- Email: [email protected]
Evaluation of preparative hydrodynamic chromatography of silica stationary phase supports
Journal of Chromatography A, 1370 (2014) 270–273
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
Short communication
Evaluation of preparative hydrodynamic chromatography of silica stationary phase supports James P. Grinias 1 , Justin M. Godinho 1 , Daniel B. Lunn, James W. Jorgenson ∗ Department of Chemistry, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599, USA
a r t i c l e
i n f o
Article history: Received 5 September 2014 Received in revised form 2 October 2014 Accepted 5 October 2014 Available online 14 October 2014 Keywords: Hydrodynamic chromatography Silica particles Liquid chromatography
a b s t r a c t Reducing the particle size distribution (PSD) of sub-2 m chromatographic packing materials can improve the performance of capillary UHPLC columns, but several size refinement methods are only partially effective in this size range. To this end, a preparative scale hydrodynamic chromatography (HDC) method was developed to size-refine C18 functionalized sub-2 m particles, but suffered from poor reproducibility and particle aggregation issues. Presented here are improvements based on the use of an ammonium hydroxide as the mobile phase. This mobile phase makes the method reproducible, decreases column conditioning requirements, and focuses on the preparation of bare silica material which allows for a wider variety of stationary phase bondings. Additionally, particle recovery for both non-porous silica size standards and bridged-ethyl hybrid (BEH) particles are detailed to highlight the advantages of this method. The data presented demonstrates the capability of this method to reduce the relative standard deviation (RSD) of the PSD of BEH particles by 33% in under 2 h with sufficient yield to pack several capillary columns. © 2014 Elsevier B.V. All rights reserved.
1. Introduction The synthesis of fully porous particles produces a wide particle size distribution (PSD), so further refinement must be carried out for improved chromatographic performance [1]. Because these particles are widely used in ultra-high pressure liquid chromatography (UHPLC) applications, reduction of the relative standard deviation (RSD) is desirable. While there have been a variety of findings regarding the impact of RSD on band broadening in different column types [2–7], two recent reports focused on capillary columns demonstrated the relationship between column performance and particle size segregation within columns packed with the same particle batch [8,9]. These two studies suggest refined particle batches may allow for improved and more consistent performance in capillary LC. Established methods used to accomplish this include sieving, sedimentation, centrifugation, elutriation, and air classification [10,11]. Unfortunately, these methods are either ineffective or very slow for particle diameters ranging from 0.5 to 2.0 m. Particle size refinement is further described in a review by Unger et al.
∗ Corresponding author. Tel.: +1 919 966 5071. E-mail address: [email protected] (J.W. Jorgenson). 1 Co-first authors. http://dx.doi.org/10.1016/j.chroma.2014.10.011 0021-9673/© 2014 Elsevier B.V. All rights reserved.
[11]. An alternative option for PSD refinement in this size range is hydrodynamic chromatography (HDC) [12]. In HDC, a particle (or molecule) has a specific elution volume related to its effective size [13–16]. As fluid moves through a channel, the velocity near the wall is slower than in the middle of the channel. Because larger particles cannot fully sample slow flow paths near the wall, they elute before the column dead time. As the particle diameter decreases, the elution volume increases and particles are separated by size. The separation mechanism can also be affected by steric effects, wall interactions, and external forces like gravity [17]. The first experimental demonstration of HDC was presented by Small, who used three packed HDC columns in series to separate a set of polystyrene latex beads for PSD analysis [18]. Since then, HDC has continued to be used for the analysis of PSD [14,19–21], but has also been demonstrated on the preparative scale with the purpose of particle refinement [12]. This preparative HDC method was reported by our laboratory and focused on decreasing the RSD of sub-2 m reversed phase packing materials [12]. This was accomplished by using a large diameter column packed with ∼34 m glass beads so that the interstitial channels effectively acted as individual HDC capillaries that enabled size refinement and recovery as opposed to just analysis. However, an empirical column conditioning method requiring four different mobile phases to suspend the reversed phase particles in acetone led to irreproducible retention times. A technique focused on
J.P. Grinias et al. / J. Chromatogr. A 1370 (2014) 270–273
particle refinement requires multiple injections for sufficient yield, and thus, shifting retention times are unacceptable and must be avoided. This follow-up communication focuses on a new, more robust preparative HDC method that uses a single mobile phase to eliminate the need for column conditioning while also preventing particle aggregation. Additionally, the actual material recovery, not explored in the previously reported preparative technique [12], is reported here. These studies further the viability and efficacy of this method for particle size refinement outside of reversed phase particles, while making operation much simpler and more reproducible. The method described here no longer focuses on a particular stationary phase, but instead size refines silica support material that can subsequently be bonded to meet nearly any selectivity requirements for a given separation.
2. Materials and methods 2.1. HDC column preparation and use 32–38 m glass beads (GP0035, Whitehouse Scientific Ltd., Chester, UK) were suspended in water. A magnet was used to remove metal fines remaining from the sieving process used during their production. The particles were dried and then dry packed into a 25 cm long × 25 mm inner diameter (i.d.) glass chromatography column (Kinesis USA, Malta, NY) by funneling them through a PTFE tube (7/16 i.d., 1/2 outer diameter (o.d.), McMasterCarr, Atlanta, GA) that was continuously raised to maintain the tube outlet approximately 1 cm above the forming bed. During packing, the side of the column was tapped with a plastic rod to settle the forming bed. After each subsequent 2 cm of bed formed, the column was tapped vertically for further consolidation. A 10 m stainless steel (SS) mesh frit (TWP, Inc., Berkeley, CA) was inserted into an adjustable endfitting (Kinesis USA, Malta, NY) placed at the inlet, which was then tightened to compress the bed and limit dead volume. A non-adjustable fitting was used at the outlet, where the same SS frit was used and secured in place using a light application of silicone sealant at the edge of the frit (DAP Products, Baltimore, MD). Flow was delivered using a Waters 600 Quaternary HPLC pump (Waters Corp., Milford, MA). During column preparation, the pump was connected to the column inlet with 0.0625 in. i.d. PTFE tubing and the column was filled with deionized water (Nanopure ultrapure water system (Barnstead International, Dubuque, IA) at 2 mL/min. Once filled, the flow rate was increased to 4 mL/min to further consolidate the packed bed. The final column length after consolidation was 22 cm. During analysis, the pump was connected to a six-port VICI electronic injector (Valco Instruments, Co., Inc.) with a 1 mL sample loop. The inlet fitting was then connected to the injector using 30 cm of 0.254 mm i.d. PEEK tubing. The column outlet was coupled to a 60 cm length of 600 m o.d., 300 m i.d. fused-silica capillary (Polymicro Technologies, Phoenix, AZ) with a detection window generated by removing a section of the polyimide coating. A Linear UV/Vis 200 detector (Thermo Scientific, Waltham, MA) was used for UV absorbance (turbidity) at 215 nm. The mobile phase was 1 mM ammonium hydroxide (ACS grade, Fisher Scientific, Hampton, NH) and the flow rate was set at 3 mL/min to ensure the system pressure did not exceed the glass tube limit of 150 psi. 60 min (∼4 column volumes) of column flushing were needed prior to column operation to bring the pH of the mobile phase measured at the outlet to ∼10. At the end of use, the column was thoroughly flushed with deionized water (∼1 h) before shutting down the system in order to reduce the pH, wash out any trapped analyte particles, and prevent column degradation.
271
2.2. HDC analysis and refinement of silica particles To test column performance, non-porous silica (NPS) size standards (Fiber Optic Center, Inc., New Bedford, MA) of 0.5 m and 1.5 m were slurried in the mobile phase at concentrations of 1.3 and 3.0 mg/mL, respectively. To each slurry, urea (ACS grade, Sigma Aldrich, St. Louis, MO) was added as a dead time marker for the column. To characterize particle recovery from the column, 10 mg/mL slurries of each NPS particle size were made from measured particle quantities. These slurries were injected in 1 mL aliquots and collected until the slurry had been fully used. The collected particles were then washed with water and centrifuged 2–3 times, dried, and weighed. Bridged-ethyl hybrid (BEH) particles of 1.0 m nominal diameter were obtained from Waters Corporation and prepared at a slurry concentration of 10 mg/mL in the mobile phase. Two fractions were collected with the separation point being at the peak maximum, 12.4 min after the injection. All slurries were sonicated prior to injection so that particles were dispersed. No particle damage was observed due to sonication. PSD values for HDC fractions of the BEH material were determined by scanning electron microscopy (SEM) with a Hitachi S-4700 cold cathode field emission SEM equipped with a Through the Lens (TTL) detector (Tokyo, Japan). Images were evaluated using Image J analysis software (http://imagej.nih.gov/ij/) to measure the particle size (statistical analysis was conducted in Microsoft Excel). Igor Pro 6.0 (Wavemetrics, Inc., Lake Oswego, OR) was used for graphical presentation and the determination of plate count (calculated from the basewidth of Gaussian peak fits).
3. Results and discussion 3.1. Separation and recovery of non-porous silica size standards by HDC In earlier work, a preparative HDC method utilized acetone as the mobile phase to size refine silica packing materials that had been bonded with C18 [12]. However, this required the column to be conditioned in a 14–16 h process with four separate mobile phases, including buffer and surfactant washes, prior to separation. This conditioning then had to be repeated after only a few hours of use. When simultaneously injecting multiple standards of varied size, separation efficiency decreased and peaks remained unresolved. This was attributed to particle aggregation occurring during the separation process, which changed the effective size of the particles over the length of the column and resulted in a broad eluted peak. In the current work, 1 mM ammonium hydroxide (pH ∼ 10) was used as a new mobile phase to solve many of these problems. Fig. 1 shows the elution of 0.5 and 1.5 m NPS particles from the HDC column and a chromatogram of the two mixed particle types. These elution times using a basic mobile phase were very reproducible when compared to the previous acetone method where they would unpredictably shift in relation to the urea dead time marker over multiple injections. Because the column reconditioning process to use acetone mobile phase is so lengthy, using only 1 mM ammonium hydroxide greatly improves not only the reproducibility but also the usable time of the column. In comparing the performance of a small dead time marker (urea), a 17 cm × 25 mm i.d. column gave a plate count (N) ∼ 3000 (reduced plate height (h) ∼ 1.6) [12] while here a 22 cm × 25 mm i.d. column had N ∼ 2600 (h∼ 2.5). Theoretical descriptions of HDC where broadening is dominated by convective mixing suggest a value for h of 1.4 [13]. These efficiency factors indicate that improvements to column packing may be one of the key factors needed for further advances to this technique [22]. Despite the column plate
272
J.P. Grinias et al. / J. Chromatogr. A 1370 (2014) 270–273
Fig. 1. Two overlaid HDC chromatograms of 0.5 (red) and 1.5 (black) m silica size standards (urea dead time marker in each run) of 1.3 and 3.0 mg/mL slurry concentrations, respectively. The blue trace shows a mixture of the two particle types injected simultaneously. The glass column was 22 cm × 25 mm diameter packed with 34 m glass beads, run at 3 mL/min in 1 mM ammonium hydroxide mobile phase. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
counts for a small dead time marker, the actual broadening of a size standard peak is more important when determining how much the PSD can be reduced. This is because broader particle peaks will have wider size distributions when a real packing material is injected. Compared to the efficiency of a urea peak, the efficiency of a ∼0.5 m NPS peak drops ∼40% with the acetone mobile phase [12] and only ∼30% with the 1 mM ammonium hydroxide mobile phase. The lower HDC separation efficiency is most likely due to particle aggregation in the acetone mobile phase. Here, the ammonium hydroxide mobile phase (pH ∼ 10) charges the particles so they repel each other throughout the HDC column. While this method could be used with bonded stationary phases (provided that the particles are well-dispersed and stable in this mobile phase) designing the technique to be used for bare silica allows for subsequent bonding with any bonded phase. This, in addition to improved reproducibility and great reduction in column conditioning time, is why ammonium hydroxide is a critical advancement to preparative scale HDC of silica particles. In the previous report, larger particles give a lower turbidity signal than smaller particles when injected onto the column [12]. This can partly be attributed to the relative number of particles in solution when preparing slurries on a weight basis (as there are ∼27 times more 0.5 m particles than 1.5 m particles per unit mass), although the detector signal does not have an exact correlation with the number of particles present. Another reason for this lower signal was due to the frit choice used in column preparation. In the previous report, the end fittings were implemented with manufacturer-provided polyethylene (PE) frits. When using
Table 1 HDC injected particle recovery. Particle size/typea
Mass injectedb (mg)
Mass recoveredc (mg)
0.5 m NPS 1.5 m NPS 0.9 m BEH Fraction 1 Fraction 2
35.4 39.0
30.4 28.8
86% 74%
N/A N/A
31.2 51.9
36% 52%
Total
87.4
76.6
88%
a
Yield
NPS, non-porous silica, BEH, bridged-ethyl hybrid. b The mass injected was calculated by taking an initial known quantity of dry particles and subtracting material recovered from injection loop overfill (material not lost to the column or wash steps). c The mass recovered does not include particles lost to the column or subsequent wash steps of collected fractions.
Fig. 2. Histograms shown for ∼100 particles sized by SEM for two HDC fractions collected from an initial 1.0 m BEH particle batch. Average size values (reported with one standard deviation) are: 1.0 m BEH starting material (A): 1.02 ± 0.24 m; Fraction 1 (B): 1.10 ± 0.18 m; Fraction 2 (C): 0.84 ± 0.16 m.
columns packed with these frits, repeated injections of particles 1.5 m or larger led to both signal reduction and increased back pressure. Examination of used frits with SEM showed significant silica particle build-up in the porous frit material. To reduce particle entrapment at the column inlet, a 10 m SS wire mesh frits were purchased separately and inserted into the fittings to replace the PE frits. With the SS frit, both the backpressure increases and signal loss due to frit clogging were greatly reduced even though the nominal pore size was less than half that of the PE frit. Rather than the earlier investigation’s dependence on the relative turbidity signal to determine the throughput of the column, here the actual particle recovery was studied. Known masses of 0.5 and 1.5 m particles were slurried in the mobile phase at a concentration of ∼10 mg/mL. Multiple injections were made and the entire eluted peak was collected for each injection. Then the particles were washed and dried to determine the particle recovery. 86% of the 0.5 m NPS particles were recovered following HDC and
J.P. Grinias et al. / J. Chromatogr. A 1370 (2014) 270–273
273
subsequent washing while 74% of the 1.5 m NPS particles were recovered (see Table 1). From these values, it can be determined that the larger particles have a lower recovery which is most likely due to trapping at the column frits. With both sizes, this reported recovery follows transfer, centrifugation, and washing of particles, procedures that can also reduce the recovery yield. While some of these particles may slowly leech onto the column over time, thorough flushing between column uses reduces carryover from day-to-day.
than smaller particles. Because a large diameter HDC column is employed, enough refined material for packing multiple capillary UHPLC columns can be obtained in a relatively short time.
3.2. Size refinement and recovery of 1.0 m BEH particles by HDC
References
While the NPS size standards described above have a relatively low RSD (<10%), sub-2 m porous particles usually have a wider PSD that can be reduced with preparative HDC by taking fractions of the eluted particle peak. To characterize such a separation, a ∼10 mg/mL slurry of 1.0 m BEH particles was prepared and several injections were made where the front and back half of the peak were collected separately. As the elution times were very reproducible, this cutoff time was also reproducible at 12.4 min after the injector was switched. In Fig. 2, histograms of the size distribution for the raw material prior to HDC (Fig. 2A) and the peak front (Fig. 2B) and peak tail (Fig. 2C) following HDC are all shown. While the raw material had an RSD value of 24%, the peak front dropped to 16% and the tail dropped to 19%. The column caused some peak tailing, due to bed structure and analyte distribution at the outlet of the column, which decreased the refinement for this portion of the peak. However, usually only the first peak fraction is used for packing material as it has had particle fines that can increase the column flow resistance [4,12] removed. The amount of material recovered for each fraction was determined by particle mass. The total recovery of the entire peak was 88% with the front half contributing 36% of the mass and the back half contributing 52% of the mass. The relative values for these two fractions again demonstrate that larger particles have a lower yield than smaller particles after being injected into the column. The total time that was needed to acquire the ∼31 mg of refined material (Fraction 1) was less than 2 h and achieved by overlapping injections as the particles elute in a narrow time window. Since it is bare silica, any number of stationary phases (compared to only C18 as with the acetone mobile phase method) could then be bonded to these particles with enough material to pack several capillary columns. 4. Conclusions In this report we have described improvements to a previously developed preparative HDC method [12] designed to reduce the PSD of sub-2 m fully porous particles. By switching from an acetone mobile phase to a basic ammonium hydroxide mobile phase, particle elution times are far more reproducible and the need for column conditioning is eliminated. Additionally, while only reversed phase particles could be separated before, now any stationary phase can be applied to the bare silica that is size-refined prior to bonding. The actual particle recovery of preparative HDC has now been characterized and was found to have yields of 75–90% of the material injected with larger particles having lower yield
Acknowledgements We would like to thank Dr. Kevin Wyndham (Waters Corporation) for kindly contributing the BEH particles used in the study and to the Waters Corporation for funding this research.
[1] J.S. Mellors, J.W. Jorgenson, Use of 1.5-m porous ethyl-bridged hybrid particles as a stationary-phase support for reversed-phase ultrahigh-pressure liquid chromatography, Anal. Chem. 76 (2004) 5441–5450. [2] I. Halász, M. Naefe, Influence of column parameters on peak broadening in high-pressure liquid chromatography, Anal. Chem. 44 (1972) 76–84. [3] C. Dewaele, M. Verzele, Influence of the particle size distribution of the packing material in reversed-phase high-performance liquid chromatography, J. Chromatogr. 260 (1983) 13–21. [4] J. Billen, D. Guillarme, S. Rudaz, J.-L. Beuthey, H. Ritchie, B. Grady, G. Desmet, Relation between the particle size distribution and the kinetic performance of packed columns. Application to a commercial sub-2 m material, J. Chromatogr. A 1161 (2007) 224–233. [5] R.A. Lieberman, Evaluation of Micron Sized Silica Based Packing Material for Ultra High Pressure Capillary Liquid Chromatography, Dept. of Chemistry, University of North Carolina at Chapel Hill, Chapel Hill, NC, 2009. [6] F. Gritti, T. Farkas, J. Heng, G. Guiochon, On the relationship between band broadening and the particle-size distribution of the packing material in liquid chromatography: theory and practice, J. Chromatogr. A 1218 (2011) 8209–8221. [7] A. Daneyko, A. Höltzel, S. Khirevich, U. Tallarek, Influence of the particle size distribution on hydraulic permeability and eddy dispersion in bulk packings, Anal. Chem. 83 (2011) 3903–3910. [8] S. Bruns, J.P. Grinias, L.E. Blue, J.W. Jorgenson, U. Tallarek, Morphology and separation efficiency of low-aspect-ratio capillary ultrahigh pressure liquid chromatography columns, Anal. Chem. 84 (2012) 4496–4503. [9] S. Bruns, E.G. Franklin, J.P. Grinias, J.M. Godinho, J.W. Jorgenson, U. Tallarek, Slurry concentration effects on bed morphology and separation efficiency of capillaries packed with sub-2 m particles, J. Chromatogr. A 1318 (2013) 189–197. [10] U.D. Neue, HPLC Columns: Theory, Technology and Practice, Wiley-VCH, Inc., New York, NY, 1997. [11] K.K. Unger, R. Skudas, M.M. Schulte, Particle packed columns and monolithic columns in high-performance liquid chromatography-comparison and critical appraisal, J. Chromatogr. A 1184 (2008) 393–415. [12] J.W. Thompson, R.A. Lieberman, J.W. Jorgenson, Hydrodynamic chromatography for the size classification of micron and sub-micron sized packing materials, J. Chromatogr. A 1216 (2009) 7732–7738. [13] A.M. Striegel, A.K. Brewer, Hydrodynamic chromatography, Ann. Rev. Anal. Chem. 5 (2012) 15–34. [14] A.M. Striegel, Hydrodynamic chromatography: packed columns, multiple detectors, and microcapillaries, Anal. Bioanal. Chem. 402 (2012) 77–81. [15] E.A. DiMarzio, C.M. Guttmann, Separation by flow, J. Polym. Sci. B: Lett. 7 (1969) 267–272. [16] E.A. DiMarzio, C.M. Guttmann, Separation by flow, Macromolecules 3 (1970) 131–146. [17] T.N. Shendruk, R. Tahvildari, N.M. Catafard, L. Andrzejewski, C. Gigault, A. Todd, L. Gagne-Dumais, G.W. Slater, M. Godin, Field-flow fractionation and hydrodynamic chromatography on a microfluidic chip, Anal. Chem. 85 (2013) 5981–5988. [18] H. Small, Hydrodynamic chromatography a technique for size analysis of colloidal particles, J. Colloid Interface Sci. 48 (1974) 147–161. [19] H. Small, M.A. Langhorst, Hydrodynamic chromatography, Anal. Chem. 54 (1982) 892A–898A. [20] R. Tijssen, J. Bos, M.E. van Kreveld, Hydrodynamic chromatography of macromolecules in open microcapillary tubes, Anal. Chem. 58 (1986) 3036–3044. [21] E. Venema, J.C. Kraak, H. Poppe, R. Tijssen, Packed-column hydrodynamic chromatography using 1-m non-porous silica particles, J. Chomatogr. A 740 (1996) 159–167. [22] S.M.F. Tavernier, E. Nies, R. Gijbels, Hard-sphere model for hydrodynamic chromatography systems, Anal. Proc. 18 (1981) 31–34.
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
Short communication
Evaluation of preparative hydrodynamic chromatography of silica stationary phase supports James P. Grinias 1 , Justin M. Godinho 1 , Daniel B. Lunn, James W. Jorgenson ∗ Department of Chemistry, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599, USA
a r t i c l e
i n f o
Article history: Received 5 September 2014 Received in revised form 2 October 2014 Accepted 5 October 2014 Available online 14 October 2014 Keywords: Hydrodynamic chromatography Silica particles Liquid chromatography
a b s t r a c t Reducing the particle size distribution (PSD) of sub-2 m chromatographic packing materials can improve the performance of capillary UHPLC columns, but several size refinement methods are only partially effective in this size range. To this end, a preparative scale hydrodynamic chromatography (HDC) method was developed to size-refine C18 functionalized sub-2 m particles, but suffered from poor reproducibility and particle aggregation issues. Presented here are improvements based on the use of an ammonium hydroxide as the mobile phase. This mobile phase makes the method reproducible, decreases column conditioning requirements, and focuses on the preparation of bare silica material which allows for a wider variety of stationary phase bondings. Additionally, particle recovery for both non-porous silica size standards and bridged-ethyl hybrid (BEH) particles are detailed to highlight the advantages of this method. The data presented demonstrates the capability of this method to reduce the relative standard deviation (RSD) of the PSD of BEH particles by 33% in under 2 h with sufficient yield to pack several capillary columns. © 2014 Elsevier B.V. All rights reserved.
1. Introduction The synthesis of fully porous particles produces a wide particle size distribution (PSD), so further refinement must be carried out for improved chromatographic performance [1]. Because these particles are widely used in ultra-high pressure liquid chromatography (UHPLC) applications, reduction of the relative standard deviation (RSD) is desirable. While there have been a variety of findings regarding the impact of RSD on band broadening in different column types [2–7], two recent reports focused on capillary columns demonstrated the relationship between column performance and particle size segregation within columns packed with the same particle batch [8,9]. These two studies suggest refined particle batches may allow for improved and more consistent performance in capillary LC. Established methods used to accomplish this include sieving, sedimentation, centrifugation, elutriation, and air classification [10,11]. Unfortunately, these methods are either ineffective or very slow for particle diameters ranging from 0.5 to 2.0 m. Particle size refinement is further described in a review by Unger et al.
∗ Corresponding author. Tel.: +1 919 966 5071. E-mail address: [email protected] (J.W. Jorgenson). 1 Co-first authors. http://dx.doi.org/10.1016/j.chroma.2014.10.011 0021-9673/© 2014 Elsevier B.V. All rights reserved.
[11]. An alternative option for PSD refinement in this size range is hydrodynamic chromatography (HDC) [12]. In HDC, a particle (or molecule) has a specific elution volume related to its effective size [13–16]. As fluid moves through a channel, the velocity near the wall is slower than in the middle of the channel. Because larger particles cannot fully sample slow flow paths near the wall, they elute before the column dead time. As the particle diameter decreases, the elution volume increases and particles are separated by size. The separation mechanism can also be affected by steric effects, wall interactions, and external forces like gravity [17]. The first experimental demonstration of HDC was presented by Small, who used three packed HDC columns in series to separate a set of polystyrene latex beads for PSD analysis [18]. Since then, HDC has continued to be used for the analysis of PSD [14,19–21], but has also been demonstrated on the preparative scale with the purpose of particle refinement [12]. This preparative HDC method was reported by our laboratory and focused on decreasing the RSD of sub-2 m reversed phase packing materials [12]. This was accomplished by using a large diameter column packed with ∼34 m glass beads so that the interstitial channels effectively acted as individual HDC capillaries that enabled size refinement and recovery as opposed to just analysis. However, an empirical column conditioning method requiring four different mobile phases to suspend the reversed phase particles in acetone led to irreproducible retention times. A technique focused on
J.P. Grinias et al. / J. Chromatogr. A 1370 (2014) 270–273
particle refinement requires multiple injections for sufficient yield, and thus, shifting retention times are unacceptable and must be avoided. This follow-up communication focuses on a new, more robust preparative HDC method that uses a single mobile phase to eliminate the need for column conditioning while also preventing particle aggregation. Additionally, the actual material recovery, not explored in the previously reported preparative technique [12], is reported here. These studies further the viability and efficacy of this method for particle size refinement outside of reversed phase particles, while making operation much simpler and more reproducible. The method described here no longer focuses on a particular stationary phase, but instead size refines silica support material that can subsequently be bonded to meet nearly any selectivity requirements for a given separation.
2. Materials and methods 2.1. HDC column preparation and use 32–38 m glass beads (GP0035, Whitehouse Scientific Ltd., Chester, UK) were suspended in water. A magnet was used to remove metal fines remaining from the sieving process used during their production. The particles were dried and then dry packed into a 25 cm long × 25 mm inner diameter (i.d.) glass chromatography column (Kinesis USA, Malta, NY) by funneling them through a PTFE tube (7/16 i.d., 1/2 outer diameter (o.d.), McMasterCarr, Atlanta, GA) that was continuously raised to maintain the tube outlet approximately 1 cm above the forming bed. During packing, the side of the column was tapped with a plastic rod to settle the forming bed. After each subsequent 2 cm of bed formed, the column was tapped vertically for further consolidation. A 10 m stainless steel (SS) mesh frit (TWP, Inc., Berkeley, CA) was inserted into an adjustable endfitting (Kinesis USA, Malta, NY) placed at the inlet, which was then tightened to compress the bed and limit dead volume. A non-adjustable fitting was used at the outlet, where the same SS frit was used and secured in place using a light application of silicone sealant at the edge of the frit (DAP Products, Baltimore, MD). Flow was delivered using a Waters 600 Quaternary HPLC pump (Waters Corp., Milford, MA). During column preparation, the pump was connected to the column inlet with 0.0625 in. i.d. PTFE tubing and the column was filled with deionized water (Nanopure ultrapure water system (Barnstead International, Dubuque, IA) at 2 mL/min. Once filled, the flow rate was increased to 4 mL/min to further consolidate the packed bed. The final column length after consolidation was 22 cm. During analysis, the pump was connected to a six-port VICI electronic injector (Valco Instruments, Co., Inc.) with a 1 mL sample loop. The inlet fitting was then connected to the injector using 30 cm of 0.254 mm i.d. PEEK tubing. The column outlet was coupled to a 60 cm length of 600 m o.d., 300 m i.d. fused-silica capillary (Polymicro Technologies, Phoenix, AZ) with a detection window generated by removing a section of the polyimide coating. A Linear UV/Vis 200 detector (Thermo Scientific, Waltham, MA) was used for UV absorbance (turbidity) at 215 nm. The mobile phase was 1 mM ammonium hydroxide (ACS grade, Fisher Scientific, Hampton, NH) and the flow rate was set at 3 mL/min to ensure the system pressure did not exceed the glass tube limit of 150 psi. 60 min (∼4 column volumes) of column flushing were needed prior to column operation to bring the pH of the mobile phase measured at the outlet to ∼10. At the end of use, the column was thoroughly flushed with deionized water (∼1 h) before shutting down the system in order to reduce the pH, wash out any trapped analyte particles, and prevent column degradation.
271
2.2. HDC analysis and refinement of silica particles To test column performance, non-porous silica (NPS) size standards (Fiber Optic Center, Inc., New Bedford, MA) of 0.5 m and 1.5 m were slurried in the mobile phase at concentrations of 1.3 and 3.0 mg/mL, respectively. To each slurry, urea (ACS grade, Sigma Aldrich, St. Louis, MO) was added as a dead time marker for the column. To characterize particle recovery from the column, 10 mg/mL slurries of each NPS particle size were made from measured particle quantities. These slurries were injected in 1 mL aliquots and collected until the slurry had been fully used. The collected particles were then washed with water and centrifuged 2–3 times, dried, and weighed. Bridged-ethyl hybrid (BEH) particles of 1.0 m nominal diameter were obtained from Waters Corporation and prepared at a slurry concentration of 10 mg/mL in the mobile phase. Two fractions were collected with the separation point being at the peak maximum, 12.4 min after the injection. All slurries were sonicated prior to injection so that particles were dispersed. No particle damage was observed due to sonication. PSD values for HDC fractions of the BEH material were determined by scanning electron microscopy (SEM) with a Hitachi S-4700 cold cathode field emission SEM equipped with a Through the Lens (TTL) detector (Tokyo, Japan). Images were evaluated using Image J analysis software (http://imagej.nih.gov/ij/) to measure the particle size (statistical analysis was conducted in Microsoft Excel). Igor Pro 6.0 (Wavemetrics, Inc., Lake Oswego, OR) was used for graphical presentation and the determination of plate count (calculated from the basewidth of Gaussian peak fits).
3. Results and discussion 3.1. Separation and recovery of non-porous silica size standards by HDC In earlier work, a preparative HDC method utilized acetone as the mobile phase to size refine silica packing materials that had been bonded with C18 [12]. However, this required the column to be conditioned in a 14–16 h process with four separate mobile phases, including buffer and surfactant washes, prior to separation. This conditioning then had to be repeated after only a few hours of use. When simultaneously injecting multiple standards of varied size, separation efficiency decreased and peaks remained unresolved. This was attributed to particle aggregation occurring during the separation process, which changed the effective size of the particles over the length of the column and resulted in a broad eluted peak. In the current work, 1 mM ammonium hydroxide (pH ∼ 10) was used as a new mobile phase to solve many of these problems. Fig. 1 shows the elution of 0.5 and 1.5 m NPS particles from the HDC column and a chromatogram of the two mixed particle types. These elution times using a basic mobile phase were very reproducible when compared to the previous acetone method where they would unpredictably shift in relation to the urea dead time marker over multiple injections. Because the column reconditioning process to use acetone mobile phase is so lengthy, using only 1 mM ammonium hydroxide greatly improves not only the reproducibility but also the usable time of the column. In comparing the performance of a small dead time marker (urea), a 17 cm × 25 mm i.d. column gave a plate count (N) ∼ 3000 (reduced plate height (h) ∼ 1.6) [12] while here a 22 cm × 25 mm i.d. column had N ∼ 2600 (h∼ 2.5). Theoretical descriptions of HDC where broadening is dominated by convective mixing suggest a value for h of 1.4 [13]. These efficiency factors indicate that improvements to column packing may be one of the key factors needed for further advances to this technique [22]. Despite the column plate
272
J.P. Grinias et al. / J. Chromatogr. A 1370 (2014) 270–273
Fig. 1. Two overlaid HDC chromatograms of 0.5 (red) and 1.5 (black) m silica size standards (urea dead time marker in each run) of 1.3 and 3.0 mg/mL slurry concentrations, respectively. The blue trace shows a mixture of the two particle types injected simultaneously. The glass column was 22 cm × 25 mm diameter packed with 34 m glass beads, run at 3 mL/min in 1 mM ammonium hydroxide mobile phase. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
counts for a small dead time marker, the actual broadening of a size standard peak is more important when determining how much the PSD can be reduced. This is because broader particle peaks will have wider size distributions when a real packing material is injected. Compared to the efficiency of a urea peak, the efficiency of a ∼0.5 m NPS peak drops ∼40% with the acetone mobile phase [12] and only ∼30% with the 1 mM ammonium hydroxide mobile phase. The lower HDC separation efficiency is most likely due to particle aggregation in the acetone mobile phase. Here, the ammonium hydroxide mobile phase (pH ∼ 10) charges the particles so they repel each other throughout the HDC column. While this method could be used with bonded stationary phases (provided that the particles are well-dispersed and stable in this mobile phase) designing the technique to be used for bare silica allows for subsequent bonding with any bonded phase. This, in addition to improved reproducibility and great reduction in column conditioning time, is why ammonium hydroxide is a critical advancement to preparative scale HDC of silica particles. In the previous report, larger particles give a lower turbidity signal than smaller particles when injected onto the column [12]. This can partly be attributed to the relative number of particles in solution when preparing slurries on a weight basis (as there are ∼27 times more 0.5 m particles than 1.5 m particles per unit mass), although the detector signal does not have an exact correlation with the number of particles present. Another reason for this lower signal was due to the frit choice used in column preparation. In the previous report, the end fittings were implemented with manufacturer-provided polyethylene (PE) frits. When using
Table 1 HDC injected particle recovery. Particle size/typea
Mass injectedb (mg)
Mass recoveredc (mg)
0.5 m NPS 1.5 m NPS 0.9 m BEH Fraction 1 Fraction 2
35.4 39.0
30.4 28.8
86% 74%
N/A N/A
31.2 51.9
36% 52%
Total
87.4
76.6
88%
a
Yield
NPS, non-porous silica, BEH, bridged-ethyl hybrid. b The mass injected was calculated by taking an initial known quantity of dry particles and subtracting material recovered from injection loop overfill (material not lost to the column or wash steps). c The mass recovered does not include particles lost to the column or subsequent wash steps of collected fractions.
Fig. 2. Histograms shown for ∼100 particles sized by SEM for two HDC fractions collected from an initial 1.0 m BEH particle batch. Average size values (reported with one standard deviation) are: 1.0 m BEH starting material (A): 1.02 ± 0.24 m; Fraction 1 (B): 1.10 ± 0.18 m; Fraction 2 (C): 0.84 ± 0.16 m.
columns packed with these frits, repeated injections of particles 1.5 m or larger led to both signal reduction and increased back pressure. Examination of used frits with SEM showed significant silica particle build-up in the porous frit material. To reduce particle entrapment at the column inlet, a 10 m SS wire mesh frits were purchased separately and inserted into the fittings to replace the PE frits. With the SS frit, both the backpressure increases and signal loss due to frit clogging were greatly reduced even though the nominal pore size was less than half that of the PE frit. Rather than the earlier investigation’s dependence on the relative turbidity signal to determine the throughput of the column, here the actual particle recovery was studied. Known masses of 0.5 and 1.5 m particles were slurried in the mobile phase at a concentration of ∼10 mg/mL. Multiple injections were made and the entire eluted peak was collected for each injection. Then the particles were washed and dried to determine the particle recovery. 86% of the 0.5 m NPS particles were recovered following HDC and
J.P. Grinias et al. / J. Chromatogr. A 1370 (2014) 270–273
273
subsequent washing while 74% of the 1.5 m NPS particles were recovered (see Table 1). From these values, it can be determined that the larger particles have a lower recovery which is most likely due to trapping at the column frits. With both sizes, this reported recovery follows transfer, centrifugation, and washing of particles, procedures that can also reduce the recovery yield. While some of these particles may slowly leech onto the column over time, thorough flushing between column uses reduces carryover from day-to-day.
than smaller particles. Because a large diameter HDC column is employed, enough refined material for packing multiple capillary UHPLC columns can be obtained in a relatively short time.
3.2. Size refinement and recovery of 1.0 m BEH particles by HDC
References
While the NPS size standards described above have a relatively low RSD (<10%), sub-2 m porous particles usually have a wider PSD that can be reduced with preparative HDC by taking fractions of the eluted particle peak. To characterize such a separation, a ∼10 mg/mL slurry of 1.0 m BEH particles was prepared and several injections were made where the front and back half of the peak were collected separately. As the elution times were very reproducible, this cutoff time was also reproducible at 12.4 min after the injector was switched. In Fig. 2, histograms of the size distribution for the raw material prior to HDC (Fig. 2A) and the peak front (Fig. 2B) and peak tail (Fig. 2C) following HDC are all shown. While the raw material had an RSD value of 24%, the peak front dropped to 16% and the tail dropped to 19%. The column caused some peak tailing, due to bed structure and analyte distribution at the outlet of the column, which decreased the refinement for this portion of the peak. However, usually only the first peak fraction is used for packing material as it has had particle fines that can increase the column flow resistance [4,12] removed. The amount of material recovered for each fraction was determined by particle mass. The total recovery of the entire peak was 88% with the front half contributing 36% of the mass and the back half contributing 52% of the mass. The relative values for these two fractions again demonstrate that larger particles have a lower yield than smaller particles after being injected into the column. The total time that was needed to acquire the ∼31 mg of refined material (Fraction 1) was less than 2 h and achieved by overlapping injections as the particles elute in a narrow time window. Since it is bare silica, any number of stationary phases (compared to only C18 as with the acetone mobile phase method) could then be bonded to these particles with enough material to pack several capillary columns. 4. Conclusions In this report we have described improvements to a previously developed preparative HDC method [12] designed to reduce the PSD of sub-2 m fully porous particles. By switching from an acetone mobile phase to a basic ammonium hydroxide mobile phase, particle elution times are far more reproducible and the need for column conditioning is eliminated. Additionally, while only reversed phase particles could be separated before, now any stationary phase can be applied to the bare silica that is size-refined prior to bonding. The actual particle recovery of preparative HDC has now been characterized and was found to have yields of 75–90% of the material injected with larger particles having lower yield
Acknowledgements We would like to thank Dr. Kevin Wyndham (Waters Corporation) for kindly contributing the BEH particles used in the study and to the Waters Corporation for funding this research.
[1] J.S. Mellors, J.W. Jorgenson, Use of 1.5-m porous ethyl-bridged hybrid particles as a stationary-phase support for reversed-phase ultrahigh-pressure liquid chromatography, Anal. Chem. 76 (2004) 5441–5450. [2] I. Halász, M. Naefe, Influence of column parameters on peak broadening in high-pressure liquid chromatography, Anal. Chem. 44 (1972) 76–84. [3] C. Dewaele, M. Verzele, Influence of the particle size distribution of the packing material in reversed-phase high-performance liquid chromatography, J. Chromatogr. 260 (1983) 13–21. [4] J. Billen, D. Guillarme, S. Rudaz, J.-L. Beuthey, H. Ritchie, B. Grady, G. Desmet, Relation between the particle size distribution and the kinetic performance of packed columns. Application to a commercial sub-2 m material, J. Chromatogr. A 1161 (2007) 224–233. [5] R.A. Lieberman, Evaluation of Micron Sized Silica Based Packing Material for Ultra High Pressure Capillary Liquid Chromatography, Dept. of Chemistry, University of North Carolina at Chapel Hill, Chapel Hill, NC, 2009. [6] F. Gritti, T. Farkas, J. Heng, G. Guiochon, On the relationship between band broadening and the particle-size distribution of the packing material in liquid chromatography: theory and practice, J. Chromatogr. A 1218 (2011) 8209–8221. [7] A. Daneyko, A. Höltzel, S. Khirevich, U. Tallarek, Influence of the particle size distribution on hydraulic permeability and eddy dispersion in bulk packings, Anal. Chem. 83 (2011) 3903–3910. [8] S. Bruns, J.P. Grinias, L.E. Blue, J.W. Jorgenson, U. Tallarek, Morphology and separation efficiency of low-aspect-ratio capillary ultrahigh pressure liquid chromatography columns, Anal. Chem. 84 (2012) 4496–4503. [9] S. Bruns, E.G. Franklin, J.P. Grinias, J.M. Godinho, J.W. Jorgenson, U. Tallarek, Slurry concentration effects on bed morphology and separation efficiency of capillaries packed with sub-2 m particles, J. Chromatogr. A 1318 (2013) 189–197. [10] U.D. Neue, HPLC Columns: Theory, Technology and Practice, Wiley-VCH, Inc., New York, NY, 1997. [11] K.K. Unger, R. Skudas, M.M. Schulte, Particle packed columns and monolithic columns in high-performance liquid chromatography-comparison and critical appraisal, J. Chromatogr. A 1184 (2008) 393–415. [12] J.W. Thompson, R.A. Lieberman, J.W. Jorgenson, Hydrodynamic chromatography for the size classification of micron and sub-micron sized packing materials, J. Chromatogr. A 1216 (2009) 7732–7738. [13] A.M. Striegel, A.K. Brewer, Hydrodynamic chromatography, Ann. Rev. Anal. Chem. 5 (2012) 15–34. [14] A.M. Striegel, Hydrodynamic chromatography: packed columns, multiple detectors, and microcapillaries, Anal. Bioanal. Chem. 402 (2012) 77–81. [15] E.A. DiMarzio, C.M. Guttmann, Separation by flow, J. Polym. Sci. B: Lett. 7 (1969) 267–272. [16] E.A. DiMarzio, C.M. Guttmann, Separation by flow, Macromolecules 3 (1970) 131–146. [17] T.N. Shendruk, R. Tahvildari, N.M. Catafard, L. Andrzejewski, C. Gigault, A. Todd, L. Gagne-Dumais, G.W. Slater, M. Godin, Field-flow fractionation and hydrodynamic chromatography on a microfluidic chip, Anal. Chem. 85 (2013) 5981–5988. [18] H. Small, Hydrodynamic chromatography a technique for size analysis of colloidal particles, J. Colloid Interface Sci. 48 (1974) 147–161. [19] H. Small, M.A. Langhorst, Hydrodynamic chromatography, Anal. Chem. 54 (1982) 892A–898A. [20] R. Tijssen, J. Bos, M.E. van Kreveld, Hydrodynamic chromatography of macromolecules in open microcapillary tubes, Anal. Chem. 58 (1986) 3036–3044. [21] E. Venema, J.C. Kraak, H. Poppe, R. Tijssen, Packed-column hydrodynamic chromatography using 1-m non-porous silica particles, J. Chomatogr. A 740 (1996) 159–167. [22] S.M.F. Tavernier, E. Nies, R. Gijbels, Hard-sphere model for hydrodynamic chromatography systems, Anal. Proc. 18 (1981) 31–34.