 0.001 0.001 0.000 0.001 0.000 0.000 0.001 0.000 0.001
0.00 0.00 0.02 0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.01 0.01 0.00 0.01 0.01 0.01
160452 73513 50095 339489 35923 61686 204753 277193 1305 1343 76553 15608 4847 1780 37544 49825
694.21 332.13 205.89 1449.85 706.70 750.88 1303.24 1847.07 94.57 130.05 572.38 113.58 176.39 80.65 201.47 263.80
0.43 0.45 0.41 0.43 1.97 1.22 0.64 0.67 7.25 9.68 0.75 0.73 3.64 4.48 0.54 0.53
16
L. Mondello et al. / J. Chromatogr. A 1105 (2006) 11–16
tion of the gas linear velocity throughout the analysis. It must be added, though, that all chiral pairs are sufficiently separated, with the exception of the sabinene enantiomers [(−)-sabinene overlaps completely with a high quantity unknown peak] and the isoborneol enantiomers which partially overlap with other low-amount impurities (peak integration was not impaired). The enantiomeric ratios relative to the 16 separated components can be derived from Table 1. As it can be seen from the reported values, there is generally a good agreement between the conventional and rapid applications. Retention time reproducibility was also evaluated in the second dimension analysis (Table 3). As it can be observed, the measured variations in terms of elution times are neglectable. Second dimension peak area reproducibilty was also measured (Table 3). Ten compounds presented a CV% value of under 1%; an additional four did not exceed the 5% mark. Only a single enantiomer pair, (−)-isoborneol and (+)-isoborneol, were characterized by a moderately high value of 7.25 and 9.68, respectively. It must be added, that these are trace-amount components and that slight peak area variations are inevitable. In general, the data reported in Table 3 may be considered as very good. It may be concluded that the more drastic analytical conditions applied in the rapid analysis did not have negative effects on the analytical performance. 4. Conclusions The present research has demonstrated the effectiveness and potential of a recently developed multidimensional gas chromatographic system. A rapid multi-sequential heart-cutting method was developed and successfully applied to the chiral analysis of rosemary essential oil. The overall analytical performance may be considered as very good. Furthermore, the instrumentation is easy to use and method optimization proved to be rather straightforward. Future investigations will be devoted to the improvement of the instrumental performance, to the
implementation of additional options such as cryo-trapping and to the employment of MS detection in conventional and fast MDGC applications. Further studies will be carried out on complex samples, in different fields, in order to define the specific advantages/disadvantages of this technique if compared to other valuable analytical tools (i.e. GC–MS, GC × GC). References [1] J.M. Davis, J.C. Giddings, Anal. Chem. 55 (1983) 418. [2] N. Ragunathan, K.A. Krock, C. Klawun, T.A. Sasaki, C.L. Wilkins, J. Chromatogr. A 856 (1999) 349. [3] S.A. Mjø, Anal. Chim. Acta 488 (2003) 231. [4] J. Dall¨uge, J. Beens, U.A.Th. Brinkman, J. Chromatogr. A 1000 (2003) 69. [5] L. Mondello, A. Casilli, P.Q. Tranchida, G. Dugo, P. Dugo, J. Chromatogr. A 1067 (2005) 235. [6] M. Harju, C. Danielsson, P. Haglund, J. Chromatogr. A 1019 (2003) 111. [7] J. Dall¨uge, L.L.P. Van Stee, X. Xu, J. Williams, J. Beens, R.J.J. Vreuls, J. Chromatogr. A 974 (2002) 169. [8] D.J. McEwen, Anal. Chem. 36 (1964) 279. [9] L. Mondello, A.C. Lewis, K.D. Bartle (Eds.), Multidimensional Chromatography, John Wiley & Sons, Chichester, England, 2002. [10] L. Mondello, M. Catalfamo, A.R. Proteggente, I. Bonaccorsi, G. Dugo, J. Agric. Food Chem. 46 (1998) 54. [11] D.R. Deans, Chromatographia 1 (1968) 18. [12] G. Schomburg, F. Weeke, F. M¨uller, M. Or´eans, Chromatographia 16 (1982) 87. [13] W. Bertsch, J. High Resolut. Chromatogr. 22 (1999) 647. [14] J. Schomburg, J. Chromatogr. A 703 (1995) 309. [15] Z. Liu, J.B. Phillips, J. Chromatogr. Sci. 29 (1991) 227. [16] B.M. Gordon, M.S. Uhrig, M.F. Borderding, H.L. Chung, W.M. Coleman III, J.F. Elder Jr., J.A. Giles, D.S. Moore, C.E. Rix, E.L. White, J. Chromatogr. Sci. 26 (1988) 174. [17] M. Lo Presti, S. Ragusa, A. Trozzi, P. Dugo, F. Visinoni, A. Fazio, G. Dugo, L. Mondello, J. Sep. Sci. 28 (2005) 273. [18] P. Sandra, C. Bicchi (Eds.), Capillary Gas Chromatography in Essential Oil Analysis Huethig, Heidelberg, 1987, p. 38. [19] A. van Es, High Speed Narrow Bore Capillary Gas Chromatography, Heidelberg, H¨uthig, 1992.
Fast enantiomeric analysis of a complex essential oil with an innovative multidimensional gas chromatographic system Luigi Mondello a,∗ , Alessandro Casilli a , Peter Quinto Tranchida a , Masanao Furukawa c , Kyoichi Komori c , Kozo Miseki c , Paola Dugo b , Giovanni Dugo a b
a Dipartimento Farmaco-chimico, Facolt` a di Farmacia, Universit`a degli Studi di Messina, viale Annunziata, 98168 Messina, Italy Dipartimento di Chimica Organica e Biologica, Facolt`a di Scienze MM.FF.NN., Universit`a di Messina, Salita Sperone, 98165 Messina, Italy c Shimadzu Corporation, 1 Nishinokyo-Kuwabaracho, Nakagyo-ku, Kyoto 604-8511, Japan
Available online 24 August 2005
Abstract The present research is focussed on the evaluation of a recently developed high performance multidimensional gas chromatographic (MDGC) system employed in the fast analysis of a series of chiral compounds contained in rosemary essential oil. The heart of the MDGC system consists in a simple transfer device for the rapid sequential re-injection of analyte “heart-cuts” from the first to the second dimension. The transfer system has no temperature restrictions, presents very low dead volumes and achieves multidimensional analysis through a pressure-balance mechanism. The MDGC set-up is characterized by two GC ovens (enabling independent temperature programming) and the possibility of mass spectrometric (MS) and/or flame ionization detection (FID). Multiple-cut conventional and fast MDGC–FID methods were developed and the results obtained compared, in order to evaluate the effectiveness of the system. In this respect, the rapid method provided the same analytical result in a greatly reduced time (approximately five times less). Furthermore, quali/quantitative data reproducibilty was very good. Fast MDGC was achieved by using micro-bore (0.1 mm I.D.) columns in both dimensions. © 2005 Elsevier B.V. All rights reserved. Keywords: Multidimensional gas chromatography; Rosemary essential oil; Chiral analysis; Transfer system; Enantiomeric ratio
1. Introduction It is well-known that single column GC is often an insufficient tool when the total separation of a complex matrix is required. Component co-elution is a common event even for low-complexity samples, as the distribution of chromatographic peaks is, generally, random [1]. In particular, major difficulties are encountered when trace-level analytes overlap with more concentrated solutes. As a consequence, the development of multidimensional instrumentation and methods is one of the foremost goals of modern-day gas chromatography. With this respect, two directions are generally followed for the attainment of more extensive and reliable analytical information: the combination of two independent GC columns (with different selectivities) or the linkage of a mass spectrometer to a gas chromatographic system.
∗
Corresponding author. Tel: +39 090 676 6536 fax: +39 090 676 6532. E-mail address: [email protected] (L. Mondello).
0021-9673/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.chroma.2005.08.013
GC–MS is an effective procedure not only for single peak identification and quantitation [2] but also for the unravelling of multi-component peaks through the use of deconvolution methods [2,3]. It is obvious, though, that MS library peak assignment is more reliable (especially for the more complex samples) when pure mass spectra relative to entirely resolved compounds are obtained. Total component separation is, therefore, always desirable, but often difficult to achieve. Comprehensive GC (GC × GC) is a remarkable chromatographic method which enables the bidimensional analysis of the entire initial sample. As such, it is characterized by a greatly enhanced peak capacity with respect to conventional multidimensional gas chromatographic (MDGC) (and to any other type of GC method) [4]. A series of outstanding separations on complex samples have been reported in the literature [5–7]. The introduction of a third mass spectrometric dimension in a comprehensive GC system may be considered the most powerful option for volatile analysis today available. Nevertheless, it may also be affirmed that GC × GC method optimization is not an easy task, especially for analysts who do not possess a great experience in this field. Furthermore, the attainment of com-
12
L. Mondello et al. / J. Chromatogr. A 1105 (2006) 11–16
prehensive GC quantitative data is dependent on commercial softwares which are sold with industrially-produced GC × GC instrumentation. An additional shortcoming is represented by the fact that the separation power of a short secondary column (generally 1 m), in some cases, may be insufficient for the resolution of all components contained in each single primary column slice. In many cases, a high number of theoretical plates is required for the thorough separation of specific fractions in the second dimension. Conventional “heart-cutting” multidimensional gas chromatography has been on the chromatographic scene for more than 40 years [8]. The analytical approach is based, essentially, on the transfer of selected solute bands from a first to a secondary column. The aim is to subject unresolved 1D compounds to a further independent separation step [9]. The second chromatographic process must not, in any way, invalidate component resolution previously achieved on the primary column. The entire procedure is enabled by the presence, between the two columns, of a specific transfer device. Column flows are diverted, basically, by exploiting two different principles: valve [10] or Deans type switching [11,12]. A thorough account of the most commonly employed interfaces is outside the scope of the present paper. Accurate descriptions and advantages/disadvantages of the most important MDGC interfaces have been reported in the literature [13,14]. It must be noted that the employment of MDGC has undergone quite a substantial decline with the vast availability of bench-top GC–MS systems and since the introduction of GC × GC in 1991 [15]. Nevertheless, heart-cutting MDGC may be considered, for some applications, the most appropriate analytical choice. In general, MDGC method optimization is much easier if compared to comprehensive GC; the combined column efficiencies, for a specific fraction, are higher with respect to GC × GC; data elaboration is simple; pure solute bands are usually delivered to the detection system (deconvolution is not necessary). Obviously, with this type of approach continuous heart-cutting over the entire sample is not possible as this would cause the loss of primary column peak separation [16]. With respect to the maximum number of cuts achievable in a single MDGC run, this is dependent on the sample-type and on the analytical conditions. The present research is based on the analytical evaluation of a new high performance MDGC system. Both conventional and rapid MDGC methods were developed and applied to the analysis of hydro-distilled rosemary essential oil. The degree of peak retention time shift during multiple-cut analysis and the analytical quali/quantitative reproducibility were determined in the more drastic of the two applications. To the authors knowledge, no previous publications on fast MDGC analysis with microbore columns have been reported.
The following pure standard components were purchased from Sigma–Aldrich (Milan, Italy): (+)-camphene; (−)camphene; (+)-sabinene; (+)--pinene; (−)--pinene; (+)limonene; (−)-limonene; (+)-camphor; (−)-camphor; (+/−)isoborneol; (+)-borneol; (−)-borneol; (+)-terpinen-4-ol; (+)-␣terpineol; (−)-␣-terpineol. The chiral compounds were diluted 1:10 (v/v) in n-hexane prior to analysis in all applications. 2.2. Instrumentation and chromatographic conditions The MDGC system consists of two Shimadzu GC-2010 gas chromatographs [from now onwards denominated as GC 1 (location of the primary column) and GC 2 (location of the secondary column], a Shimadzu GCMS-QP2010 quadrupole mass spectrometer (not employed in the present research) and a Shimadzu AOC-20i autoinjector (Shimadzu Corporation, Kyoto, Japan). GC 1 and GC 2 present a split/splitless injection system (the injection system in GC 2 was not used) and a flame ionization detector [from now onwards denominated FID 1 (situated in GC 1) and FID 2 (situated in GC 1)]. GC 1 and GC 2 are connected by means of a heated transfer line (200 ◦ C). A schematic of the transfer device, which is partly located in GC 1, is shown in Fig. 1a (stand-by position) and b (cut position). The interface consists of an advanced pressure control (APC) unit [and a pressure sensor (PS)] which supplies carrier gas, at constant
2. Experimental 2.1. Samples and sample preparation Hydro-distilled rosemary essential oil was purchased from Sigma–Aldrich (Milano, Italy). The sample was diluted 1:10 (v/v) in n-hexane prior to analysis in all applications.
Fig. 1. (a) A schematic of the tranfer system in stand-by position; (b) a schematic of the tranfer system in cut position (for abbreviations refer to text).
L. Mondello et al. / J. Chromatogr. A 1105 (2006) 11–16
pressure (defined as P), to a three-way solenoid valve (defined as V). The valve is linked to two stainless steel tube branches that are bridged by a fused silica restrictor (R1). The latter, positioned in the flow pathway, causes a pressure drop indicated as P. A second restrictor (R2), which is located within GC 1, acts as a connection between the first and second column. R2, as can be seen in the figure, is also subjected to flow rates that derive from the two stainless tube branches (P and P − P). When V is in the normally opened position the transfer system is in the “stand-by” status. In this situation, the pressure (supplied by the APC unit) at the entrance of the secondary column (P) is higher than the pressure (supplied by the APC unit) at the end of the primary column (P − P). As a consequence, the access of the first dimension gas flow to the second column is prevented and is, thus, directed solely to FID 1. The FID 2 flame is maintained by a gas flow that derives from the APC unit pressure P. On the contrary, when V is in the normally-closed position the transfer system is in the “cut” status. In this case, the pressure (supplied by the APC unit) at the entrance of the primary column (P) is higher than the pressure (supplied by the APC unit) at the head of the secondary column (P − P). As a consequence, the first dimension mobile phase flows freely to the second column and to FID 2. The FID 1 flame is maintained by a gas flow that derives from the APC unit pressure P. MDGC method calculations were carried out with the support of a MDGC Calculation software (Shimadzu, Kyoto, Japan). The MDGC system was controlled by a MDGC software (Shimadzu, Kyoto, Japan). Data were acquired by a GC solution software (Shimadzu, Kyoto, Japan). The experimental conditions in the conventional MDGC application were as follows: column 1 was a SE 52 (5% diphenyl + 95% polydimethylsiloxane) 25 m × 0.25 mm I.D. × 0.25 m; column 2 was a DEtTBuSililBETA-086 (diethyltert-butyl-silil--cyclodextrin) 25 m × 0.25 mm I.D. × 0.25 m (MEGA, Legnano, Italy). Column 1 was passed through the transfer line and connected to column 2 by using a press-fit. The transfer device was connected to FID 1 by means of a 0.25 m × 0.15 mm I.D. retention gap (Restek, Bellefonte, PA, USA). An inlet pressure of 250 kPa (He) was applied to the transfer device by means of an APC unit. Injection mode GC 1 (250 ◦ C): split (25:1). Injection volume: 1 L. Carrier gas: He. Constant inlet pressure: 345 kPa. Initial linear velocity column 1: 36 cm/s. Initial linear velocity column 2: 86 cm/s. Tempera◦ ture program GC 1: 50 ◦ C to 280 ◦ C at 3 C/min. Temperature ◦ ◦ program GC 2: 45 C (12 min) to 180 C at 2 ◦ C/min. FID 1 (280 ◦ C): H2 flow: 50.0 mL/min; air flow: 400.0 mL/min; make up: 0.0 mL/min. Sampling rate: 80 ms. FID 2 (210 ◦ C): H2 flow: 50.0 mL/min; air flow: 400.0 mL/min; make up: 50.0 mL/min. Sampling rate: 40 ms. The experimental conditions in the fast MDGC application were as follows: column 1 was a SE 52 (5% diphenyl + 95% polydimethylsiloxane) 10 m × 0.1 mm I.D. × 0.1 m; column 2 was a DEtTBuSililBETA-086 (diethyl-tert-butyl-silil-cyclodextrin) 10 m × 0.1 mm I.D. × 0.1 m (MEGA, Legnano, Italy). Column 1 was passed through the transfer line and connected to column 2 by using a press-fit. The transfer device was connected to FID 1 by means of a 0.5 m × 0.1 mm I.D.
13
retention gap (Restek, Bellefonte, PA, USA). An inlet pressure of 450 kPa (He) was applied to the transfer device by means of an APC unit. Injection mode GC 1 (250 ◦ C): split (100:1). Injection volume: 0.5 L. Carrier gas: H2 . Constant inlet pressure: 600 kPa. Initial linear velocity column 1: 50 cm/s. Initial linear velocity column 2: 131 cm/s. Temperature program GC 1: 50 ◦ C to 280 ◦ C at 10 ◦ C/min. Temperature program ◦ GC 2: 45 ◦ C (2.5 min) to 180 C (8 min) at 10 ◦ C/min. FID 1 (280 ◦ C): H2 flow: 20.0 mL/min; air flow: 400.0 mL/min; make up (He): 40.0 mL/min. Sampling rate: 20 ms. FID 2 (210 ◦ C): H2 flow: 50.0 mL/min; air flow: 400.0 mL/min; make up (He): 30.0 mL/min. Sampling rate: 20 ms. 3. Results and discussion As seen previously, the MDGC set-up is characterized by the presence of two separate carrier gas inlets connected to the injection system and, via APC, to the transfer system. It is obvious that the influence of each inlet pressure, in terms of column flow rates and interface pressure variations, would require a considerable expenditure of time to define. As such, a dedicated user-friendly software was employed for these calculations during MDGC method optimization in all applications (see Section 2.2). A conventional MDGC method was developed and applied to the analysis of rosemary essential oil. This specific matrix is particularly suited for the testing of a bidimensional GC system as it contains several chiral components [17]. Helium was used as carrier gas (at constant pressure) with an initial linear velocity of 36 cm/s applied in the first dimension and 86 cm/s in the secondary. The primary column flow rate may be considered as optimum, while the chiral column value is rather high. In general, it may be affirmed that the MDGC system configuration enables rather good (but not optimum) flow rates if both columns are considered. Prior to “heart-cutting”, a “stand-by” analysis was carried out in order to define the retention time cutting windows of the chiral components to be re-analyzed in the second dimension. A 12.5 min chromatographic expansion (the entire analysis time was 34 min) relative to this application is illustrated in Fig. 2 (upper chromatogram). As it can be seen, the retention times windows of 8 peaks corresponding to 9 chiral pairs [(+/−)-sabinene and (+/−)--pinene co-elute] were defined. Peak identification (in all applications) was carried out by using pure standard components and GC–MS data previously obtained on a rosemary essential oil sample [17]. The eight peaks were then cut and then transferred to the second dimension. It must be added that, in some cases, only a portion of the entire peak (i.e. limonene) was diverted onto the second column. It is well known that in quality control-aimed chiral applications on essential oils the most important parameters to be defined are enantiomeric ratios [13]. Consequently, the quantitative transfer of all optically active compounds is not necessary (except obviously for low amount components). In the case of highly concentrated components, the transfer of only small peak portions avoids second dimension column overloading. The second dimension chiral result for the eight transferred peaks is reported in a 26 min chromatographic expansion illustrated in Fig. 2 (lower chromatogram). It is evident that all the
L. Mondello et al. / J. Chromatogr. A 1105 (2006) 11–16
14
Fig. 2. Upper chromatogram: a 12.5 min chromatographic expansion relative to the conventional MDGC rosemary oil application with the transfer system in the stand-by position; peak identification: (1S,4R)-(−)-camphene and (1R,4S)-(+)-camphene (peak 1), (1S,5S)-(−)--pinene and (1R,5R)-(+)-pinene (peak 2), (1S,5S)-(−)-sabinene and (1R,5R)-(+)-sabinene (peak 3), (S)-(−)-limonene and (R)-(+)-limonene (peak 4), (1S,4S)-(−)-camphor and (1R,4R)-(+)-camphor (peak 5), (1R,2R,4R)-(−)-isoborneol and (1S,2S,4S)(+)-isoborneol (peak 6), (1S,2R,4S)-(−)-borneol and (1R,2S,4R)-(+)-borneol (peak 7), (R)-(−)-terpinen-4-ol and (S)-(+)-terpinen-4-ol (peak 8), (S)-(−)-␣terpineol and (R)-(+)-␣-terpineol (peak 9). Lower chromatogram: a 26 min second dimension chromatographic expansion relative to the conventional MDGC rosemary oil application. Peak numbering is the same as above.
ously in both dimensions. As a consequence, a 6 min isothermal stand-time (approximately retention time of the first cut) was added to the original chiral column temperature program (a method previously optimized in single column chiral analysis). At this point a fast MDGC method was developed and applied to the same matrix. A twin set of 0.1 mm I.D. micro-columns was installed; the configuration of the transfer system remained unaltered. The effectiveness of the micro-bore column approach in achieving fast high resolution GC has been several times reported [18,19]. The objective in the present investigation was to reproduce the conventional analytical result in a much shorter time. Hydrogen was employed as mobile phase with an initial linear velocity of 50 cm/s applied in the pre-micro-bore column and 131 cm/s in the analytical micro-bore column. The first dimension flow rate may be considered as near-to-optimum, while the chiral column value is more than double the ideal value. It must be added, though, that the use of micro-bore columns enables the application of higher-than-optimum mobile phase velocities with little loss in terms of column efficiency [18]. A “stand-by” rapid MDGC application is illustrated in a 4.5 min chromatographic expansion reported in the upper chromatogram illustrated in Fig. 3 (the entire analysis time was 9.3 min). Although all peaks of interest are well-separated, it must be noted that, with respect to the conventional application, there is a slight loss in terms of peak resolution and sensitivity. The former is probably due to the better operational conditions (in terms of flow rates) in the primary conventional column,
18 enantiomers are well-resolved apart from (+)-sabinene which partially overlaps with an unidentified high quantity compound. Nevertheless, peak integration was possible for this enantiomer. The enantiomeric ratios relative to the 18 components are listed in Table 1. The overall analytical run-time requested (elution time of the last component of interest on the second column) was 43 min. It must be added that data acquisition started simultaneTable 1 Percentage amounts of the most abundant enantiomers (out of each pair) analyzed by using conventional and fast MDGC Enantiomer
%Amount (conventional MDGC)
%Amount (fast MDGC)
(−)-Camphene (−)--Pinene (−)-Sabinene (+)-Limonene (−)-Camphor (+)-Isoborneol (−)-Borneol (−)-Terpinen-4-ol (+)-␣-Terpineol
68.6 87.1 74.1 62.6 57.0 52.9 84.6 67.8 58.2
68.6 87.1 – 63.2 57.5 50.7 83.1 73.1 57.0
Fig. 3. Upper chromatogram: a 4.5 min chromatographic expansion relative to the fast MDGC rosemary oil application with the transfer system in the standby position; peak identification: tricyclene (peak A); a-phellandrene (peak B); unknown (peak C); a-terpinolene (peak D); bornyl acetate (peak E). Lower chromatogram: a 5 min second dimension chromatographic expansion relative to the fast MDGC rosemary oil application. For numbered peaks refer to Fig. 2 legend.
L. Mondello et al. / J. Chromatogr. A 1105 (2006) 11–16
15
Table 2 Fast MDGC first dimension chromatogram: mean retention times (min) and peak areas, standard deviations, coefficients of variation and retention time shifts for five components (not transferred in the second column) in five consecutive MDGC analyses on rosemary essential oil Component
Mean retention time
s
CV%
Retention time shift (s)
Mean areas
s
CV%
Tricyclene (A) ␣-Phellandrene (B) Unknown (C) ␣-Terpinolene (D) Bornyl acetate (E)
1.535 2.282 2.895 3.239 5.826
0.001 0.001 0.000 0.001 0.001
0.04 0.02 0.00 0.02 0.01
−0.06 −0.24 −0.72 −0.24 −0.18
4161 5053 10662 6697 28424
10.88 123.61 50.72 47.07 148.13
0.26 2.47 0.47 0.70 0.52
while the latter is due to the introduction onto the primary fast column of a lower amount of sample (micro-bore columns are characterized by a low sample capacity). Resolution (RS ) for several peak pairs across the chromatogram was measured and compared to values derived from the conventional application. Although the RS values generally decreased in the rapid MDGC application, the observed reductions were never over 30%. The peaks defined by letters are components that were not be subjected to chiral analysis; they were selected in order to evaluate the degree of retention shift during multiple heart-cutting. It is obvious that in any multidimensional gas chromatography system this type of occurrence must be, at the most, neglectable. After the “stand-by” application, the eight peaks were cut and then transferred to the second dimension. Except for peak 6, only internal peak slices were diverted onto the analytical column. This, in order to avoid column overloading during the rapid chiral separation. Five consecutive MDGC analysis were carried out in order to evaluate the fast MDGC analytical reproducibility. The retention time reproducibility and the degree of peak shift were evaluated for the five untransferred components (Table 2). As it can be seen from the CV% values, the retention time reproducibility was excellent. The degree of peak shift was quite acceptable with a maximum value of −0.72 s observed after the first three cuts. The observed peak shift after the subsequent five cuts (peaks 4–8) was very low (−0.18 s measured at peak E). It must be emphasized that, although this negative
aspect had little or no effect on the final analytical performance, future research will be dedicated to the elimination of this minor problem. Peak area reproducibility was also evaluated; the correspondent analytical data is also reported in Table 2. As it can be observed, the CV% values are very low (a maximum value of 2.47), demonstrating a satisfactory peak area reproducibility under the fast operational conditions. The MDGC chiral result is illustrated in Fig. 3 (lower chromatogram). The final peak elutes at approximately 8.7 min, with a speed gain of nearly a factor of five if compared to the conventional application. The greater presence of impurities (clearly visible in the chromatogram) in this chiral analysis must be emphasized. This negative feature is connected to the slightly reduced peak resolution attained in the first dimension. As such, a higher amount of undesired analytes were diverted together with the chiral solutes onto the secondary capillary. In general, peak resolution decreases, while peak sensitivity is quite satisfactory (it is essentially maintained if compared to the conventional analysis). Both of these aspects are related to the high linear velocity in the chiral column, which reduces both column efficiency and solute band broadening (higher and narrower peaks are delivered to FID 2). Resolution, for several peak couples, was again calculated and compared to values derived from the conventional chromatogram. RS losses were much higher in the first 5 min part of the chiral chromatogram, if compared to later parts. This, as a direct consequence of the gradual reduc-
Table 3 Fast MDGC second dimension chromatogram: mean retention times (min) and peak areas, standard deviations and coefficients of variation for chiral components in five consecutive MDGC analyses on rosemary essential oil Component
Mean retention time
s
CV%
Mean areas
s
CV%
(−)-Camphene (+)-Camphene (+)--Pinene (−)--Pinene (−)-Limonene (+)-Limonene (+)-Camphor (−)-Camphor (−)-Isoborneol (+)-Isoborneol (−)-Borneol (+)-Borneol (−)-Terpinen-4-ol (+)-Terpinen-4-ol (−)-␣-Terpineol (+)-␣-Terpineol
4.099 4.280 4.459 4.533 5.755 5.927 6.813 6.879 7.253 7.335 7.464 7.553 8.048 8.091 8.538 8.670
0.000 0.000 0.001 0.001 0.001 0.000 0.001 0.001 0.001 0.000 0.001 0.000 0.000 0.001 0.000 0.001
0.00 0.00 0.02 0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.01 0.01 0.00 0.01 0.01 0.01
160452 73513 50095 339489 35923 61686 204753 277193 1305 1343 76553 15608 4847 1780 37544 49825
694.21 332.13 205.89 1449.85 706.70 750.88 1303.24 1847.07 94.57 130.05 572.38 113.58 176.39 80.65 201.47 263.80
0.43 0.45 0.41 0.43 1.97 1.22 0.64 0.67 7.25 9.68 0.75 0.73 3.64 4.48 0.54 0.53
16
L. Mondello et al. / J. Chromatogr. A 1105 (2006) 11–16
tion of the gas linear velocity throughout the analysis. It must be added, though, that all chiral pairs are sufficiently separated, with the exception of the sabinene enantiomers [(−)-sabinene overlaps completely with a high quantity unknown peak] and the isoborneol enantiomers which partially overlap with other low-amount impurities (peak integration was not impaired). The enantiomeric ratios relative to the 16 separated components can be derived from Table 1. As it can be seen from the reported values, there is generally a good agreement between the conventional and rapid applications. Retention time reproducibility was also evaluated in the second dimension analysis (Table 3). As it can be observed, the measured variations in terms of elution times are neglectable. Second dimension peak area reproducibilty was also measured (Table 3). Ten compounds presented a CV% value of under 1%; an additional four did not exceed the 5% mark. Only a single enantiomer pair, (−)-isoborneol and (+)-isoborneol, were characterized by a moderately high value of 7.25 and 9.68, respectively. It must be added, that these are trace-amount components and that slight peak area variations are inevitable. In general, the data reported in Table 3 may be considered as very good. It may be concluded that the more drastic analytical conditions applied in the rapid analysis did not have negative effects on the analytical performance. 4. Conclusions The present research has demonstrated the effectiveness and potential of a recently developed multidimensional gas chromatographic system. A rapid multi-sequential heart-cutting method was developed and successfully applied to the chiral analysis of rosemary essential oil. The overall analytical performance may be considered as very good. Furthermore, the instrumentation is easy to use and method optimization proved to be rather straightforward. Future investigations will be devoted to the improvement of the instrumental performance, to the
implementation of additional options such as cryo-trapping and to the employment of MS detection in conventional and fast MDGC applications. Further studies will be carried out on complex samples, in different fields, in order to define the specific advantages/disadvantages of this technique if compared to other valuable analytical tools (i.e. GC–MS, GC × GC). References [1] J.M. Davis, J.C. Giddings, Anal. Chem. 55 (1983) 418. [2] N. Ragunathan, K.A. Krock, C. Klawun, T.A. Sasaki, C.L. Wilkins, J. Chromatogr. A 856 (1999) 349. [3] S.A. Mjø, Anal. Chim. Acta 488 (2003) 231. [4] J. Dall¨uge, J. Beens, U.A.Th. Brinkman, J. Chromatogr. A 1000 (2003) 69. [5] L. Mondello, A. Casilli, P.Q. Tranchida, G. Dugo, P. Dugo, J. Chromatogr. A 1067 (2005) 235. [6] M. Harju, C. Danielsson, P. Haglund, J. Chromatogr. A 1019 (2003) 111. [7] J. Dall¨uge, L.L.P. Van Stee, X. Xu, J. Williams, J. Beens, R.J.J. Vreuls, J. Chromatogr. A 974 (2002) 169. [8] D.J. McEwen, Anal. Chem. 36 (1964) 279. [9] L. Mondello, A.C. Lewis, K.D. Bartle (Eds.), Multidimensional Chromatography, John Wiley & Sons, Chichester, England, 2002. [10] L. Mondello, M. Catalfamo, A.R. Proteggente, I. Bonaccorsi, G. Dugo, J. Agric. Food Chem. 46 (1998) 54. [11] D.R. Deans, Chromatographia 1 (1968) 18. [12] G. Schomburg, F. Weeke, F. M¨uller, M. Or´eans, Chromatographia 16 (1982) 87. [13] W. Bertsch, J. High Resolut. Chromatogr. 22 (1999) 647. [14] J. Schomburg, J. Chromatogr. A 703 (1995) 309. [15] Z. Liu, J.B. Phillips, J. Chromatogr. Sci. 29 (1991) 227. [16] B.M. Gordon, M.S. Uhrig, M.F. Borderding, H.L. Chung, W.M. Coleman III, J.F. Elder Jr., J.A. Giles, D.S. Moore, C.E. Rix, E.L. White, J. Chromatogr. Sci. 26 (1988) 174. [17] M. Lo Presti, S. Ragusa, A. Trozzi, P. Dugo, F. Visinoni, A. Fazio, G. Dugo, L. Mondello, J. Sep. Sci. 28 (2005) 273. [18] P. Sandra, C. Bicchi (Eds.), Capillary Gas Chromatography in Essential Oil Analysis Huethig, Heidelberg, 1987, p. 38. [19] A. van Es, High Speed Narrow Bore Capillary Gas Chromatography, Heidelberg, H¨uthig, 1992.