- Email: [email protected]
Use of greatly-reduced gas flows in flow-modulated comprehensive two-dimensional gas chromatography-mass spectrometry
Journal of Chromatography A, 1359 (2014) 271–276
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
Use of greatly-reduced gas flows in flow-modulated comprehensive two-dimensional gas chromatography-mass spectrometry夽 Peter Q. Tranchida a , Flavio A. Franchina a , Paola Dugo a,b,c , Luigi Mondello a,b,c,∗ a
Dipartimento di Scienze del Farmaco e dei Prodotti per la Salute, Università di Messina, viale Annunziata, 98168 Messina, Italy Centro Integrato di Ricerca (C.I.R.), Università Campus Bio-Medico, Via Álvaro del Portillo, 21–00128 Roma, Italy Chromaleont s.r.l. A start-up of the University of Messina, c/o Dipartimento di Scienze del Farmaco e Prodotti per la Salute, Università di Messina, viale Annunziata, 98168 Messina, Italy b c
a r t i c l e
i n f o
Article history: Received 2 June 2014 Received in revised form 15 July 2014 Accepted 17 July 2014 Available online 25 July 2014 Keywords: Comprehensive two-dimensional GC GC × GC Flow modulation Method optimization
a b s t r a c t The present research is specifically based on the use of greatly-reduced gas flows, in flow-modulator (FM) comprehensive two-dimensional gas chromatography systems. In particular, focus of the present research is directed to FM devices characterized by an accumulation stage, and a much briefer reinjection step. It has been widely accepted that the operation of such FM systems requires high gas flows (≥20 mL/min), to re-inject the gas-phase contents of sample (or accumulation) loops, onto the second column. On the contrary, it will be herein demonstrated that much lower gas flows (≈ 6–8 mL/min) can efficiently perform the modulation step of re-injection. The possibility of using such improved operational conditions is given simply by a fine optimization of the processes of accumulation and re-injection. The application of lower gas flows not only means that second-dimension separations are carried out under better analytical conditions, but, even more importantly, greatly reduces problems which arise when using mass spectrometry (i.e., sensitivity and instrumental pumping capacity). © 2014 Elsevier B.V. All rights reserved.
1. Introduction Comprehensive 2D GC (GC × GC) separations are performed on two columns in sequence, with a transfer system (modulator) located somewhere between them. The function of the modulator, which can be either flow or cryogenic, is to accumulate, and re-inject chromatographic bands from the first onto the second column. For more details on GC × GC the reader is directed to the literature [1,2]. In terms of modulation, a series of cryogenic systems have been developed, gradually conquering the GC × GC scene [3]. However, cryogenic systems are characterized by a main disadvantage, namely the high costs of purchase and operation. Consequently, the concept and development of flow modulators is of high interest. The most interesting flow-modulator (FM) device (in our opinion), was first described in in 2006 [4], and consisted of: three deactivated fused-silica columns, two microvolume T-unions and
夽 Presented at the 13th International Symposium on Hyphenated Techniques in Chromatography and Separation Technology, Bruges, Belgium, 29 - 31 January 2014. ∗ Corresponding author. Tel.: +39 090 6766536; fax: +39 090 358220. E-mail address: [email protected] (L. Mondello). http://dx.doi.org/10.1016/j.chroma.2014.07.054 0021-9673/© 2014 Elsevier B.V. All rights reserved.
a two-way solenoid valve (located outside the GC oven), connected to an auxiliary flow source (Fig. 1). The output ports of the solenoid valve were connected to the unions by using two fused-silica segments. One of the T-unions was linked to the first-dimension outlet, while the other directed the flow to the second column. A fused-silica segment acted as sample loop, and bridged the two unions. When the modulator was in the “accumulation” mode, the auxiliary flow (20 mL/min) was directed to the “second-dimension” union, and the primarycolumn effluent (17 L/s; 1.02 mL/min) flowed through the loop; it is noteworthy that the accumulation period (1.4 s) was lower than the time necessary for the effluent to reach the bottom union (about 1.5 s). When the solenoid valve was switched for a brief period (100 ms), the auxiliary flow flushed the content of the loop onto the head of the second column. The high flow exiting the modulator was split between two columns. It is noteworthy that Amirav worked on the development of an altogether similar GC × GC flowmodulation concept (in combination with mass spectrometry with supersonic molecular beams) in 2006, and submitted an application for a patent [5]. The work of Seeley et al. highlighted a main disadvantage, namely the generation of excessively-high flows in the second analytical column. In many following works, based on the use of such a model, high flows were used to flush the accumulation loop [6–8].
272
P.Q. Tranchida et al. / J. Chromatogr. A 1359 (2014) 271–276
Fig. 1. The single-loop, dual-stage FM device introduced by Seeley et al. in 2006 (Reproduced with permission from Am. Lab) [4].
The only commercial flow modulator (Agilent Technologies) is based on the modulator proposed by Seeley [9], and is characterized by an integrated configuration. The accumulation chamber is of fixed volume, and is located within a stainless steel plate. A series of investigations, focused on the use of the Agilent modulator, have been published, with all reporting the use of high secondary-column flows [10,11]. In FM GC × GC-MS experiments, using a wafer-type interface, Tranchida et al. found that sensitivity was much higher in stand-alone GC-MS applications [12]. In that research, the majority of the gas flow, exiting the modulator, was diverted to waste to avoid exceeding the pumping capacity of the MS system. In a recent work, Seeley and co-workers reported the use of a high-speed Deans switch, exploited as GC × GC modulator [13]. Even though reduced secondary column gas flows (i.e., 2 mL/min) were generated, the interface was characterized by a main disadvantage, namely a low duty cycle (i.e., 0.1). Such works highlight the need for high duty-cycle modulators, with gas flow requirements which are compatible with commonly-used MS systems. It will be herein demonstrated that such an objective is possible through fine optimization of the modulation conditions. 2. Materials and methods 2.1. Samples The C9 and C10 n-alkanes were kindly provided by Supelco (Milan, Italy), and diluted in n-hexane prior to injection. The fragrance was kindly provided by an industry (the name is not given for secrecy reasons), and was diluted 1:10 (v/v) using a 1:1 (v/v) mixture of n-hexane and ethyl acetate. 2.2. FM GC × GC-quadrupole MS analyses All FM comprehensive two-dimensional GC-quadrupole (quad) MS applications were carried out on a system consisting of two
independent Shimadzu GC2010 gas chromatographs (GC1 and GC2), and a QP-2010 Ultra quadrupole mass spectrometer (Kyoto, Japan). Data were acquired using the GCMS solution software (Shimadzu). Bidimensional chromatograms, in all applications, were generated by using the ChromSquare software v. 2.0 (Shimadzu Europe, Duisburg, Germany). The two GC ovens were linked through a heated transfer line. The primary GC (GC1) was equipped with an AOC-20i auto-injector and a split-splitless injector (310 ◦ C). The primary column (situated in GC1), an SLB-5ms [(silphenylene polymer, practically equivalent in polarity to poly(5% diphenyl/95% methylsiloxane)] 30 m × 0.25 mm ID × 0.25 m df , was connected to position 1 of the 7-port interface (SGE, Ringwood, Victoria, Australia), after passing through the heated transfer line. An SPB-50 8 m × 0.32 mm ID × 0.20 m capillary segment [poly(50% diphenyl/50% dimethyl siloxane] was connected to position 6 of the interface. First and second-dimension columns were kindly supplied by Supelco. Position 7 of the interface was blocked by using an adequate nut. An external 40.9 L loop (20 cm × 0.71 mm OD × 0.51 mm ID stainless steel tubing) was employed. GC1 and GC2 temperature programs (fragrance application): 80 ◦ C–310 ◦ C at 3 ◦ C/min. Initial He head pressure (constant linear velocity): 82.5 kPa. Initial auxiliary (APC: advanced pressure control) He pressure (constant linear velocity): 46 kPa. Injection volume: 2 L; split ratio: 1:20. Modulation period: 5.0 s (accumulation period 4.5 s/injection period 0.5 s). The experimental conditions for the C9 n-alkane experiment are reported in R&D. Quadrupole MS conditions: ionization mode: electron ionization (70 eV). Mass range: 45–360 m/z; scan frequency: 25 Hz. Mass spectral database matching was carried out by using the FFNSC 2.0 database (Shimadzu), equipped with linear retention index (LRI) information.
P.Q. Tranchida et al. / J. Chromatogr. A 1359 (2014) 271–276
273
Fig. 2. Scheme reporting an efficient FM process: (A) accumulation step; (B) re-injection step; (C) modulated C9 n-alkane.
2.3. GC-quad MS analyses The GC-quadMS fragrance application was carried out on a system consisting of a Shimadzu GC2010 gas chromatograph, and a QP-2010 Ultra quadrupole mass spectrometer. Data were acquired using the GCMS solution software. The GC was equipped with an AOC-20i auto-injector and a split-splitless injector (310 ◦ C). The column employed was that used as primary column in the comprehensive 2D GC-MS applications. GC temperature program: 80 ◦ C–310 ◦ C at 3 ◦ C/min. Initial He head pressure (constant linear velocity = 30 cm/s): 34.3 kPa. Injection volume: 2 L; split ratio: 1:20. Quadrupole MS conditions: ionization mode: electron ionization (70 eV). Mass range: 45-360 m/z; scan frequency: 2.5 Hz. Mass spectral library matching was carried out by using the FFNSC 2.0 database. 3. Results and discussion 3.1. FM GC × GC-quadMS: use of greatly-reduced gas flows Finely-tuned optimization is of the highest importance in flowmodulated experiments based on the use of an accumulation loop. Specifically, attention must be devoted to the linear velocity of the isolated chromatography band inside the loop during the accumulation period, to avoid breakthrough. Breakthrough occurs when an isolated chromatography band reaches the loop outlet, prior to the re-injection pulse. Another important and simple concept is that an FM re-injection period of 100 ms, with a flow of 24 mL/min, will be equivalent to that of a re-injection period of 400 ms, with a flow of 6 mL/min. A fundamental aspect is that the length of the re-injection pulse must not be too long, otherwise the primary-column stop-flow conditions will not be maintained. The main consequence of such an occurrence will be that first-dimension effluent will start filling the loop before the end of the re-injection pulse.
In the present research, a 7-port wafer chip, described in previous work, was employed [7]. The seventh port was closed because its function, namely to divert part of the flow exiting the modulator to waste, was of no use in the present investigation. The modulation loop dimensions were 20 cm × 0.51 mm ID, corresponding to a volume of 40.9 L. A case of efficient flow modulation will now be described and shown (Fig. 2), using C9 linear alkane as test analyte. In the scheme, the first (0.25 mm ID) and second (0.32 mm ID) column, and loop, are shown with the same apparent width. For reasons of clarity, it will be herein assumed that the analyte band (or heart cut) within the loop is not characterized by a concentration gradient. However, such an assumption is not correct, because the analyte concentration within the loop is directly related to that of the same compound at the outlet of the first column [14]. The isothermal GC temperature (in both ovens) was 120 ◦ C, while the injector gauge pressure (pinj ) was 84.1 kPa and the auxiliary gauge pressure (paux ) was 40 kPa. The first-dimension + loop flow rate, during accumulation, was approx. 0.44 mL/min or 7.33 L/s. The average linear velocity within the loop was calculated to be 3.4 cm/s. If an accumulation period of 4.5 s is applied, then at the end of the accumulation step the isolated chromatography band will occupy a loop length of 15.3 cm, or a 31.3 L volume (Fig. 2A). During the passage from the first dimension to the loop, the length of the analyte band will decrease due to the differences in ID. The same concept, namely an increase in the length of the analyte band, will occur during the passage from the loop to the second column. The extent to which such length variations occur is not within the scope of the present work. During the re-injection step, the loop + second-dimension flow rate, was approx. 6.1 mL/min (or 101.33 L/s), corresponding to a loop linear velocity of 46.8 cm/s, and a second-dimension one of about 180 cm/s. A re-injection step of 500 ms was applied, which was sufficient to push the isolated band for a length of 23.4 cm, hence entirely out of the loop (Fig. 2B). The result of such modulation conditions, for
274
P.Q. Tranchida et al. / J. Chromatogr. A 1359 (2014) 271–276
Table 1 Peak identification for the compounds indicated in Fig. 3a-b; peak widths (4) and s/n values in the GC × GC-MS experiment; s/n values in the GC-MS application; s/n factor difference (s/n diff.) in the GC × GC-MS application. n◦
ID
4 (ms) GC × GC
s/n GC × GC
s/n GC
s/n diff.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
␣-pinene Sabinene -pinene Limonene ␥-terpinene Dihydromyrcenol Linalool 6,6-dimethoxy-2,5,5-trimethyl-hex-2-ene Gardenol Linalyl acetate cis-2-tert-butyl-cyclohexanol acetate trans-2-tert-butyl-cyclohexanol acetate Calone Cashmeran Cedrol trans-methyl-dihydrojasmonate ␥-Iso E Super Amberlyn Isopropyl-tetradecanoate Galaxolide Ethylene brassylate p-methoxy-octyl-cinnamate
800 800 960 1120 960 1120 960 1120 960 1120 960 800 800 960 640 1600 960 960 1080 1600 1440 960
18 13 65 342 40 186 58 156 15 167 29 6 6 26 7 885 34 27 90 773 271 54
11 6 33 237 28 142 45 82 25 100 20 3 8 24 5 536 24 20 66 414 150 51
1.7 2.0 2.0 1.4 1.4 1.3 1.3 1.9 0.6 1.7 1.4 1.8 0.7 1.1 1.6 1.7 1.4 1.3 1.4 1.9 1.8 1.1
C9 n-alkane, is shown in Fig. 2C. The base peak width (4), for the highest modulated peak, was 780 ms. It is clear that if the FM loop-optimization conditions are not finely tuned, then poorly-shaped modulated peaks will appear. A typical example, among others, consists in breakthrough: if the accumulation loop gas velocity is too high, then the isolated effluent band will occupy the entire loop length, plus a segment of second column, by the end of the accumulation period. The main consequence is that the analyte breakthrough will reach the detector in a steady stream, before the re-injected analyte band, generating peak fronting, or a gentle baseline rise prior to the main modulation signal. 3.2. A real world application An FM GC × GC-quadMS application was carried out on a commercial fragrance, under the conditions reported in Section 2.2. Obviously, the experiment was performed with efficient modulation conditions, using the 41-L loop. The (constant) average linear velocity (ALV) in the first dimension was circa 11 cm/s, corresponding to a flow at the beginning of the experiment of 0.44 mL/min; the ALV inside the loop corresponded to 2.9 cm/s, generating analyte bands of about 13 cm at the end of the accumulation period. The primary-column gas velocity was less than the ideal value (≈30 cm/s), and was related mainly to the necessity of a very low intra-loop gas velocity. Such a factor will enable the application of “normal” GC × GC modulation periods (e.g., 4–6 s), with no breakthrough. However, higher first-dimension gas flows can be used simply by reducing the accumulation period. The 500-ms re-injection step, at an ALV within the loop of 52.5 cm/s, pushed the solute bands for a length of 26.2 cm. The flow rate, through the “loop + second column” during re-injection, was calculated to be approximately 7.9 mL/min (at the beginning of the analysis). Separations in the second dimension were carried out at an AVL of approx. 200 cm/s. Two expansions, illustrating the FM GC × GC-quadMS fragrance application, are illustrated in Fig. 3a-b. The 22 compounds, numbered in Fig. 3, belong to different chemical classes and are herein considered to evaluate the analytical performance. It is noteworthy that peak tailing is evident for the more abundant compounds. Such an issue has been recently discussed by Griffith et al., who proposed an alternative flow-modulation approach (“reversed-flow”) to circumvent (in part) such a problem
[15]. Identification was performed through MS database matching, and the fact that it was only tentative is not important for the scope of the present investigation. In terms of peak widths (4), these ranged between 640 and 1600 ms, with an average value of 1031 ms (Table 1). Such peak width values were expected, due to the dimensions of the second column. It is highly presumable that more than one wrap-around occurred per peak. However, wraparound in many instances can be considered an advantage, because it enables the occupation of the dead-time space. The applied acquisition frequency (25 Hz) was more-thansufficient for proper peak re-construction. Signal-to-noise (s/n) values were also calculated (the most intense modulated peak was considered), and are reported in Table 1. At this point, with the scope of evaluating FM GC × GC-quadMS sensitivity, the primary column was connected directly to the MS ion source. The fragrance was analyzed under the same injection (sample volume and split ratio) and linear temperature program conditions. The applied He pressure generated an average linear velocity of 30 cm/s, corresponding to an initial flow of 0.66 mL/min. The applied MS spectral production frequency was 2.5 Hz, and enabled similar peak re-construction, in terms of number of data points per peak. Under such experimental conditions, it was found that sensitivity was generally lower than in the FM GC × GCquadMS application (Table 1). Specifically, for twenty compounds GC × GC-quadMS s/n values were increased by factors in the 1.1–2.0 range. Only for two analytes, namely gardenol and calone, the GC × GC-quadMS sensitivity was lower (by factors 0.6 and 0.7, respectively). The reasons for such a different behaviour can be related to the modulation phase and to the solute-stationary phase (second column) interactions (both analytes are in the polar part of the chromatogram). However, from the data reported in Table 1 it is possible to derive a general tendency (because modulation phase is random), inasmuch that for the 22 compounds the average sensitivity enhancement was by a factor of 1.5. Such a result was a great improvement compared to previous research, in which it was found that sensitivity, in stand-alone GC-MS applications, was found to be 3–4 times higher [12]. C9 and C10 alkanes were analyzed through GC-quadMS and FM GC × GC-quadMS analysis, under the conditions applied in the fragrance applications (one-dimensional traces are shown in Fig. 4). The GC × GC sensitivity enhancement factors were 1.7 and 1.9 for C9 and C10 , respectively. As can be observed, signal intensities
P.Q. Tranchida et al. / J. Chromatogr. A 1359 (2014) 271–276
Fig. 3. a-b Two expansions relative to a FM GC × GC-qMS analysis of a fragrance. For peak identification refer to Table 1.
275
276
P.Q. Tranchida et al. / J. Chromatogr. A 1359 (2014) 271–276
(x10,000,000) 4.0 3.0 2.0 1.0 0.0 4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
min
Fig. 4. GC-qMS and GC × GC-qMS traces relative to the temperature-programmed analysis of C9 and C10 alkanes. The first eluting peak corresponds to C9 (the modulated trace was moved from its original position).
were much higher in the GC × GC experiment (C9 : 39,000,000 vs. 2,700,000; C10 : 47,000,000 vs. 3,100,000), due to the rapid seconddimension elution conditions. However, the use of a 25-Hz spectral production frequency generated much more background noise compared to the GC-quadMS experiment. 4. Conclusions A fine flow-modulation optimization process has been herein applied, with the main benefit consisting in the use of lower modulation flows, compared to those normally employed in past research. In fact, it has been demonstrated that FM GC × GCMS experiments can be efficiently performed using 6–8 mL/min second-dimension flows. Such flows were compatible with the quadMS system employed, eliminating the need to divert a substantial part of the flow to waste (or to another detector). Consequently, there was no loss in sensitivity due to second-dimension flow splitting. It can be anticipated that the proposed model will open new opportunities in the use of flow modulation in GC × GC-MS analysis. The results described were obtained on a flow modulator with an external accumulation loop; the only commercially-available flow modulator (Agilent Technologies) is based on such an approach, though with an internal accumulation chamber. Although the proposed model can probably be extended to that device, such an assumption would need to be confirmed experimentally. Acknowledgements The authors thank Shimadzu and Supelco Corporations for the continuous support. References [1] Z. Liu, J.B. Phillips, Comprehensive two-dimensional gas chromatography using an on-column thermal modulator interface, J. Chromatogr. Sci. 29 (1991) 227–231.
[2] H.J. Cortes, B. Winniford, J. Luong, M. Pursch, Comprehensive two dimensional gas chromatography review, J. Sep. Sci. 32 (2009) 883–904. [3] M. Edwards, A. Mostafa, T. Górecki, Modulation in comprehensive twodimensional gas chromatography: 20 years of innovation, Anal. Bioanal. Chem. 401 (2011) 2335–2349. [4] J.V. Seeley, N.J. Micyus, J.D. McCurry, S.K. Seeley, Comprehensive twodimensional gas chromatography with a simple fluidic modulator, Am. Lab. 38 (2006) 24–26. [5] A. Amirav, Pulsed flow modulation gas chromatography mass spectrometry with supersonic molecular beams method and apparatus, Israel Pat. Appl. 176724 (2006). [6] M. Poliak, M. Kochman, A. Amirav, Pulsed flow modulation comprehensive two-dimensional gas chromatography, J. Chromatogr. A 1186 (2008) 189–195. [7] P.Q. Tranchida, F.A. Franchina, P. Dugo, L. Mondello, A flow-modulated comprehensive gas chromatography–mass spectrometry method for the analysis of fatty acid profiles in marine and biological samples, J. Chromatogr. A 1255 (2012) 171–176. [8] J.V. Seeley, S.K. Seeley, E.K. Libby, Z.S. Breitbach, D.W. Armstrong, Comprehensive two-dimensional gas chromatography using a high-temperature phosphonium ionic liquid column, Anal. Bioanal. Chem. 390 (2008) 323–332. [9] D. Quimby, J.D. McCurry, W.M. Norman, Capillary flow technique for gas chromatography: reinvigorating a mature analytical discipline, LCGC The Peak (2007, (April)) 7–15. ˇ [10] J. Krupˇcík, R. Gorovenko, I. Spanik, P. Sandra, D.W. Armstrong, Flow-modulated comprehensive two-dimensional gas chromatography with simultaneous flame ionization and quadrupole mass spectrometric detection, J. Chromatogr. A 1280 (2013) 104–111. [11] Q. Gu, F. David, F. Lynen, K. Rumpel, G. Xu, P. De Vos, P. Sandra, Analysis of bacterial fatty acids by flow modulated comprehensive two-dimensional gas chromatography with parallel flame ionization detector/mass spectrometry, J. Chromatogr. A 1217 (2010) 4448–4453. [12] P.Q. Tranchida, F.A. Franchina, M. Zoccali, I. Bonaccorsi, F. Cacciola, L. Mondello, A direct sensitivity comparison between flow-modulated comprehensive 2D and 1D GC in untargeted and targeted MS-based experiments, J. Sep. Sci. 36 (2013) 2746–2752. [13] A. Ghosh, C.T. Bates, S.K. Seeley, J.V. Seeley, High speed Deans switch for low duty cycle comprehensive two-dimensional gas chromatography, J. Chromatogr. A 1291 (2013) 146–154. [14] P.Mc A. Harvey, R.A. Shellie, Factors affecting peak shape in comprehensive two-dimensional gas chromatography with non-focusing modulation, J. Chromatogr. A 1218 (2011) 3153–3158. [15] J.F. Griffith, W.L. Winniford, K. Sun, R. Edam, J.C. Luong, A reversed-flow differential flow modulator for comprehensive two-dimensional gas chromatography, J. Chromatogr. A 1226 (2012) 116–123.
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
Use of greatly-reduced gas flows in flow-modulated comprehensive two-dimensional gas chromatography-mass spectrometry夽 Peter Q. Tranchida a , Flavio A. Franchina a , Paola Dugo a,b,c , Luigi Mondello a,b,c,∗ a
Dipartimento di Scienze del Farmaco e dei Prodotti per la Salute, Università di Messina, viale Annunziata, 98168 Messina, Italy Centro Integrato di Ricerca (C.I.R.), Università Campus Bio-Medico, Via Álvaro del Portillo, 21–00128 Roma, Italy Chromaleont s.r.l. A start-up of the University of Messina, c/o Dipartimento di Scienze del Farmaco e Prodotti per la Salute, Università di Messina, viale Annunziata, 98168 Messina, Italy b c
a r t i c l e
i n f o
Article history: Received 2 June 2014 Received in revised form 15 July 2014 Accepted 17 July 2014 Available online 25 July 2014 Keywords: Comprehensive two-dimensional GC GC × GC Flow modulation Method optimization
a b s t r a c t The present research is specifically based on the use of greatly-reduced gas flows, in flow-modulator (FM) comprehensive two-dimensional gas chromatography systems. In particular, focus of the present research is directed to FM devices characterized by an accumulation stage, and a much briefer reinjection step. It has been widely accepted that the operation of such FM systems requires high gas flows (≥20 mL/min), to re-inject the gas-phase contents of sample (or accumulation) loops, onto the second column. On the contrary, it will be herein demonstrated that much lower gas flows (≈ 6–8 mL/min) can efficiently perform the modulation step of re-injection. The possibility of using such improved operational conditions is given simply by a fine optimization of the processes of accumulation and re-injection. The application of lower gas flows not only means that second-dimension separations are carried out under better analytical conditions, but, even more importantly, greatly reduces problems which arise when using mass spectrometry (i.e., sensitivity and instrumental pumping capacity). © 2014 Elsevier B.V. All rights reserved.
1. Introduction Comprehensive 2D GC (GC × GC) separations are performed on two columns in sequence, with a transfer system (modulator) located somewhere between them. The function of the modulator, which can be either flow or cryogenic, is to accumulate, and re-inject chromatographic bands from the first onto the second column. For more details on GC × GC the reader is directed to the literature [1,2]. In terms of modulation, a series of cryogenic systems have been developed, gradually conquering the GC × GC scene [3]. However, cryogenic systems are characterized by a main disadvantage, namely the high costs of purchase and operation. Consequently, the concept and development of flow modulators is of high interest. The most interesting flow-modulator (FM) device (in our opinion), was first described in in 2006 [4], and consisted of: three deactivated fused-silica columns, two microvolume T-unions and
夽 Presented at the 13th International Symposium on Hyphenated Techniques in Chromatography and Separation Technology, Bruges, Belgium, 29 - 31 January 2014. ∗ Corresponding author. Tel.: +39 090 6766536; fax: +39 090 358220. E-mail address: [email protected] (L. Mondello). http://dx.doi.org/10.1016/j.chroma.2014.07.054 0021-9673/© 2014 Elsevier B.V. All rights reserved.
a two-way solenoid valve (located outside the GC oven), connected to an auxiliary flow source (Fig. 1). The output ports of the solenoid valve were connected to the unions by using two fused-silica segments. One of the T-unions was linked to the first-dimension outlet, while the other directed the flow to the second column. A fused-silica segment acted as sample loop, and bridged the two unions. When the modulator was in the “accumulation” mode, the auxiliary flow (20 mL/min) was directed to the “second-dimension” union, and the primarycolumn effluent (17 L/s; 1.02 mL/min) flowed through the loop; it is noteworthy that the accumulation period (1.4 s) was lower than the time necessary for the effluent to reach the bottom union (about 1.5 s). When the solenoid valve was switched for a brief period (100 ms), the auxiliary flow flushed the content of the loop onto the head of the second column. The high flow exiting the modulator was split between two columns. It is noteworthy that Amirav worked on the development of an altogether similar GC × GC flowmodulation concept (in combination with mass spectrometry with supersonic molecular beams) in 2006, and submitted an application for a patent [5]. The work of Seeley et al. highlighted a main disadvantage, namely the generation of excessively-high flows in the second analytical column. In many following works, based on the use of such a model, high flows were used to flush the accumulation loop [6–8].
272
P.Q. Tranchida et al. / J. Chromatogr. A 1359 (2014) 271–276
Fig. 1. The single-loop, dual-stage FM device introduced by Seeley et al. in 2006 (Reproduced with permission from Am. Lab) [4].
The only commercial flow modulator (Agilent Technologies) is based on the modulator proposed by Seeley [9], and is characterized by an integrated configuration. The accumulation chamber is of fixed volume, and is located within a stainless steel plate. A series of investigations, focused on the use of the Agilent modulator, have been published, with all reporting the use of high secondary-column flows [10,11]. In FM GC × GC-MS experiments, using a wafer-type interface, Tranchida et al. found that sensitivity was much higher in stand-alone GC-MS applications [12]. In that research, the majority of the gas flow, exiting the modulator, was diverted to waste to avoid exceeding the pumping capacity of the MS system. In a recent work, Seeley and co-workers reported the use of a high-speed Deans switch, exploited as GC × GC modulator [13]. Even though reduced secondary column gas flows (i.e., 2 mL/min) were generated, the interface was characterized by a main disadvantage, namely a low duty cycle (i.e., 0.1). Such works highlight the need for high duty-cycle modulators, with gas flow requirements which are compatible with commonly-used MS systems. It will be herein demonstrated that such an objective is possible through fine optimization of the modulation conditions. 2. Materials and methods 2.1. Samples The C9 and C10 n-alkanes were kindly provided by Supelco (Milan, Italy), and diluted in n-hexane prior to injection. The fragrance was kindly provided by an industry (the name is not given for secrecy reasons), and was diluted 1:10 (v/v) using a 1:1 (v/v) mixture of n-hexane and ethyl acetate. 2.2. FM GC × GC-quadrupole MS analyses All FM comprehensive two-dimensional GC-quadrupole (quad) MS applications were carried out on a system consisting of two
independent Shimadzu GC2010 gas chromatographs (GC1 and GC2), and a QP-2010 Ultra quadrupole mass spectrometer (Kyoto, Japan). Data were acquired using the GCMS solution software (Shimadzu). Bidimensional chromatograms, in all applications, were generated by using the ChromSquare software v. 2.0 (Shimadzu Europe, Duisburg, Germany). The two GC ovens were linked through a heated transfer line. The primary GC (GC1) was equipped with an AOC-20i auto-injector and a split-splitless injector (310 ◦ C). The primary column (situated in GC1), an SLB-5ms [(silphenylene polymer, practically equivalent in polarity to poly(5% diphenyl/95% methylsiloxane)] 30 m × 0.25 mm ID × 0.25 m df , was connected to position 1 of the 7-port interface (SGE, Ringwood, Victoria, Australia), after passing through the heated transfer line. An SPB-50 8 m × 0.32 mm ID × 0.20 m capillary segment [poly(50% diphenyl/50% dimethyl siloxane] was connected to position 6 of the interface. First and second-dimension columns were kindly supplied by Supelco. Position 7 of the interface was blocked by using an adequate nut. An external 40.9 L loop (20 cm × 0.71 mm OD × 0.51 mm ID stainless steel tubing) was employed. GC1 and GC2 temperature programs (fragrance application): 80 ◦ C–310 ◦ C at 3 ◦ C/min. Initial He head pressure (constant linear velocity): 82.5 kPa. Initial auxiliary (APC: advanced pressure control) He pressure (constant linear velocity): 46 kPa. Injection volume: 2 L; split ratio: 1:20. Modulation period: 5.0 s (accumulation period 4.5 s/injection period 0.5 s). The experimental conditions for the C9 n-alkane experiment are reported in R&D. Quadrupole MS conditions: ionization mode: electron ionization (70 eV). Mass range: 45–360 m/z; scan frequency: 25 Hz. Mass spectral database matching was carried out by using the FFNSC 2.0 database (Shimadzu), equipped with linear retention index (LRI) information.
P.Q. Tranchida et al. / J. Chromatogr. A 1359 (2014) 271–276
273
Fig. 2. Scheme reporting an efficient FM process: (A) accumulation step; (B) re-injection step; (C) modulated C9 n-alkane.
2.3. GC-quad MS analyses The GC-quadMS fragrance application was carried out on a system consisting of a Shimadzu GC2010 gas chromatograph, and a QP-2010 Ultra quadrupole mass spectrometer. Data were acquired using the GCMS solution software. The GC was equipped with an AOC-20i auto-injector and a split-splitless injector (310 ◦ C). The column employed was that used as primary column in the comprehensive 2D GC-MS applications. GC temperature program: 80 ◦ C–310 ◦ C at 3 ◦ C/min. Initial He head pressure (constant linear velocity = 30 cm/s): 34.3 kPa. Injection volume: 2 L; split ratio: 1:20. Quadrupole MS conditions: ionization mode: electron ionization (70 eV). Mass range: 45-360 m/z; scan frequency: 2.5 Hz. Mass spectral library matching was carried out by using the FFNSC 2.0 database. 3. Results and discussion 3.1. FM GC × GC-quadMS: use of greatly-reduced gas flows Finely-tuned optimization is of the highest importance in flowmodulated experiments based on the use of an accumulation loop. Specifically, attention must be devoted to the linear velocity of the isolated chromatography band inside the loop during the accumulation period, to avoid breakthrough. Breakthrough occurs when an isolated chromatography band reaches the loop outlet, prior to the re-injection pulse. Another important and simple concept is that an FM re-injection period of 100 ms, with a flow of 24 mL/min, will be equivalent to that of a re-injection period of 400 ms, with a flow of 6 mL/min. A fundamental aspect is that the length of the re-injection pulse must not be too long, otherwise the primary-column stop-flow conditions will not be maintained. The main consequence of such an occurrence will be that first-dimension effluent will start filling the loop before the end of the re-injection pulse.
In the present research, a 7-port wafer chip, described in previous work, was employed [7]. The seventh port was closed because its function, namely to divert part of the flow exiting the modulator to waste, was of no use in the present investigation. The modulation loop dimensions were 20 cm × 0.51 mm ID, corresponding to a volume of 40.9 L. A case of efficient flow modulation will now be described and shown (Fig. 2), using C9 linear alkane as test analyte. In the scheme, the first (0.25 mm ID) and second (0.32 mm ID) column, and loop, are shown with the same apparent width. For reasons of clarity, it will be herein assumed that the analyte band (or heart cut) within the loop is not characterized by a concentration gradient. However, such an assumption is not correct, because the analyte concentration within the loop is directly related to that of the same compound at the outlet of the first column [14]. The isothermal GC temperature (in both ovens) was 120 ◦ C, while the injector gauge pressure (pinj ) was 84.1 kPa and the auxiliary gauge pressure (paux ) was 40 kPa. The first-dimension + loop flow rate, during accumulation, was approx. 0.44 mL/min or 7.33 L/s. The average linear velocity within the loop was calculated to be 3.4 cm/s. If an accumulation period of 4.5 s is applied, then at the end of the accumulation step the isolated chromatography band will occupy a loop length of 15.3 cm, or a 31.3 L volume (Fig. 2A). During the passage from the first dimension to the loop, the length of the analyte band will decrease due to the differences in ID. The same concept, namely an increase in the length of the analyte band, will occur during the passage from the loop to the second column. The extent to which such length variations occur is not within the scope of the present work. During the re-injection step, the loop + second-dimension flow rate, was approx. 6.1 mL/min (or 101.33 L/s), corresponding to a loop linear velocity of 46.8 cm/s, and a second-dimension one of about 180 cm/s. A re-injection step of 500 ms was applied, which was sufficient to push the isolated band for a length of 23.4 cm, hence entirely out of the loop (Fig. 2B). The result of such modulation conditions, for
274
P.Q. Tranchida et al. / J. Chromatogr. A 1359 (2014) 271–276
Table 1 Peak identification for the compounds indicated in Fig. 3a-b; peak widths (4) and s/n values in the GC × GC-MS experiment; s/n values in the GC-MS application; s/n factor difference (s/n diff.) in the GC × GC-MS application. n◦
ID
4 (ms) GC × GC
s/n GC × GC
s/n GC
s/n diff.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
␣-pinene Sabinene -pinene Limonene ␥-terpinene Dihydromyrcenol Linalool 6,6-dimethoxy-2,5,5-trimethyl-hex-2-ene Gardenol Linalyl acetate cis-2-tert-butyl-cyclohexanol acetate trans-2-tert-butyl-cyclohexanol acetate Calone Cashmeran Cedrol trans-methyl-dihydrojasmonate ␥-Iso E Super Amberlyn Isopropyl-tetradecanoate Galaxolide Ethylene brassylate p-methoxy-octyl-cinnamate
800 800 960 1120 960 1120 960 1120 960 1120 960 800 800 960 640 1600 960 960 1080 1600 1440 960
18 13 65 342 40 186 58 156 15 167 29 6 6 26 7 885 34 27 90 773 271 54
11 6 33 237 28 142 45 82 25 100 20 3 8 24 5 536 24 20 66 414 150 51
1.7 2.0 2.0 1.4 1.4 1.3 1.3 1.9 0.6 1.7 1.4 1.8 0.7 1.1 1.6 1.7 1.4 1.3 1.4 1.9 1.8 1.1
C9 n-alkane, is shown in Fig. 2C. The base peak width (4), for the highest modulated peak, was 780 ms. It is clear that if the FM loop-optimization conditions are not finely tuned, then poorly-shaped modulated peaks will appear. A typical example, among others, consists in breakthrough: if the accumulation loop gas velocity is too high, then the isolated effluent band will occupy the entire loop length, plus a segment of second column, by the end of the accumulation period. The main consequence is that the analyte breakthrough will reach the detector in a steady stream, before the re-injected analyte band, generating peak fronting, or a gentle baseline rise prior to the main modulation signal. 3.2. A real world application An FM GC × GC-quadMS application was carried out on a commercial fragrance, under the conditions reported in Section 2.2. Obviously, the experiment was performed with efficient modulation conditions, using the 41-L loop. The (constant) average linear velocity (ALV) in the first dimension was circa 11 cm/s, corresponding to a flow at the beginning of the experiment of 0.44 mL/min; the ALV inside the loop corresponded to 2.9 cm/s, generating analyte bands of about 13 cm at the end of the accumulation period. The primary-column gas velocity was less than the ideal value (≈30 cm/s), and was related mainly to the necessity of a very low intra-loop gas velocity. Such a factor will enable the application of “normal” GC × GC modulation periods (e.g., 4–6 s), with no breakthrough. However, higher first-dimension gas flows can be used simply by reducing the accumulation period. The 500-ms re-injection step, at an ALV within the loop of 52.5 cm/s, pushed the solute bands for a length of 26.2 cm. The flow rate, through the “loop + second column” during re-injection, was calculated to be approximately 7.9 mL/min (at the beginning of the analysis). Separations in the second dimension were carried out at an AVL of approx. 200 cm/s. Two expansions, illustrating the FM GC × GC-quadMS fragrance application, are illustrated in Fig. 3a-b. The 22 compounds, numbered in Fig. 3, belong to different chemical classes and are herein considered to evaluate the analytical performance. It is noteworthy that peak tailing is evident for the more abundant compounds. Such an issue has been recently discussed by Griffith et al., who proposed an alternative flow-modulation approach (“reversed-flow”) to circumvent (in part) such a problem
[15]. Identification was performed through MS database matching, and the fact that it was only tentative is not important for the scope of the present investigation. In terms of peak widths (4), these ranged between 640 and 1600 ms, with an average value of 1031 ms (Table 1). Such peak width values were expected, due to the dimensions of the second column. It is highly presumable that more than one wrap-around occurred per peak. However, wraparound in many instances can be considered an advantage, because it enables the occupation of the dead-time space. The applied acquisition frequency (25 Hz) was more-thansufficient for proper peak re-construction. Signal-to-noise (s/n) values were also calculated (the most intense modulated peak was considered), and are reported in Table 1. At this point, with the scope of evaluating FM GC × GC-quadMS sensitivity, the primary column was connected directly to the MS ion source. The fragrance was analyzed under the same injection (sample volume and split ratio) and linear temperature program conditions. The applied He pressure generated an average linear velocity of 30 cm/s, corresponding to an initial flow of 0.66 mL/min. The applied MS spectral production frequency was 2.5 Hz, and enabled similar peak re-construction, in terms of number of data points per peak. Under such experimental conditions, it was found that sensitivity was generally lower than in the FM GC × GCquadMS application (Table 1). Specifically, for twenty compounds GC × GC-quadMS s/n values were increased by factors in the 1.1–2.0 range. Only for two analytes, namely gardenol and calone, the GC × GC-quadMS sensitivity was lower (by factors 0.6 and 0.7, respectively). The reasons for such a different behaviour can be related to the modulation phase and to the solute-stationary phase (second column) interactions (both analytes are in the polar part of the chromatogram). However, from the data reported in Table 1 it is possible to derive a general tendency (because modulation phase is random), inasmuch that for the 22 compounds the average sensitivity enhancement was by a factor of 1.5. Such a result was a great improvement compared to previous research, in which it was found that sensitivity, in stand-alone GC-MS applications, was found to be 3–4 times higher [12]. C9 and C10 alkanes were analyzed through GC-quadMS and FM GC × GC-quadMS analysis, under the conditions applied in the fragrance applications (one-dimensional traces are shown in Fig. 4). The GC × GC sensitivity enhancement factors were 1.7 and 1.9 for C9 and C10 , respectively. As can be observed, signal intensities
P.Q. Tranchida et al. / J. Chromatogr. A 1359 (2014) 271–276
Fig. 3. a-b Two expansions relative to a FM GC × GC-qMS analysis of a fragrance. For peak identification refer to Table 1.
275
276
P.Q. Tranchida et al. / J. Chromatogr. A 1359 (2014) 271–276
(x10,000,000) 4.0 3.0 2.0 1.0 0.0 4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
min
Fig. 4. GC-qMS and GC × GC-qMS traces relative to the temperature-programmed analysis of C9 and C10 alkanes. The first eluting peak corresponds to C9 (the modulated trace was moved from its original position).
were much higher in the GC × GC experiment (C9 : 39,000,000 vs. 2,700,000; C10 : 47,000,000 vs. 3,100,000), due to the rapid seconddimension elution conditions. However, the use of a 25-Hz spectral production frequency generated much more background noise compared to the GC-quadMS experiment. 4. Conclusions A fine flow-modulation optimization process has been herein applied, with the main benefit consisting in the use of lower modulation flows, compared to those normally employed in past research. In fact, it has been demonstrated that FM GC × GCMS experiments can be efficiently performed using 6–8 mL/min second-dimension flows. Such flows were compatible with the quadMS system employed, eliminating the need to divert a substantial part of the flow to waste (or to another detector). Consequently, there was no loss in sensitivity due to second-dimension flow splitting. It can be anticipated that the proposed model will open new opportunities in the use of flow modulation in GC × GC-MS analysis. The results described were obtained on a flow modulator with an external accumulation loop; the only commercially-available flow modulator (Agilent Technologies) is based on such an approach, though with an internal accumulation chamber. Although the proposed model can probably be extended to that device, such an assumption would need to be confirmed experimentally. Acknowledgements The authors thank Shimadzu and Supelco Corporations for the continuous support. References [1] Z. Liu, J.B. Phillips, Comprehensive two-dimensional gas chromatography using an on-column thermal modulator interface, J. Chromatogr. Sci. 29 (1991) 227–231.
[2] H.J. Cortes, B. Winniford, J. Luong, M. Pursch, Comprehensive two dimensional gas chromatography review, J. Sep. Sci. 32 (2009) 883–904. [3] M. Edwards, A. Mostafa, T. Górecki, Modulation in comprehensive twodimensional gas chromatography: 20 years of innovation, Anal. Bioanal. Chem. 401 (2011) 2335–2349. [4] J.V. Seeley, N.J. Micyus, J.D. McCurry, S.K. Seeley, Comprehensive twodimensional gas chromatography with a simple fluidic modulator, Am. Lab. 38 (2006) 24–26. [5] A. Amirav, Pulsed flow modulation gas chromatography mass spectrometry with supersonic molecular beams method and apparatus, Israel Pat. Appl. 176724 (2006). [6] M. Poliak, M. Kochman, A. Amirav, Pulsed flow modulation comprehensive two-dimensional gas chromatography, J. Chromatogr. A 1186 (2008) 189–195. [7] P.Q. Tranchida, F.A. Franchina, P. Dugo, L. Mondello, A flow-modulated comprehensive gas chromatography–mass spectrometry method for the analysis of fatty acid profiles in marine and biological samples, J. Chromatogr. A 1255 (2012) 171–176. [8] J.V. Seeley, S.K. Seeley, E.K. Libby, Z.S. Breitbach, D.W. Armstrong, Comprehensive two-dimensional gas chromatography using a high-temperature phosphonium ionic liquid column, Anal. Bioanal. Chem. 390 (2008) 323–332. [9] D. Quimby, J.D. McCurry, W.M. Norman, Capillary flow technique for gas chromatography: reinvigorating a mature analytical discipline, LCGC The Peak (2007, (April)) 7–15. ˇ [10] J. Krupˇcík, R. Gorovenko, I. Spanik, P. Sandra, D.W. Armstrong, Flow-modulated comprehensive two-dimensional gas chromatography with simultaneous flame ionization and quadrupole mass spectrometric detection, J. Chromatogr. A 1280 (2013) 104–111. [11] Q. Gu, F. David, F. Lynen, K. Rumpel, G. Xu, P. De Vos, P. Sandra, Analysis of bacterial fatty acids by flow modulated comprehensive two-dimensional gas chromatography with parallel flame ionization detector/mass spectrometry, J. Chromatogr. A 1217 (2010) 4448–4453. [12] P.Q. Tranchida, F.A. Franchina, M. Zoccali, I. Bonaccorsi, F. Cacciola, L. Mondello, A direct sensitivity comparison between flow-modulated comprehensive 2D and 1D GC in untargeted and targeted MS-based experiments, J. Sep. Sci. 36 (2013) 2746–2752. [13] A. Ghosh, C.T. Bates, S.K. Seeley, J.V. Seeley, High speed Deans switch for low duty cycle comprehensive two-dimensional gas chromatography, J. Chromatogr. A 1291 (2013) 146–154. [14] P.Mc A. Harvey, R.A. Shellie, Factors affecting peak shape in comprehensive two-dimensional gas chromatography with non-focusing modulation, J. Chromatogr. A 1218 (2011) 3153–3158. [15] J.F. Griffith, W.L. Winniford, K. Sun, R. Edam, J.C. Luong, A reversed-flow differential flow modulator for comprehensive two-dimensional gas chromatography, J. Chromatogr. A 1226 (2012) 116–123.