Gilson, IncMiddleton, WIwww.gilson.com

Preparative HPLC: Factors and Parameters that

Directly Affect Recovery of Collected Fractions

Gary Scharrer and Joan Stevens, Ph.D.

AbstractPurification of synthetic, natural and biological compounds in any quantity usually requires the use of preparative HPLC. Collecting fractions from the preparative HPLC is the common approach to achieving the purified compound(s) of interest. Collection of fractions however is not trivial and is affected by many conditions and parameters within preparative HPLC. In the perfect world chromatographers would be able to optimize for each compound needing to be purified, however this not the case. The present study will examine a multitude of factors and parameters that directly affect fraction collection recovery. Some of the parameters to be discussed are column performance, mobile phase constituents, detector setting, and collection parameters such as collection via level versus slope and delay volumes. The data presented will offer guidelines for all chromatographers purifying compounds by preparative HPLC and can be implemented into existing systems.

Outline• Preparative Systems• Fraction Collection Factors

– Column Performance– Mobile Phase Constituents– Detector Settings– Peak Level Collection– Slope Collection– Delay Volume

• Summary

Prep System Configurations

System A

• Gilson 215 Liquid Handler/Fraction Collector, equipped w/175-mm Z-arm, 819 Injection Module with 5.0-mL SS loop and bevel-tip probe (269x1.3x0.8 mm ID)

• Gilson 333/34 HPLC Pump, equipped with H3 pump heads (flow rates up to 200 mL/min.)

• Gilson 155 UV/Vis Dual-wavelength Detector (0.2-mm pathlength, 1.0 AUFS)

819 Injector

Figure 1

System B

• Gilson 215 Liquid Handler/Fraction Collector, equipped w/175-mm Z-arm, 845Z Injection Module w/5.0-mL SS loop, and bevel-tip probe (269x1.3x0.8mm ID)

• Gilson 322 HPLC Pump, equipped with H2 heads (flow rate up to 30 mL/min.)• Gilson 155 UV/Vis Dual-wavelength Detector (0.2-mm pathlength, 1.0 AUFS)

Prep System Configuration845Z Injector

Figure 2

System/Injectors Used In Study

• 819 Injector – System A– Dispenses sample through an injection port to the

sample loop• 845Z Injector – System B

– Aspirates sample directly into the sample loop(Z-Mounted Valve)

• Above systems/injectors were both used in this study.

Factors that Affect FC Recovery• Column Performance

– Compound Polarity• Peak shape due to polarity can directly influence fraction

collection parameters. The following data includes compounds of high to low polarity (caffeine, ethyl paraben, and biphenyl).

– Gradient versus Isocratic Elution• Peak broadening due to elution type can affect time and

efficiency of fractionation. Typically gradients are used to shorten purification time and give better peak shapes.

FC Factors (cont.)• Mobile Phase Constituents

– Compound polarity in gradient program selection

– Injection Technique• Standard

– Sample is dispensed into sample loop and is injected into the mobile phase via a switching valve.

• Sandwich – Sample is placed between plugs of solvent, to protect

from pre-column precipitation, prior to injection. The protected sample is then dispensed and injected as in the standard technique.

FC Factors (Cont.)• Detector Settings

• Peak Width– Set at 0 to acquire raw data with no smoothing

effects.• Sensitivity

– Level of detection desired. This parameter should not be confused with fraction collection sensitivity.

– Amount of fraction collected with be affected by increasing or decreasing this parameter when using peak level for collection.

Collection Parameters – Peak Level

Level

Start Collection

Stop Collection

Disadvantages• Loss of Compound• No separation of multiple

peaks

Figure 3. Peak Level is the fraction collection parameter which uses a specified input set at a percentage of full-scale to start and endcollection.

Affects of Peak Level Changes

Level

Level

All peaks above levelare collected. Peaks below level

not collected.Recovery reduced

Figure 4. As peak level decreases the recovery increases, any peaks that are at or above peak level setting are collected.

0

20

40

60

80

100

120

2 5 10 25 50 100

Peak Level

Reco

very

(%)

CaffeineEthyl ParabenBiphenyl

Peak Level vs. Recovery

Graph 1. As peak level increases, recovery starts to drop. If the peak level is set too high, a large percentage could be lost to waste, and if set too low, peak purity could be affected, by collecting closely eluting compounds.

Peak Level = 100Peak Level = 50

Peak Level =25Peak Level = 10

Peak Level = 5

Collection Parameters – Peak Slope

peak width – peak width time at½ height

peak sensitivity-% height of smallestpeak

Peak Slope is the parameter which uses peak sensitivity and peak width to determine peak collection

Peak sensitivity is the height of the shortest peak, measuredIn % of full scale. The peak sensitivity should be slightly greaterThan the highest baseline noise.

Peak Width is the duration in minutes, of a typical Gaussian peak at half-height. The peak width should be slightly longer than the width of the narrowest peak.

Slop

e

Minutes

Figure 5

Slope vs. RecoverySensitivity vs. Recovery

0.08 Peak Width

0

20

40

60

80

100

120

2 5 10 15 20

Peak Sensitivity (%)

Rec

over

y (%

)

CaffeineEthyl ParabenBiphenyl

Graph 2 & 3. Slope is the combination of peak width and sensitivity. Above shows the impact of each factor. As the sensitivity increases, recovery decreases and as the peak width measured at ½ height increases recovery increases.

Peak Width vs RecoveryPeak Sensitiv ity @ 5%

0

20

40

60

80

100

120

0.02 0.04 0.06 0.08 0.1

Peak W idth @ half he ightR

ecov

ery

(%)

CaffeineEthyl ParabenBiphenyl

Collection Parameters (Cont.)– Delay Volume (The Parameter used to

compensate for the delay of the compound to reach the fraction collector from the detector). Factors which affect delay volume are:

• Flow rate• System Plumbing

– Distance from detector cell to fraction collection valve– Tubing Diameter

Delay Volume

Fraction Collector

DetectorColumn

The volume it takes the sampleto travel from the detector to the fraction collector dispense head.

Tube Volume Equation Vtotal=Tubing Length (in.) x (Tubing ID(in.)2)/2 x 3.14

To determine the delay volume accurately thevolume of all of the flow components between the detector and fraction collection vessel must be added to get the total delay volume.

Figure 6

Affects of Delay in Fraction Collection

Fraction Collection Start

Fraction Collection EndActual Peak Collection Start

Actual Peak Collection End

Delay Volume(Actual Peak Collection Start – Fraction Collection Start)

Fraction CollectedFraction Lost

Compound Recovered

Figure 7. This figure illustrates the large amount of compound which could belost if the delay volume is not considered. This also illustrates the illusion ofcollecting a pure fraction while obtaining possible impurity peaks.

Affects of Tubing Dimensions on Delay Volume

Detector

Graph 4. Delay retention between the detector and collection valve using 20ft x 0.04 in. ID tubing, flow rate @ 25 mL/min. Compounds used where caffeine, ethyl paraben, and biphenyl.

CollectionValve

0.26 min0.26 min

0.26 min

Minutes

1.0 2.0 3.0 4.0 5.0 6.0 7.0

mV

olts

1000

2000

0

Inner Diameter[inch] Inner Diameter [mm]

Volume[per cm in µL]

Volume[per inch in µL]

0.50 mm 5.146

11.577

20.581

0.75 mm

1.00 mm

0.020" 2.026

0.030" 4.558

0.040" 8.103

Calculation of Delay Volume

Determine items to be included in the delay volume calculation

• Outlet detector flow cell tubing• Inlet tubing from coupler to 3-way collection valve• Collection valve

• The low mount collection valve is immediately before the collection vessel no additional volume components are necessary.

Add all of the above volumes together to obtain a total delay volume.

It is important to verify your delay volume calculation prior to injecting and collecting a valuable sample. A 2.5 mg/mL solution of Methylene Blue dye in water could be used a visual test solution.

Gilson Detector Prep Outlet Tubing (11.58 µL/inch x 24 inches) 277.8 µL

3-way valve – preparative for 215, FC204 114 µL

Affects of Tubing Size on Delay Volume

0

1

2

3

4

5

6

7

8

5 10 15 20

Tubing Le ngth (ft)

Del

ay V

olum

e (m

L)

Graph 5. Delay volume is dependent on tubing length and inner diameter or the volume of mobile phase between the detector and the collection vessel. This graph illustrates the effect of these factors on the delay volume.

0.04 in ID

0.03 in ID

0.02 in ID

Delay Volume vs. Recovery

Graph 6. If the delay volume is set incorrectly, the recoveries could be drastically affected.

5ft x 0.03in. ID

0

20

40

60

80

100

120

140

0.75 1 1.5 1.9

Caffeine

Ethyl paraben

Biphenyl

Delay Volume (mL)

% R

ecov

ery

Summary

• Fraction collection goals:– High recovery– Pure sample

• Consider all parameters as critical for optimum recovery– Chromatography– Detector settings including wavelength– Collection Parameters– Delay volume or time

• Incorrect or ignored parameters can cause recovery losses of over 50%.• With 69 of 80 human therapeutic proteins on the market in 2003 sold at prices

over $10,000/gram, losses of 50% would be $5000/gram. Can you afford to ignore these factors?