ARTICLE pubs.acs.org/ac
Evaluation of a Medium-Polarity Ionic Liquid Stationary Phase in the Analysis of Flavor and Fragrance Compounds Carla Ragonese,† Danilo Sciarrone,† Peter Quinto Tranchida,† Paola Dugo,†,‡ Giovanni Dugo,† and Luigi Mondello*,†,‡ † ‡
Dipartimento Farmaco-Chimico, Facolta di Farmacia, Universita di Messina, Viale Annunziata, 98168Messina, Italy Universita Campus-Biomedico, Via Alvaro del Portillo, 21, 00128 Roma, Italy ABSTRACT: The present research is focused on the evaluation, in terms of efficiency and polarity, of a recently introduced gas chromatography (GC) column, coated with a 1,12-di(tripropylphosphonium) dodecane bis(trifluoromethansulfonyl) amide ionic-liquid stationary phase (SLB-IL59) and its application to the analysis of a complex essential oil. The ionic liquid column demonstrated very good efficiency, in terms of plate number, and a polarity close to that of the 100% poly(ethyleneglycol) stationary phase. In this preliminary evaluation, the SLB-IL59 30 m column was subjected to bleeding measurements, by means of conventional gas chromatography/ mass spectrometry (GC/MS) and, in addition, of comprehensive 2D GC. The SLB-IL59 column (30 m 0.20 μm df, 0.25 mm i.d.) was then evaluated in the analysis of typical essential oil constituents, in the form of pure standard compounds. Resolution toward several analytes was measured and the results were compared to those obtained with both apolar [silphenylene polymer, equivalent to poly(5% diphenyl/95% dimethylsiloxane)] and medium-polarity [100% poly(ethyleneglycol)] stationary phases, namely, the most common columns employed in the analysis of essential oils; peak symmetry, for different essential oil constituents, was also measured and expressed through tailing factors (at 10% of peak height). The final part of the investigation was devoted to the GC/MS analysis of lemon essential oil, with GC-flame ionization detection (FID) used for quantification. Linear retention indices of all the identified compounds were determined, and the data obtained were compared to those attained on the apolar and “wax” columns. The results obtained in the present investigation reveal the great potential of this novel stationary phase, as a medium-polarity alternative, in the analysis of essential oils.
E
ssential oils represent a valuable product widely employed in several fields such as the flavor, fragrance, and food industries. The volatile fraction of essential oils is usually rather complex and can be composed of hundreds of compounds, mainly monoterpenes, sesquiterpenes, and their oxygenated counterparts, which are separated with great difficulty in a single gas chromatography (GC) analysis. It is common to analyze essential oils on both apolar and polar columns for qualitative and quantitative purposes.1 Nonpolar columns are preferred in the identification of unknown compounds achieved with a mass spectrometer (MS) detector with the support of linear retention index (LRI) data. LRI values, which often present a fluctuating retention behavior on polar columns, are calculated using a reference series of n-alkanes. Polar columns, on the other hand, are often used in order to unravel coelutions that can possibly occur on apolar capillaries. The increasing demands derived from the food and flavor industries for natural and genuine products has increased the requirements of advanced analytical techniques; in this concern, different GC techniques have been exploited for the analysis of essential oils, such as enantio-GC/MS,2 fast GC methodologies, 36 classical multidimensional GC, 79 comprehensive 2D GC (GC GC),10 and GC-isotope ratio MS;11 with regards r 2011 American Chemical Society
to the liquid chromatography (LC) field, different approaches have been reported such as LCMS12 and comprehensive 2D LC (LC LC).13 Although advanced methods can expand the space for separation and/or enhance selectivity, it is also true that many such approaches are often characterized by a high operational complexity, requiring skillful analysts. In almost all separationscience situations, stationary phase selectivity is always a key point in conventional as well as in multidimensional chromatography. Recently, a new class of compounds has received increasing attention for their exploitation as GC stationary phases, namely, room temperature ionic liquids. Room temperature ionic liquids represent a class of organic nonmolecular solvents generally constituted of an organic cation containing N- or P- (i.e., alkyl imidazolium, phosphonium) counterbalanced by an anion of organic or inorganic nature and that are liquid at 20 °C.14 Ionic liquids have been widely employed in several fields of chemistry. Specific properties such as low volatility, high thermal stability, excellent selectivity toward specific chemical classes, and Received: August 2, 2011 Accepted: September 7, 2011 Published: September 08, 2011 7947
dx.doi.org/10.1021/ac202012u | Anal. Chem. 2011, 83, 7947–7954
Analytical Chemistry moreover good wetting abilities on the inner wall of fused silica capillaries make this class of compounds highly suitable as GC stationary phases.15,16 Indeed, IL columns have already been applied in a number of applications, such as polyaromatic hydrocarbons,17,18 chlorinated pesticides,17,19 essential oils,20 fatty acid methyl esters,2124 and flavors and fragrances21,25 as well as in classical multidimensional GC,26 GC GC,27,28 and comprehensive 3D GC systems.29 In the present research, a novel ionic-liquid (SLB-IL59) GC column [1,12-di(tripropylphosphonium)dodecane bis(trifluoromethansulfonyl)amide], which has recently become commercially available, was first evaluated in terms of efficiency, polarity, and bleeding measurement; in a second phase, the applicability of the IL column, in essential oil analysis, was evaluated by calculating the resolution and peak symmetry for selected pure standard compounds and the results were compared with those obtained using both an apolar [silphenylene polymer, equivalent to poly(5% diphenyl/95% dimethylsiloxane)] and a (mediumpolarity) 100% poly(ethyleneglycol) stationary phase. Finally, a real-world sample, namely, a lemon essential oil, was subjected to quali/quantitative analysis on the IL column and on the two wellestablished capillaries; peak identification was achieved by coinjection of pure standards and by means of a dedicated mass spectral database.
’ EXPERIMENTAL SECTION Samples and Sample Preparation. A naphthalene standard was used as the probe for the Golay plot construction; a 1000 ppm solution was prepared in n-hexane and stored at 4 °C prior to analysis. Benzene, n-butanol, 2-pentanone, nitropropane, and pyridine, used for the calculation of the McReynolds constants, as well as n-alkane standards (decane, undecane, dodecane, tridecane, and tetradecane) were injected neat by means of a 7 μm polydimethylsiloxane solid-phase microextraction (SPME) fiber (Supelco, Bellefonte, PA). Limonene, p-cymene, linalool, 1,8-cineole, (Z)-β-ocimene, (E)-caryophyllene, geranial, neral, geranyl and neryl acetate standards were used for resolution and peak symmetry evaluation; mixtures of standards (1000 ppm each) were prepared in n-hexane and stored at 4 °C prior to analysis. β-Pinene, camphene, myrcene, γ-terpinene, α-terpinene, α-phellandrene, terpinolene, octanal, linalool, nonanal, citronellal, terpinen-4-ol, (E)-β-farnesene, α-terpineol, camphor, carvone and α-bisabolol were used for identification; mixtures of standards (1000 ppm each) were prepared in n-hexane and stored at 4 °C prior to analysis. All the above cited pure standard compounds were kindly provided by Supelco-Sigma/Aldrich (Bellefonte, PA). Lemon essential oil was kindly supplied by Simone Gatto SpA (Messina, Italy). The sample was diluted in n-hexane (1:10) and stored at 4 °C prior to analysis. Instrumentation. All gas chromatography-flame ionization detection (GC-FID) analyses were carried out on a GC2010 (Shimadzu, Milan, Italy) gas chromatograph equipped with a splitsplitless injector, an AOC-20i autosampler, and a flame ionization detector. All data were collected by the GC Solution software (Shimadzu). All GC/MS analyses were carried out on a GCMS-QP2010 instrument (Shimadzu), equipped with a splitsplitless injector, an AOC-20i autosampler, and a quadrupole MS detector. All data were collected by the GCMS Solution software (Shimadzu). Mass spectral searching was performed using the FFNSC 1.3 database (Shimadzu).
ARTICLE
For bleeding measurements, under comprehensive 2D GC/MS conditions, applications were carried out on a Shimadzu GC GC/MS system consisting of one GC2010 gas chromatograph and a QP2010 Plus quadrupole mass spectrometer. The analytical column was connected, via an SGE SilTite mini-union (Ringwood, Victoria, Australia), to uncoated tubing (3 m 0.25 mm i.d.), used only in this specific bleeding application, to perform dual-stage loop-type modulation (the modulator is under license from Zoex Corporation, Houston, TX) and to transfer effluent to the mass spectrometer. Cryogenic modulation was applied every 6 s. Data were collected by the GCMS Solution software and elaborated using the Chromsquare 1.4 software (Shimadzu Europe, Duisburg, Germany). Operating Conditions. GC-FID analyses, for the construction of the Golay plot on the SLB-IL59 column, were carried out at an isothermal temperature (100 °C) using helium as the carrier gas (10100 cm/s). Injection volume and mode, 1.0 μL; split (100:1). GC-FID analyses for the determination of McReynolds constant were carried out at an isothermal temperature (120 °C) and at a helium linear velocity of 40 cm/s. Injection was performed by means of a 7 μm poly(dimethylsiloxane) solid phase microextraction (SPME) fiber in order to avoid coelution with the solvent. For the analyses of volatile compounds, the SLB-IL59 (1,12-di(tripropylphosphonium) dodecane bis(trifluoromethansulfonyl) amide) 30 m 0.20 μm d f, 0.25 mm i.d. column (Sigma-Aldrich/Supelco) was operated under programmedtemperature conditions: 50300 °C at 3 °C/min. Helium was used as the carrier gas (30 cm/s). Injection volume and mode, 1.0 μL; split (100:1). The Supelcowax 10 (100% polyethyleneglycol) 30 m 0.25 μm df, 0.25 mm i.d. column (Sigma-Aldrich/Supelco) was operated under programmed-temperature conditions: 50280 °C at 3 °C/min. Injection volume and mode, 1.0 μL; split (100:1). Helium was used as the carrier gas (30 cm/s). The SLB-5ms [silphenylene polymer, equivalent to poly(5% diphenyl/95% dimethylsiloxane)] 30 m 0.25 μm df, 0.25 mm i.d. column (Sigma-Aldrich/Supelco) was operated under programmed-temperature conditions: 50300 °C at 3 °C/min. Injection volume and mode, 1.0 μL; split (100:1). Helium was used as the carrier gas (30 cm/s). FID parameters in all applications: makeup gas was N2 at a flow rate of 50 mL/min; H2 flow rate was 50 mL/min; the air flow rate was 400 mL/min. FID sampling rate: 80 ms. MS parameters in all applications: mass range 40400 amu; scan speed, 2000 amu/s. Ionization mode, electron ionization (70 eV); ion source temperature, 200 °C; interface temperature, 250 °C.
’ RESULTS AND DISCUSSION Efficiency and Polarity Evaluation. The preliminary evaluation of efficiency on the 30 m SLB-IL59 (0.2 μm df, 0.25 mm i.d.) column demonstrated that this novel IL column possesses a minimum height equivalent to one theoretical plate (HETP, 0.255 mm) and total plate number (117 806) values in good agreement with that predicted by theory for a conventional capillary column; the coating efficiency (CE), obtained by calculating the ratio between the theoretical and experimental minimum HEPT values, equaled 84%. The phase polarity was calculated by means of McReynolds constants (ΔI). The latter represent the Kovats indices of 7948
dx.doi.org/10.1021/ac202012u |Anal. Chem. 2011, 83, 7947–7954
Analytical Chemistry
ARTICLE
Table 1. McReynolds Constant Values and Polarity Numbers Attained on the Evaluated Columns ΔI SLB-5
ΔI
ΔI SLB-
ΔI
ΔI SLB-
ms
Supelcowax
IL59
TCEP
IL100
benzene n-butanol
33 30
334 509
338 505
622 871
602 853
2-pentanone
55
375
549
772
884
nitropropane
91
601
649
1072
1017
pyridine
43
505
583
957
1081
P (total)
251
2324
2624
4294
4437
polarity number (PN)
6
52
59
97
100
probes
selected solutes, each describing a given interaction with the stationary phase. In particular, the McReynolds constant represents the difference between the Kovats index of a specific solute, determined at an isothermal temperature, and the Kovats index measured on a squalane stationary phase. The sum of the first five probes (P) measures the overall polarity of the phase.30 Furthermore, in order to classify GC capillary columns, a scale of polarity was employed in the present study. The following procedure was developed by our group and proposed and applied by Supelco for the designation of ionic liquid GC columns.31 The polarity of a given column, expressed as polarity number (PN) is calculated from PNx ¼ ðPx =PSLB-IL100 Þ 100 where Px represents the sum of the first five indices calculated on the given stationary phase, while PSLB-IL100 is the sum of the first five indices calculated on the SLB-IL100 column; the latter column is coated with an ionic liquid stationary phase, namely, dicationic 1,9-di(3-vinyl-imidazolium)nonane bis(trifluoromethyl) sulfonyl imidate, and possesses a polarity value higher than the 1,2,3-tris(2-cyanoetoxy)propane phase (TCEP) and with a P value was equal to 100. The McReynolds constants and PN values, obtained on the SLB-IL59 phase as well as on the capillary columns used for comparative purposes, are listed in Table 1. It must be noted that the PN data has been attained in previous research, is currently available for consultation,31 and herein reported for the benefit of the reader. According to the McReynolds classification, the SLB-IL59 phase showed a polarity much higher than the SLB-5ms one and comparable (slightly higher) to that of Supelcowax 10; hence, the SLB-IL59 phase can be defined as a medium or medium-high polarity one. Consequently, in the present study the IL column was utilized as a polar column in essential oil analysis; in particular, the analytical performance was evaluated in the analysis of genuine lemon oil and the results were compared to those obtained on the SLB-5ms and on the Supelcowax stationary phases. Bleed Evaluation. Since the aim of the present research is the evaluation of the IL stationary phase as a polar alternative, in the analysis of essential oils, bleeding measurements were considered of the greatest importance. Again, the results derived were compared to those attained on the “wax” phase. A considerable advantage of the IL phase, over the “wax” column, is the higher maximum operational temperature (300 °C vs 280 °C). Bleed evaluations were carried out by means of GC/MS by measuring the noise intensity at the higher operating temperature on both columns: a value of 3 105 pA was obtained on the IL column at 300 °C, while the noise intensity at 280 °C, on the
Figure 1. Elution order of typical essential oil terpenes on (1) SLB-5 ms, (2) Supelcowax 10, and (3) SLB-IL59. Peak assignment: (A) p-cymene, (B) limonene, (C) (Z)-β-ocimene and (D) 1,8-cineole.
other capillary, was 1.85 107 pA; such results have already been reported in previous studies.28 Lower bleeding results in better sensitivity, and more accurate quantitation, due to a better signalto-noise (S/N) ratio. The major source of noise in any chromatographic system is that related to chemical noise, mainly column bleed. From a GC/MS identification point of view, since column bleed is a source of nonsolute fragment ions, a reduced bleed will result in better spectral matching against a database of reference spectra. In an additional study, bleeding was evaluated using a GC GC/MS system for both the “wax” and IL phase. No seconddimension column was utilized in the GC GC/MS system but just the 30 m column connected to a 3 m uncoated capillary, the latter used for loop modulation and to transfer the column bleed to the ion source. Bleeding was measured calculating the area of a 5 min rectangle that included the noise signal at higher operating temperature of both columns: bleeding values of 1.8 108 pA s at 280 °C and 4.9 106 pA s at 300 °C were attained on the “wax” and IL phases, respectively. In both “bleed” experiments, the signal intensities measured on the IL stationary phase were 2 orders of magnitude lower. The higher signal intensities observed in the GC GC/MS experiments were due to cryogenic focusing during the modulation process. Separation of Specific Essential Oil Constituents. p-Cymene, 1,8-cineole, limonene and (Z)-β-ocimene were utilized to 7949
dx.doi.org/10.1021/ac202012u |Anal. Chem. 2011, 83, 7947–7954
Analytical Chemistry
ARTICLE
Table 2. Values of Resolution Calculated for Selected Pairs (See Figure 1) SLB-5 ms
Supelcowax 10
SLB-IL59
RsAB
2.1
28.0
29.8
RsBD RsBC
3.4 1.5
4.0 14.2
18.6 9.2
RsCD
1.8
9.9
11.3
RsAD
5.2
23.6
6.7
RsAC
3.5
14.5
21.3
evaluate the resolving power of both the apolar and the two polar columns. Such compounds were chosen because they are a challenging group to resolve using the apolar capillary; the latter is, as aforementioned, the most commonly used stationary phasetype in flavor and fragrance analysis. Figure 1 reports the chromatograms obtained for the separation of the standard components on the three columns; as can be easily observed, the two polar columns showed, as expected, higher selectivity toward this group of compounds with respect to the apolar column. In fact, it is well-known that these analytes coelute almost entirely and are difficult to quantify on an apolar column, while they are separated on a polar column, generally a 100% poly(ethyleneglycol) phase. As can be noticed, the elution order differs between the apolar and the polar capillaries, due obviously to the different solutestationary phase interactions. Additionally, even if the “wax” and IL columns possess a similar polarity, they showed a reversed elution order for (Z)-β-ocimene and 1, 8-cineole. For a question of simplicity, resolution values were calculated according to the elution order on the SLB-5ms phase, hence for the following couples: p-cymenelimonene (RsAB), limonene(Z)-β-ocimene (RsBC), (Z)-β-ocimene1,8-cineole (RsCD), limonene1,8-cineole (RsBD), p-cymene1,8-cineole (RsAD), and p-cymene(Z)-β-ocimene (RsAC). Table 2 lists the resolution values obtained, equal or lower than 2.1 (for vicinal compounds) on the apolar stationary phase; resolution values for the same pairs resulted in much higher values on the polar columns evaluated, as could be expected. The resolution value, for the externally located compounds (i.e., with the lowest and highest retention times), equaled 5.2 on the apolar (RsAD), 28.0 on the “wax” (RsAB), and 29.8 on the SLB-IL59 (RsAB) columns; hence, the higher resolution for the entire class of compounds was obtained on the ionic liquid stationary phase. A direct comparison between the 100% poly(ethyleneglycol) and the IL phases enabled the following conclusions: higher resolution between p-cymene and limonene (AB), and a slightly better separation between (Z)-β-ocimene and 1,8-cineole (CD) were achieved on the IL column. For the BC, BD, AC, and AD pairs, it has to be noted that the two polar columns showed a reversed elution order for 1,8-cineole (D) and (Z)-β-ocimene (C) resulting in a much higher resolution value between limonene and 1,8-cineole (BD), higher resolution between p-cymene and (Z)-β-ocimene (AC), slightly lower resolution between limonene and (Z)-β-ocimene (BC), and much lower resolution between p-cymene and 1,8-cineole (AD) on the ionic liquid phase. Limonene, linalool, (E)-caryophyllene, geranial, neral, and neryl acetate were chosen to evaluate peak symmetry and capacity factors for the Supelcowax 10 and SLB-IL59 columns (Figure 2). Each standard was chosen as representative for a specific class of compounds: limonene for monoterpene hydrocarbons,
Figure 2. Analysis of standard volatile compounds on the SLB-IL59 (A) and the Supelcowax 10 (B) columns. For peak assignment see Table 3.
Table 3. Tailing Factor (TF) and Capacity Factor Values Calculated on the SLB-IL59 and Supelcowax 10 Columns SLB-IL59 no.
Supelcowax 10
TF
k0
TF
k0
1
limonene
1.01
2.4
0.95
4.6
2
linalool
0.99
8.5
0.95
12.7
3
(E)-caryophyllene
1.01
10.5
0.91
13.6
4 5
neryl acetate neral
1.01 1.03
14.4 15.9
0.97 1.05
16.7 15.8
6
geranial
1.06
16.7
1.07
16.9
linalool for monoterpene alcohols, (E)-caryophyllene for sesquiterpene hydrocarbons, neral and geranial for monoterpenes aldehydes, and neryl acetate for monoterpene esters. Table 3 reports results showing that the tailing factors on the IL stationary phase ranged from 1.06 to 0.99: the highest values, namely, 1.06 and 1.03, were obtained for geranial and neral (monoterpene aldehydes), respectively, unveiling a slight degree of tailing. On the 100% poly(ethyleneglycol) stationary phase, the values ranged from 1.07 for geranial to 0.91 for (E)-caryophyllene, thus demonstrating a slight degree of tailing (stronger retention) for the monoterpene aldehyde and a slight degree of fronting (weaker retention) for the sesquiterpene hydrocarbon. With regards to capacity factors, the values listed in Table 3 show a faster elution on the SLB-IL59 column for the monoterpene hydrocarbons, monoterpene alcohols and sesquiterpene hydrocarbons, while both the monoterpene ester (neryl acetate) and aldehydes (neral and geranial) present comparable retention on both phases. Analysis of a Lemon Essential Oil. The final step of the present evaluation study was to test the IL phase in the analysis of a genuine essential oil. In particular, a lemon essential oil was 7950
dx.doi.org/10.1021/ac202012u |Anal. Chem. 2011, 83, 7947–7954
Analytical Chemistry
ARTICLE
Figure 3. Chromatogram relative to the GC/MS analysis of a lemon essential oil sample on the SLB-5ms column. For peak assignment see Table 4.
Figure 4. Chromatogram relative to the GC/MS analysis of a lemon essential oil sample on the SLB-IL59 column. For peak assignment see Table 4.
subjected to analysis on the apolar and the polar columns. Sicilian lemon essential oil is a valuable citrus oil, the complexity of which has been highlighted in recent research.32 Figures 3, 4, and 5 show the chromatograms relative to the GC/MS analyses of a lemon essential oil on the SLB-5ms, SLB-IL59, and Supelcowax 10 columns, respectively. The identification of 41 compounds, reported in Table 4, was achieved on the SLB-IL59 column by means of an MS database33 (see the Experimental Section) and by the coinjection of pure standard compounds. As can be seen, almost all the volatile constituents of the lemon essential oil were well-resolved on the SLB-IL59 phase while, as expected, the nonpolar column was characterized by a certain degree of coelution between p-cymene, limonene, 1,8-cineole, (Z)-β-ocimene, and β-phellandrene (peaks 15, 7, 14, 11, and 9). Such overlapping could possibly lead to incorrect identification and quantification of these components. With regards to the results obtained on the Supelcowax 10 column, it can be noticed
that 1,8-cineole still coelutes with limonene, a pair resolved only on the IL column. Besides some less important coelutions, it must be emphasized that, as can be seen in Figures 4 and 5, a higher degree of separation among neral, geranial, neryl and geranyl acetate (peaks 39, 37, 35 and 36) was attained on the IL phase. Furthermore, on the Supelcowax 10 column, neryl acetate (peak 35) overlaps completely with β-bisabolene (peak 30), while they are well separated on the ionic liquid column. It must be added that a partial coelution also occurred in the IL separation, namely, between transsabinene hydrate (a trace-amount compound) and linalool (peaks 17 and 18). However, the overall separation performance and, thus, selectivity of the IL column, in this application type, was better. The number of coelutions was certainly higher using the two well-established capillaries. For all the assigned peaks, IL linear retention indices were determined by using n-alkanes as the reference standard; the LRI values are listed in Table 4. 7951
dx.doi.org/10.1021/ac202012u |Anal. Chem. 2011, 83, 7947–7954
Analytical Chemistry
ARTICLE
Figure 5. Chromatogram relative to the GC/MS analysis of a lemon essential oil sample on the Supelcowax 10 column. For peak assignment see Table 4.
The availability of LRI information is a valid tool during the identification process of essential oil constituents; terpenoids, infact, produce very similar spectra after fragmentation; thus, LRI values can be used to discriminate between similar MS database matches. However, on polar columns the use of series of n-alkanes as reference standards is not recommended because of their low solubility in polar stationary phases. In general, the stability of LRI values, calculated using n-alkanes on a “wax” phase, is very low; sometimes, variations of up to 100 units can occur. In the present work, LRI values were calculated for the compounds identified on the ionic liquid column, and their variability was evaluated over a 6 month period (n = 30). The average LRI values and standard deviations are listed in Table 4, showing values never higher than (7.3 (limonene). It must be noted that the LRI values were calculated on the real-world sample; thus, in the case of limonene, the higher SD value could be justified by its high relative amount (∼70%) that generated a very broadened peak. The same experimental procedure was conducted on both the apolar and polar stationary phase, with standard deviations never over (5.9 on the former column, while a maximum SD value of (18.8 was obtained on the latter capillary; such results demonstrate that the stability of linear retention indices calculated on the ionic liquid stationary phase is comparable to that observed on the apolar phase. In the final step of the study, the lemon essential oil sample was subjected to GC-FID analysis in order to quantify the identified compounds. Quantitative analysis was carried out, using nonane as internal standard, by means of the following equation: weightx % ¼ ððAx =Ais Þ Cis RRFÞ=weight oil 100 where Ax is the peak area of the analyte; Ais is the peak area of the internal standard; Cis is the internal standard concentration (g/g) and RRF is the relative response factor, taking into account the response factors calculated for each chemical class. Analyte amounts are expressed as g/100 g of oil. The results in terms of concentration (g/100 g) and their relative standard deviations, calculated on three replicates, are reported in Table 5. It must be emphasized that, in the GC-FID analysis, under the same conditions of gas linear velocity (30 cm/s), resolution was
Table 4. Peak Identification, Experimental LRI Values, and Their (() Standard Deviations (SD) (n = 30) on the SLBIL59 Column peak
compound
LRIexp. SD peak
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
α-thujene α-pinene camphene β-pinene sabinene myrcene limonene α-terpinene (Z)-β-ocimene γ-terpinene β-phellandrene terpinolene (E)-β-ocimene 1,8-cineole p-cymene octanal trans-sabinene hydrate linalool nonanal (Z)-α-bergamotene (E)-α-bergamotene
992 1000 1038 1078 1088 1118 1141 1147 1173 1185 1190 1201 1208 1222 1247 1392 1486 1490 1509 1519 1540
0.6 0.7 0.6 1.9 1.1 0.8 7.3 4.7 1.6 4.2 2.1 1.4 1.1 3.2 0.9 2.5 2.6 2.2 2.4 1.6 2.2
22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41
compound cis-limonene oxide citronellal trans-limonene oxide terpinen-4-ol (E)-caryophyllene (E)-β-farnesene (Z)-α-bisabolene trans-β-bergamotene β-bisabolene (E)-α-bisabolene α-terpineol bicyclogermacrene camphor neryl acetate geranyl acetate neral carvone geranial spathulenol α-bisabolol
LRIexp. SD 1572 1583 1589 1589 1596 1623 1627 1633 1647 1680 1692 1715 1741 1811 1845 1889 1930 1940 2162 2220
2.7 2.6 2.9 2.5 3.6 4.4 4.4 7.1 6.2 3.6 3.0 1.1 2.9 3.6 3.6 4.4 5.4 4.8 3.8 3.7
generally improved with respect to the GC-MS experiment. Such a phenomenum is related to the different pressure conditions across the column; in fact, under lower pressure conditions (i.e., GC/MS), the optimum linear velocity is increased due to the higher values of the diffusion coefficients of the solute in the mobile phase. Consequently, while the He velocity used in the GC-FID application can be considered as ideal, the same cannot be affirmed for the GC/MS experiment. For example, in the case of limonene/p-cymene on the apolar stationary phase and neryl acetate/β-bisabolene on the “wax” one, a complete coelution was observed in the GC/MS analysis, while a slight degree of separation was attained in the GC-FID experiment; moreover, the trans-sabinene hydrate/linalool pair, which underwent partial overlapping in the IL GC/MS application, was baseline resolved in the GC-FID one (data not shown). 7952
dx.doi.org/10.1021/ac202012u |Anal. Chem. 2011, 83, 7947–7954
Analytical Chemistry
ARTICLE
Table 5. Quantitative Data (g/100 g of Lemon Essential Oil) Relative to the Apolar, “Wax”, and IL Columns and Relative Standard Deviations (n = 3)a . peak no
compound name
SLB-5ms
RSD %
Supelcowax 10
RSD %
SLB-IL59
RSD %
1
α-thujene
0.38
3.9
0.38
7.9
0.38
3.9
2
α-pinene
1.72
1.7
1.73
2.2
1.71
1.8
3 4
camphene β-pinene
0.05 10.96
8.2 0.9
0.05 11.04
6.6 4.5
0.05 10.97
6.3 1.1
5
sabinene
1.75
2.5
1.75
3.8
1.76
2.3
6
myrcene
1.48
3.0
1.51
4.4
1.49
3.0
7
limonene
67.72
1.4
67.19
1.3
67.14
1.3
8
α-terpinene
0.13
9.5
0.12
8.4
0.13
6.9
0.07
3.7
0.07
5.7
8.41
1.7
8.37
1.7
8.42
1.7
4.2 3.9
0.38 0.31
3.9 2.9
5.2
9
a
conc (g/100 g)
(Z)-β-ocimene
10
γ-terpinene
11 12
β-phellandrene terpinolene
0.31
3.2
0.37 0.31
13
(E)-β-ocimene
0.12
8.2
0.13
14
1,8-cineole
0.13
3.2
0.07
4.3 3.1
15
p-cymene
0.40
7.2
0.39
4.0
0.38
16
octanal
0.09
4.3
0.10
5.7
0.09
4.3
17
trans-sabinene hydrate
0.02
3.3
0.01
3.2
0.01
7.1
18
linalool
0.12
4.9
0.13
2.6
0.14
2.2
19 20
nonanal (Z)-α-bergamotene
0.11 0.03
3.8 7.7
0.09 0.03
6.4 5.6
0.11 0.03
2.8 6.5 3.2
21
(E)-α-bergamotene
0.37
4.0
0.37
5.6
0.38
22
cis-limonene oxide
0.01
1.2
0.01
3.6
0.02
5.6
23
citronellal
0.11
3.7
0.10
6.8
0.11
1.9
24
trans-limonene oxide
0.01
4.5
0.01
4.6
0.02
6.3
25
terpinen-4-ol
0.03
2.4
0.02
2.5
0.03
1.6
26
(E)-caryophyllene
0.19
4.7
0.20
4.9
0.19
4.2
27 28
(E)-β-farnesene (Z)-α-bisabolene
0.04 0.05
1.7 2.2
0.03 0.05
6.1 4.1
0.04 0.05
4.2 6.2 4.5
29
trans-β-bergamotene
0.02
3.2
0.02
4.3
0.02
30
β-bisabolene
0.58
3.4
0.64
6.2
0.58
2.6
31
(E)-α-bisabolene
0.01
3.0
0.01
3.3
32
α-terpineol
0.18
5.1
0.17
4.0
33
bicyclogermacrene
0.05
2.2
0.05
1.2
34
camphor
0.05
2.2
0.05
3.1
0.05
1.6
35 36
neryl acetate geranyl acetate
0.71 0.61
3.9 3.5
0.80 0.61
8.3 4.9
0.72 0.61
3.5 3.1
37
neral
0.88
3.1
0.85
3.5
38
carvone
0.04
4.7
39
geranial
1.51
3.0
1.48
40
spathulenol
0.02
6.7
0.02
41
α-bisabolol
0.03
7.4
0.03
0.17
5.5
0.89
2.7
0.05
4.3
5.4
1.51
3.0
4.6
0.02
3.3
6.9
0.03
3.2
Compounds are reported on the basis of the SLB-IL59 elution order.
From a quantitative point of view, incomplete separation can lead to incorrect analyte quantification. Infact, as shown in Table 5, with comparison of the results obtained on the different stationary phases, it is evident how quantitative data can be affected by the degree of coelution, such as in the case of neryl acetate and β-bisabolene: the values attained on the SLB-IL59 and SLB-5ms columns are in good agreement, while there is a relative error higher than 10% on the Supelcowax 10 column, leading to an overestimation of both neryl acetate (0.80% vs 0.72%) and
β-bisabolene (0.64% vs 0.58%). Furthermore, neither carvone nor bicyclogermacrene could be quantified on the Supelcowax 10 column as they both coelute with neryl acetate.
’ CONCLUSIONS The ionic liquid column evaluated in the present research has demonstrated to be a valuable tool for the achievement of an excellent separation performance, in the field of flavor and fragrance 7953
dx.doi.org/10.1021/ac202012u |Anal. Chem. 2011, 83, 7947–7954
Analytical Chemistry compounds. The results herein described have proven that the IL column possesses a degree of polarity comparable to that of the 100% polyethylenglycol stationary phase, combined with a higher thermal stability (300 °C vs 280 °C). Furthermore, the stability of the IL linear retention indices, obtained using n-alkanes, was very similar to that of the apolar stationary phase. Consequently, this novel medium-polarity ionic liquid stationary phase could provide a valid alternative to the well-established apolar column in the analysis of flavor and fragrance constituents, with the possibility to use LRI data as a discrimination factor in GC/ MS applications. Moreover, the overall performance of the IL phase, compared to the widely used “wax” one, has demonstrated to be superior in the analysis of essential oils, both in identification and quantification terms.
’ AUTHOR INFORMATION Corresponding Author
*Phone: +39-090-6766536. Fax +39-090-358220. E-mail:
[email protected].
’ ACKNOWLEDGMENT The Project was funded by the Italian Ministry for the University and Research (MUR) with a PNR 2005-2007 Project no. RBIP06SXMR “Sviluppo di metodologie innovative per l’analisi di prodotti agroalimentari”. The authors gratefully thank Shimadzu and Sigma-Aldrich/Supelco Corporations for their continuous support.
ARTICLE
(18) Anderson, J. L.; Armstrong, D. W. Anal. Chem. 2003, 75, 4851–4858. (19) Zapadlo, M.; Krupcík, J.; Kovalczuk, T.; Majek, P.; Spanik, I.; Armstrong, D. W.; Sandra, P. J. Chromatogr., A 2011, 1218, 746–751. (20) Qi, M.; Armstrong, D. W. Anal. Bioanal. Chem. 2007, 388, 889–899. (21) Payagala, T.; Zhang, Y.; Wanigasekara, E.; Huang, K.; Breitbach, Z. S.; Sharma, P. S.; Sidisky, L. M.; Armstrong, D. W. Anal. Chem. 2009, 81, 160–173. (22) Ragonese, C.; Tranchida, P. Q.; Dugo, P.; Dugo, G.; Sidisky, L. M.; Robillard, M. V.; Mondello, L. Anal. Chem. 2009, 81, 5561–5568. (23) Ragonese, C.; Tranchida, P. Q.; Sciarrone, D.; Mondello, L. J. Chromatogr., A 2009, 1216, 8992–8997. (24) Delmonte, P.; Kia, A.-R. F.; Kramer, J. K.G.; Mossoba, M. M.; Sidisky, L.; Rader, J. I. J. Chromatogr., A 2011, 1218, 545–554. (25) Huang, K.; Han, X.; Zhang, X.; Armstrong, D. W. Anal. Bioanal. Chem. 2007, 389, 2265–2275. (26) Sciarrone, D.; Tranchida, P. Q.; Ragonese, C.; Schipilliti, L.; Mondello, L. J. Sep. Sci. 2010, 33, 594–599. (27) Seeley, J. V.; Seeley, S. K.; Libby, E. K.; Breitbach, Z. S.; Armstrong, D. W. Anal. Bioanal. Chem. 2008, 390, 323–332. (28) Purcaro, G.; Tranchida, P. Q.; Ragonese, C.; Conte, L.; Dugo, P.; Dugo, G.; Mondello, L. Anal. Chem. 2010, 82, 8583–8590. (29) Siegler, W. C.; Crank, J. A.; Armstrong, D. W.; Synovec, R. E. J. Chromatogr., A 2010, 1217, 3144–3149. (30) McNair, H. M.; Miller, J. M. Basic Gas Chromatography, 1st ed.; Wiley Inter-Science: New York, 1998. (31) http://www.sigmaaldrich.com/etc/medialib/docs/Supelco/ Posters/1/ionic_liquid_gc_columns.Par.0001.File.tmp/ionic_liquid_ gc_columns.pdf (32) Dugo, P.; Ragonese, C.; Russo, M.; Sciarrone, D.; Santi, L.; Cotroneo, A.; Mondello, L. J. Sep. Sci. 2010, 33, 3374–3385. (33) Shimadzu, FFNSC 1.3 GC-MS library.
’ REFERENCES (1) Dugo, G.; Mondello, L. Citrus Oil: Composition, Advanced Analytical Techniques, Contaminants and Biological Activity, 1st ed.; Taylor & Francis Group, LLC, CRC Press: London, 2010. (2) Liberto, E.; Cagliero, C.; Sgorbini, B.; Bicchi, C.; Sciarrone, D.; D’Acampora Zellner, B.; Mondello, L.; Rubiolo, P. J. Chromatogr., A 2008, 1195, 117–126. (3) Mondello, L.; Casilli, A.; Tranchida, P. Q.; Cicero, L.; Dugo, P.; Dugo, G. J. Agric. Food Chem. 2003, 51 (19), 5602–5606. (4) Veriotti, T.; Sacks, R. Anal. Chem. 2002, 74, 5635–5640. (5) Mondello, L.; Sciarrone, D.; Casilli, A.; Tranchida, P. Q.; Dugo, P.; Dugo, G. J. Sep. Sci. 2007, 30, 1905–1911. (6) Tranchida, P. Q.; Mondello, M.; Sciarrone, D.; Dugo, P.; Dugo, G.; Mondello, L. J. Sep. Sci. 2008, 31 (14), 2634–2369. (7) Sciarrone, D.; Schipilliti, L.; Ragonese, C.; Tranchida, P. Q.; Dugo, P.; Dugo, G.; Mondello, L. J. Chromatogr., A 2010, 1217, 1101–1105. (8) Lo Presti, M.; Sciarrone, D.; Crupi, M. L.; Costa, R.; Ragusa, S.; Dugo, G.; Mondello, L. Flavour Fragrance J. 2008, 23, 249–257. (9) Sciarrone, D.; Ragonese, C.; Carnovale, C.; Piperno, A.; Dugo, P.; Dugo, G.; Mondello, L. J. Chromatogr., A 2010, 1217, 6422–6427. (10) Tranchida, P. Q.; Dugo, P.; Dugo, G.; Mondello, L. J. Chromatogr., A 2004, 1054, 3–16. (11) Schipilliti, L.; Tranchida, P. Q.; Sciarrone, D.; Russo, M.; Dugo, P.; Dugo, G.; Mondello, L. J. Sep. Sci. 2010, 33, 617–625. (12) Dugo, P.; Mondello, L.; Dugo, L.; Stancanelli, R.; Dugo, G. J. Pharm. Biomed. Anal. 2000, 24 (1), 147–154. (13) Francois, I.; de Villiers, A.; Tienpont, B.; David, F.; Sandra, P. J. Chromatogr., A 2008, 1178, 33–42. (14) Welton, T. Chem. Rev. 1999, 99, 2071–2083. (15) Armstrong, D. W.; He, L.; Liu, L.-S. Anal. Chem. 1999, 71, 3873–3876. (16) Dhanesar, S. C.; Coddens, M. E.; Poole, C. F. J. Chromatogr. 1985, 349, 249–265. (17) Anderson, J. L.; Armstrong, D. W. Anal. Chem. 2005, 77, 6453–6462. 7954
dx.doi.org/10.1021/ac202012u |Anal. Chem. 2011, 83, 7947–7954
