Post on 18-Mar-2018
transcript
Spherical Carbon Core-Porous Nanodiamond/Polymer Shell
Particles for Reversed-Phase HPLC
Journal: Analytical Chemistry
Manuscript ID: Draft
Manuscript Type: Article
Date Submitted by the Author:
n/a
Complete List of Authors: Wiest, Landon; Brigham Young University, Chemistry and Biochemistry Jensen, David; Brigham Young University, Chemistry and Biochemistry Hung, Chuan-Hsi; Brigham Young University, Chemistry and Biochemistry Olsen, Betsy; Brigham Young University, Chemistry and Biochemistry Davis, Robert; Brigham Young University, Physics and Astronomy Vail, Michael; US Synthetic Corporation Dadson, Andrew; US Synthetic, R&D; US Synthetic Corporation
Linford, Matthew; Brigham Young University, Chemistry and Biochemistry
ACS Paragon Plus Environment
Submitted to Analytical Chemistry
1
Spherical Carbon Core-Porous Nanodiamond/Polymer
Shell Particles for Reversed-Phase HPLC
Landon A. Wiesta, David S. Jensen
a, Chuan-Hsi Hung
a, Rebecca E. Olsen
a, Robert C. Davis
b, Michael
A. Vailc, Andrew Dadson
c, Matthew R. Linford
a*.
a Department of Chemistry and Biochemistry, Brigham Young University, Provo, UT 84602, USA
b Department of Physics & Astronomy, Brigham Young University, Provo, UT 84602, USA
c US Synthetic Corporation, Orem, UT 84058, USA
* Correspondence should be sent to mrlinford@chem.byu.edu or to the above address
ABSTRACT We report the development of chromatographic phases created by coating spherical 3 µm carbon particles with poly(allylamine) (PAAm) and nanodiamond for high-performance liquid chromatography (HPLC). These hybrid particles are prepared via layer-by-layer (LbL) deposition of PAAm and nanodiamond, resulting in core-shell (pellicular) particles. The starting material (core carbon particles) is first immersed in a solution of PAAm, which leads to the self-limiting adsorption of this polymer. After washing, the particles are immersed in a slurry of nanodiamond, which leads to its self-limiting deposition. Alternating, self-limiting, depositions of PAAm and nanodiamond are continued until the desired thickness of a porous PAAm/nanodiamond shell is formed around the carbon core. Finally, the core-shell particles are simultaneously functionalized and cross-linked with a mixture of 1,2-epoxyoctadecane and 1,2,7,8-diepoxyoctane to create a mechanically stable C18 phase. Core-shell particles are characterized by scanning electron microscopy (SEM), and their surface area, pore diameter, and volume are determined using the Brunauer-Emmett-Teller (BET) method. Particle size distribution (PSD) measurements are also obtained. Columns packed with these ca. 4 µm particles showed efficiencies of 56 000 N/m for n-butylbenzene. Van Deemter studies were performed, although the C term from this analysis and the particle size distribution pointed to particle agglomeration. The particles show considerable stability at high pH (11.3 and even 13) over extended periods of time. At pH 11.3, a ca. 5% loss in k was observed after 1 600 column volumes of mobile phase passed through the column. This experiment was followed by one at pH 13.0 in which a ca. 1% loss in k was observed over a 1 000 column volume period, suggesting considerable stability at high pH for this phase. When a narrower particle size distribution was obtained, the Van Deemter curve showed the expected lowering of its C term. The Supporting Information of this work contains a MATLAB program, which conveniently plots Van Deemter curves, including the individual contributions of the A, B, and C terms, the residuals to the fit, the values of A, B, and C, the rmse and R2 values for the fit, and the optimal values of u and H.
Page 1 of 23
ACS Paragon Plus Environment
Submitted to Analytical Chemistry
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
2
1. INTRODUCTION
Silica is the workhorse of modern liquid chromatography.1-2 Accordingly, its surface has been
extensively studied and modified, which has led to a broad array of available functionalities for the
chromatographer. However, despite its flexibility, silica lacks stability at both high and low pHs.
Indeed, for many silica-based chromatographic materials, the useful window of pH stability lies between
ca. 3 and 8,3-4 although if better silane ligands and/or modified silica, such as silica-polymer hybrids,5
are used this range can be extended. Nevertheless, there remains some question regarding the long-term
stability of these materials, especially at elevated pH, where a loss of column efficiency, increase in back
pressure, and ultimately bed collapse may occur.5-6
A motivation for creating HPLC stationary phases/supports that are stable at high pH exists in
the pharmaceutical industry. For example, McCalley and coworkers7-8 expressed the difficulty of
separating basic compounds under reversed-phase conditions because these analytes usually exist in
their protonated state under the pH conditions appropriate for most silica-based columns. (Protonated
species are typically poorly retained under reversed-phase conditions.9) Thus, high pHs (at least high
enough to deprotonate amines) would be advantageous in such separations. McCalley further observes
that of all synthesized pharmaceutical compounds, 70% are bases. Because the pKa values of most
amines are ca. 9.5 to 11, and at least in aqueous solution, a solution pH must be at least one pH unit
above the pKa value of an acidic moiety to have 90% of these groups deprotonated, there is a need for a
chromatographic material that can withstand a pH where the basic groups on analytes would be largely
deprotonated. (Note that in this work when we refer to the “pKa of an amine”, we are, technically
speaking, referring to the pKa of the conjugate acid of that amine.)
Nonsiliceous phases, based on organic polymers, titania, alumina, and graphitic carbon, have
been developed and are usually stable at high pH, but often lack efficiency.10-14 Peter Carr and
coworkers developed zirconia as both a normal and a reversed phase material,15-18 where this material
Page 2 of 23
ACS Paragon Plus Environment
Submitted to Analytical Chemistry
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
3
also has stability over a wide pH range.12,19-20 However, due to Lewis acid sites on the surface,
undesirable secondary interactions occur with certain analytes.21-23 These nonspecific interactions, in
conjunction with the difficulty associated with functionalizing the zirconia surface,18 help explain why
zirconia-based supports phases have not become more mainstream products.
Porous graphitic carbon is an important material that is marketed commercially by more than one
firm, e.g., Hypercarb™ by ThermoFisher.24-25 It is stable at extreme pH values and also elevated
temperatures, although its selectivity is different from standard reversed phases and noticeable tailing is
observed with many analytes. While these differences/limitations have prevented it from being more
widely adopted, there is currently a great deal of interest in this material because of its stability.
Diamond has also been studied as a support/stationary phase in liquid chromatography.26 For
example, Nesterenko and coworkers employed sintered, microdispersed detonation nanodiamond for
normal phase separations27 and ion exchange chromatography.28 They performed baseline separations of
various compounds using their normal phase material and achieved 15 400 plates/m. Their peaks
showed considerable asymmetry, especially at longer retention times. In more recent work they have
achieved 45 300 plates/m.29
Work in the Linford group at Brigham Young University has focused on the chemical
modification of diamond and its subsequent use in SPE and HPLC. Our first separations were
performed on PAAm-coated 50-70 µm diamond particles,30-31 where solid-phase extraction (SPE) of
lipids was demonstrated. It was later shown that various alkyl and a perfluoroalkyl isocyanate would
react with the PAAm-coated diamond through urea linkages.32 SPE of pesticides from water was
performed on the resulting C18 phase. We also demonstrated that deuterium-terminated diamond
(DTD)33 would react with di-tert-amyl peroxide, and that DTD could be further functionalized with
polymers by radical polymerization reactions.34 The resulting diamond materials could also be used for
SPE. In general, these materials were stable under extreme pH conditions.31-32
Page 3 of 23
ACS Paragon Plus Environment
Submitted to Analytical Chemistry
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
4
Clearly, however, the nonporous particles employed in these original studies would never be
suitable for HPLC, even if they could be “shrunk” to an appropriate size. To remedy this problem,
pellicular particles35 were formed by coating irregularly shaped 1.7 µm diamond particles with bilayers
of PAAm and nanodiamond in a layer-by-layer (LbL) fashion.36 These PAAm-coated core-shell particles
were then reacted with 1,2-epoxyoctadecane, creating a reversed-phase. This phase was able to separate
pesticides (cyanazine and diazinon) and various alkyl benzenes. An efficiency of 54 800 plates/m was
obtained with diazinon, which was a solid improvement over previous diamond-based materials.
Unfortunately, after an extended period of time, the reversed-phase material began to degrade. Another
PAAm/nanodiamond pellicular phase, this time cross-linked/functionalized with 1,2,5,6-
diepoxycyclooctane, was then prepared, and this material showed considerably improved stability, albeit
lower plate counts. However, even with this improved stability, back pressures were high for all of the
particles made with irregular diamond particles.
While this earlier work was an advance, a variety of issues needed to be addressed regarding
these diamond-based particles. First, it would be important to find a spherical, inert support to serve as
the core for these particles, where this material needed to be amenable to functionalization with our
layer-by-layer chemistry. A spherical core would also be important because our irregular diamond
particles appeared to be difficult to pack and the A terms in their Van Deemter curves would most likely
be high. In addition, it was imperative to find a way to stabilize the reversed-phase with some sort of
cross linker so that these phases would be mechanically stable over a longer period of time.
In this work, we address these and other issues, showing the development of phases created by
coating spherical 3 µm carbon particles with layers of PAAm and nanodiamond. The two types of
stationary phases described herein were both reversed-phases (C18), but one phase was simultaneously
cross-linked and functionalized. As expected, the non-cross-linked phase showed low mechanical
stability, but the cross-linked, reversed-phase material showed good stability over an extended period of
time, and at high pH. Particular improvements over our last study36 include:
Page 4 of 23
ACS Paragon Plus Environment
Submitted to Analytical Chemistry
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
5
1. The use of spherical carbon particles instead of irregular diamond particles.
2. That plate counts per meter for these new core-shell particles are higher than for the previous particles, in spite of the fact that the new particles are larger.
3. The use of two epoxides (a monofunctional epoxide and a bifunctional epoxide) in the functionalization/cross linking of the PAAm/nanodiamond layers, where the monofunctional epoxide provides C18 chains and the bifunctional epoxide provides cross linking.
4. A demonstration that these new particles are stable at pH 11.3 and pH 13.
5. Reduced back pressures, i.e., higher possible flow rates, which allows (for the first time in our work on diamond-containing pellicular particles) acquisition of Van Deemter curves. Analysis of this data is also presented.
6. Reproducible pressure-flow curves.
7. A demonstration that by appropriate particle preparation, relatively tight particle size distributions can be obtained, which translates into the expected flattening of the Van Deemter curves for core-shell particles at increased flow rates.
Note that the Supporting Information of this paper contains a short computer program in
MATLAB that plots and analyzes Van Deemter data. In particular, this program plots the data, along
with the contributions of the A, B, and C terms to the Van Deemter expression, and also the residuals to
the fit. It further provides values of the rmse and R2 for the fit, and the optimal values of u and H.
Page 5 of 23
ACS Paragon Plus Environment
Submitted to Analytical Chemistry
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
6
2. EXPERIMENTAL
2.1 Reagents and Materials: water (18 MΩ resistance, filtered using a Milli-Q Water System,
Millipore, Billerica, MA), poly(allylamine), average 17 000 Mw (20 wt. % in water, Sigma Aldrich, St.
Louis, MO), poly(allylamine), average 65 000 Mw (20 wt. % in water, Sigma Aldrich, St. Louis, MO),
methanol, (HPLC grade Fisher Scientific, Fair Lawn, NJ), cyclohexanol (Reagent grade, Fisher
Scientific), xylenes (ACS grade, Mallinkrodt Baker Inc. Phillipsburg, NJ), isopropyl alcohol (ChromAR,
Mallinkrodt Baker), acetonitrile (HPLC grade, EMD, Gibbstown, NJ), triethylamine (Mallinkdrodt
Baker), tetramethylammonium hydroxide (25 wt. % solution in water, Sigma Aldrich, St. Louis, MO),
1,2-epoxyoctadecane (technical grade, 90%, Alfa Aesar, Ward Hill, MA), 1,2,7,8-diepoxyoctane (97%,
Sigma Aldrich), spherical carbon (3 µm mean size, Supelco, St. Louis, MO), Triton X-100
(Electrophoresis grade, Fisher Scientific), benzenoid hydrocarbon kit (Sigma Aldrich), nanodiamond
(8.32 wt. %, 0.0-0.1 nm, Advanced Abrasives Corporation, Pennsauken, NJ), empty HPLC Columns (30
mm X 4.6 mm ID with 0.5 µm frits, Restek U.S., Bellefonte, PA), 50 mL Centrifuge Tubes (Sarstedt,
Inc., Newton, NC).
2.2 Instrumentation: Our HPLC consisted of a dual wavelength detector (Model No. 2487), a
binary HPLC pump (Model No. 1525), and a column oven (Model Number 5CH) all from Waters
Corporation, Milford MA. Column Packer: Pack-in-a-Box, 10 000 psi pump (Chrom Tech, Inc., Apple
Valley, MN). Secondary Scanning Microscope, Philips XL30 ESEM FEG (FEI Corporation, Hillboro,
OR). Surface Area Analyzer, Micromeritics TriStar II, (Micromeretics Instrument Corporation,
Norcross, GA). Particle Size Distribution Analyzer, Beckman Coulter LS 13 320 Multi-Wavelength
Particle Size Analyzer (Beckman Coulter, Inc., Brea, CA). Particle size distributions were obtained by
placing drops of a suspension of our analyte in an analysis bath.
2.3 SEM Stub Preparation. Samples were prepared by placing a slurry of particles directly on
an SEM stub and the resulting samples were dried in an oven. Imaging was done under high-vacuum
conditions with an accelerating potential of 5 kV and a spot size of 3. (This is an arbitrary number
Page 6 of 23
ACS Paragon Plus Environment
Submitted to Analytical Chemistry
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
7
commonly used in SEM that has no units. This number represents the size of the aperture that allows
electrons through for imaging)
2.4 Surface Area Measurements. Surface area measurements were performed by taking
Brunauer−Emmett−Teller (BET) isotherm measurements using a Micromeretics instrument. Specific
surface areas of the samples were determined from N2 adsorption at 77 K (Micromeritics TriStar II). The
samples were degassed at 200 °C for 12 hours prior to data collection.
2.5 Particle Preparation. For particle preparation, we used a Layer-by-Layer (LbL) procedure
that was similar to that performed by Saini et al on diamond core particles.36 About 0.5 g of spherical,
carbon particles, 3 µm in diameter, were suspended in 40 mL of a 1:1 water:methanol (H2O:MeOH)
mixture containing 3.3 mL of 65 000 Mw poly(allylamine) (PAAm) solution. These particles were
stirred for 24 h in this solution. The particles were then placed in a 50 mL screw cap plastic centrifuge
tube, centrifuged at 5 000 rpm and rinsed three times with the 1:1 H2O:MeOH solution. We next added
1.5 mL of nanodiamond slurry (8.32 wt. %) to the PAAm coated particles that were suspended in ca. 40
mL of the rinse solution. The solution with the partially coated particles and nanodiamond was shaken
by hand for 5 minutes and allowed to settle for one minute. It was then centrifuged and rinsed twice
with a 1:1 MeOH:H2O mixture to remove excess nanodiamond from the particles. To these particles,
now coated with a layer of PAAm and nanodiamond, were added 1.5 mL of a 7.5 wt % aqueous solution
of PAAm (17 000 Mw). The particles were again agitated by hand for 5 min and allowed to settle for
one minute. Excess PAAm was removed by centrifuging and rinsing three times with the same
MeOH:H2O mixture. Deposition of nanodiamond (8.32 wt. %) and PAAm (17 000 Mw) was
subsequently performed in alternating steps until the desired thickness of the porous shell was reached,
terminating in a PAAm coating. Ultimately we aimed for a 0.5 µm shell, which was measured
periodically during particle growth by scanning electron microscopy.
2.6 Particle Functionalization. Core-shell particles made through deposition of 30
PAAm/nanodiamond bilayers, and terminated with a PAAm coating, were rinsed three times in
Page 7 of 23
ACS Paragon Plus Environment
Submitted to Analytical Chemistry
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
8
isopropanol and three times in 1:1 cyclohexanol:xylenes. The particles were then placed in ca. 15 g of
the cyclohexanol:xylenes solution to which functionalizing agents were added. To prepare a non-cross-
linked phase we added 10 wt. % of 1,2-epoxyoctadecane. This was reacted with the particles in a round
bottom flask, which was fitted with a water-cooled condenser and heated to 130 °C for 54 h. For the
cross-linked phase we added both 10 wt. % of 1,2-epoxyoctadecane and 0.5 wt. % of 1,2,7,8-
diepoxyoctane. The diepoxide served as the cross-linker. The reaction conditions were the same in the
preparation of the cross-linked and the non-cross-linked particles.
The reactions were allowed to cool to room temperature. Excess functionalizing reagent was
removed by rinsing and centrifuging three times with the cyclohexanol:xylenes solution, three times
with isopropanol, and three times with a 1% (v/v) aqueous solution of Triton X-100.
2.7 Particle Sieving. After functionalization, the particle size distribution of the particles was
measured. There were agglomerates around 100 µm, so the particles, in a 1% (v/v) Triton X-100
aqueous solution, were passed through a 40 µm sieve. This removed most of the large agglomerates,
however the particle size distribution was still far from uniform. After sieving, the particles were
collected back into a small volume after sieving by centrifugation.
2.8 Particle Optimization. In an effort to improve the particle size distribution, core particles
were sonicated prior to PAAm and nanodiamond deposition (Sonifier Cell Disruptor, Heat Systems Co.,
Model: W1850, Melvile, N.Y.). Other than this initial sonication, the particles were formed,
functionalized, cross-linked, and tested in the same manner as the previous batch of cross-linked
particles.
2.9 Column Packing. Packing was performed by suspending the particles in 12 mL of an
aqueous solution of Triton X-100 (1% (v/v)). The Triton solution was also used as the pushing solution
during packing. The slurry was poured into the packing chamber which had a 30 mm x 4.6 mm ID
column attached at its end. The maximum packing pressure was set at 7 000 psi (8500 psi for the
Page 8 of 23
ACS Paragon Plus Environment
Submitted to Analytical Chemistry
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
9
improved (sonicated) particles). Packing occurred over a 25 min. period and pressure was released
gradually over a 30 minute period.
2.10 HPLC. All mobile phases were composed of water and acetonitrile (ACN) and were auto-
mixed. The pH of the water was set at 11.3 by addition of 0.1% (v/v) triethylamine, or pH 13.0 by
addition of tetramethylammonium hydroxide. The column was kept 35.0 °C for all experiments. A
mobile phase of 60:40 ACN:H2O was used for most separations. Analytes were injected using a 20 µL
sample loop. All separations were performed under isocratic conditions.
2.10.1 Van Deemter Curve Acquisition. A mobile phase composed of 60:40 ACN: H2O at pH
11.3 was used to obtain a Van Deemter curve for the column packed with the cross-linked particles.
Twelve different velocities increasing by 0.1 mL/min from 0.1–1.2 mL/min were used for the van
Deemter plot. The analytes used were benzene, ethylbenzene, n-propylbenzene and n-butylbenzene.
2.10.2 Stability Tests. To test the overall stability and pH stability of the cross-linked material,
two stability tests were performed. The first test was run under the following conditions: flow rate: 0.5
mL/min, mobile phase composition: 60:40 ACN:H2O with 0.1 % (v/v) triethylamine added to the
aqueous portion of the mobile phase to set the pH at 11.3, temperature at 35 °C. The test occurred over
a period of 1 600 column volumes. The analytes used for this test were from a benzenoid hydrocarbon
kit and included benzene, ethylbenzene, n-butylbenzene and n-hexylbenzene.
The stability test of the stationary phase at high pH (pH 13) was performed in a similar manner
to the stability test at pH 11.3 and was undertaken on the same column. The mobile phase composition
was 60:40 ACN:H2O, and 1 % (v/v) tetramethylammonium hydroxide was added to the aqueous
component to create a pH 13 mobile phase. The column temperature was maintained at 35.0 °C. The
test occurred over a period of 1 000 column volumes. The analytes used for this test mixture were the
same as in the previous stability test.
Page 9 of 23
ACS Paragon Plus Environment
Submitted to Analytical Chemistry
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
10
After the stability tests, the HPLC system was flushed with ACN or MeOH and water for many
minutes to remove the corrosive material that might potentially damage the pump and/or detector flow
cell. The columns were also flushed with the same mobile phase and were stored under MeOH between
uses.
Page 10 of 23
ACS Paragon Plus Environment
Submitted to Analytical Chemistry
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
11
3. RESULTS AND DISCUSSION
3.1 Non Cross-Linked Reversed Phase. The first batch of core-shell particles investigated was
not cross-linked. The primary amines from the PAAm in the shell layer were only reacted with
monofunctional 1,2-epoxyoctadecane resulting in a C18 (reversed) phase. Chromatography was
performed on this column using alkyl benzene analytes with a mobile phase composition of 60:40
ACN:H2O, with the aqueous portion set at pH 11.3 by addition of 0.1 % (v/v) triethylamine (TEA). The
flow rate was 0.5 mL/min, with the column temperature maintained at 35.0 °C. Under all conditions
explored, peaks showed a large amount of fronting regardless of analyte concentration. This may be due
to non-uniform column packing. Moreover, the non-cross-linked column showed a rapid increase in
back pressure over a short period of time which indicated mechanical instability of this material.
At one point, the flow rate on the column was doubled from 0.5 mL/min to 1.0 mL/min. Upon
returning to the original flow rate, the back pressure had increased significantly from 2 040 to 3 620 psi.
After this experiment, the backpressure steadily increased over a 6 hour period to 4570 psi. At this
point, the experiment was terminated. We had previously observed mechanical instability with non-
cross-linked phases in our lab, so we opted for a different approach that included cross-linking with the
hope that a more mechanically stable phase could be created.
3.2 Cross-Linked, Reversed Phase
3.2.1 Pressure-flow relationship and reversed phase character. To determine the effect of cross-
linking, the column was reacted with 1,2-epoxyoctadecane, under the same conditions as described
above, but with the exception that a cross-linker, 1,2,7,8-diepoxyoctane, was added to the reaction
solution. The resulting cross-linked stationary phase was then packed under the same conditions and in a
column of the same dimensions as the previous column. From the chromatography, it was immediately
clear that this phase was less hydrophobic than the non-cross-linked phase, which would be consistent
Page 11 of 23
ACS Paragon Plus Environment
Submitted to Analytical Chemistry
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
12
with the incorporation of the diepoxide into the stationary phase. (Obviously, the diepoxide is less
hydrophobic than 1,2-epoxyoctadecane.)
Figure 1. Reversed phase separation of (1) benzene, (2) ethylbenzene, (3) n-butylbenzene, (4) n-hexylbenzene. Mobile phase composition was 40:60 H2O:ACN with 0.1 (v/v) % triethylamine, setting the pH of the mobile phase at 11.3. Flow rate: 0.5 mL/min.
For example, under the same conditions used with the non-cross-linked column (mobile phase and
pressure), the last eluting peak, n-hexylbenzene, eluted about 1.5 min earlier. Figure 1 shows the
chromatogram of this and other alkyl benzenes on this cross-linked column. There were also immediate
indications that the material would be stable over a longer period of time, as evidenced by our ability to
increase and decrease repeatedly and reproducibly the mobile phase velocity. A plot of the resulting
linear relationship between pressure and flow is shown in Figure 2.
Page 12 of 23
ACS Paragon Plus Environment
Submitted to Analytical Chemistry
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
13
As a further note of comparison, under the same initial conditions, the back pressure for the
cross-linked column was 940 psi, while the starting pressure for the non-cross-linked column was 2040
psi. These results for the non cross-linked particles suggest mechanical instability during packing, which
would lead to clogging of the frit or the interstitial spaces between the particles by fines, possibly
sloughed off the particle surfaces during column packing.
3.2.2 Surface area, Pore Size and Volume. The surface area of the cross-linked particles was
44.2 m2/g by BET isotherm measurements. Note that because diamond is twice as dense as silica, this
number should be multiplied by two if it is to be directly compared to the surface area of silica-based
columns.36 The particles had a mean pore size of 280 Å and a pore volume of 0.356 cm3/g.
Page 13 of 23
ACS Paragon Plus Environment
Submitted to Analytical Chemistry
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
14
3.2.3 Stability at pH 11.3. The first stability test performed on the cross-linked column was a
long-term study that took place over 1 600 column volumes at pH 11.3. An analyte mixture containing
benzene, ethylbenzene, n-butylbenzene and n-hexylbenzene in acetonitrile was employed. The flow rate
was 0.5 mL/min and the retention factor, k, was recorded as a function of column volumes. Accordingly,
Figure 3 shows the average value of k for these analytes as a function of column volumes, where the
initial average value of k was set to 100%.
Figure 3. Decrease in k as a function of column volumes of solvent passed through the column. Over a 1 600 column volume period, a 5% decrease in k was observed.
The trial ran over a 26.6 h period and resulted in a decrease in k of ca. 5%. This phase shows the
greatest stability of any HPLC phase we have created to date.
3.2.4 Stability at pH 13.0. A second stability test using the same analyte mixture was performed
at pH 13.0 on this same column. The mobile phase composition was 60:40 ACN:H2O with the aqueous
portion set at pH 13.0 by addition of 1 % (v/v) tetramethylammonium hydroxide. The flow rate for this
stability test was 0.5 mL/min, and the column temperature was 35.0 °C. As shown in Figure 4, over this
Page 14 of 23
ACS Paragon Plus Environment
Submitted to Analytical Chemistry
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
15
period of time only a 1% loss in retention factor was observed. To our knowledge no other silica-based
or hybrid phase column has this degree of stability.
3.2.5 Van Deemter study. The reasonably low backpressures of this column opened the
possibility of varying the flow rates to obtain a Van Deemter curve. For this study, the mobile phase
was the same as that used for the first stability test (pH 11.3), the analyte mixture consisted of benzene,
ethylbenzene, n-propylbenzene and n-butylbenzene in acetonitrile, and measurements were taken every
0.1 mL/min from 0.1 to 1.2 mL/min. Figure 5 shows the resulting Van Deemter curve obtained by
plotting the average value of H for the diffrent analytes vs. the mobile phase velocity.
Page 15 of 23
ACS Paragon Plus Environment
Submitted to Analytical Chemistry
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
16
In this study, our best plate counts (56 000 N/m) were obtained for n-butylbenzene at flow rates
of 0.4-0.6 mL/min, giving an optimal plate height of 20.5 µm. It is significant that these plate counts are
higher than those for phases previously created in our lab,36 despite previous phases having smaller
particle sizes. These results appear to be on the edge of industrial viability. In the data collected for this
study, nearly ideal peak symmetries for the analytes were observed: the peak symmetries for the
benzene, ethylbenzene, n-propylbenzene and n-butylbenzene peaks were 1.03, 1.05, 0.99, and 1.03,
respectively (see Figure 6).
Page 16 of 23
ACS Paragon Plus Environment
Submitted to Analytical Chemistry
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
17
Figure 6. Reversed phase separation of (1) benzene, (2) ethylbenzene, (3) n-propylbenzene, (4) n-butylbenzene giving peak symmetries of 1.03, 1.05, 0.99, and 1.03 respectively. This chromatogram is an example of the separations performed during the Van Deemter study. This chromatogram was obtained at a flow rate of 0.5 mL/min
1 2
3
4
3.2.6 Particle Size Distribution Measurements and Scanning Electron Microscopy. Although we
were pleased with the results from the Van Deemter study, we were concerned about the higher than
desired reduced plate heights. While good reduced plate heights are typically around 2, we had a
reduced plate height of 5 based on our projected particle size (4 µm). We were also surprised to see that
our C term had contributed so significantly to our overall plate height since we had created a phase
based on a core-shell particle.
To obtain greater insight into these problems, we measured the particle size distribution (PSD) of
our particles. And despite starting with a powder with a 3 µm average particle size and a shell thickness
of 0.5 µm (4 µm total), our measurements showed that we had a mean particle size of 14.0 µm and a
D90/10 (skewness) of 3.9 after functionalization. This less-than-ideal PSD is shown in Figure 7, which
indicates a need for particle size optimization. Scanning electron microscopy also gave some indication
of the presence of agglomerates.
Page 17 of 23
ACS Paragon Plus Environment
Submitted to Analytical Chemistry
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
18
Image A in Figure 8 shows a single particle, about 4 µm in diameter. Image B shows what may be a
large agglomerate, and Image C indicates some necking between some of the particles. However, we
must admit that it is not always a trivial matter to differentiate between drying-down effects in sample
preparation and actual agglomeration.
Page 18 of 23
ACS Paragon Plus Environment
Submitted to Analytical Chemistry
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
19
3.2.7 Particle Optimization. Our next goal was to create a new batch of particles with the same
cross-linked/C18 functionality, but with fewer agglomerates. In this effort, the particles were sonicated
after the first PAAm coating, but before LbL deposition. After particle formation, a substantially
improved PSD was obtained (see Figure 9), and the mean particle size of this batch was 5 µm. The
column was characterized as before, and the resulting Van Deemter curve showed the expected
flattening of its C term. Whereas the C term for the previous particles was 16.8 for n-butylbenzene, the
C term for the sonicated particles was 4.84. Unfortunately, the A term for this new column/material
increased, which suggests that our packing procedure still needs to be optimized.
Page 19 of 23
ACS Paragon Plus Environment
Submitted to Analytical Chemistry
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
20
Figure 9. Single analyte Van Deemter curves obtained at 35 °C using n-butylbenzene. A) PSD and Van Deemter curve for particles that were not sonicated prior to nanodiamond deposition. The C term for this column was high, presumably due to agglomerates. B) PSD and Van Deemter curve for particles that were sonicated prior to the first nanodiamond deposition. The C term was greatly reduced, although the A term was high, which was attributed to unoptimized column packing.
4. CONCLUSIONS
We report spherical core particles surrounded by PAAm-nanodiamond shells for HPLC. In this
study, we first developed a non-cross-linked C18 reversed-phase material. The resulting columns
appeared to be unstable, had lower efficiencies, and showed a significant increase in back-pressure over
a short period of time, as had our previous non-cross-linked columns.36
Our next attempt to create a stable phase included the addition of a cross-linker (1,2,7,8-
diepoxyoctane) during functionalization. The back pressures of this column were the lowest observed
for any of our diamond-based core-shell particles to date. The pressure-flow behavior appeared to be
completely reversible and allowed us to obtain a Van Deemter curve for this phase. The optimal flow
rates were between 0.4-0.6 mL/min, resulting in a plate height of 18.6 µm (56 000 N/m) for our best
analyte, n-butylbenzene. Unfortunately, our A and C terms were higher than we would have liked.
Page 20 of 23
ACS Paragon Plus Environment
Submitted to Analytical Chemistry
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
21
Not only does the cross-linked phase show the best efficiencies yet seen for a diamond-based
phase (56 000 N/m), but it also exhibits good stability under extreme pH conditions. The stability test
performed at pH 11.3 showed a 5% decrease in retention factor, k, over a 1 600 column volume period.
The stability test performed on the same column at pH 13.0 showed a 1 % decrease in k over a 1 000
column volume period. To the best of our knowledge, the cross-linked stationary phase shows greater
pH stability than any other commercially available silica or hybrid column.
Improvement in our particle size distribution was accomplished by sonication prior to LbL
deposition. This resulted in an improved C term, presumably because fewer agglomerates were present
in the newly packed column. On the other hand, the A term was higher. This is attributed to an
unoptimized column packing procedure. Provided that column packing could be optimized, the
efficiencies of columns packed with particles created using the sonication procedure could be much
higher.
SUPPORTING INFORMATION AVAILABLE: A MATLAB program, which conveniently plots Van
Deemter curves, including the individual contributions of the A, B, and C terms, the residuals to the fit,
the values of A, B, and C, the rmse and R2 values for the fit, and the optimal values of u and H.
ACKNOWLEDGMENTS We thank Paul Ross from Supelco for providing us with the spherical carbon
particles that served as our core material, Michael Standing at BYU for operating the SEM and
obtaining fantastic images of our particles, and US Synthetic for continued funding and support.
Page 21 of 23
ACS Paragon Plus Environment
Submitted to Analytical Chemistry
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
22
REFERENCES (1) Miller, J. M. Chromatography: Concepts and Contrasts; 2nd ed.; John Wiley & Sons, Inc: Hoboken, NJ, 2005. (2) Snyder, L. R.; Kirkland, J. J.; Glajch, J. L. Practical HPLC Method Development; 2nd ed.; John Wiley: New York, 1997. (3) Kirkland, J. J.; Glajch, J. L.; Farlee, R. D. Anal. Chem. 1989, 61 (1), 2-11. (4) Nawrocki, J. J. Chromatogr., A 1997, 779 (1-2), 29-71. (5) Wyndham, K. D.; O'Gara, J. E.; Walter, T. H.; Glose, K. H.; Lawrence, N. L.; Alden, B. A.; Izzo, G. S.; Hudalla, C. J.; Iraneta, P. C. Analytical Chemistry 2003, 75 (24), 6781-6788. (6) Kirkland, J. J.; Henderson, J. W.; De Stefano, J. J.; van Straten, M. A.; Claessens, H. A. J. Chromatogr., A 1997, 762 (1-2), 97-112. (7) McCalley, D. V. J. Chromatogr., A 2010, 1217 (6), 858-880. (8) McCalley, D. V. Advances in Chromatography, Vol 46 2008, 46 305-350. (9) Gagliardi, L. G.; Castells, C. B.; Ràfols, C.; Rosés, M.; Bosch, E. Analytical Chemistry 2006, 78 (16), 5858-5867. (10) Haky, J. E.; Vemulapalli, S.; Wieserman, L. F. J. Chromatogr., A 1990, 505 (2), 307-318. (11) Knox, J. H.; Kaur, R. P. Adv. Chromatogr. 1997, 37 73-119. (12) Nawrocki, J.; Rigney, M.; McCormick, A.; Carr, P. W. J. Chromatogr., A 1993, 657 (2), 229-282. (13) Tanaka, N.; Araki, M. Adv. Chromatogr. 1989, 30 81-122. (14) Trüdinger, U.; Müller, G.; Unger, K. K. J. Chromatogr., A 1990, 535 111-125. (15) Nawrocki, J.; Dunlap, C.; McCormick, A.; Carr, P. W. J. Chromatogr., A 2004, 1028 (1), 1-30. (16) Sun, L.; Carr, P. W. Anal. Chem. 1995, 67 (20), 3717-3721. (17) Sun, L.; McCormick, A. V.; Carr, P. W. J. Chromatogr., A 1994, 658 (2), 465-473. (18) Nawrocki, J.; Dunlap, C.; Li, J.; Zhao, J.; McNeff, C. V.; McCormick, A.; Carr, P. W. J.
Chromatogr., A 2004, 1028 (1), 31-62. (19) Cazes, J. Encyclopedia of Chromatography; 1st ed.; Marcel Dekker: New York, 2004. (20) Rigney, M. P.; Weber, T. P.; Carr, P. W. J. Chromatogr., A 1989, 484 273-291. (21) Blackwell, J. Chromatographia 1993, 35 (3), 133-138. (22) Blackwell, J. A.; Carr, P. W. J. Chromatogr., A 1991, 549 59-75. (23) Blackwell, J. A.; Carr, P. W. J. Liq. Chromatogr. Relat. Technol. 1991, 14 (15), 2875-2889. (24) Pereira, L. J. Liq. Chromatogr. Relat. Technol. 2008, 31 (11), 1687-1731. (25) West, C.; Elfakir, C.; Lafosse, M. J. Chromatogr., A 2010, 1217 (19), 3201-3216. (26) Nesterenko, P.; Haddad, P. Anal. Bioanal. Chem. 2010, 396 (1), 205-211. (27) Nesterenko, P. N.; Fedyanina, O. N.; Volgin, Y. V. Analyst 2007 (132), 403-405. (28) Nesterenko, P. N.; Fedyanina, O. N.; Volgin, Y. V.; Jones, P. J. Chromatogr., A 2007, 1155 (1), 2-7. (29) Nesterenko, P. N.; Fedyanina, O. N. J. Chromatogr., A 2010, 1217 (4), 498-505. (30) Saini, G.; Gates, R.; Asplund, M. C.; Blair, S.; Attavar, S.; Linford, M. R. Lab Chip 2009, 9 (12), 1789-1796. (31) Saini, G.; Yang, L.; Lee, M. L.; Dadson, A.; Vail, M. A.; Linford, M. R. Anal. Chem. 2008, 80 (16), 6253-6259. (32) Saini, G.; Wiest, L. A.; Herbert, D.; Biggs, K. N.; Dadson, A.; Vail, M. A.; Linford, M. R. J. Chromatogr., A 2009, 1216 (16), 3587-3593. (33) Yang, L.; Lua, Y. Y.; Tan, M.; Scherman, O. A.; Grubbs, R. H.; Harb, J. N.; Davis, R. C.; Linford, M. R. Chem. Mater. 2007, 19 (7), 1671-1678. (34) Yang, L.; Vail, M. A.; Dadson, A.; Lee, M. L.; Asplund, M. C.; Linford, M. R. Chem.
Mater. 2009, 21 (19), 4359-4365.
Page 22 of 23
ACS Paragon Plus Environment
Submitted to Analytical Chemistry
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
23
(35) Kirkland, J. J.; Langlois, T. J.; DeStefano, J. J. Am. Lab. 2007, 8 (39), 18-21. (36) Saini, G.; Jensen, D. S.; Wiest, L. A.; Vail, M. A.; Dadson, A.; Lee, M.
L.;Shutthanandan, V.; Linford, M. R. Anal. Chem. 2010, 82 (11), 4448-4456.
Page 23 of 23
ACS Paragon Plus Environment
Submitted to Analytical Chemistry
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960