- Email: [email protected]
Miniaturised electrically actuated high pressure injection valve for portable capillary liquid chromatography
Talanta 180 (2018) 32–35
Contents lists available at ScienceDirect
Talanta journal homepage: www.elsevier.com/locate/talanta
Miniaturised electrically actuated high pressure injection valve for portable capillary liquid chromatography Yan Lia, Kirsten Paceb, Pavel N. Nesterenkoa, Brett Paulla, Roger Stanleyc, Mirek Mackaa,
T
⁎
a
School of Physical Sciences and Australian Centre for Research on Separation Science (ACROSS), University of Tasmania, Hobart 7001, Australia LabSmith, 6111 Southfront Rd, Suite E, Livermore, CA 94551, USA c Centre for Food Innovation, University of Tasmania, Locked Bag 1370, Launceston 7250, Australia b
A R T I C L E I N F O
A B S T R A C T
Keywords: Miniaturised HPLC injection valve Capillary LC Portable analysis Portable chromatography Analytical separations Biogenic amines
A miniaturised high pressure 6-port injection valve has been designed and evaluated for its performance in order to facilitate the development of portable capillary high performance liquid chromatography (HPLC). The electrically actuated valve features a very small size (65 × 19 × 19 mm) and light weight (33 g), and therefore can be easily integrated in a miniaturised modular capillary LC system suited for portable field analysis. The internal volume of the injection valve was determined as 98 nL. The novel conical shape of the stator and rotor and the spring-loaded rotor performed well up to 32 MPa (4641 psi), the maximum operating pressure investigated. Suitability for application was demonstrated using a miniaturised capillary LC system applied to the chromatographic separation of a mixture of biogenic amines and common cations. The RSD (relative standard deviation) values of retention times and peak areas of 6 successive runs were 0.5–0.7% and 1.8–2.8% for the separation of biogenic amines, respectively, and 0.1–0.2% and 2.1–3.0% for the separation of cations, respectively. This performance was comparable with bench-top HPLC systems thus demonstrating the applicability of the valve for use in portable and miniaturised capillary HPLC systems.
1. Introduction Increasing capabilities for electronic miniaturisation over the last few decades have accelerated research towards affordable miniaturised, portable, and in-field deployable analytical instrumentation. Numerous studies on portable instruments have been reported in the area of separation science including capillary electrophoresis (CE) [1] and gas chromatography (GC) [2]. Miniaturisation of liquid chromatography (LC), however, has not yet received as much attention as either CE or GC due to the many technical challenges such as high operating pressures and the environmental and safety issues associated with use of organic solvents [3–10]. Most components of portable LC systems that have been reported have been primarily fabricated in-house and therefore, apart from a very few commercialised systems [4,7], are not widely available to other users. Two reviews on portable LC were published in 2015 [11,12]. Both reviews have summarised and discussed the advances in miniaturisation of LC components (column, pump, injector and detector). From these reviews, it is apparent that there has not been significant progress in the miniaturisation of HPLC injection valves. Recently, our research group reported upon a miniaturised modular
⁎
medium pressure capillary LC system [9]. This system was based on low-cost ‘off-the-shelf’ components, assembled on a breadboard of a commercially available flexible microfluidic platform. Although almost all the components used in this system were miniaturised, the sample introduction system used a standard stainless steel injection valve, which was the only commercially available capillary/nano injection valve on the market at that time. Importantly, the weight of this nanoinjection valve itself represented some 66% of the entire LC system weight (0.86 kg out of 1.30 kg). Clearly, the lack of miniaturised high pressure injection valves is one of the most critical bottlenecks in the development of miniaturised and portable LC systems. Low solvent consumption is inherently of critical importance in miniaturised and portable LC systems. They are therefore normally designed as capillary/nano LC, using low (nL range) injection volumes. For nano LC systems, where less than 20 nL injection volume is required, the only available injection valves are those with an integrated internal sample loop. However, this type of injection valve lacks flexibility as changing the sample injection volume requires a change of the entire injection valve. Therefore, injection valves using external sample loops present the best option for capillary LC systems when the required injection volume is more than 100 nL [12].
Correspondence to: School of Physical Sciences and Australian Centre for Research on Separation Science, University of Tasmania, Private Bag 75, Hobart 7001, Australia. E-mail address: [email protected] (M. Macka).
https://doi.org/10.1016/j.talanta.2017.11.061 Received 22 October 2017; Received in revised form 25 November 2017; Accepted 27 November 2017 Available online 02 December 2017 0039-9140/ © 2017 Elsevier B.V. All rights reserved.
Talanta 180 (2018) 32–35
Y. Li et al.
(ca. 38 nL) hold-up volume for gradient formation. Thermo Fisher Scientific IonPac capillary columns (Sunnyvale, USA; 0.4 × 250 mm) CS17 were used for separation studies. The internal volume and injection carry-over study, and the demonstration of separation of biogenic amines used a Knauer 2600 UV detector (Knauer, Berlin, Germany) containing a 45 nL flow cell. The operating backpressure study and the separation of cations demonstration used a TraceDec® contactless conductivity (C4D) detector (Innovative Sensor Technologies GmbH, Strasshof an der Nordbahn, Austria) with an inserted capillary i.d. of 100 µm. The detectors were interfaced with the miniaturised LC system through an eDAQ (Denistone East, New South Wales, Australia) integrated potentiostat system E466 converter for data acquisition. In this work, all data were recorded at 20 Hz sampling frequency. The presented chromatograms were processed online by a mains digital filter (50 Hz) and a low-pass digital filter with a cut-off frequency of 1 Hz resulting in negligible effect on peak heights. The ICS-5000 IC was controlled with Chromeleon 7.1 chromatography management software (Thermo Scientific Dionex). The miniaturised LC system was controlled using uProcess™ software (LabSmith, Livermore CA, USA), and data acquisition and processing was carried out using Chart software (eDAQ).
Herein, the performance of a novel miniaturised high pressure 6port injection valve was investigated, and demonstrated within a miniaturised LC system. The valve was developed and provided by LabSmith (Livermore CA, USA), which now has been commercially available. It has a highly compact design making it currently the smallest and lightest commercially available high pressure LC injection valve. Characterisation of the injection valve included the determination of the internal volume, the carry-over between injections, and the consistency of function at a range of operating backpressure values. Its analytical performance parameters as an injection valve were then demonstrated in the separation of biogenic amines and cations using a miniaturised LC system. 2. Experimental Details of chemicals and reagents, and methods and procedures are in the Supplementary Information (SI, Experimental). 2.1. Instrumentation The uProcess™ AV303-C360 miniaturised injection valve contains a conical spring-loaded rotor (SI (Supplementary information), Fig. S1, Fig. S2 and Fig. S3). The rotor and stator are made entirely from VESPEL®, as a solvent compatible polymer material offering the capability of metal-free analysis. The summarised specifications are listed in Table S1, and more detailed descriptions can be found from the manufacturer specifications [13]. Design considerations are given in references accompany the SI (Fig. S2 and Fig. S3) and Section 3.1. The valve can be assembled into the LabSmith uProcess™ microfluidic system. The injection valve internal volume determination and operating backpressure were studied using an ICS-5000 ion chromatograph (IC) (Thermo Scientific Dionex, Sunnyvale, USA). The restrictors for testing the injection valve operation at different backpressures and maximum operating pressure were made of polyimide coated fused silica capillary (Polymicro Tech; Phoenix, AZ, USA). Polyimide coated fused silica capillary (Polymicro Tech; Phoenix, AZ, USA) was also used for the external sample loop. A miniaturised LC system reported previously [9] (Fig. 1 and Fig. S1) was employed for the separation of biogenic amines and cations. In brief, the system was based on a LabSmith uProcess™ microfluidic system ("Microfluidic Fluid Control & Connectors" [14]) as the backbone, with LabSmith and other ‘off-the-shelf’ components assembled on a breadboard. The pumping system was formed by four miniaturised syringe pumps, with 5, 20 and 100 µL syringe volume options (LabSmith, Livermore CA, USA). Two pairs of pumps were used for each mobile phase to enable both a continuous pumping as well as gradient elution capability. Two microfluidic switching valves were used for connecting the two pairs of syringe pumps, which were then connected through a Y-connector, thus providing an exceptionally low
3. Results and discussion 3.1. Injection valve design considerations The two-position, six-port valve (LabSmithAV303-C360) was designed for compact size, low swept volume, and low power consumption. The valve wetted interface consists of a body with six machined ports and a rotating stem with three machined grooves. The depth of the grooves controls the valve swept volume. The stem configuration used for this application had a total volume of 100 nL as the internal volume. The ports connect to 360 µm OD capillary via a swage-type fitting. The stem and body are precision machined of Vespel. The valve design aims to achieve high pressure via preload of the stem's machined face against the valve body face. The design has a low sealing surface area (less than 20 mm 2), resulting in minimal torque required to rotate the stem. This in turn allows the use of a miniature and widely-available gear motor to rotate the stem into position. This design has several advantages over other automated micro-valve designs, including small size, low cost, and low power consumption. The motor used for the injection valve is a 300:1 brushed DC micro gear motor (10 mm × 12 mm × 25 mm, Zhengke Motor Company, Zhejiang, China) and is actuated between 0 and 5 V (via pulse-width modulation). The port-to-stem alignment is achieved via machined stop position. The fork rotates clockwise or counter-clockwise to push the valve stem arm until it seats at a stop position. When the stem arm reaches the seating position, the motor controller senses an increase in current and stops the motor. A compliant rubber gasket is used between the stem arm and motor fork to cushion the motor impact when it reaches the end stop. The controller detects the resulting motor current rise and quickly removes power to the motor to reduce stresses in the motor gear box. The valve is connected to a compact controller manifold via a 6-pin flat flex cable. The controller manifold (LabSmith part number 4VM01 or 4VM02) supports simultaneous or sequenced control of up to 4 valves. During valve actuation, the manifold draws approximately 85 mA. The valve and the controller can be integrated into a miniaturised LC system as previously reported (Fig. 1 and Fig. S1) [9]. In this application, the gear motor was designed to have a lifetime of > 100,000 actuations. This compares to a service life of 1000,000 actuations for the wetted portion of the valve. To account for this lifetime discrepancy, the valve motor assembly is designed to be easily removed and replaced in the field. This replacement requires no tools, and can be done while the valve is mounted to a breadboard and
Fig. 1. An image of the miniaturised LC system.
33
Talanta 180 (2018) 32–35
Y. Li et al.
without disconnecting the capillary connections to the ports. 3.2. Internal volume and carry-over The internal volume of the 6-port injector is composed of the volume of the groove in the rotor, and the volume of the connecting bore holes [15] (SI, Fig. S2 C and D, as labelled in red). The internal volume combined with the volume of an external sample loop then gives the resulting injection volume. The internal volume of an injector is the minimum injection loading capacity and significantly affects the injection performance as it is the main factor of extra-column variance causing band broadening [16–19]. Low carry-over between injections is also required to minimise error. Therefore, the internal volume and carry-over of the miniaturised injector were determined experimentally. A linear dependence between the peak area (uracil, 1 mg L−1) and the volume of external sample loop was obtained (SI, Fig. S5). The internal volume was determined by extrapolating the linear relationship between peak area and sample volume back to a peak area equal to zero (SI, Fig. S5, calibration curve intercept with the x-axis). Finally, the experimentally determined internal volume was 98 nL. This was in excellent agreement with the value given by the manufacturer of 100 nL. The 98 nL internal volume is lower or comparable to the values for other commercial and research 6-port capillary injectors (130 -336 nL; Table. S1 in SI). The carry-over study showed that no carry-over effect could be observed, even with excessively high concentrations of solute (500 mg L−1 uracil). This potential source of error has therefore been eliminated.
Fig. 2. Performance of the miniaturised injection valve evaluated by injection of 448 nL of sodium chloride solution (10 mg L−1) at different backpressures, 0.07 MPa (10 psi), 1.6 MPa (232 psi) and 30 MPa (4350 psi). Deionised water at 10 µL min−1. The results demonstrate no significant difference between the injected plug shapes. Conditions: Detection: C4D detector (voltage − 12 dB, gain 50%). For details of conditions see Experimental and Fig. S4 in SI. 2 2 σtot ≈ σinj .
Fig. 2 shows the performance of the miniaturised injection valve with a 448 nL injection volume (98 nL internal + 350 nL external) evaluated under widely differing backpressures of 0.07 MPa (10 psi), 1.6 MPa (232 psi) and 30 MPa (4350 psi). These represent low pressure, medium pressure and high pressure LC, respectively. The peak width (w) was obtained by extrapolating the tangents to the inflection points down to the baseline. Taking the width as the length of a sample plug in the capillary/column, then according to the flow rate (10 µL min−1), the volume of the sample plug could be calculated as 525 nL, 542 nL and 467 nL at 0.07 MPa (10 psi), 1.6 MPa (232 psi) and 30 MPa (4350 psi), respectively. These represent an overall difference in plug volume across the entire pressure range of less than 15%. The %RSD values for peak areas at 0.07 MPa, 1.6 MPa and 30 MPa was 0.33%, 0.29% and 1.08%. The high peak reproducibility therefore demonstrated excellent performance of the miniaturised injection valve.
3.3. Maximum operating pressure and injection performance under different backpressures The maximum operating backpressure of an injector is a key factor that determines its applicability to high pressure capillary systems. The effect of the backpressure on injector performance is therefore an important parameter to evaluate. According to Darcy's Law [20], the backpressure should rise linearly with increasing flow rate. The maximum operating pressure was therefore determined experimentally to be at least 32 MPa (4641 psi) (SI, Fig. S6) as shown by no pressure drop or leakage observed over 24 h operation. This was the maximum pressure of the IC system. The demonstrated operating pressure of the injector is suitable for the majority of applications. To study the injection performance of an injection valve requires that extra-column volumes contributed by other components within an LC system be eliminated or minimised. The effect of extra-column volumes on injection performance with capillary LC has been well characterised [15,19]. In general, each component in a LC system can be assumed to contribute to the peak variance independently, and the sum 2 of these contributions is the total variance (σtot ):
3.4. Demonstration in miniaturised capillary chromatography The overall performance of the miniaturised injector integrated in the miniaturised capillary LC system [9] was demonstrated via the separation of biogenic amines (Fig. 3A) and common cations (Fig. 3B). The peak asymmetry (B/A, w0.1) factor was found to be 1.17–1.40 and 0.77–1.29 for biogenic amines and cations, respectively. RSD value of retention times and peak areas of 6 successive runs were 0.5–0.7% and 1.8–2.8% for the separation of biogenic amines, respectively, 0.1–0.2% and 2.1–3.0% for the separation of cations, respectively. These are comparable to the previous results where RSD value of peak areas were 3.3–4.8% [9] when using the same miniaturised capillary LC system with a standard sized HPLC nano-injector. The miniaturised injection valve therefore performed as well as a standard injector valve when incorporated into a small portable LC system. This is the last component of LC systems to be scaled down. It thus enables studies to be initiated on portable applications where immediate results are required using in-field or in-factory analysis or where samples are unstable to transportation to distant laboratories. From the practical point of view miniaturisation also enables minimisation of component costs and minimisation of waste solvents. For wide utility, it is also critical that all materials in the valve having contact with the mobile phase should have a high degree of inertness in respect to possible components of mobile phases. The contact components of the valve are entirely of VESPEL® which is a polyimide plastic material widely used in GC, HPLC, and other switching valves [21]. It therefore has excellent thermal, chemical and mechanical stability and good organic solvent tolerance, which has been shown in our previous
2 2 2 2 2 σtot = σcol + σcap + σinj + σdet 2 2 2 2 where σcol , σcap , σinj and σdet is the variance contribution of the column, connecting capillaries, injection system and detection system, respectively. For this research on-capillary detection was used at the shortest possible distance (55 mm) to minimise the effect from connecting ca2 pillaries. Hence, the σcap can be regarded as negligible relative to other 2 is nil, there is also contributions. For on-capillary detection, where σdet no contribution due to detector volume broadening. A recent study by Aggarwal et al. [15], showed that for capillary LC, above a certain flow 2 rate (in their study, 0.2 µL min−1), the σcol became marginal. In this current study, the injector was connected directly to on-capillary de2 tection without a column. Thus σcol is nil and there is no variance contribution due to column volume. So finally, the extra-column variance in our system was:
34
Talanta 180 (2018) 32–35
Y. Li et al.
backpressures and maximum operating pressure. The rotor and stator of the injection valve is made from polymer (VESPEL®) making the injection valve suitable for metal-free HPLC, as well as compatible with common HPLC solvents, which is important in reversed phase HPLC. The potential of this miniaturised injection valve for portable LC is demonstrated in a miniaturised capillary LC system for the separations of model analytes of biogenic amines and common cations. The results show a comparable reproducibility to the previous results obtained from the same miniaturised capillary LC system with a standard nano injector. Acknowledgements Mirek Macka gratefully acknowledges the Australian Research Council Future Fellowship (FT120100559). Appendix A. Supplementary material Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.talanta.2017.11.061. References [1] M. Ryvolová, M. Macka, M. Ryvolová, J. Preisler, M. Macka, Trends Anal. Chem. 29 (2010) 339–353. [2] C. Henry, Anal. Chem. 69 (1997) 195A–200A. [3] T. Otagawa, J.R. Stetter, S. Zaromb, J. Chromatogr. 360 (1986) 252–259. [4] G.I. Baram, J. Chromatogr. A 728 (1996) 387–399. [5] V.M. Tulchinsky, S. Technologies, A.B. Park, 2, pp. 281–285, 1998. [6] A. Ishida, M. Fujii, T. Fujimoto, S. Sasaki, I. Yanagisawa, H. Tani, M. Tokeshi, Anal. Sci. 31 (2015) 1163–1169. [7] S. Sharma, A. Plistil, R.S. Simpson, K. Liu, P.B. Farnsworth, S.D. Stearns, M.L. Lee, J. Chromatogr. A 1327 (2014) 80–89. [8] S. Sharma, A. Plistil, H.E. Barnett, H.D. Tolley, P.B. Farnsworth, S.D. Stearns, M.L. Lee, Anal. Chem. 87 (2015) 10457–10461. [9] Y. Li, M. Dvorak, P.N. Nesterenko, R. Stanley, N. Nuchtavorn, L.K. Krcmova, J. Aufartova, M. Macka, Anal. Chim. Acta 896 (2015) 166–176. [10] B. Yang, M. Zhang, T. Kanyanee, B.N. Stamos, P.K. Dasgupta, Anal. Chem. 86 (2014) 11554–11561. [11] S. Sharma, L.T. Tolley, H.D. Tolley, A. Plistil, S.D. Stearns, M.L. Lee, J. Chromatogr. A 1421 (2015) 38–47. [12] C.E. Nazario, M.R. Silva, M.S. Franco, F.M. Lanças, J. Chromatogr. A 1421 (2015) 18–37. [13] 〈http://labsmith.com/Spec_Sheets/LabSmith_AV201-AV202-AV303_Spec_Sheet_ 1015.pdf〉, 20/04/. [14] 〈http://products.labsmith.com/fluid-control-and-connectors/#.WFt52xt95v1〉, 20/04/. [15] P. Aggarwal, K. Liu, S. Sharma, J.S. Lawson, H.D. Tolley, M.L. Lee, J. Chromatogr. A 1380 (2015) 38–44. [16] R. Scott, C. Simpson, J. Chromatogr. Sci. 20 (1982) 62–66. [17] R.C. Simpson, J. Chromatogr. A 691 (1995) 163–170. [18] M.D. Foster, M.A. Arnold, J.A. Nichols, S.R. Bakalyar, J. Chromatogr. A 869 (2000) 231–241. [19] A. Prüß, C. Kempter, J. Gysler, T. Jira, J. Chromatogr. A 1016 (2003) 129–141. [20] H. Darcy, Dalmont, Paris, 647, 1856. [21] 〈http://www2.dupont.com/Vespel/en_US/assets/downloads/vespel_s/VPEA10948-00-A0311.pdf〉, 20/04/.
Fig. 3. A. Isocratic separation of biogenic amines (L-DOPA 6 mg L−1, norfenefrine 8 mg L−1, phenylephrine 12 mg L−1and 5-hydroxytryptophan 12 mg L−1) using a miniaturised capillary chromatographic system. A Conditions: Eluent 25 mM methanesulfonic acid; Detection: 280 nm photometric detection Isocratic separation of cations (Li 4 mg L−1, NH4 8 mg L−1, K 8 mg L−1, dimethylamine 15 mg L−1and trimethylamine 15 mg L−1). B. Conditions: Eluent 6 mM methanesulfonic acid; Detection: C4D detector, voltage −12 dB, gain 50%. Flow rate 6 µL min−1; column: IonPac CS17 (250 × 0.4 mm i.d.); injection volume: 448 nL. For more detail see the Experimental section in SI.
publication, where 100% methanol and acetonitrile were used as mobile phase in switching valves made from the same material [9]. 4. Conclusions This study evaluates a new miniaturised high pressure 6-port injection valve, and demonstrates its successful integration and use in a miniaturised, low cost modular capillary LC. The evaluated injection valve is inherently superior in terms of its small size and low weight to other injection valves as its highly compact design making it currently the smallest and lightest commercially available high pressure LC injection valve. However, it also shows a comparable level of performance to that of a standard HPLC nano-injection valve, with regard to its internal volume, carry-over, performance under different
35
Contents lists available at ScienceDirect
Talanta journal homepage: www.elsevier.com/locate/talanta
Miniaturised electrically actuated high pressure injection valve for portable capillary liquid chromatography Yan Lia, Kirsten Paceb, Pavel N. Nesterenkoa, Brett Paulla, Roger Stanleyc, Mirek Mackaa,
T
⁎
a
School of Physical Sciences and Australian Centre for Research on Separation Science (ACROSS), University of Tasmania, Hobart 7001, Australia LabSmith, 6111 Southfront Rd, Suite E, Livermore, CA 94551, USA c Centre for Food Innovation, University of Tasmania, Locked Bag 1370, Launceston 7250, Australia b
A R T I C L E I N F O
A B S T R A C T
Keywords: Miniaturised HPLC injection valve Capillary LC Portable analysis Portable chromatography Analytical separations Biogenic amines
A miniaturised high pressure 6-port injection valve has been designed and evaluated for its performance in order to facilitate the development of portable capillary high performance liquid chromatography (HPLC). The electrically actuated valve features a very small size (65 × 19 × 19 mm) and light weight (33 g), and therefore can be easily integrated in a miniaturised modular capillary LC system suited for portable field analysis. The internal volume of the injection valve was determined as 98 nL. The novel conical shape of the stator and rotor and the spring-loaded rotor performed well up to 32 MPa (4641 psi), the maximum operating pressure investigated. Suitability for application was demonstrated using a miniaturised capillary LC system applied to the chromatographic separation of a mixture of biogenic amines and common cations. The RSD (relative standard deviation) values of retention times and peak areas of 6 successive runs were 0.5–0.7% and 1.8–2.8% for the separation of biogenic amines, respectively, and 0.1–0.2% and 2.1–3.0% for the separation of cations, respectively. This performance was comparable with bench-top HPLC systems thus demonstrating the applicability of the valve for use in portable and miniaturised capillary HPLC systems.
1. Introduction Increasing capabilities for electronic miniaturisation over the last few decades have accelerated research towards affordable miniaturised, portable, and in-field deployable analytical instrumentation. Numerous studies on portable instruments have been reported in the area of separation science including capillary electrophoresis (CE) [1] and gas chromatography (GC) [2]. Miniaturisation of liquid chromatography (LC), however, has not yet received as much attention as either CE or GC due to the many technical challenges such as high operating pressures and the environmental and safety issues associated with use of organic solvents [3–10]. Most components of portable LC systems that have been reported have been primarily fabricated in-house and therefore, apart from a very few commercialised systems [4,7], are not widely available to other users. Two reviews on portable LC were published in 2015 [11,12]. Both reviews have summarised and discussed the advances in miniaturisation of LC components (column, pump, injector and detector). From these reviews, it is apparent that there has not been significant progress in the miniaturisation of HPLC injection valves. Recently, our research group reported upon a miniaturised modular
⁎
medium pressure capillary LC system [9]. This system was based on low-cost ‘off-the-shelf’ components, assembled on a breadboard of a commercially available flexible microfluidic platform. Although almost all the components used in this system were miniaturised, the sample introduction system used a standard stainless steel injection valve, which was the only commercially available capillary/nano injection valve on the market at that time. Importantly, the weight of this nanoinjection valve itself represented some 66% of the entire LC system weight (0.86 kg out of 1.30 kg). Clearly, the lack of miniaturised high pressure injection valves is one of the most critical bottlenecks in the development of miniaturised and portable LC systems. Low solvent consumption is inherently of critical importance in miniaturised and portable LC systems. They are therefore normally designed as capillary/nano LC, using low (nL range) injection volumes. For nano LC systems, where less than 20 nL injection volume is required, the only available injection valves are those with an integrated internal sample loop. However, this type of injection valve lacks flexibility as changing the sample injection volume requires a change of the entire injection valve. Therefore, injection valves using external sample loops present the best option for capillary LC systems when the required injection volume is more than 100 nL [12].
Correspondence to: School of Physical Sciences and Australian Centre for Research on Separation Science, University of Tasmania, Private Bag 75, Hobart 7001, Australia. E-mail address: [email protected] (M. Macka).
https://doi.org/10.1016/j.talanta.2017.11.061 Received 22 October 2017; Received in revised form 25 November 2017; Accepted 27 November 2017 Available online 02 December 2017 0039-9140/ © 2017 Elsevier B.V. All rights reserved.
Talanta 180 (2018) 32–35
Y. Li et al.
(ca. 38 nL) hold-up volume for gradient formation. Thermo Fisher Scientific IonPac capillary columns (Sunnyvale, USA; 0.4 × 250 mm) CS17 were used for separation studies. The internal volume and injection carry-over study, and the demonstration of separation of biogenic amines used a Knauer 2600 UV detector (Knauer, Berlin, Germany) containing a 45 nL flow cell. The operating backpressure study and the separation of cations demonstration used a TraceDec® contactless conductivity (C4D) detector (Innovative Sensor Technologies GmbH, Strasshof an der Nordbahn, Austria) with an inserted capillary i.d. of 100 µm. The detectors were interfaced with the miniaturised LC system through an eDAQ (Denistone East, New South Wales, Australia) integrated potentiostat system E466 converter for data acquisition. In this work, all data were recorded at 20 Hz sampling frequency. The presented chromatograms were processed online by a mains digital filter (50 Hz) and a low-pass digital filter with a cut-off frequency of 1 Hz resulting in negligible effect on peak heights. The ICS-5000 IC was controlled with Chromeleon 7.1 chromatography management software (Thermo Scientific Dionex). The miniaturised LC system was controlled using uProcess™ software (LabSmith, Livermore CA, USA), and data acquisition and processing was carried out using Chart software (eDAQ).
Herein, the performance of a novel miniaturised high pressure 6port injection valve was investigated, and demonstrated within a miniaturised LC system. The valve was developed and provided by LabSmith (Livermore CA, USA), which now has been commercially available. It has a highly compact design making it currently the smallest and lightest commercially available high pressure LC injection valve. Characterisation of the injection valve included the determination of the internal volume, the carry-over between injections, and the consistency of function at a range of operating backpressure values. Its analytical performance parameters as an injection valve were then demonstrated in the separation of biogenic amines and cations using a miniaturised LC system. 2. Experimental Details of chemicals and reagents, and methods and procedures are in the Supplementary Information (SI, Experimental). 2.1. Instrumentation The uProcess™ AV303-C360 miniaturised injection valve contains a conical spring-loaded rotor (SI (Supplementary information), Fig. S1, Fig. S2 and Fig. S3). The rotor and stator are made entirely from VESPEL®, as a solvent compatible polymer material offering the capability of metal-free analysis. The summarised specifications are listed in Table S1, and more detailed descriptions can be found from the manufacturer specifications [13]. Design considerations are given in references accompany the SI (Fig. S2 and Fig. S3) and Section 3.1. The valve can be assembled into the LabSmith uProcess™ microfluidic system. The injection valve internal volume determination and operating backpressure were studied using an ICS-5000 ion chromatograph (IC) (Thermo Scientific Dionex, Sunnyvale, USA). The restrictors for testing the injection valve operation at different backpressures and maximum operating pressure were made of polyimide coated fused silica capillary (Polymicro Tech; Phoenix, AZ, USA). Polyimide coated fused silica capillary (Polymicro Tech; Phoenix, AZ, USA) was also used for the external sample loop. A miniaturised LC system reported previously [9] (Fig. 1 and Fig. S1) was employed for the separation of biogenic amines and cations. In brief, the system was based on a LabSmith uProcess™ microfluidic system ("Microfluidic Fluid Control & Connectors" [14]) as the backbone, with LabSmith and other ‘off-the-shelf’ components assembled on a breadboard. The pumping system was formed by four miniaturised syringe pumps, with 5, 20 and 100 µL syringe volume options (LabSmith, Livermore CA, USA). Two pairs of pumps were used for each mobile phase to enable both a continuous pumping as well as gradient elution capability. Two microfluidic switching valves were used for connecting the two pairs of syringe pumps, which were then connected through a Y-connector, thus providing an exceptionally low
3. Results and discussion 3.1. Injection valve design considerations The two-position, six-port valve (LabSmithAV303-C360) was designed for compact size, low swept volume, and low power consumption. The valve wetted interface consists of a body with six machined ports and a rotating stem with three machined grooves. The depth of the grooves controls the valve swept volume. The stem configuration used for this application had a total volume of 100 nL as the internal volume. The ports connect to 360 µm OD capillary via a swage-type fitting. The stem and body are precision machined of Vespel. The valve design aims to achieve high pressure via preload of the stem's machined face against the valve body face. The design has a low sealing surface area (less than 20 mm 2), resulting in minimal torque required to rotate the stem. This in turn allows the use of a miniature and widely-available gear motor to rotate the stem into position. This design has several advantages over other automated micro-valve designs, including small size, low cost, and low power consumption. The motor used for the injection valve is a 300:1 brushed DC micro gear motor (10 mm × 12 mm × 25 mm, Zhengke Motor Company, Zhejiang, China) and is actuated between 0 and 5 V (via pulse-width modulation). The port-to-stem alignment is achieved via machined stop position. The fork rotates clockwise or counter-clockwise to push the valve stem arm until it seats at a stop position. When the stem arm reaches the seating position, the motor controller senses an increase in current and stops the motor. A compliant rubber gasket is used between the stem arm and motor fork to cushion the motor impact when it reaches the end stop. The controller detects the resulting motor current rise and quickly removes power to the motor to reduce stresses in the motor gear box. The valve is connected to a compact controller manifold via a 6-pin flat flex cable. The controller manifold (LabSmith part number 4VM01 or 4VM02) supports simultaneous or sequenced control of up to 4 valves. During valve actuation, the manifold draws approximately 85 mA. The valve and the controller can be integrated into a miniaturised LC system as previously reported (Fig. 1 and Fig. S1) [9]. In this application, the gear motor was designed to have a lifetime of > 100,000 actuations. This compares to a service life of 1000,000 actuations for the wetted portion of the valve. To account for this lifetime discrepancy, the valve motor assembly is designed to be easily removed and replaced in the field. This replacement requires no tools, and can be done while the valve is mounted to a breadboard and
Fig. 1. An image of the miniaturised LC system.
33
Talanta 180 (2018) 32–35
Y. Li et al.
without disconnecting the capillary connections to the ports. 3.2. Internal volume and carry-over The internal volume of the 6-port injector is composed of the volume of the groove in the rotor, and the volume of the connecting bore holes [15] (SI, Fig. S2 C and D, as labelled in red). The internal volume combined with the volume of an external sample loop then gives the resulting injection volume. The internal volume of an injector is the minimum injection loading capacity and significantly affects the injection performance as it is the main factor of extra-column variance causing band broadening [16–19]. Low carry-over between injections is also required to minimise error. Therefore, the internal volume and carry-over of the miniaturised injector were determined experimentally. A linear dependence between the peak area (uracil, 1 mg L−1) and the volume of external sample loop was obtained (SI, Fig. S5). The internal volume was determined by extrapolating the linear relationship between peak area and sample volume back to a peak area equal to zero (SI, Fig. S5, calibration curve intercept with the x-axis). Finally, the experimentally determined internal volume was 98 nL. This was in excellent agreement with the value given by the manufacturer of 100 nL. The 98 nL internal volume is lower or comparable to the values for other commercial and research 6-port capillary injectors (130 -336 nL; Table. S1 in SI). The carry-over study showed that no carry-over effect could be observed, even with excessively high concentrations of solute (500 mg L−1 uracil). This potential source of error has therefore been eliminated.
Fig. 2. Performance of the miniaturised injection valve evaluated by injection of 448 nL of sodium chloride solution (10 mg L−1) at different backpressures, 0.07 MPa (10 psi), 1.6 MPa (232 psi) and 30 MPa (4350 psi). Deionised water at 10 µL min−1. The results demonstrate no significant difference between the injected plug shapes. Conditions: Detection: C4D detector (voltage − 12 dB, gain 50%). For details of conditions see Experimental and Fig. S4 in SI. 2 2 σtot ≈ σinj .
Fig. 2 shows the performance of the miniaturised injection valve with a 448 nL injection volume (98 nL internal + 350 nL external) evaluated under widely differing backpressures of 0.07 MPa (10 psi), 1.6 MPa (232 psi) and 30 MPa (4350 psi). These represent low pressure, medium pressure and high pressure LC, respectively. The peak width (w) was obtained by extrapolating the tangents to the inflection points down to the baseline. Taking the width as the length of a sample plug in the capillary/column, then according to the flow rate (10 µL min−1), the volume of the sample plug could be calculated as 525 nL, 542 nL and 467 nL at 0.07 MPa (10 psi), 1.6 MPa (232 psi) and 30 MPa (4350 psi), respectively. These represent an overall difference in plug volume across the entire pressure range of less than 15%. The %RSD values for peak areas at 0.07 MPa, 1.6 MPa and 30 MPa was 0.33%, 0.29% and 1.08%. The high peak reproducibility therefore demonstrated excellent performance of the miniaturised injection valve.
3.3. Maximum operating pressure and injection performance under different backpressures The maximum operating backpressure of an injector is a key factor that determines its applicability to high pressure capillary systems. The effect of the backpressure on injector performance is therefore an important parameter to evaluate. According to Darcy's Law [20], the backpressure should rise linearly with increasing flow rate. The maximum operating pressure was therefore determined experimentally to be at least 32 MPa (4641 psi) (SI, Fig. S6) as shown by no pressure drop or leakage observed over 24 h operation. This was the maximum pressure of the IC system. The demonstrated operating pressure of the injector is suitable for the majority of applications. To study the injection performance of an injection valve requires that extra-column volumes contributed by other components within an LC system be eliminated or minimised. The effect of extra-column volumes on injection performance with capillary LC has been well characterised [15,19]. In general, each component in a LC system can be assumed to contribute to the peak variance independently, and the sum 2 of these contributions is the total variance (σtot ):
3.4. Demonstration in miniaturised capillary chromatography The overall performance of the miniaturised injector integrated in the miniaturised capillary LC system [9] was demonstrated via the separation of biogenic amines (Fig. 3A) and common cations (Fig. 3B). The peak asymmetry (B/A, w0.1) factor was found to be 1.17–1.40 and 0.77–1.29 for biogenic amines and cations, respectively. RSD value of retention times and peak areas of 6 successive runs were 0.5–0.7% and 1.8–2.8% for the separation of biogenic amines, respectively, 0.1–0.2% and 2.1–3.0% for the separation of cations, respectively. These are comparable to the previous results where RSD value of peak areas were 3.3–4.8% [9] when using the same miniaturised capillary LC system with a standard sized HPLC nano-injector. The miniaturised injection valve therefore performed as well as a standard injector valve when incorporated into a small portable LC system. This is the last component of LC systems to be scaled down. It thus enables studies to be initiated on portable applications where immediate results are required using in-field or in-factory analysis or where samples are unstable to transportation to distant laboratories. From the practical point of view miniaturisation also enables minimisation of component costs and minimisation of waste solvents. For wide utility, it is also critical that all materials in the valve having contact with the mobile phase should have a high degree of inertness in respect to possible components of mobile phases. The contact components of the valve are entirely of VESPEL® which is a polyimide plastic material widely used in GC, HPLC, and other switching valves [21]. It therefore has excellent thermal, chemical and mechanical stability and good organic solvent tolerance, which has been shown in our previous
2 2 2 2 2 σtot = σcol + σcap + σinj + σdet 2 2 2 2 where σcol , σcap , σinj and σdet is the variance contribution of the column, connecting capillaries, injection system and detection system, respectively. For this research on-capillary detection was used at the shortest possible distance (55 mm) to minimise the effect from connecting ca2 pillaries. Hence, the σcap can be regarded as negligible relative to other 2 is nil, there is also contributions. For on-capillary detection, where σdet no contribution due to detector volume broadening. A recent study by Aggarwal et al. [15], showed that for capillary LC, above a certain flow 2 rate (in their study, 0.2 µL min−1), the σcol became marginal. In this current study, the injector was connected directly to on-capillary de2 tection without a column. Thus σcol is nil and there is no variance contribution due to column volume. So finally, the extra-column variance in our system was:
34
Talanta 180 (2018) 32–35
Y. Li et al.
backpressures and maximum operating pressure. The rotor and stator of the injection valve is made from polymer (VESPEL®) making the injection valve suitable for metal-free HPLC, as well as compatible with common HPLC solvents, which is important in reversed phase HPLC. The potential of this miniaturised injection valve for portable LC is demonstrated in a miniaturised capillary LC system for the separations of model analytes of biogenic amines and common cations. The results show a comparable reproducibility to the previous results obtained from the same miniaturised capillary LC system with a standard nano injector. Acknowledgements Mirek Macka gratefully acknowledges the Australian Research Council Future Fellowship (FT120100559). Appendix A. Supplementary material Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.talanta.2017.11.061. References [1] M. Ryvolová, M. Macka, M. Ryvolová, J. Preisler, M. Macka, Trends Anal. Chem. 29 (2010) 339–353. [2] C. Henry, Anal. Chem. 69 (1997) 195A–200A. [3] T. Otagawa, J.R. Stetter, S. Zaromb, J. Chromatogr. 360 (1986) 252–259. [4] G.I. Baram, J. Chromatogr. A 728 (1996) 387–399. [5] V.M. Tulchinsky, S. Technologies, A.B. Park, 2, pp. 281–285, 1998. [6] A. Ishida, M. Fujii, T. Fujimoto, S. Sasaki, I. Yanagisawa, H. Tani, M. Tokeshi, Anal. Sci. 31 (2015) 1163–1169. [7] S. Sharma, A. Plistil, R.S. Simpson, K. Liu, P.B. Farnsworth, S.D. Stearns, M.L. Lee, J. Chromatogr. A 1327 (2014) 80–89. [8] S. Sharma, A. Plistil, H.E. Barnett, H.D. Tolley, P.B. Farnsworth, S.D. Stearns, M.L. Lee, Anal. Chem. 87 (2015) 10457–10461. [9] Y. Li, M. Dvorak, P.N. Nesterenko, R. Stanley, N. Nuchtavorn, L.K. Krcmova, J. Aufartova, M. Macka, Anal. Chim. Acta 896 (2015) 166–176. [10] B. Yang, M. Zhang, T. Kanyanee, B.N. Stamos, P.K. Dasgupta, Anal. Chem. 86 (2014) 11554–11561. [11] S. Sharma, L.T. Tolley, H.D. Tolley, A. Plistil, S.D. Stearns, M.L. Lee, J. Chromatogr. A 1421 (2015) 38–47. [12] C.E. Nazario, M.R. Silva, M.S. Franco, F.M. Lanças, J. Chromatogr. A 1421 (2015) 18–37. [13] 〈http://labsmith.com/Spec_Sheets/LabSmith_AV201-AV202-AV303_Spec_Sheet_ 1015.pdf〉, 20/04/. [14] 〈http://products.labsmith.com/fluid-control-and-connectors/#.WFt52xt95v1〉, 20/04/. [15] P. Aggarwal, K. Liu, S. Sharma, J.S. Lawson, H.D. Tolley, M.L. Lee, J. Chromatogr. A 1380 (2015) 38–44. [16] R. Scott, C. Simpson, J. Chromatogr. Sci. 20 (1982) 62–66. [17] R.C. Simpson, J. Chromatogr. A 691 (1995) 163–170. [18] M.D. Foster, M.A. Arnold, J.A. Nichols, S.R. Bakalyar, J. Chromatogr. A 869 (2000) 231–241. [19] A. Prüß, C. Kempter, J. Gysler, T. Jira, J. Chromatogr. A 1016 (2003) 129–141. [20] H. Darcy, Dalmont, Paris, 647, 1856. [21] 〈http://www2.dupont.com/Vespel/en_US/assets/downloads/vespel_s/VPEA10948-00-A0311.pdf〉, 20/04/.
Fig. 3. A. Isocratic separation of biogenic amines (L-DOPA 6 mg L−1, norfenefrine 8 mg L−1, phenylephrine 12 mg L−1and 5-hydroxytryptophan 12 mg L−1) using a miniaturised capillary chromatographic system. A Conditions: Eluent 25 mM methanesulfonic acid; Detection: 280 nm photometric detection Isocratic separation of cations (Li 4 mg L−1, NH4 8 mg L−1, K 8 mg L−1, dimethylamine 15 mg L−1and trimethylamine 15 mg L−1). B. Conditions: Eluent 6 mM methanesulfonic acid; Detection: C4D detector, voltage −12 dB, gain 50%. Flow rate 6 µL min−1; column: IonPac CS17 (250 × 0.4 mm i.d.); injection volume: 448 nL. For more detail see the Experimental section in SI.
publication, where 100% methanol and acetonitrile were used as mobile phase in switching valves made from the same material [9]. 4. Conclusions This study evaluates a new miniaturised high pressure 6-port injection valve, and demonstrates its successful integration and use in a miniaturised, low cost modular capillary LC. The evaluated injection valve is inherently superior in terms of its small size and low weight to other injection valves as its highly compact design making it currently the smallest and lightest commercially available high pressure LC injection valve. However, it also shows a comparable level of performance to that of a standard HPLC nano-injection valve, with regard to its internal volume, carry-over, performance under different
35