Since the mid 1990s, researchers and scientists have understood the inherent advantages of purifying small, drug-like molecules (molecular weight less than 1500 daltons) with supercritical fluid chromatography (SFC). Among these advantages are speed and environmental impact. SFC utilizes carbon dioxide (CO2) as the main component in the mobile phase.
The low viscosity and high diffusion coefficient of carbon dioxide allow for a much wider optimal flow rate range compared to that of the liquids used in HPLC. At the high end of this range, the linear velocity (Van Deemter optimal) of the SFC mobile phase is three to five times that of HPLC, equating to SFC being at least three times faster than HPLC, assuming that the column dimensions are the same.
SFC provides an environmentally friendly alternative to HPLC. Carbon dioxide (recycled from industries such as petrol chemical) typically composes between 60 and 95 percent of the mobile phase. When the purification is completed, the researcher is left with between 60 and 95 percent less solvent compared to HPLC, leading to thousands of dollars saved per purification due to the rising cost of solvents in today's market. The lower solvent amount necessary for SFC purifications also equates to less time and energy required to dry the sample. Because of these and other factors, SFC is considered a green technology.
Many of the hurdles associated with SFC, such as carbon dioxide delivery and back pressure regulation, have been addressed and overcome during the past 10 years, yet one challenge has continued to plague the community. When the eluent passes the SFC system’s back pressure regulator, the carbon dioxide in the mobile phase expands approximately 500 fold, making high-purity, high-yield fraction collection with standard HPLC equipment impossible.
To address this problem, the Modular SFC CFC-2 centrifugal fraction utilizes centrifugal force to separate the CO2 from the sample and any organic component of the mobile phase. The gaseous CO2 is directed through an exhaust hose typically to a cold trap, and then to the nearest hood. Heavier liquid molecules, along with dissolved and/or precipitated sample components, are collected in standard lab glassware in the CFC-2 rotor. To collect the next fraction, a distributor-like mechanism mounted to the rotor redirects the elution tube to the next fraction vessel while the rotor continues to spin. When not collecting, the instrument diverts the flow to waste.
The CFC-2 connects to the eluent flow from the outlet of the back pressure regulator of any commercially available SFC instrument. Fraction collection can be triggered from any SFC detector that is capable of sending out an analog signal (including the abundance output of a mass spectrometer). The next fraction command can be activated according to time, threshold and/or slope.
When using destructive detectors such as a mass spectrometer and/or an ELSD, a flow splitter is used to sample the main flow, which is directed to the CFC-2, and thus enabling collection of stacked injections and mass directed collection. Figure 1 shows the typical flow path and electronic connection for CFC-2 peak detection.
.jpg)
Note: while the figure depicts the autosampler injection going into the combined methanol and CO2 stream, some manufacturers set up the flow path so that the injection goes directly into the modifier stream and is then combined with the CO2 at the head of the column. Click to enlarge
|
A unique attribute of the CFC-2 is the Centri-Fan, which provides the analyst with the ability to dry fractions simultaneously during collection. Fan blades inside the Centri-Fan direct the rotor chamber gas at up to 50 L/min across the surface of the liquid contained within each fraction vessel. The gas flow repeatedly shears the top layer off the solvent until the fraction is completely dry.
The centrifugal force created by the spinning rotor ensures that no material is lost during the drying process. This was confirmed by studies where fine, dry powder was subjected to this blowdown process with no loss of weight. The drying rate for methanol, when using a four 250-mL bottle rotor configuration, is greater than 70 mL/hr per container for a total of more than 280 mL/hr.
To demonstrate the effectiveness of the centrifugal force of the CFC-2 to maintain sample integrity, two recovery tests were performed at different flow rates, while different techniques were used for calculating recovery.
Test 1
Instrumentation
SuperPure Discovery Series SFC, Thar Technologies and CFC-2 Centrifugal Fraction Collector, Modular SFC, Inc.
Method
Total Flow Rate: 4mL/min
Modifier: Methanol
Gradient Conditions: Isocratic at 25% modifier
Injection Volume: 50uL
Column Oven: 35o C
Column: 2 Ethylpyridine 4.6 mm x 250 mm
Detector: 220 nm
Back Pressure Regulator: Isobaric at 150 Bar
CFC-2 Method:
Max tube time = 4 min
Delay time = 0.01 min
Collection Window 1 = 1.00 min to 2.50 min
Collection Window 2 = 3.25 min to 4.75 min
Reset = 5 min
CFC-2 to SFC Connection
The CFC-2 was connected to the outlet of the back pressure regulator via 0.030 in. ID PEEK tubing. Stereo cables with RCA jacks were used to connect the detector signal, start signal and chart mark signal.
Sample Preparation
Astemizol, Cortizone, and Sulfamethoxazole were dissolved in methanol at a concentration of 10mg/mL for each compound. Each compound was also dissolved in methanol separately at ~5mg/mL to check elution order.
Analytical Note
While Astemizol and Cortizone produced different retention times when injected separately, it was found that they nearly co-eluted when in solution together. Therefore, for the purpose of this experiment, Astemizol and Cortizone were treated as a single peak (peak 1). Sulfamethoxazole was collected as peak 2.
Sample Purification
Four injections (50uL each) were made of the ~ 10mg/mL solution of each compound and the fractions collected. Peak 1 was collected in tube 1 for each injection. Peak 2 was collected in tube 3 for each injection. For purity and recovery study, Sulfamethoxazole was chosen. (Figure 2)
Amount Purified
4 x 50 uL = 200 uL = 0.200 mL
0.200 mL x 10 mg/mL = 2.0 mg of Sulfamethoxazole
Peak 2 produced a volume of approximately 3 mL for the four combined injections. This is due to solvent evaporation during collection. Peak 2 was then diluted to 10 mL for purity and recovery testing. This resulted in a theoretical concentration of 0.200mg/mL. A 50 uL volume of this sample was injected producing the following results:
The injection produced a peak area of 4.46. There was no measurable trace of peak 1 indicating a purity of nearly 100%. (Figure 3)
Recovery Testing
The original standard solution containing 10 mg/mL of each compound was diluted 50 fold by adding 20 uL of the standard solution to 980 uL of methanol. This produced a 0.200mg/mL solution. This was injected at 50uL and the area calculated. (Figure 4)
The peak area for Sulfamethoxazole in the standard solution was 4.19.
Recovery Calculation
((Peak Area for recovered fraction) / (Peak area for standard solution)) x 100 = % Recovery
4.46 / 4.19 x 100 = 106% Recovery
Recovery Discussion
The excess recovery can probably be attributed to the difficulty of pipetting uL volumes of methanol during dilution of the standard solution.
Test 2
Instrumentation
Berger Prep SFC System, Thar Technologies
CFC-2 Centrifugal Fraction Collector, Modular SFC, Inc.
Method
Total Flow Rate: 50mL/min
Modifier: Methanol
Gradient Conditions: 5 to 65% modifier over 6 minutes
Injection Volume: 0.500 mL
Column Oven: 35oC
Column: Diol 21.2 mm x 150 mm
UV Detector: 240 nm
Back Pressure Regulator: Isobaric at 140 Bar, Isothermal at 60 degrees Celsius
CFC-2 Method:
Max tube time = 1.0 min
Delay time = 0.00 min
UV threshold = 12% full scale for start of peak; 8% full scale for end of peak
Stack = 1
Auto = 24
End Run = 6 minutes
CFC-2 Rotor: 24 position, 16x150mm test tubes
CFC-2 Rotor speed: 1100 rpm
CFC-2 to SFC Connection
The CFC-2 was connected to the outlet of the back pressure regulator via 0.040 in. ID PEEK tubing. Stereo cables with RCA jacks were used to connect the detector signal, start signal and chart mark signal.
Sample Preparation
Carbamazepine was selected as the compound in test two. Carbamazepine was dissolved in methanol to yield a concentration of 73 mM. The solution was prepared by an analytical chemist employed by the company were testing was performed.
Sample Purification
Carbamazepine was loaded into three consecutive positions of a 96 deep well plate and placed on the deck of the Berger Prep SFC System. The CFC-2 was loaded with pre-pared 16 x 150mm test tubes. The three injections were made at 0.50mL each.
After collection, the partially dried tubes were transferred to a Genevac rotary evaporation system and allowed to spin for an additional three hours so as to eliminate the potential of the solvent contributing to the sample weight. The fractions were completely dry before weighing. The fractions were weighed using an automated weighing system.
Analytical Note
The first peak in the chromatogram did not yield any mass and was therefore attributed to solvent effect. (Figure 5)
Pos |
Sample plus tube |
Tare Weight |
Recovered Amount |
Theoretical |
Recovery |
AVG Recovery |
2 |
11566.92 |
11559.36 |
7.56 |
8.62 |
87.69 |
91.98 |
4 |
11506.05 |
11498.16 |
7.89 |
8.62 |
91.52 |
|
6 |
11522.85 |
11514.51 |
8.34 |
8.62 |
96.74 |
|
Table 1. RecoveryConclusion
The CFC-2 provides SFC users the ability to collect fractions in an HPLC like fashion while achieving suitable recovery and purity values. It should be noted that almost five percent of the peak area occurs in the tailing region of the peak after the chart mark trace indicates that collection was stopped. Setting an ending peak threshold much less than eight percent or reducing injection mass to reduce tailing probably would have increased average recovery to above 95%.
The available Centri-Fan accessory simultaneously dries fractions during collection. By combining the fraction collection and drying steps, purifications can be completed in less time and with less effort compared to traditional SFC collection techniques.
Acknowledgements
The author would like to thank William Farrell and David Masters-Moore of Pfizer in La Jolla, CA as well as Frank Riley and Todd Zelesky of Pfizer in Groton, CT for all the technical and analytical help during the process of generating the data used in this publication.
Michael A. Burns is the vice president at Modular SFC. He may be contacted at ChromatographyTechniques@advantagemedia.com.