Journal of Chromatography A, 1325 (2014) 109–114
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
A “plug-and-use” approach towards facile fabrication of capillary columns for high performance nanoflow liquid chromatography Zhiliang Xiao, Lin Wang, Ya Liu, Qiuquan Wang, Bo Zhang ∗ Department of Chemistry and the MOE Key Laboratory of Spectrochemical Analysis and Instrumentation, College of Chemistry and Chemical Engineering, Xiamen University, Xiamen 361005, China
a r t i c l e
i n f o
Article history: Received 20 October 2013 Received in revised form 28 November 2013 Accepted 1 December 2013 Available online 11 December 2013 Keywords: Column technology Chromatography Proteomics High throughput screening Capillary column
a b s t r a c t Capillary columns used for nanoflow liquid chromatography play an important role in modern proteomics. High quality columns are needed to provide high peak capacity and highly reproducible separations. This is extremely important when multiple separations were compared in parallel in searching for potential biomarkers. Herein, we introduce a “plug-and-use” fritting technology for fabrication of high quality and highly reproducible capillary columns. Due to the identical length, good permeability, and stability of the prefabricated frits adopted, the capillary columns presented excellent performance consistency in terms of retention time, peak width as well as peak capacity at a column-to-column level (relative standard deviations, RSDs, at 0.4–0.9%, 2.1–3.6%, and 2.7%, respectively, n = 6) for separations of complex mixtures of protein digest. For capillary columns packed with 5 m particles, high separation efficiency was demonstrated by the minimum plate height of 11 m, approaching the theoretical performance limit of such material. For separations of protein digests, the columns demonstrated excellent peak capacities of 110 and 300 for 20 and 360 min gradients, respectively. The simple fabrication, good performance as well as consistent quality of such columns provide a reliable tool for high throughput separations requiring the use of multiple high performance capillary columns in parallel. © 2013 Elsevier B.V. All rights reserved.
1. Introduction In 2010, Nature Biotechnology published guidelines for column chromatography by Human Proteome Organization’s Proteomics Standards Initiative [1], which highlighted the demand for high quality microcolumns towards the standardization of proteomic analysis. Capillary column used for nanoflow liquid chromatography (nanoLC) has become a key separation tool for biomolecules [2,3]. Over the past ten years or so, many developments and improvements for microcolumn technology have been introduced, both in capillary [4–12], and microchip formats [13–18], for normal pressure and recently also ultra high pressure nanoLC separations [19–22]. As an indispensible consumptive material in modern proteomics, capillary columns need to be quality-controlled to a high standard. In practice, however, quality control studies were seldom seen over the microcolumns fabricated and used. A common practice seen in many proteomics laboratories is that, one packs a single column and uses it for nanoLC–mass spectrometry (MS) experiment straight away. The column is used until its performance degradation or column breakage, and then a new capillary is
∗ Corresponding author. Tel.: +86 592 2188691; fax: +86 592 2188691. E-mail address: [email protected] (B. Zhang). 0021-9673/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.chroma.2013.12.002
packed and used. Although MS as an information-rich detector can tolerate, to a great extent, the inconsistency of separation performance between columns, the column-to-column irreproducibility may become a crucial issue when large-scale screening proteomics based on parallel use of multiple columns [16–18,23–26] is performed, especially when optical detection is adopted. At the turn of the century, the introduction of capillary array electrophoresis using 96 or 384 capillaries greatly pushed forward the pace of genomic discovery [27]. From the viewpoint of proteome-wide screening for biological discovery [28,29], there needs innovative developments in separation platforms and related consumable devices. It is to this end that high quality capillary columns, with excellent performance consistency at a column-tocolumn level, are demanded by, but not limited to, high throughput proteomics. To date, the most routinely used chromatographic medium for microscale bioseparations is particulate packed capillary columns [2,3], although monolithic [30–35], and open-tubular capillary columns [36], have also been introduced. The aim of the present work is to develop a packed column technology with high standard reproducibility to support high performance microcolumn separations of complex mixtures. In resolving this issue, there are two aspects one needs to take into account: fritting and packing. Although column packing plays an important role in column performance [37,38], the main technical challenge is fabrication of high
110
Z. Xiao et al. / J. Chromatogr. A 1325 (2014) 109–114
Fig. 1. A two-step “plug-and-use” fritting approach. Step 1: A large permeable porous (perfusive) silica bead is tapped onto one end of a capillary. Step 2: The captured silica bead was forced into the capillary. In the photos, the dark dots are the perfusive silica beads intentionally left in to show the relative size of the beads and the capillary tube. The perfusive silica bead has a nominal outer diameter of ∼110 m and the capillary’s inner diameter is 100 m.
quality frits inside capillary tube [39]. In this study, we introduce in a “plug-and-use” approach, based on prefabricated frits with good permeability and predetermined short length, to facilitate column preparation. We will also interrogate the columns’ quality as well as performance consistency in separations of complex mixtures.
2. Experimental 2.1. Materials and apparatus Polyimide-coated fused silica capillaries were purchased from Yongnian Reafine Chromatography (Hebei China). The porous silica particles ∼110 m in diameter with large throughpores about 1 m, to be used as prefabricated single particle frits, were provided by X-tec (Bromborough, UK). The packing material Ultimate XB-C18 ˚ was obtained from Welch Materials Inc. (Shanghai, (5 m, 300 A) China). Thiourea, NH4 HCO3 , methyl-, ethyl-, propyl-, and butylbenzenes of analytical grade, dithiothreitol (DTT), iodoacetamide (IAA) trifluoroacetic acid (TFA), trypsin of sequencing grade, standard protein cytochrome C, lysozyme, ovalbumin, bovine serum albumin, and transferrin were purchased from Sigma–Aldrich (St. Louis, MO). Acetonitrile and acetone of HPLC grade were provided by Merck (Darmstadt, Germany). An Elite P230 high pressure pump from Dalian Elite Analytical Instruments (Dalian, China) was used for column packing.
2.2. Protein digestion Complex peptide mixtures were prepared by tryptic digestion of standard proteins in solution. Generally, proteins were solubilized in 8 M urea, 50 mM NH4 HCO3 . Then, the sample was reduced by DTT and alkylated by IAA. Finally, trypsin was added at
a protein-to-enzyme ratio of 50:1, the digestion was incubated at 37 ◦ C over night. 2.3. Nanoflow liquid chromatography NanoLC experiments were carried out on an Ultimate 3000 nanoLC system (Thermo-Dionex, Amsterdam, The Netherlands), equipped with an autosampler and a variable wavelength UV–vis detector with a 3 nL flow cell. A 4 nL Valco nanovolume injector (VICI AG, Schenkon, Switzerland) was used for column performance evaluation under isocratic condition. For large volume injections under gradient elution, the autosampler with a 1 L loop was adopted. 2.4. “Plug-and-use” fritting and column packing A 25 cm long, fused silica capillary (100 m I.D., 365 m O.D.) was chosen as the column tubing. As shown in Fig. 1, one end of the capillary was tapped into a micro centrifuge tube, in which a small number of ∼110 m perfusive silica beads were deposited. A single perfusive silica bead can be captured at the head of the capillary (Step 1, Fig. 1), the single bead was then pushed into the capillary by pressing the end of the capillary against a plane surface (Step 2, Fig. 1). This two-step process can be monitored and confirmed by observation under a microscope. This single silica bead served as the outlet frit of the column. The capillary column was slurry-packed under high pressure. The packing material was suspended in acetone at a concentration of 2 mg/mL and ultrasonicated for 15 min. The slurry was loaded into a reservoir (4.6 mm I.D., 15 cm long) attached to a high pressure pump. The one end fritted capillary tube was connected to the reservoir via the open end. Pressure was increased gradually (upto 6000 psi) until the column was packed. The column was cut to a desired length after the packing system was fully depressurized. Finally, another single perfusive silica bead was forced into the cut end serving as the inlet frit of the column. Before use, the column was mounted on to the
Z. Xiao et al. / J. Chromatogr. A 1325 (2014) 109–114
111
Fig. 2. Single perfusive bead-fritted column based on keystone effect. A, formation of keystone effect between the bead, fines, and capillary wall, B, six columns fabricated and their end frits using single perfusive beads, C, SEM of a single perfusive silica bead and D, SEM of a perfusive bead as an end frit inside a capillary. The capillary tube has an inner diameter of 100 m.
nanoLC system and fully equilibrated with the mobile phase under high pressure. For this study, twelve capillary columns, six for 15 cm long and the other six 20 cm long, both 100 m I.D., packed with Ultimate ˚ were fabricated and used for performance XB-C18, 5 m, 300 A, investigations. The 20 cm long columns were mainly used for Van Deemter curve evaluation in isocratic mode, and the 15 cm long columns were used for protein digest separations performed in gradient elution mode. 3. Results and discussion 3.1. “Plug-and-use” approach for capillary column fabrication The “plug-and-use” frit is formed using prefabricated single perfusive silica beads, as we first introduced for electrochromatography (CEC) [40,41]. The beads have diameters of ∼110 m, as shown in Fig. 2. A single bead was forced into one end of the capillary and lodged in the place due to keystone effect [6,40–44]. As we demonstrated before [40,41], this keystone effect was formed between the single silica bead and the fines and the capillary wall (Fig. 2A). The keystone effect is like a stone arch bridge in real life, which can sustain heavy weight on its top but without pier underneath [6]. In this case (Fig. 2A), the flow pressure applied on the single bead was largely converted to pressures against capillary wall via keystone effect. At the same time, since the single bead is perfusive (i.e., rich in throughpores) [40,41], liquid flow can go through the bead itself. In this way, the perfusive silica bead can work effectively as a frit: it sustains high hydraulic pressure, allows solvent go though and holds particulate material inside the column. Previously, Mann’s group reported formation of keystone effect
based on particulate packing material at tapered capillary end [6]. In our case, the keystone effect was enabled straight after “plug-in” of a single perfusive bead. The capillary was used as native, no end shrinking (as for the frit-less columns) [6,44–46] or chemical reaction (as for Kasil [7,9,11,21,22] or polymer monolithic frits [47–50]) was needed. In practice, such “plug-and-use” fritting operation takes almost no labor or time, which greatly facilitates capillary column fabrication. Six capillary columns fabricated through this “plug-and-use” approach are presented in Fig. 2B. The end frits based on single silica beads are clearly visualized. They have essentially the same topology and more importantly, consistently short length of ∼100 m. This is crucial for separation reproducibility and performance consistency between columns, as discussed below. In this “plug-and-use” approach, a capillary column can be fabricated within one hour according to our experience, while the most time-consuming part is column packing. This feature significantly improved column production throughput. Most importantly, it makes column quality control in laboratory realistically feasible: one can fabricate a batch of capillary columns for certain set of experiments and all go through a quality control process before use. 3.2. Operational aspects Our previous work has proved such frit’s feasibility and stability in electrochromatography [40,41]. Obviously, CEC separation is driven by electroosmosis rather than hydraulic pressure [51]. In the present study, the top concern is that whether the frits can sustain high hydraulic pressure in pressure-driven liquid chromatography. Experimental result has shown that the “plug-and-use” fritted columns can be run up to 6000 psi (40 MPa) without problem,
112
Z. Xiao et al. / J. Chromatogr. A 1325 (2014) 109–114
the frits were stable during the whole time span, revealing excellent long-term mechanical strength of the frits (and thus the columns) under pressure-driven chromatography. The columns were also tested for sudden pressure drop. Since the columns were fritted at both ends, no leakage of packing material was observed, the columns showed excellent stability against pressure pulsations. While many in-house fabricated capillary columns used for proteomics analysis [6–11] were only fritted (or tapered) at the outlet end, and the inlet end was left open. These one-end-fritted columns may be subject to the risk of packing material loss when experiencing sudden pressure fluctuation during running.
3.3. Column performance and consistency
Fig. 3. Van Deemter curve evaluation. Capillary column: Ultimate XB-C18, 5 m, ˚ 200 mm × 100 m i.d.; mobile phase: 60% ACN; injection volume: 4 nL; UV 300 A, detection: 214 nm. A retained neutral analyte, butylbenzene, was used as the standard.
which is the pressure uplimit of the Dionex nanoLC system. In fact, the columns have been exposed to 6000 psi during packing. The columns were also subjected to repeated routine use and storage for 12 month under both isocratic and gradient elution conditions. All
To interrogate the quality of the columns fabricated, we used a mixture of neutral analytes (alkylbenzenes) to evaluate the columns’ separation efficiency. With a retained analyte (butylbenzene) as the probe, Van Deemter curve was drawn in linear velocity range 0.3–2.0 mm/s (corresponding to volumetric flow rate 100–650 nL/min), as shown in Fig. 3. The optimum plate height (10.9 m, equivalent to 92,000 plates/m, observed at linear velocity 1.1 mm/s) is approaching 10 m, the theoretical limit of 5 m packing material [52]. This is an important sign of the good quality of the column. It is well documented that [39,53], the frit length, as an additional section to the capillary column bed,
Fig. 4. Separations of tryptic digest of Cytochrome C on six capillary columns. A, column-to-column reproducibility. B, 35 repeated injections performed on one column. ˚ 150 mm × 100 m i.d.; mobile phase A, H2 O + 0.05% TFA, B, ACN + 0.05% TFA, 5–50% B in 20 min; injection volume: 0.2 L; Capillary columns: Ultimate XB-C18, 5 m, 300 A, flow rate: 350 nL/min; UV detection: 214 nm.
Z. Xiao et al. / J. Chromatogr. A 1325 (2014) 109–114
113
˚ 150 mm × 100 m i.d.; mobile phase A, H2 O + 0.05% TFA, B, Fig. 5. Long gradient separation of tryptic digest of five proteins. Capillary column: Ultimate XB-C18, 5 m, 300 A, ACN + 0.05% TFA, 5–50% B in 360 min; proteins: cytochrome C, lysozyme, ovalbumin, bovine serum albumin, and transferrin; injection volume: 0.5 L; flow rate: 350 nL/min; UV detection: 214 nm.
contributes significantly to the column bed’s overall nonuniformity, and leads to decreased separation efficiency. Compared with other frits reported (commonly at lengths around 1000–5000 m) [11,21,22,47–50], the consistently short frit length (∼100 m) in this “plug-and-use” strategy significantly diminished frit effect and improved overall uniformity of the packed bed. With the same neutral analytes in isocratic elution mode, excellent column-to-column reproducibility of retention times was observed (RSD < 1%, Fig. S1 and Table S1 in Supplementary data). The focus of our investigation is column performance consistency in gradient separation of complex mixtures. A tryptic digest of cytochrome C was used for this evaluation. As shown in Fig. 4A, highly consistent chromatograms were recorded on all the six columns. To represent peaks of different eluting time and size, five marker peaks (Fig. 4A) were picked up and traced for retention time and peak width at half height (Tables S2 and S3 in Supplementary data). Between all the six columns, the five marker peaks gave excellent reproducibility of RSD = 0.4–0.9% for retention times, and RSD = 2.1–3.6% for peak widths at half height: both marked an extremely stable column-to-column consistency. In literature, an RSD
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
A “plug-and-use” approach towards facile fabrication of capillary columns for high performance nanoflow liquid chromatography Zhiliang Xiao, Lin Wang, Ya Liu, Qiuquan Wang, Bo Zhang ∗ Department of Chemistry and the MOE Key Laboratory of Spectrochemical Analysis and Instrumentation, College of Chemistry and Chemical Engineering, Xiamen University, Xiamen 361005, China
a r t i c l e
i n f o
Article history: Received 20 October 2013 Received in revised form 28 November 2013 Accepted 1 December 2013 Available online 11 December 2013 Keywords: Column technology Chromatography Proteomics High throughput screening Capillary column
a b s t r a c t Capillary columns used for nanoflow liquid chromatography play an important role in modern proteomics. High quality columns are needed to provide high peak capacity and highly reproducible separations. This is extremely important when multiple separations were compared in parallel in searching for potential biomarkers. Herein, we introduce a “plug-and-use” fritting technology for fabrication of high quality and highly reproducible capillary columns. Due to the identical length, good permeability, and stability of the prefabricated frits adopted, the capillary columns presented excellent performance consistency in terms of retention time, peak width as well as peak capacity at a column-to-column level (relative standard deviations, RSDs, at 0.4–0.9%, 2.1–3.6%, and 2.7%, respectively, n = 6) for separations of complex mixtures of protein digest. For capillary columns packed with 5 m particles, high separation efficiency was demonstrated by the minimum plate height of 11 m, approaching the theoretical performance limit of such material. For separations of protein digests, the columns demonstrated excellent peak capacities of 110 and 300 for 20 and 360 min gradients, respectively. The simple fabrication, good performance as well as consistent quality of such columns provide a reliable tool for high throughput separations requiring the use of multiple high performance capillary columns in parallel. © 2013 Elsevier B.V. All rights reserved.
1. Introduction In 2010, Nature Biotechnology published guidelines for column chromatography by Human Proteome Organization’s Proteomics Standards Initiative [1], which highlighted the demand for high quality microcolumns towards the standardization of proteomic analysis. Capillary column used for nanoflow liquid chromatography (nanoLC) has become a key separation tool for biomolecules [2,3]. Over the past ten years or so, many developments and improvements for microcolumn technology have been introduced, both in capillary [4–12], and microchip formats [13–18], for normal pressure and recently also ultra high pressure nanoLC separations [19–22]. As an indispensible consumptive material in modern proteomics, capillary columns need to be quality-controlled to a high standard. In practice, however, quality control studies were seldom seen over the microcolumns fabricated and used. A common practice seen in many proteomics laboratories is that, one packs a single column and uses it for nanoLC–mass spectrometry (MS) experiment straight away. The column is used until its performance degradation or column breakage, and then a new capillary is
∗ Corresponding author. Tel.: +86 592 2188691; fax: +86 592 2188691. E-mail address: [email protected] (B. Zhang). 0021-9673/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.chroma.2013.12.002
packed and used. Although MS as an information-rich detector can tolerate, to a great extent, the inconsistency of separation performance between columns, the column-to-column irreproducibility may become a crucial issue when large-scale screening proteomics based on parallel use of multiple columns [16–18,23–26] is performed, especially when optical detection is adopted. At the turn of the century, the introduction of capillary array electrophoresis using 96 or 384 capillaries greatly pushed forward the pace of genomic discovery [27]. From the viewpoint of proteome-wide screening for biological discovery [28,29], there needs innovative developments in separation platforms and related consumable devices. It is to this end that high quality capillary columns, with excellent performance consistency at a column-tocolumn level, are demanded by, but not limited to, high throughput proteomics. To date, the most routinely used chromatographic medium for microscale bioseparations is particulate packed capillary columns [2,3], although monolithic [30–35], and open-tubular capillary columns [36], have also been introduced. The aim of the present work is to develop a packed column technology with high standard reproducibility to support high performance microcolumn separations of complex mixtures. In resolving this issue, there are two aspects one needs to take into account: fritting and packing. Although column packing plays an important role in column performance [37,38], the main technical challenge is fabrication of high
110
Z. Xiao et al. / J. Chromatogr. A 1325 (2014) 109–114
Fig. 1. A two-step “plug-and-use” fritting approach. Step 1: A large permeable porous (perfusive) silica bead is tapped onto one end of a capillary. Step 2: The captured silica bead was forced into the capillary. In the photos, the dark dots are the perfusive silica beads intentionally left in to show the relative size of the beads and the capillary tube. The perfusive silica bead has a nominal outer diameter of ∼110 m and the capillary’s inner diameter is 100 m.
quality frits inside capillary tube [39]. In this study, we introduce in a “plug-and-use” approach, based on prefabricated frits with good permeability and predetermined short length, to facilitate column preparation. We will also interrogate the columns’ quality as well as performance consistency in separations of complex mixtures.
2. Experimental 2.1. Materials and apparatus Polyimide-coated fused silica capillaries were purchased from Yongnian Reafine Chromatography (Hebei China). The porous silica particles ∼110 m in diameter with large throughpores about 1 m, to be used as prefabricated single particle frits, were provided by X-tec (Bromborough, UK). The packing material Ultimate XB-C18 ˚ was obtained from Welch Materials Inc. (Shanghai, (5 m, 300 A) China). Thiourea, NH4 HCO3 , methyl-, ethyl-, propyl-, and butylbenzenes of analytical grade, dithiothreitol (DTT), iodoacetamide (IAA) trifluoroacetic acid (TFA), trypsin of sequencing grade, standard protein cytochrome C, lysozyme, ovalbumin, bovine serum albumin, and transferrin were purchased from Sigma–Aldrich (St. Louis, MO). Acetonitrile and acetone of HPLC grade were provided by Merck (Darmstadt, Germany). An Elite P230 high pressure pump from Dalian Elite Analytical Instruments (Dalian, China) was used for column packing.
2.2. Protein digestion Complex peptide mixtures were prepared by tryptic digestion of standard proteins in solution. Generally, proteins were solubilized in 8 M urea, 50 mM NH4 HCO3 . Then, the sample was reduced by DTT and alkylated by IAA. Finally, trypsin was added at
a protein-to-enzyme ratio of 50:1, the digestion was incubated at 37 ◦ C over night. 2.3. Nanoflow liquid chromatography NanoLC experiments were carried out on an Ultimate 3000 nanoLC system (Thermo-Dionex, Amsterdam, The Netherlands), equipped with an autosampler and a variable wavelength UV–vis detector with a 3 nL flow cell. A 4 nL Valco nanovolume injector (VICI AG, Schenkon, Switzerland) was used for column performance evaluation under isocratic condition. For large volume injections under gradient elution, the autosampler with a 1 L loop was adopted. 2.4. “Plug-and-use” fritting and column packing A 25 cm long, fused silica capillary (100 m I.D., 365 m O.D.) was chosen as the column tubing. As shown in Fig. 1, one end of the capillary was tapped into a micro centrifuge tube, in which a small number of ∼110 m perfusive silica beads were deposited. A single perfusive silica bead can be captured at the head of the capillary (Step 1, Fig. 1), the single bead was then pushed into the capillary by pressing the end of the capillary against a plane surface (Step 2, Fig. 1). This two-step process can be monitored and confirmed by observation under a microscope. This single silica bead served as the outlet frit of the column. The capillary column was slurry-packed under high pressure. The packing material was suspended in acetone at a concentration of 2 mg/mL and ultrasonicated for 15 min. The slurry was loaded into a reservoir (4.6 mm I.D., 15 cm long) attached to a high pressure pump. The one end fritted capillary tube was connected to the reservoir via the open end. Pressure was increased gradually (upto 6000 psi) until the column was packed. The column was cut to a desired length after the packing system was fully depressurized. Finally, another single perfusive silica bead was forced into the cut end serving as the inlet frit of the column. Before use, the column was mounted on to the
Z. Xiao et al. / J. Chromatogr. A 1325 (2014) 109–114
111
Fig. 2. Single perfusive bead-fritted column based on keystone effect. A, formation of keystone effect between the bead, fines, and capillary wall, B, six columns fabricated and their end frits using single perfusive beads, C, SEM of a single perfusive silica bead and D, SEM of a perfusive bead as an end frit inside a capillary. The capillary tube has an inner diameter of 100 m.
nanoLC system and fully equilibrated with the mobile phase under high pressure. For this study, twelve capillary columns, six for 15 cm long and the other six 20 cm long, both 100 m I.D., packed with Ultimate ˚ were fabricated and used for performance XB-C18, 5 m, 300 A, investigations. The 20 cm long columns were mainly used for Van Deemter curve evaluation in isocratic mode, and the 15 cm long columns were used for protein digest separations performed in gradient elution mode. 3. Results and discussion 3.1. “Plug-and-use” approach for capillary column fabrication The “plug-and-use” frit is formed using prefabricated single perfusive silica beads, as we first introduced for electrochromatography (CEC) [40,41]. The beads have diameters of ∼110 m, as shown in Fig. 2. A single bead was forced into one end of the capillary and lodged in the place due to keystone effect [6,40–44]. As we demonstrated before [40,41], this keystone effect was formed between the single silica bead and the fines and the capillary wall (Fig. 2A). The keystone effect is like a stone arch bridge in real life, which can sustain heavy weight on its top but without pier underneath [6]. In this case (Fig. 2A), the flow pressure applied on the single bead was largely converted to pressures against capillary wall via keystone effect. At the same time, since the single bead is perfusive (i.e., rich in throughpores) [40,41], liquid flow can go through the bead itself. In this way, the perfusive silica bead can work effectively as a frit: it sustains high hydraulic pressure, allows solvent go though and holds particulate material inside the column. Previously, Mann’s group reported formation of keystone effect
based on particulate packing material at tapered capillary end [6]. In our case, the keystone effect was enabled straight after “plug-in” of a single perfusive bead. The capillary was used as native, no end shrinking (as for the frit-less columns) [6,44–46] or chemical reaction (as for Kasil [7,9,11,21,22] or polymer monolithic frits [47–50]) was needed. In practice, such “plug-and-use” fritting operation takes almost no labor or time, which greatly facilitates capillary column fabrication. Six capillary columns fabricated through this “plug-and-use” approach are presented in Fig. 2B. The end frits based on single silica beads are clearly visualized. They have essentially the same topology and more importantly, consistently short length of ∼100 m. This is crucial for separation reproducibility and performance consistency between columns, as discussed below. In this “plug-and-use” approach, a capillary column can be fabricated within one hour according to our experience, while the most time-consuming part is column packing. This feature significantly improved column production throughput. Most importantly, it makes column quality control in laboratory realistically feasible: one can fabricate a batch of capillary columns for certain set of experiments and all go through a quality control process before use. 3.2. Operational aspects Our previous work has proved such frit’s feasibility and stability in electrochromatography [40,41]. Obviously, CEC separation is driven by electroosmosis rather than hydraulic pressure [51]. In the present study, the top concern is that whether the frits can sustain high hydraulic pressure in pressure-driven liquid chromatography. Experimental result has shown that the “plug-and-use” fritted columns can be run up to 6000 psi (40 MPa) without problem,
112
Z. Xiao et al. / J. Chromatogr. A 1325 (2014) 109–114
the frits were stable during the whole time span, revealing excellent long-term mechanical strength of the frits (and thus the columns) under pressure-driven chromatography. The columns were also tested for sudden pressure drop. Since the columns were fritted at both ends, no leakage of packing material was observed, the columns showed excellent stability against pressure pulsations. While many in-house fabricated capillary columns used for proteomics analysis [6–11] were only fritted (or tapered) at the outlet end, and the inlet end was left open. These one-end-fritted columns may be subject to the risk of packing material loss when experiencing sudden pressure fluctuation during running.
3.3. Column performance and consistency
Fig. 3. Van Deemter curve evaluation. Capillary column: Ultimate XB-C18, 5 m, ˚ 200 mm × 100 m i.d.; mobile phase: 60% ACN; injection volume: 4 nL; UV 300 A, detection: 214 nm. A retained neutral analyte, butylbenzene, was used as the standard.
which is the pressure uplimit of the Dionex nanoLC system. In fact, the columns have been exposed to 6000 psi during packing. The columns were also subjected to repeated routine use and storage for 12 month under both isocratic and gradient elution conditions. All
To interrogate the quality of the columns fabricated, we used a mixture of neutral analytes (alkylbenzenes) to evaluate the columns’ separation efficiency. With a retained analyte (butylbenzene) as the probe, Van Deemter curve was drawn in linear velocity range 0.3–2.0 mm/s (corresponding to volumetric flow rate 100–650 nL/min), as shown in Fig. 3. The optimum plate height (10.9 m, equivalent to 92,000 plates/m, observed at linear velocity 1.1 mm/s) is approaching 10 m, the theoretical limit of 5 m packing material [52]. This is an important sign of the good quality of the column. It is well documented that [39,53], the frit length, as an additional section to the capillary column bed,
Fig. 4. Separations of tryptic digest of Cytochrome C on six capillary columns. A, column-to-column reproducibility. B, 35 repeated injections performed on one column. ˚ 150 mm × 100 m i.d.; mobile phase A, H2 O + 0.05% TFA, B, ACN + 0.05% TFA, 5–50% B in 20 min; injection volume: 0.2 L; Capillary columns: Ultimate XB-C18, 5 m, 300 A, flow rate: 350 nL/min; UV detection: 214 nm.
Z. Xiao et al. / J. Chromatogr. A 1325 (2014) 109–114
113
˚ 150 mm × 100 m i.d.; mobile phase A, H2 O + 0.05% TFA, B, Fig. 5. Long gradient separation of tryptic digest of five proteins. Capillary column: Ultimate XB-C18, 5 m, 300 A, ACN + 0.05% TFA, 5–50% B in 360 min; proteins: cytochrome C, lysozyme, ovalbumin, bovine serum albumin, and transferrin; injection volume: 0.5 L; flow rate: 350 nL/min; UV detection: 214 nm.
contributes significantly to the column bed’s overall nonuniformity, and leads to decreased separation efficiency. Compared with other frits reported (commonly at lengths around 1000–5000 m) [11,21,22,47–50], the consistently short frit length (∼100 m) in this “plug-and-use” strategy significantly diminished frit effect and improved overall uniformity of the packed bed. With the same neutral analytes in isocratic elution mode, excellent column-to-column reproducibility of retention times was observed (RSD < 1%, Fig. S1 and Table S1 in Supplementary data). The focus of our investigation is column performance consistency in gradient separation of complex mixtures. A tryptic digest of cytochrome C was used for this evaluation. As shown in Fig. 4A, highly consistent chromatograms were recorded on all the six columns. To represent peaks of different eluting time and size, five marker peaks (Fig. 4A) were picked up and traced for retention time and peak width at half height (Tables S2 and S3 in Supplementary data). Between all the six columns, the five marker peaks gave excellent reproducibility of RSD = 0.4–0.9% for retention times, and RSD = 2.1–3.6% for peak widths at half height: both marked an extremely stable column-to-column consistency. In literature, an RSD
