Nexera UC Supercritical Fluid Extraction/Supercritical Fluid Chromatography System

Please tell us how you became interested in SFC technology and what made you decide to use the system.

Our team mainly analyzes and purifies synthetic compounds. First, we had started to use an SFC system for preparative purposes. We purchased a semi-preparative SFC system in 2010 expecting higher throughput with less solvent consumption. At that time, we had been planning to increase the number of units if we were able to verify the usability of the preparative SFC system on achiral compounds. However, the system often failed, and we were unable to use it in many cases. Around 2013, we had a project that required a separation and a purification of the chiral metabolites. We could easily separate some metabolites with SFC while they could not be separated by either normal or reversed-phase LC. We ultimately decided to purchase the SFC systems for analytical purposes because we thought we would encounter such compounds again, and a more reliable analysis would also be required.

This experience put the thought into your head that the SFC system would also perform on the different compounds as well, didn’t it? 

Yes. We thought it would be great if we could easily use SFC as an option when handling compounds that are hard to separate using an LC system.

Do you have any other ideas about the benefits of SFC?

It’s already well-known that SFC has the excellent ability to separate chiral compounds. Additionally, I’ve been feeling that SFC addresses many potentials not only for giving the orthogonal property to the LC, regarding the chromatographic selectivity, retention, and elution. Although I have worked with an SFC system for half a year, many cases have surprised me since they showed dramatically different chromatographic selectivity from the traditional reversed-phase chromatography. On the other hand, I often hear the phrase "normal phase-like" as a determination of SFC. However, I have an idea that SFC is obviously different from normal phase chromatography.

Do you mean that a combination of the normal phase column and SFC provides results that are different from normal phase LC?

Yes, I do. It was unbelievable that two compounds that eluted at the void and the last part of the gradient in a reversed-phase LC appeared at the same position in SFC. I do not think this result can be obtained either by a normal phase LC or by a reversed-phase LC with different types of columns. Furthermore, it’s very helpful that DMSO can be used as the sample solvent. This aspect is also different from normal phase chromatography since DMSO, which has a retention on a normal phase column, causes peak-broadening in normal phase mode.

SFC gave similar retention times for the compounds that have weak and strong retentions in the reversed-phase mode.

I know that the close retention times are not desirable from the point of view of the chromatographic separation. However, what I would like to point out is I believe that SFC may analyze some compounds that cannot be separated by an LC system because the separation selectivity of SFC is significantly different from that of LC. Considering the synthetic compounds, I expect that SFC analysis may play better than reversed-phase LC because it provides a proper retention and a good peak shape being independent of the hydrophobicity or polarity of the compounds. Additionally, the idea of "unified chromatography (UC)," which allows using it from the supercritical to the 100% modifier solvent without any complicated change to the platform, might expand analysis possibilities for a wider range of samples.

After using an SFC system for six months, what do you think of its usability and the theory behind it as compared to an LC?

I expected that SFC procedures would be quite different from those of LC because a CO₂ cylinder is used. However, once we understood the mechanism, the procedures were not so different from those of the Nexera X2 and Nexera XR we regularly use. The appearance is not very different either. I rather feel that the operations are easier than LC systems. Also, the SFC consumes a very small amount of mobile phase. As a result, for analytical scale, both the CO₂ and the modifier reduce at a lower rate. We check the remaining mobile phase less frequently as compared to the LC system.

This is an encouraging message to LC users who hesitate to use an SFC system. Can you think of any negative points about the SFC as compared to the LC and GC?

I think a major problem is that its use is limited by laws and regulations in Japan. In Japan, the types of usable columns are limited, which is frustrating. Another problem is that any generic method, such as the combination of an ODS column and water-acetonitrile condition in a reversed phase LC, has not been found in SFC yet. What I mean is a method to separate any compounds at a certain level instead of separating (only) target components. There is an ordinary scheme in a reversed-phase LC method. We first check separation using an ODS column, arrange the gradient somewhat, and then change the column and mobile phase conditions if necessary. In SFC, there are no analytical conditions that can provide a certain level of separation to try first. If such a condition is found, people may be more interested in trying SFC. Currently, we need to perform scouting first, which may be a problem with SFC.

In other words, when you have more options, optimization is more difficult.

It is. By the way, may I ask the original purpose of developing the SFE through the collaborative research?

What kind of applications do you want to challenge using the SFE system?

When we saw the online SFE system, we first thought of how we would use it to purify synthetic compounds. As with LC systems, we often hear about the troublesome steps in the purification procedure; which solvent should be used to dissolve the samples, how well they are dissolved, how sample solvents negatively affect chromatogram, and how precipitation occurs because of the solubility difference between the mobile phase and the sample solvents, which causes clogging or stops the instrument from operating. We thought using an online SFE as a sample introduction device into a preparative SFC system would be good because only extracted and dissolved compounds would be loaded on the column. Therefore, there is no need to consider the sample solvent and the risk of clogging. In SFE, the loading amount can be controlled by the extraction time instead of the injection volume. This aspect allows the compounds that have poor solubility into CO₂ increasing in the loading amount by extending the extraction time. Additionally, we might also be able to collect the entire amount of the target compound by repeating the extraction process. For these reasons, I expect that we will be able to use the SFE system without worrying about any problems related to sample solvents. I did a search about this idea and found several reports in other countries as well.

In contrast, there is another case that the remaining compound on the residue after the extraction can be the pure product. It would be interesting if we could place a crude material in an SFE vessel and wash it with CO₂ to obtain a purified product in a dry condition when the vessel is open.

Shimadzu develops SFE systems to extract and analyze target compounds. But you are thinking about washing target compounds instead of extracting them, which means that you extract impurities and then purify the remaining portion?

Unfortunately, the Nexera UC is an analytical scale system. We can only evaluate the possibility of SFE modules for purification currently. Even so, we have learned that washing compounds with CO₂ can be applied depending on samples through our past experiences.
Of course, the purification performance should depend on the solubility of the target component and other impurities to CO₂. I mean, all "extraction", "clean-up", and "purification" functions are essential for the truly variable purification platform. I hope that Shimadzu will develop this type of system.

Thank you very much for very good comments and suggestions.
Would you tell us why you chose Shimadzu’s Nexera UC instead of other similar products?

Because SFC systems are mainly used for screening with multiple conditions, we were conscious about the burden for changing columns frequently. Therefore, we focused on specifications of the column oven, the number of columns that can be installed, and the ease of column replacement. Although we looked at other companies’ products, we ultimately selected the Nexera UC because Shimadzu's products use flexible piping to make column replacement easy, meaning complicated column replacement work would not be an issue anymore.

The final deciding factor was that an LC and an SFC can be utilized in the same platform with the Nexera UC. We hesitated to purchase a dedicated SFC system only to discover that it would be useless for our purpose. I also feel that switching back and forth between an LC and an SFC is desirable, as the combination of this technique with an MS detector would become more common in the near future. Two individual systems of LC and SFC simply require double the space. Additionally, I think it would be a big burden to change the systems and piping again to connect an MS system to an LC system at certain times, and to an SFC system at others.

We also appreciate the mobile phase blending function. The current standard modifier is 10 mM ammonium formate methanol solution. We blend methanol and 100 mM ammonium formate methanol solution by a ratio of 9:1. We are using the instrument for a wide range of purposes because the measurement can be performed on an LC without changing configuration settings.

Is the operating rate of the instrument high?

Yes. Since we started to use the Nexera UC, we have been showing the results acquired by SFC every time we have an opportunity. I think people's awareness about SFC has been changing. Separating the analysis system and semi-preparative system improved the efficiency of preparative SFC itself.

The system you purchased can perform regular LC, SFC, and on-line SFE. Which of these do you use the most?

We mainly use it for SFC because the LC analysis can be performed by other LC systems. We have so many things to try on the SFC and SFE currently. We basically use the system for SFC and try it for SFE when we have time to spare. We want to try using it while switching back and forth between SFC and LC as soon as possible.

What do you think about SFE usability? Is it different from existing chromatographs?

When it comes to usability, the procedures are not very complicated. Once we understand the flow diagram and create methods, all we need to think about is how to optimize the static and dynamic extractions procedure. One thing we should be aware of is that the system volume increases by 180 µL when an SFE is added to a regular liquid injection autosampler. The suitable volume or shape of the extraction vessels to obtain higher extraction efficiency may vary according to the sample form. I hope this will be improved.

Thank you for sharing your opinion. We will try to make improvements, including a better extraction vessel.
Does the Nexera UC have any outstanding features?

I had heard that SFC technology does not provide very high reproducibility. I have been using the system for six months constantly worrying about this. However, we never experienced poor reproducibility. Also, we obtained the same result as that had been observed one or two months before. This convinced us that the SFC can be used just like the LC, which is an excellent feature of the Nexera UC system.

You mean that the reproducibility is better than your first worry, don’t you? The basic structure of the pump is the same as that of regular LC. The only difference is the pump head that delivers CO₂. Feeding back the temperature on the pump head helps to provide the good reproducibility. The other key technology is reducing the pressure fluctuation on the back pressure regulator (BPR).  It influences the reproducibility. There is almost no pressure variation in Shimadzu's BPR, which also contributes to excellent reproducibility.

Do you think that it’s possible to resolve the worse S/N ratio than what’s achieved with an LC? Is it an unavoidable issue because of the principle of this technique? I am sure the system can become more attractive to LC users in development and quality control departments if this problem is solved.

Indeed. The influence of the refractive index is more significant in an SFC because the pressure-resistant cells are used. Shimadzu LabSolutions has a function to compensate the chromatogram by subtracting what is obtained in the region with no absorption. Noise can be reduced in the same way. Of course, we’d like to plan to develop the detector cell with a lower refractive index to make the system provide a S/N ratio as good as that of LC.

What do you think about the software? Would you tell us your impression about the method scouting?

I appreciate that method scouting significantly reduces the work for method management because columns and mobile phases are switched very frequently. The fact that we can use the method scouting function through LabSolutions, which we generally use, is very helpful for us. The method scouting function allows us to create just one method file even if we need to modify many parameters, such as the gradient program, columns and their positions, and the blending ratio of the mobile phases.

What do you think about Shimadzu's support system?

I am satisfied because support is provided quickly. As this system was the first SFC system here, at Fuji Gotemba Research Laboratories, we were concerned about having troubles. However, we were given thorough explanations about the important points. After installation, we still had several unclear matters, for example, how to detect leaking, because the system uses supercritical fluid and CO₂ gas instead of liquid. But Shimadzu gave us precise instructions on the phone and sometimes visited us. Those actions wiped out our worries.

Thank you for the compliment. Could you share your opinions about Shimadzu overall and any requests to us?

It is beneficial to share our views with a Japanese company like Shimadzu because the voices of chromatography users in Japan can influence product development. I believe our opinions led to the development of excellent add-on software such as method scouting, OpenSolution, and SharedBrowser as well as the improvement of measurement and analysis usability with LabSolutions. The unique analytical technologies such as iPDeA and iDReC can be available for the data files acquired in the past, just by updating to the later version. The new peak detection algorithm of iPReC Finder provides better peak integration that looks like the person does it manually. It results in fewer data processing steps with higher satisfaction. I had the opportunity to observe the development work of iPReC Finder. It was clear that the experienced development personnel were devoted to creating the best system so that the users do not have to configure settings each time. I was moved by their devotion and the younger development personnel's motivation and enthusiasm. I also feel that we currently communicate well with the staff in charge of the instrument development.

However, I am concerned that an attitude that satisfies customers' requests too much can be a negative aspect as well. For example, when two users have similar but different needs or they have needs that conflict with each other, you may add two similar functions to fulfill both needs. Fulfilling all of those requests may be difficult; adding new functions and new windows to software, or sometimes completely novel platforms or software, should be considered. Satisfying specific requests from users with a powerful voice or company-specific problems should not take priority over more general requests from users with a smaller voice. I hope Shimadzu fairly considers the requests of all users and clearly shows its direction while closely listening to users' opinions.

Thank you very much. Do you have any other requests?

I am waiting for Shimadzu to release a preparative SFC system. It would be great if it is equipped with an SFE system.

You mean an SFE system as an autosampler?

Exactly. I also think that the parallel screening, often applied in other countries, will help to improve the throughput.

I understand your requests. Thank you so much for your important ideas.

The Nexera UC has the potential to provide new solutions to customers that have a problem with analysis using an HPLC because it sometimes demonstrates separation characteristics different from existing HPLC systems.
The three-in-one system adopted by Chugai Pharmaceutical can be used as SFC, UHPLC, and SFE units. We will support customers continuously in order to help them use our systems for various purposes and develop products that satisfy customers' requests.