The (Long and Winding) Road to LPGC-MS

LPGC, when coupled to MS, is a fast and robust alternative to traditional GC-MS – but the concept of LPGC is not new; in fact, the low-pressure route towards faster GC has been known to analytical chemists since the 1960s. However, several challenges have stood in the way of its widespread adoption. In 2000, Jaap de Zeeuw injected (no pun intended) new life into LPGC with a simple solution – a restrictor that maintained positive inlet pressure for a wide-bore column.

Advocates of the technique have had to navigate a long and winding road fraught with obstacles, including technical challenges, commercial pressures, and dismissiveness from the analytical community. Today, LPGC-MS is accessible to all via a commercialized kit.

Here, three trailblazers on LPGC’s journey – Jaap de Zeeuw, Hans-Gerd Janssen, and Steve Lehotay – share how the technique managed to persevere despite the hurdles and discuss where it might go next.
 

Meet the LPGC Experts

Jaap de Zeeuw

Jaap de Zeeuwis currently an international specialist in GC at Restek Corporation. With 41 years of experience in GC capillary technology, he has developed many PLOT columns, developed the first bonded wax column, published more than 100 articles in the field of GC, and is the inventor of fast low-pressure GC using restriction at the inlet. Jaap also made the world’s longest fused silica capillary, for which a Guinness World Record was granted. In 2016, he developed a new technique for coating PLOT columns based on SPIN deposition. Jaap has several hobbies, such as gardening, playing music, scouting, and helping out with his wife’s B&B, and he is currently working on a program for teaching creative thinking to technical/analytical people.

Hans-Gerd Janssen

Hans-Gerd Janssen has an MSc and PhD degree in analytical chemistry from the Eindhoven University where he studied and later worked in Carel Cramers’ group. After working at Eindhoven University for almost 10 years as assistant and associate professor, he joined Unilever in 1999. Janssen has written more than 200 publications on theory and method development in chromatography and MS. From 2004 to 2019, he was a part-time professor at Amsterdam University, next to his position as senior scientist in analytical chemistry, focusing on food analysis, at Unilever R&D Wageningen. In 2019, he accepted a part-time professorship at Wageningen University where his research focuses on recognition-based analytical chemistry. Janssen holds positions in the editorial advisory boards of several journals and has served in various scientific and organizing committees for international conferences.

Steven J Lehotay

Steven J Lehotay is a lead scientist with the USDA Agricultural Research Service at the Eastern Regional Research Center in Wyndmoor, Pennsylvania, USA. Since 1992, he has conducted scientific investigations and method development research involving improvement in the analysis of pesticides, veterinary drugs, and other contaminants in food and environmental samples. Steven is the co-inventor of the “quick, easy, cheap, effective, rugged, and safe” (QuEChERS) sample preparation approach. And he has been awarded with numerous honors, including the AOAC International Harvey W Wiley Award. According to the Stanford c-score metric, he resides among the top 0.19 percent of published analytical chemists. 

Disclaimer: Mention of brand or firm name does not constitute an endorsement by the USDA above others of a similar nature not mentioned. Lehotay’s views in this article are solely his own and do not represent the views or position of the USDA or any other entity.

When did you first encounter LPGC?

Jaap: I guess I should kick this one off! The idea first occurred to me in around 1997 when I was thinking about MS and how it commonly uses long 0.25 mm columns. This is a logical choice when you need positive pressure in the injection port and the MS is running under vacuum – you need that length to have sufficient restriction. But, as I’m always looking for a challenge, I wanted to see if it was possible to use a wide-bore column with MS rather than the typical narrow-bore.

Based on the Van Deemter equation, I realized that – theoretically – at lower pressure a much higher optimal linear velocity could be obtained. I checked the literature for vacuum GC but saw that the setup being trialed was often challenging because the vacuum had to be created in the injection system. Instead, I proposed using pliers to restrict the flow on the inlet side of a metal 0.53 mm i.d. capillary. This approach led to a similar separation but an approximately nine times faster run time – and that simple idea was the basis of what we’ve continued to build upon with LPGC-MS to this day!

Steve: Let’s see, my first email to Jaap was on February 25, 2000 – I had come across the title of a presentation that he made in Gifu, Japan, in November 1999, and Aviv Amirav had given me Jaap’s contact info. In the email, I described how I had already made plans for a summer student, Katerina Mastovska, to investigate LPGC-MS with a quadrupole MS instrument. Interestingly, though, I first heard about “subambient pressure GC-MS” in 1989 from Mark Hail of Richard Yost’s group at the University of Florida when we were both graduate students there.

Hans-Gerd: Well, the general concept has been around for a while – Carel Cramers, Piet Leclercq, and Jack Rijks started research into speeding up GC separations at Eindhoven University sometime in the 1970s. Cees Schutjes, the first PhD student in the field, defended his thesis in 1983, which included a theoretical treatise of the Golay equation. From this equation the increase of the mobile phase diffusion coefficients immediately followed as one route towards faster GC. When I started as a student in the Cramers’ group in 1986, the thesis of Schutjes was standard information we all had to study. We also studied vacuum outlet conditions for their higher speed. But it was only around 1998 that we started doing experiments in low pressure GC, jointly with Jaap.

Did you immediately recognize the benefits of this technique?

Steve: I’ve always felt fortunate that I learned chromatography from John Dorsey at the University of Florida at the time. Being a great teacher, Dorsey always started each chromatography class with questions about his previous lecture. After spending weeks on the theory of chromatography with an emphasis to optimize separations and peak resolution, Dorsey began class one day by drawing a chromatogram on the chalkboard of two peaks with excellent resolution about two minutes apart. He then asked, “What’s wrong with this separation?” He brushed aside the aspersions that his drawn peaks weren’t perfectly Gaussian, and no one in the class saw a problem. He announced, “It’s wasting time!”

I’ve never forgotten that moment or that concept, but it seems too many others have. I think too many chromatographers and mass spectrometristsforget that we are all analytical chemists. The specialist’s mindset fixates on the power of chemical separations and not enough on practical matters of sample preparation, throughput, ease, cost, ruggedness, validation, and – above all – robustness. When I learned about Rapid-MS in 2000, I knew immediately that the restriction capillary was a brilliant idea and a solution for fast GC-MS.Having entered the “real-world” of pesticide residue monitoring in 1992, I recognized the benefits of LPGC very quickly. I was taught to always use a guard column in chromatography, and the idea to use the guard column also as a restrictor in LPGC was an elegant solution that I wish I had considered first!

Jaap: I immediately recognized the value of LPGC-MS, but some of those around me did not initially. Once I came up with my “simple solution,” I proposed the idea to the management of Varian – the company I was working for at the time. They didn’t take interest at first, so I wasn’t able to do any experimentation. I then spoke to some of my esteemed colleagues – namely Carel Cramers, Aviv Amirav, and Hans-Gerd Janssen – and Varian became more interested with this expert backing. However, they decided it was more of an “academic” pursuit best left to the University of Eindhoven, and this is when Hans-Gerd and I began working together.

The initial protocol, based on the 0.53 mm capillary that had to be squeezed on one side using pliers, seemed really promising so we eventually filed for a patent – the initial design used a short 0.10 mm ID restriction (ca. 60 cm) coupled with a 10 m x 0.53 mm capillary. We worked with over 20 external groups who all came back with results showing the speed benefit of the columns. On top of this, it was clear that the technology would fit into many different application areas.

At this point, Varian took ownership of the product and decided to introduce it exclusively for the ion trap MS, with application in environmental trace analysis of pesticides and PCBs. I think this was a mistake – it took three years before this “Rapid MS” offering was also made commercially available because it didn’t gain the expected revenue. Notably, the Rapid-MS instrument had a slower data acquisition rate than others, and I think this put a lot of people off.

Hans-Gerd: It’s a good point Jaap makes here. The larger instrument and column manufacturers are sometimes a bit risk averse, so even though Varian eventually recognized the value of LPGC, it took a while to develop. Generally, I think the larger companies try to target scientists at conferences and hope that they will start to use the new technology and spread the word. That works, but is not a rapid route.

At Eindhoven University, we always had a massive interest in fast GC. Over the years we have helped many labs convert their regular GC to a faster run. Although we were mostly following the route of narrow-bore columns, it was clear that other options like very short columns, columns packed with very fast particles, or low-pressure outlet conditions also had their unique advantages. LPGC was always an option that people liked because it required only very small modifications to the equipment. Sometimes narrow columns are the preferred route, sometimes LPGC is better, sometimes both approaches work. But I’d say we very much recognized the benefits of LPGC-MS when it came along.

What additional developments led to the technique as it is today?

Jaap: In the first three years, there was a lot of noise in the market. Lots of people saw its potential and wanted to experiment with it – one such person in particular was Steve. During this period, I recognized some limitations in terms of the restriction lifetime/maintenance and the coupling, but never got the chance to work on that at Varian. However, when I joined Restek in 2008, I finally had the opportunity to try out different ways of making the restriction.

I came up with another simple solution based on making the coupling with PressFit and positioning this inside the injector body. This meant the coupling and restriction were always at high temperature and in an inert atmosphere. A publication was written and patent filed, but it never made it to a commercial product.

Steve: My colleagues and I had issues with the very narrow restriction capillaries in the Rapid-MS product. We ended up simply connecting two commercially available columns from any vendor (5m, 0.18 mm i.d. guard/restrictor capillary with a 15 m, 0.53 mm i.d., 1 µm film thickness analytical column), which provided both more robustness and theoretical plates. I reached out to all the GC vendors for 20 years about LPGC, and in 2021, Restek finally commercialized a product using our column dimensions.

Aside from this, there have been many technological advances in the past two decades to continually improve upon the performance and features of LPGC-MS. Most notably, commercial triple quadrupole MS/MS instruments were introduced, which provided greater targeted analyte detectability (both sensitivity and selectivity) and faster data acquisition speeds. High-resolution MS instruments have also been introduced, for which LPGC is compatible. Improvements of QuEChERS and analyte protectants streamlined sample preparation and improved peak shapes in GC. The development of a light and reliable capillary column union also helped make LPGC more practical for shipping and installation.
 

Who should consider LPGC-MS?

Steve: I think everyone that is using GC for analysis should consider LPGC-MS. My lab has used it routinely for nearly 20 years now. Why? Firstly, megabore columns are preferable in routine monitoring using GC because of their much greater sample loadability and robustness. LPGC is therefore very useful for rapid, sensitive, and robust analysis of pesticides, environmental contaminants, and pretty much any GC-amenable analyte that isn’t too volatile. Secondly, LPGC as a product can be used as is in general applications, but, as a technique, it also possesses more parameters for investigation and innovation than standard GC-MS. For example, Amirav, Fialkov, and I recently published a paper (1) using resistive heating in LPGC-MS with a short 0.25 mm i.d. capillary column that achieved multiresidue analysis in <1 min. My long-term research plans are to implement such methods to enable ultra-fast monitoring without having to ship samples to a lab, for example. In the short term, one of my projects is to systematically evaluate LPGC-MS in food safety applications using different column dimensions.

Jaap: What we basically do with LPGC is trade efficiency for speed, but using a robust solution. As long as components that elute at the same retention time can be separated by MS, LPGC is useful. Of course, if isobaric components elute together it will not work. In that case, we need separation by chromatography using highly selective stationary phases, or use other means, such as software tools, to meet the analytical need.

It’s also worth noting that there are other ways to speed up MS separations, like working at higher flow, extreme fast programming, and using short, smaller diameter capillary columns. But these do not provide the robustness and loadability of the vacuum GC solution.

Hans-Gerd: LPGC is not unique – it is one of approximately 10 routes towards faster GC. But it is a rather simple technique. In generic terms, I would think it is the preferred method if you have simple separations not requiring a high peak capacity and are using MS detection, while sample preparation is not too much of an issue. In such situations, LPGC can increase your sample throughput up to five or ten-fold.

So why haven’t more people adopted LPGC?

Hans-Gerd: It’s a good question, and we have to be honest here. We clearly thought that all labs would move to faster GC, but nowadays we see that the vast majority of them are still using the classical columns with run times between 30 and 60 minutes. I see two reasons for that. First, I think the need for very fast separation is limited. With a regular run time of say 45 minutes you can easily do 30 samples per day. Many labs will not have that many samples. And we should also not forget that the sample preparation, data interpretation, and paperwork that comes with 30 samples per day can be significant. The GC run itself is usually not the rate-limiting step, and for many the benefits of faster GC are simply not worth the investment and risk.

A second reason for the limited acceptance of fast GC is the strong overpromise in literature and by some manufacturers. With fast GC there is always a price to pay. Narrow-bore columns offer you a faster analysis speed at the expense of a slightly reduced system reliability. LPGC only works for simple separations. Fast temperature programming reduces your separation power a lot. These limitations have not always been communicated honestly.

Jaap: As I’ve mentioned, some were slow to recognize the value of this technology and I think this has been the biggest limiting step to progress. Even when people (or instrument companies) did recognize the benefits of LPGC, they couldn’t see the value in pursuing it commercially. In particular, Varian didn’t fully recognize the impact of LPGC having low data acquisition rates.

Steve: Right. Varian made the initial mistake of treating Rapid-MS more as an “introduction device” to their ion trap MS detector, which was not ideal for LPGC due to its slow (250 ms) data acquisition rate. Secondly, the product was introduced before it was fully studied and optimized, leading to narrow marketing and less-than-ideal column dimensions. Another crucial mistake was the over-pricing of the product. Justifiably, the primary goal of a company is to make money – meeting customer needs is only pursued if it serves this primary goal – but if more customers had demanded LPGC, then vendors would have taken more notice. Even so, the 20-year time frame of Jaap’s US patent held by Varian and then Agilent certainly put a dampener on commercialization by others until now.

Hans-Gerd also makes good points. Furthermore, LPGC does not work for MS techniques that do not operate with the ion source under vacuum conditions. Also, many volatiles are already analyzed quickly in standard GC, thus LPGC is not going to provide as much gain compared to analyses that currently take 15-60 min. Otherwise, LPGC-MS trades a small degree of separation efficiency for speed, sensitivity, and robustness.

Misaligned incentivization is another problem. For example, nearly all academicians choose to study what rewards and/or interests them the most. The “novelty” of the idea and technique usually motivates them, and they tend to seek applications to suit their preferred tool, not the other way around. Grants, patents, and citations usually drive their choices. Company scientists are similar in that they spend an inordinate amount of their time on niche applications that are difficult for those customers who complain the most loudly. In my view, they should put more focus on improving efficiency and performance of their most profitable applications that are taken for granted. Moreover, the scientific publishers and media tend to highlight “sexy” topics, which are promoted by those scientists on their editorial boards.

I can give several other reasons more people haven’t adopted LPGC. There have been informed criticisms over the years, such as the need for more separation efficiency, analyses of volatiles, or problems with column bleed. But one of my biggest frustrations is to notice dismissiveness, disbelief, and disinformation from those who have never studied or tried LPGC themselves.

What would you like to see for the future of LPGC?

Jaap: In theory, the technique could allow even faster separations depending on the temperature programming speed of the instrument. The 15 m column chosen here could also be replaced for a shorter column – like a 7 m x 0.32 mm or even a 3 m x 0.25 mm column – as long as vacuum conditions inside the separation capillary are created. However, there will of course be other challenges for using such short capillaries, including sample introduction and focusing.

Steve: Do you know what may be even better than LPGC-MS? Fast GC-MS using supersonic molecular beams (SMB) – also known as “Cold-EI.” In SMB-MS, the column outlet is not under vacuum, thus LPGC is not possible using that detector. However, column flow rate can be increased to 32 mL/min, for example, to provide rapid, high-quality analyses. I would urge anyone in this field to view the application notes and publications from Aviv Amirav. His long and winding road has also been fraught with multiple bad timings (company consolidations) and human foibles.

Hans-Gerd: A drawback of LPGC is that you need to buy a special “thing”; the restrictor. A big step forward would be LPGC without the need for a restrictor at all. This would require the gas inlet system of the GC to be able to work with sub-ambient pressures while still avoiding the ingress of air via the split exit – but it should be technically feasible.

Also with regards to the future, Steve’s work on faster integration methods (summation integration) should be mentioned. The GC run time might be relevant for the total duration of your analysis, but in terms of costs it is not the main contributor. LPGC combined with simpler, fully automated, and more reliable methods for peak integration is an ideal combination.

Steve: Thanks for mentioning the summation function integration. Indeed, you are right to point out some of the drawbacks of LPGC, so I want to summarize how we’ve been overcoming the current limitations of this technique: i) we have done high-throughput and easy sample prep with QuEChERS (and now QuEChERSER) since 2003; ii) we also have been using analyte protectants since 2003 to improve peak shapes and separations for somewhat polar analytes; iii) for the past decade, we’ve been increasing selectivity of detection by using MS/MS (for targeted analytes only); and iv) since about 2015, we’ve used summation integration (for targeted analytes only, too).

You also mentioned the “thing” needed for LPGC – that “thing” is merely an appropriate guard column and union, which are not unusual items. I suppose ferrules for the megabore column is another “thing,” but standard columns, liners, septa, ferrules, nuts, and so on are also “things” that by the same logic should preclude anybody from doing any analyses at all! In any case, Restek now sells that “thing” in the same way as any other item, and LPGC has always been available as a custom item.

Hans-Gerd: Excellent remarks Steve! And it’s this type of discussion that highlights, I believe, our passion for this often overlooked technique. I know only a few users of LPGC, but I am not aware of anyone who tried it and gave up. The technique works and is reliable and there are certainly more people who could benefit from it.

Jaap: I couldn’t agree more.Essentially, LPGC-MS can speed up a lot of conventional MS applications where analysis time is important. It should be of interest to anyone using MS, in my opinion, but the best way to get the word out there is to show the data and let experienced chromatographers speak up.

Steve: As hundreds of people involved in GC analysis can attest, I have discussed LPGC in nearly every encounter with them for more than 20 years! My inclusion of a few slides about LPGC has been a staple in most of my presentations, to the point that some people are sick of hearing about it (and I am definitely sick of talking about it!). I’ve emailed many gurus of GC over the years about LPGC, and when I noticed research or review articles in which LPGC should have been mentioned, I sometimes emailed key publications about LPGC to the authors to inform them of their oversight.

As a US federal civil servant, I have no business or financial relationships with anyone about my work, and my motivation is to help others improve their chemical analyses and lab operations. I admire Jaap de Zeeuw for inventing the restrictor approach, and I’ve wished for 20 years that my lab could simply purchase pre-connected LPGC columns with the dimensions of our choosing. Now that this option is available, there is one less excuse for those who haven’t tried LPGC-MS.

Hans-Gerd: This has been a trying journey, that’s for sure! But we’ve all come away from it having grown and developed as analytical scientists along the way. For us, the success of LPGC is an excellent example of a cooperation between three parties: academics who develop theoretical concepts, hardware developers that create the tools to put these ideas into practice, and users who re-define the workflows in their laboratory to maximally benefit from the new development. Many successful new methods are the result of such three-party interactions.