PARTICLE SIZE CONSIDERATIONS OF SUPERFICIALLY POROUS PARTICLESSUPERFICIALLY POROUS PARTICLES

Joseph J. DeStefano, Stephanie A. Schuster, Robert S. Bichlmeir, and William L. Johnson

Advanced Materials Technology, Inc., 3521 Silverside Road, Suite 1‐K, Wilmington, DE 19810

ABSTRACT

Columns of fused-core superficially porous particles demonstrate higherColumns of fused core, superficially porous particles demonstrate higherefficiency than columns packed with totally porous particles of similar size. Thisresult likely is because of superior eddy and longitudinal diffusion properties(smaller van Deemter A and B terms) observed for core-shell particles resultingf ti ll ti l i di t ib ti d hi h d itfrom exceptionally narrow particle size distributions and higher densitycontributing to an ability to form homogeneous packed beds. The achievementof efficiencies comparable to those obtained using columns of sub-2-μm totallyporous particles without the need for ultra high pressure instrumentation is ap p g pdistinct advantage of fused-core columns. Fused-core particles with a widerange of uniform particle sizes were synthesized to allow the preparation ofstable, efficient packed columns for this study. These columns were used tomeasure the effects of particle size on chromatographic performance andmeasure the effects of particle size on chromatographic performance andpressure requirements. This report describes the effect of particle size on severalfactors of separation importance, including plate heights, reduced plate heights,pressure, and ease of use. Depending on the intended application, certainp p g ppparticle sizes are recommended for use over others. The performance of largerfused-core particles exceeds even the high expectations observed for smallersuperficially porous particles, likely because larger particles form morehomogeneous packed beds than smaller particles The utility of larger andhomogeneous packed beds than smaller particles. The utility of larger andsmaller superficially porous particles are compared and evaluated for high-speedand high-resolution applications.

HALO® Fused-Core ParticlesHALO Fused Core Particles

particle size

shell thickness

SEM of HALO Fused-core

Graphical representation of HALO F dFused-core Fused-core

CHARACTERISTICS OF PARTICLES USED IN THIS STUDY

Total Particle Size Shell Thickness Core Diameter Surface Area

2/2.0 microns 0.5 micron 1.0 micron 100 m2/g

2.7 microns 0.5 micron 1.7 microns 125 m2/g

4.1 microns 0.55 micron 3.0 microns 105 m2/g

4.6 microns 0.6 micron 3.4 microns 95 m2/gg

Test Conditions:TEST CHROMATOGRAM

Test Conditions:Columns: 3.0 x 50 mmMobile Phase: 60/40 ACN/H2OFlow Rate: variedTemperature: ambient (ca. 22 oC)Detector: semi-micro UV @ 254nm

95

@Peak Identities:

1. Uracil (t0 marker)2. Phenol3. 4-Cl-1-nitrobenzene4. Naphthalene (k values 3.3 to 3.8)

55

75 12

3

4

15

35

‐50.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6

Time minutesTime, minutes

Plate Height Vs Velocity Plots

7.0

7.5

8.0

4.6 µm

mic

rom

eter

s

5 5

6.0

6.5 4.1 µm

Pla

te H

eigh

t,

4 .5

5.0

5.5

2.7 µm

3.0

3.5

4.0 2.0 µm

Data fitted to Knox equation

The Plate Heights of columns packed with particles of different

Linear M obile Phase Velocity, m m /sec

1 2 3 4 5 6 7

sizes, as expected, get smaller as the particle size gets smaller.

Reduced Plate Height Vs Velocity Plots

2.4

2.6

g yH

eigh

t

2.0

2.2

2.0 µm

Red

uced

Pla

te H

1.6

1.8 2.7 µm4.1 µm

R

1.2

1.4

4.6 µm

Linear Mobile Phase Velocity, mm/sec

1 2 3 4 5 6 71.0

Data fitted to Knox equation

Reduced Plate Heights (h = H/dp) get smaller as the particle size is increased, indicating more homogeneity in packed beds for the larger particles.

Overlay of Size Distributions of Individual Batches of 2.7 µm and 4.1 µm Fused-core Particles

Individual batches of Fused-core particles exhibit very narrow particle size distributions.

Actual Size Distributions of a 50/50 Mixtureof 2.7 and 4.1-Micron Particles

By Number

0 1 2 3 4 5 6 7 8Particle Size, microns

By Volume

Particle Size, microns

• Mixtures of different sized particles can be measured by Coulter Counter and give different values if assessed by number or volume.M t f ti l i b l i t ti f th• Measurement of particle size by volume is more representative of the contents of a column than is size by number.

Plate Height Vs Velocit Plots Incl ding Mi ed Particles

8 .0

Plate Height Vs Velocity Plots Including Mixed Particleset

ers

6 .5

7 .0

7 .5 4.1 µm

Hei

ght,

mic

rom

e

5 .5

6 .050%/50%2.7/4.1 µm

Pla

te

4 .0

4 .5

5 .0

2.7 µm

L inear M ob ile P hase V e loc ity , m m /sec

1 2 3 4 5 6 7

3 .5 Data fitted to Knox equation

50/50 mix of 2.7 and 4.1 µm particles produces a column with plate heights intermediate to the components of the mixture.

Comparison of Single-Sized Particles and a MixtureComparison of Single-Sized Particles and a Mixture

Packing Particles dp90a dp10b dp90/dp10

Plate Number(N)*

Plate Height (H)*

Pressure (bar)*

Plates per bar

4 1 µm 4 30 3 89 1 11 8590 5 58 34 2534.1 µm 4.30 3.89 1.11 8590 5.58 34 253

2.7 µm 3.02 2.69 1.12 11950 4.19 128 93

50/50 mix 4.1/2.7 µm 4.26 2.63 1.62 10650 4.69 94 113

* Values taken at optimum flow* Values taken at optimum flow

• High efficiency is still possible for a mixed bed of particles of

a dp90 indicates that 90% of the particles are smaller than the given sizeb dp10 indicates that 10% of the particles are smaller than the given size

High efficiency is still possible for a mixed bed of particles of different sizes, but the pressure increase is not favorable.

Plate Number Per Unit of Pressure(Plate number and pressures taken at optimum flow for each particle size)(Plate number and pressures taken at optimum flow for each particle size)

5 50/50 mix of 2.7 and 4.1 µm113

3

4 4.6 µm

4.1 µm

269

253

1

22.7 µm

2.0 µm

93

69

0 50 100 150 200 250 300

1

Plate number per bar of pressure

69

• The plate number available per unit of pressure greatly increases as the particle size is increased.

• The column containing a mixture of large and small particles is not a good compromise of performance features because pressure changes faster thancompromise of performance features because pressure changes faster than plate number with changes in average particle size.

Comparison of 5μm and 2.7μm particle separation

0.5 mL/min, 40C, 1 L of ginseng root extractA: 10 mM Amm. Formate/0.1% formic acid, B: ACN 20–50% in 10.0 min.254 nm, Agilent 1100

65 30

Expanded view

25

35

45

55 2.1 x 50 mm, HALO-5, C18Pressure: 66 bar

15

20

25

‐5

5

15

0 1 2 3 4 5 6 7 8 9 10

104 4.2 4.4 4.6 4.8 5

25

30

35

45

55

65 2.1 x 50 mm, 2.7 µm HALO, C18Pressure: 150 bar

10

15

20

25

‐5

5

15

2510

4.5 4.7 4.9 5.1 5.3 5.5

0 1 2 3 4 5 6 7 8 9 10

• Use shorter columns for faster separations• For shorter columns, improved resolution is observed when smaller particles are used

1.7972.486

3 831

Fused-Core, 2.1 x 150 mm, 5-m SPP0.5 mL/min, 30C, 3 L of ~50:50 ACN/water DNPH standard mix

Use 5 µm SPP Columns to Achieve High Resolution at Low Pressure

0.020

Volts

3.505

3.831

4.322

6.3806.679

7.127

0.5 mL/min, 30 C, 3 L of 50:50 ACN/water DNPH standard mixA: water, B: 80/20 ACN/THF 45% B, hold 5 min, 45–80% in 20 min.360 nm, bandwidth 8 nm; 40 Hz

Worst Case  Pressure~189 bar

000

0.010

0.553

4.088

5.648

5.926

8.384

9.546

9.99910.791

12.221

12.627

2.874 4.082

6 310

Totally Porous, 2.1 x 150 mm, 1.8-m0.5 mL/min, 30C, 3 L of ~50:50 ACN/water DNPH standard mix A: water, B: 80/20 ACN/THF

0 2 4 6 8 10 12 14 16Time (min)

0.0

00.020

Volts

3.928

5.8076.310

7.058

8.820

9.2869.565 10.009

12.825

13.860

15.393

45% B, hold 5 min, 45–80% in 20 min.360 nm, bandwidth 8 nm; 40 Hz Worst Case  Pressure

~855 bar

0.000

0.010

1.519

6.681 8.583

11.393

12.345

13.738

14.961

15.393

0 2 4 6 8 10 12 14 16Time (min)

PEAK IDENTITIES (in order): Formaldehyde-2,4-DNPH, Acetaldehyde-2,4-DNPH ,Acetone-2,4-DNPH,Acrolein-2,4-DNPH,Propionaldehyde-2,4-DNPH, Crotonaldehyde-2,4-DNPH, 2-Butanone-2,4-DNPH, Methacrolein-2,4-DNPH, Butyraldehyde-2,4-DNPH, Benzaldehyde-2,4-DNPH,Valeraldehyde-2,4-DNPH,m-Tolualdehyde-2,4-DNPH, and Hexaldehyde-2,4-DNPH Note: Small peaks preceding labeled aliphatic aldehyde peaks are minor geometric isomers (syn/anti).

Use 5-µm Fused-Core Particles for Ballistic Gradients

60 0.392

0.450

0.493

0.639

5-µm Fused-Core C18, 2.1 x 50 mm2.0 mL/min, 40C, 1 L of 8 flavonoids in MeOH

Max pressure: 210 bar

40m

AU

0.3770.504

0 514

, , A: 10 mM Amm. Formate/0.1% formic acid, B: ACN 5–60% B in 1.0 min.254 nm, bandwidth 8 nm; 80 Hz; Agilent 1200 SL; Low Delay Volume Configuration ~ 120 µL; ADVR ON

Max pressure: 210 bar

20m 0.514

0

0.0 0.2 0.4 0.6 0.8 1.0Time (min)

Analytes, in elution order: hesperidin, myricetin , quercetin, naringenin & apigenin(coeluted) hesperetin kaempferol biochanin(coeluted), hesperetin, kaempferol, biochanin

• Larger particles enable faster flow rates at lower back pressures

Conclusions

1. Fused-core particles can be synthesized with different particle diameters to produce columns with high efficiencies.

2. Columns packed with Fused-core particles of different diameters show p pimprovement in Plate Heights as the particle size is reduced.

3. Reduced Plate Heights (column efficiencies normalized for particle size) show the opposite effect – with columns of larger particles having smaller Reduced Plate Heights than smaller particles – indicating larger particles g g gmay be easier to pack into homogeneous beds.

4. Mixtures of larger and smaller particles can provide good column performance but the resulting pressure makes this a bad compromise.

5. Narrow particle size distributions for Fused-core packings may not be solely responsible for the exceptionally high performance of columns packed with superficially porous particles.

6. Longer columns of 5µm Fused-core particles are recommended for high resolution separations while short columns of 5µm Fused-core particles are recommended for very fast gradients.

Acknowledgements

Special thanks to Jason Lawhorn for providing the Coulter analysis data and to Jack Kirkland for discussions and help in organizing and graphing theto Jack Kirkland for discussions and help in organizing and graphing the information used in this poster. Additional thanks to Thomas Waeghe for the applications using HALO-5 columns.