Advances in Small Molecule Analysis

 New Adventures in GC–IMS for Nontargeted Screening of Food and Environmental Samples

Essentials of LC Troubleshooting, Part III

Characterizing RNA Modifications in Small-Volume Samples and Single Cells

 Sample Preparation Strategies 

A Tool for Lipid Characterization of HepG2 Cells

SFC Coupled with Drift Time Ion Mobility QTOF Mass Spectrometry 

Kevin A. Schug is a Full Professor and Shimadzu Distinguished Professor of Analytical Chemistry in the Department of Chemistry & Biochemistry at The University of Texas (UT) at Arlington. He joined the faculty at UT Arlington in 2005 after completing a Ph.D. in Chemistry at Virginia Tech under the direction of Prof. Harold M. McNair and a post-doctoral fellowship at the University of Vienna under Prof. Wolfgang Lindner. Research in the Schug group spans fundamental and applied areas of separation science and mass spectrometry. Schug was named the LCGC Emerging Leader in Chromatography in 2009, and most recently has been named the 2012 American Chemical Society Division of Analytical Chemistry Young Investigator in Separation Science awardee.

In the past, I have openly lamented the seeming lack of National Science Foundation (NSF) support for research into fundamental separation science. I have not been alone in that sentiment. This past fall, I was heartened to receive a NSF grant (CHE-2108767) to support explorations into how data science techniques could be used to advance complex on-line extraction and analysis systems. I am willing to eat my past words to some degree. While I think more federal research support could always be provided, with this project support, we are investigating fundamental relationships between the structures of molecules and their interaction with different materials, in the context of on-line supercritical fluid extraction–supercritical fluid chromatography (SFE–SFC).

We are working together with the Center on Stochastic Modeling, Optimization, and Statistics (COSMOS) in the Department of Industrial Engineering at the University of Texas at Arlington. We approached this group, led by Profs. Victoria Chen and Jay Rosenberger, when we realized our attempts at using standard multivariate optimization for SFE–SFC(1) might fall short in the long term. SFE–SFC is an extremely powerful analytical system, but it is quite complex in the number of variables, which need to be considered and optimized for new applications.

While I do not intend to drill down into the nuts and bolts of the full project details, our intent is to combine machine learning and surrogate optimization techniques to efficiently reach optimal SFE–SFC conditions for a wide range of applications. A big challenge is being able to generate a methodology, that could work across a wide range of molecule types and materials. SFE and SFC are known to be applicable across much of the space that both gas chromatography and liquid chromatography can analyze. The materials under consideration are those that might contain the molecules from which they need to be extracted (such as sample materials), as well as the SFC stationary phases, which are used to both trap and separate the molecules.

Deriving parameters associated with a particular molecule that are predictive of its physicochemical properties, as well as its interactions with a wide variety of materials, is also challenging. On the one side, much can be done by determining linear solvation energy relationships for solutes that have detailed property descriptors, but the sets of molecules where these descriptors are known is limited, and the descriptors are not easy to determine for other molecules of interest. On the other hand, properties that are easily calculable, such as pKa or log P, provide limited prediction capacity when the molecules are present in complex systems.

With that in mind, we have decided to pursue machine learning methods, which can encode the actual chemical structure of the compound and correlate it with measured properties. This type of work is being pursued in drug discovery and synthesis, but as of yet, only to a very limited degree in analytical chemistry.

In order to determine the best encoding strategy for the two- and three-dimensional chemical structures, we have been exploring the potential to predict vacuum ultraviolet (VUV) absorption spectra. Previously, we have investigated the use of computational chemistry methods for predicting VUV spectra (2,3,4). Techniques that use, for example, time-dependent density functional theory, do an okay job, but they do not often produce the fine spectral structure we can observe in experimental gas phase VUV spectra. Using a variety of deep learning methods, we have had good success now in predicting VUV spectra from chemical structures, and even vice versa. This should create a powerful tool to aid both VUV detection for gas chromatography, as well as provide a framework for us to advance our work to optimize SFE–SFC.

We will also use a surrogate optimization technique to study and optimize the on-line extraction and trapping variable. Surrogate optimization is an enhanced-response surface methodology. It incorporates a wider range of functions, in order to handle more complex response surfaces. Our team has been working on the code. For that initial work, we have focused on modeling electrospray response of different analytes, which is a bit simpler and less instrument intensive than jumping into the SFE–SFC optimization. This also builds off of previous related research, we have reported in the past (5).

Working with the students and faculty from COSMOS has been a great experience. While I am far from able to code or decode any Python, it has been extremely enlightening to gain a better appreciation for cutting-edge data science. The biggest challenge with such a collaboration is communication. As we are each experts in our domains, trying to bridge the gap requires discussions that often drop down to very basic fundamentals in each of our fields. But, as we begin to be able to better speak each other’s language, the potential for advancement in both of our fields has become clear.

Now, as some students graduate from the team, it will be interesting to see what kind of employment they can find. I have no doubt that analytical chemists with some data science background will be snatched up. I wonder more about the data scientists with some analytical chemistry background. I performed a cursory search on some job sites but did not find a glut of opportunities. Are analytical chemistry companies considering how they might utilize a hard-core data scientist? I would think that this should be quite an active area of hiring. I can attest to the exceptional level of data science expertise these industrial engineers possess as they leave the COSMOS program. I would like to, with this blog, promote them to the analytical chemistry community. Let me know if you are hiring in this area, or if you know someone or some company who is.

A.P. Wicker, K. Tanaka, M. Nishimura, V Chen, T. Ogura, W. Hedgepeth, and K.A. Schug, 

http://www.chromatographyonline.com/lcgc-blog-theoretical-computations-aid-vacuum-ultraviolet-spectroscopic-gas-chromatography-detection

J.X. Mao, P. Walsh, P. Kroll, and K.A. Schug, 

 is a Full Professor and the Shimadzu Distinguished Professor of Analytical Chemistry in the Department of Chemistry & Biochemistry at The University of Texas (UT) at Arlington. He is also a Partner in Medusa Analytical, LLC. Research in the Schug group at UT Arlington spans fundamental and applied areas of separation science, spectroscopy, and mass spectrometry. Schug was named the 

 Emerging Leader in Chromatography in 2009 and the 2012 American Chemical Society Division of Analytical Chemistry Young Investigator in Separation Science. He is a fellow of both the U.T. Arlington and U.T. System-Wide Academies of Distinguished Teachers.

Articles

We are investigating fundamental relationships between the structures of molecules and their interaction with different materials, in the context of on-line supercritical fluid extraction–supercritical fluid chromatography (SFE–SFC).

The LCGC Blog: Working with Data Scientists to Improve On-line Chemical Extraction and Analysis

Acoustic ejection mass spectrometry (AEMS) has recently emerged as the premier ultrahigh-throughput mass spectrometric methodology for drug discovery and related fields. The ultrahigh analytical speed (~1 s/sample) of AEMS has significantly enhanced the efficiency of many high throughput applications. As a result, a data processing and reviewing tool with a matching speed is in high demand for the large amount of data generated, especially for applications such as quality control (QC) of compound collections and high throughput chemistry, where full-scan MS data required convoluted subsequent peak extraction and evaluation. In this study, we demonstrated the feasibility of a tool developed specifically for this purpose. The process using the tool involved automated splitting of the full scan data to correlate well positions with each signal peak, extraction of expected mass traces, and subsequent peak integration. Data evaluation based on verification rules, such as detected mass accuracy, isotopic pattern, and signal-to-noise ratio (S/N), enabled a comprehensive assessment of sample quality that was complemented by visualization in the form of a plate heat map generated from the selected rules. The tool demonstrated fast and straightforward data review and reporting and, more importantly, at a matching speed of sample analysis by acoustic ejection mass spectrometry. The choice of data processing and storage over the cloud further facilitated results sharing among data users.

The recent emergence and utilization of acoustic ejection mass spectrometry (AEMS) in various miniaturized drug discovery campaigns has significantly increased the throughput and turnaround of these applications (1–3). The enhanced efficiency is driven by the ultrafast analytical speed of AEMS of ~1 s/sample, the direct sampling of a dimethyl sulfoxide (DMSO) solution, and the negligible sample consumption of a few nanoliters. As a result, the high-speed, chromatography-free analysis generates an unprecedented amount of data from a range of high throughput assays arrayed in high-density microtiter plates, shifting the bottleneck of the high throughput workflow from conventionally sample analysis to data review. For quantitative AEMS assays, such as bioanalysis of cytochrome P450 or transporter inhibition, where a single probe substrate is detected across plates by multiple reaction monitoring (MRM)–mass spectrometry (MS), the data review is usually straightforward, involving mainly MRM peak integration and subsequent intensity comparison across wells or plates if needed (4). However, for qualitative or semiquantitative analysis using full scan MS detection, the data review is much more involved. Examples of such analysis include reaction monitoring in high throughout chemistry (HTC) (5), compound integrity check in high throughput screening (HTS) (6,7), and quality control (QC) in compound management and storage (8). The full scan data review on these occasions requires generating an extracted ion chromatogram (XIC) from the complex total ion chromatogram (TIC), for each well of the microtiter plate typically containing discrete targeted analytes to help identify the presence or absence of the expected compounds of interest. Subsequently, a verification needs to be established for each sample based on both target identity (qualitative) and amount (quantitative). Manual or even semiautomatic execution of this process is tedious and time-consuming, more so for full scan data acquired on quadrupole MS when the unit resolution alone is inadequate for definitive identity confirmation, compared to full scan data acquired on high resolution MS (5). The requirement for speedy full scan data processing propelled by the ultrafast AEMS has made the automated data processing tool in high demand, especially when better understanding of AEMS mechanism and exploration of novel operation modes continue to push for even faster data acquisition speed (9,10). In addition, the availability of an automated full scan data processing tool can facilitate the utilization of full scan AEMS for nonsubstrate-based or compound-based quantitative assays, such as in vitro metabolic stability, to eliminate the MRM optimization required for these assays (11). Without any commercially available tools in place, significant effort and resources have been spent on the development and maintenance of custom-built analytical tools for full scan data processing (12), or combinations of commercial software supported by home-built scripts (13). These user-built tools, mainly developed for specific applications, lack the comprehensive considerations required for various full scan functions from expansive user groups.

We have previously reported a fully vendor-supported, semiautomated full scan data review platform, using a research version data splitting tool and commercially available software (8). Although the platform has provided the basic functions required for the full scan AEMS data review, the lack of needed speed still made data processing the bottleneck of the entire analytical process. In this platform, the correlation between AEMS signal peaks in the data file with their respective well positions in the sample plate had to be made manually by assigning the retention time for the first ejection, which sometimes could be confusing and unreliable. Because the software was initially designed for processing high resolution MS (HRMS) data, it was not entirely applicable to quadrupole full scan data for which more stringent data survey and verification rules were needed to compensate for the unit resolution by quadrupole MS. In addition, the lack of any plate-based visualization capability in the tool, such as plate heat maps, compelled users to perform plotting and graphing in excel using exported results.

In this study, we evaluated a commercially available, automated full scan AEMS data review tool developed and supported by the AEMS instrument vendor Sciex. The tool specifically incorporated verification rules to support processing of full scan AEMS data acquired on quadrupole MS, in addition to HRMS. The process using the tool included data splitting using an add-on app as an integral part of the system operating software for aligning signal peaks in the data file with the corresponding well positions in the sample plate. It also included cloud-based data processing with unique features such as isotopic pattern comparison of target compounds between the observed and theoretical spectra (as opposed to confirmation of just protonated molecules previously) and background subtraction for eliminating matrix interference. The tool allowed for results visualization in the format of a heat map for a plate-based survey of the sample characteristics. The effectiveness of the tool, including rates for false positive or false negative, data processing speed, user friendliness and versatility for AEMS full scan data, were evaluated.

A Sciex Echo MS system was used for all of the sample analysis and data acquisition. The system comprised of an acoustic droplet ejection module coupled to a triple quadrupole mass spectrometer 6500+. An open-port interface (OPI), designed to capture and transport the ejected samples, was mounted between the acoustic dispenser and an OptiFlow ion source equipped on the MS instrument. The OPI carrier solvent used was 0.1% (v/v) formic acid (J.T. Baker) in methanol (MilliporeSigma), delivered by a gear pump. The system was operated by the Sciex OS operating software with both the MS instrument and acoustic dispenser built into the system hardware configuration. The ion source gas 1 (nebulizing gas) was set at a fixed value of 90 psi, and the carrier solvent flow was fine-tuned (typically around 0.43 mL/min) to achieve the proper formation of a “stable vortex” at the inlet of OPI, commanded by a balanced flow of the carrier solvent and nebulizing gas. The balanced solvent and gas flows allowed for the most efficient transportation of ejected samples through the OPI sample transfer tube to the MS for detection. The ion source gas 2 was set at 50 psi, curtain gas at 20 psi, and the source temperature at 300 °C. The MS was operated under a positive electrospray ionization (ESI) mode at a spray voltage of 5500V, and full-scan data (

 150–800) were acquired at a scan speed of 2000 Da/s. All samples from an entire plate (or user assigned wells from an AEMS run) were acquired into a single MS data file. An ejection volume of 5 nL was employed, with a pause time of 1 s between each sample ejection. The sample plates used were the Echo Qualified 384-Well flat-bottom plate or the LDV COC MS 1536-well plate (Beckman Coulter).

In setting up sample analysis, a “marker well,” typically containing a compound of a known identity, was ejected at the beginning of the sample sequence. Unlike other samples or wells, the marker well underwent a unique ejection pattern containing three ejection events, a single-droplet ejection, a continuous droplet ejection at a low frequency of 10 Hz for 1.5 s, followed by another single-droplet ejection. The recognition of this ejection pattern of the marker well by the data processing algorithm enabled a determination of the delay time from the acoustic ejection event to the MS signal, which was then used to correlate all other MS signal peaks with their appropriate sample well positions. The user may also choose to incorporate a blank matrix well in the plate to generate its mass spectrum that can be used for background subtraction in later data processing. The blank matrix spectrum may also be generated on a separate analysis occasion and saved in a separate data file.

The tool used for full scan AEMS data processing consisted of a Splitter, a Sciex OS add-on app, and OneOmics, a web-based data review suite. The workflow using the tool included four major steps: TIC data file splitting, XIC extraction, sample verification, and plate visualization (Figure 1). The post-acquisition TIC data file contained all of the ejections from a run and was first split into individual subfiles, with each containing a TIC peak from one ejection. The split subfiles were then uploaded to the cloud via the WIFFDB Manager app in the OneOmics Suit, followed by data processing and review by the Echo MS app. For peak extraction, a Microsoft Excel file was first browsed into the app tabulated with the well position, expected target compound formula and charge agent (H

 and others) for up to five targets per well. The expected 

 value was then searched in the full scan data based on the extracting window set by the user, typically ±0.5 amu of the targeted mass calculated from the formula and charge agent provided for quadrupole data, to produce XIC for each ejection. From the XIC generated, the full scan mass spectrum can be visualized for any selected wells of interest. With a user-defined window (for instance, for ±10 amu of the targeted mass), the observed mass spectrum may be overlayed with the theoretical isotopic distribution calculated from the formula of the target compound. A score of correlation was then deduced from the isotopic pattern comparison for each compound well. Combined with the detected mass accuracy, as well as the signal-to-noise ratio (S/N) of the XIC peak for respective wells, the sample quality was evaluated for verification purposes. The results can then be visualized in the format of a plate heat map, and color-coded for each well based on their verification status before being exported in the form of an Excel file.

Figure 1: Workflow using a full-scan data processing tool.

AEMS analysis achieved an ultrafast analytical speed of seconds or subseconds per sample, using acoustic technology for direct liquid sampling coupled with MS detection. Unlike the conventional LC–MS analysis where each injection is typically acquired to a single data file, AEMS analysis has a single MS file remained open for the acquisition of all ejections from an applicable microtiter plate, either 384-well or 1536-well. This way, the overhead associated with opening and closing the individual acquisition files for each ejection is eliminated to achieve an overall maximized analytical speed. Therefore, the resulting data file may contain hundreds or more signal peaks, which presents challenges in the downstream data processing in terms of appropriately correlating the microtiter plate well positions with their corresponding signal peaks in TIC. This matching or correlating process could be perplexing if the correlation is solely based on finding the expected mass in the respective wells, especially when the wells under survey may be absent of the expected mass or present with the same expected mass. In AEMS, the scanning of the marker well was designed to eliminate that confusion via the unique signal pattern of a continuous infusion bracketed by two single-droplet ejections. The confirmation of the marker well at both the beginning and end of the sample sequence provided unequivocal association of its retention time in the TIC to the marker well position, which enabled a definitive correlation of all other well positions with their respective signal peaks based on the delay time set across each ejection during acquisition (Figure 2a). The provision of the marker well retention time in the Splitter then ensured an appropriate assignment of post-split subfiles to their respective well positions in the sample plate (Figure 2b).

Figure 2: Pre-splitting TIC and post-splitting subfiles: (a) TIC with marker well scanned at the beginning and end of the sample sequence and (b) post-splitting subfiles.

From the post-split subfiles, XIC was generated based on the imported compound information and user-defined search criteria (Figure 3a). Simultaneous search and extraction of multiple targets in each well offered the much-needed versatility for applications like HTC to help identify not only the expected products, but also the remaining starting materials and potential side products (Figure 3b). The choice of different charge agents, such as proton adduct and sodium adduct for the target compounds, enabled a comprehensive survey of the targets in the full scan data and minimized potential false negatives in cases where unexpected adducts were formed. Although generating XICs multiple times with one charge agent in each could be implemented as previously reported (8), it would significantly slow down the data extraction process.

Figure 3: XIC and corresponding mass spectrum (a) XIC and (b) full scan mass spectrum of the corresponding XIC.

In the data review tool, based on the XIC, the observed mass spectrum was then displayed and overlaid with the theoretical isotopic distribution of the target compound, enabling a direct visualization of the correlation between the two (Figure 4). A correlation score, or “average ratio offset,” was also calculated for a more quantitative comparison. For each target compound, the observed intensity ratios of the found isotopic mass 

 were compared respectively with the corresponding theoretical ratios calculated from the formula, and the average ratio difference from the comparison was defined as the average ratio offset. The lower the offset was, the better the overlay was between the observed and theoretical isotopic distribution; hence, the higher the confidence was in the match to the expected target compound. The average ratio offset offered additional assurance in identity confirmation compared to the simple matching of the protonated molecule, which was especially helpful for quadrupole full scan data where unit resolution was inadequate in differentiating isobaric interferences, which could shift the observed peak mass centroid and affect the apparent isotope ratios. In this case, a poor average ratio offset from the targeted mass would raise a red flag for potentially false positive findings. In addition to its utility for small molecule identification in HTC or compound QC, the isotopic distribution overlay was also an effective approach for identifying trace metal chelated organic complexes, which typically displayed characteristic isotopic fingerprints unique to the chelating metal involved (14). We have demonstrated the utility of this approach in the investigation evaluating chelator effectiveness and chelating ratios between thiourea (TU, C

OS) and palladium (Figure 5). The metal-chelated organic complexes have been demonstrated to produce false positives in HTS campaigns (15), whereas an ultrafast AEMS scanning of the HTS plate would effectively detect the complexes to avoid misleading HTS results.

Figure 4: Overlay of observed target mass with theoretical isotopic distribution and background subtraction; blue trace: observed mass spectrum, red trace: theoretical isotopic distribution of target compound, and grey trace: background mass spectrum.

Figure 5: Metal-chelated organic complex, with a 2:1 chelating complex of TU (chelator)/Pd and an overlay of observed isotopic distribution with theoretical; blue trace is the observed mass spectrum, and the fuscia trace is the theoretical isotopic distribution.

Direct MS analysis of small molecules in biological assays or chemical reactions by AEMS can be subject to interferences by the matrix that is otherwise removed by chromatography in LC–MS analysis. The presence of interfering background ions in the matrix, especially when their isotopic distribution fully or partially overlapped with that of the target compound, may result in a false positive target identification or an overestimation of target quantity. Subtracting common background ions from those of the analytes was an effective approach to minimize the interference, which was especially helpful for full scan data acquired on quadrupole MS because it lacked the resolving power offered by HRMS (Figure 4).

In the data processing and reviewing tool, a results table was in place upon generation of XICs and tabulated with parameters for qualitative target evaluation, including expected formula, detected mass, absolute mass error compared to the theoretical, average ratio offset. Other parameters for quantitative target evaluation, such as XIC intensity and S/N ratio, were also listed in the table. Among them, three parameters were used for sample verification in both qualitative and quantitative terms based on thresholds set by the user, including absolute mass error, average ratio offset, and S/N ratio. The targeted compound was only verified when criteria for all three were met, with a green-coded verification status listed in the results table; otherwise, a yellow-coded status was displayed to indicate that further investigation was needed. The sample quality of the entire plate may be visualized in the form of a plate heat map, with wells color-coded based on the detected XIC intensity (Figure 6). At user discretion, samples failing verification may be excluded from the plate heat map, as illustrated by an “x” in the heat map. When any of the verification rules were modified, the results table and plate heat map were automatically updated, making this interactive feature especially convenient while greatly enhancing the speed of the data review process.

Figure 6: Results visualization in format of plate heat map.

The tool we evaluated has demonstrated feasibility for AEMS full scan data processing and review, with well-designed features for reliable data extraction, compound identification, sample verification, and results visualization. The unique scan feature of the marker well in the sample plate ensured proper extraction of appropriate mass from the corresponding well positions. The identification and verification features were especially beneficial for full scan data acquired on unit-resolution quadrupole MS, which could more likely generate false positives or false negatives compared to data obtained on HRMS. Enabled by the user selectable verification rules, the comprehensive assessment of the sample quality in both qualitative and quantitative terms was complemented by results visualization in the form of a plate heat map and an exportable results table. The tool demonstrated fast and straightforward data processing and review. More importantly, this process was done at a matching speed to that of sample analysis by AEMS, effectively eliminating the bottleneck of the analytical process. The choice of data processing and storage over the cloud further facilitated results sharing among data users.

We would like to thank Dr. Anthony Cauley and Dr. Yong Zhang of Bristol-Myers Squibb, Small Molecule Drug Discovery, for generating the HTC samples for this study.

(1) H. Zhang, C. Liu, W.Y. Hua, L.P. Ghislain, J.H. Liu, L. Aschenbrenner, S. Noell, et al, 

https://doi.org/10.1021/acs.analchem.1c01137

(4) A. Wagner, J. Zhang, C. Liu, T.R. Covey, T.V. Olah, H.N. Weller, and W.Z. Shou, 

https://doi.org/10.1021/acs.analchem.0c03006

(5) K.J. DiRico, W.Y. Hua, C. Liu, J.W. Tucker, A.S. Ratnayake, M.E. Flanagan, et al, 

https://doi.org/10.1021/acsmedchemlett.0c00066

(6) R.P. Simon, T.T. Habe, R. Ries, M. Winter, Y.T. Wang, A. Fernandez-Montalvan, D. Bischoff, et al, 

https://doi.org/10.1177/24725552211028135

(7) X.J. Wen, C. Liu, L. Ghislain, K. Tovar, V. Shah, S.J. Stout, S. Cifelli, et al, 

https://doi.org/10.1021/acs.analchem.0c04312

(8) J. Zhang, Y. Zhang, C. Liu, T. Covey, J. Nielsen, S. Li, H. Weller, and W. Shou, 

(9) C. Liu, G.J. Van Berkel, D.M. Cox, and T.R. Covey, 

https://doi.org/10.1021/acs.analchem.0c02999

(10) T.T. Habe, C. Liu, T.R. Covey, R.P. Simon, W. Reindl, F.H. Buttner, et al, 

https://doi.org/10.1021/acs.analchem.0c01632

(11) T. Hollenbeck, S. Thibodeaux, P. White, L. Westling, A. Chong, and J. Isbell, 

https://doi.org/10.1016/j.xphs.2020.07.025

(12) J.W. Sawicki, A.R. Bogdan, P.A. Searle, N. Talaty, and S.W. Djuric, 

(13) S.S. Lin, S. Dikler, W.D. Blincoe, R.D. Ferguson, R.P. Sheridan, Z.W. Peng, D.V. Conway, et al, 

(15) J.C. Hermann, Y.S. Chen, C. Wartchow, J. Menke, L. Gao, S.K. Gleason, et al, 

 are with Bristol Myers Squibb in their Research and Early Development department, in Princeton, New Jersey. 

 are with Sciex, in Concord, Ontario. Direct correspondence to: 

Acoustic ejection mass spectrometry (AEMS) has recently emerged as the premier ultrahigh-throughput mass spectrometric methodology for drug discovery and related fields.

A Full Scan Data Review Tool to Match the Speed of Acoustic Ejection Mass Spectrometry

The Column spoke to Philipp Weller from the Institute of Instrumental Analytics and Bioanalytics at the Mannheim University of Applied Sciences, Germany, about the benefits of nontargeted screening approaches in food safety and environmental analysis using gas chromatography–ion mobility spectroscopy (GC–IMS), and the best way to approach data processing and model building with the generated data.

 spoke to Philipp Weller from the Institute of Instrumental Analytics and Bioanalytics at the Mannheim University of Applied Sciences, Germany, about the benefits of nontargeted screening approaches in food safety and environmental analysis using gas chromatography–ion mobility spectroscopy (GC–IMS), and the best way to approach data processing and model building with the generated data.

Q. Nontargeted screening approaches are gaining in popularity, particularly for the identification of contaminants in the environment or regarding food safety. What advantages does gas chromatography–ion mobility spectroscopy (GC–IMS) offer over alternative analytical techniques?

 First of all, GC–IMS provides equally high, or even superior, sensitivity compared to GC–mass spectrometry (MS) for polar or medium polar compounds, which are commonly found in volatile organic compound (VOC) fingerprints of fermentations, food stuffs, or environmental samples. In our experiments using GC–IMS, we can often skip sample enrichment procedures, such as solid-phase microextraction (SPME), which is typically not possible for GC–MS—at least not for electron ionization (EI) mode. Since we have GC–IMS–MS prototypes running for most of our experiments, we can directly compare sensitivities between the two. What we found is that for a number of smaller, highly polar molecules, such as carboxylic acids, esters, aldehydes, and ketones, there is around a 10‑fold improvement in sensitivity. This is mainly due to the low-energy ionization, which happens as a secondary effect of the primary 3H ionization of the buffer gas (typically N2) in the form of protonated water clusters. This is, however, also one limitation of the principle because molecules with low proton affinities, such as alkanes, will feature much lower sensitivities than, for example, aldehydes or ketones. As such, it is decisive about the compounds that are relevant when using GC–IMS in contaminant analysis.

The second aspect is the simplicity of both the hardware and the low demand for infrastructure. The IMS cells are extraordinarily robust and simply built, with no complex moving parts, such as turbo pumps, which makes the technology a perfect match for industrial applications or a point-of-care use.

The third aspect is size and weight. IMS cells may be built in a very compact manner, including all required electronics, without having to sacrifice robustness or sensitivity.

Q. One challenge of IMS analysis is interference from widespread ionization, which results in low selectivity. Your paper mentions that the addition of a suitable dopant can overcome these limitations (1). How easy is it to mitigate this issue and do you have any advice when choosing a dopant?

 In our experience, most of the interference that occurs during
GC–IMS analysis is a result of suboptimal chromatography rather than from IMS ionization; if the user optimizes the GC pre-separation as perfectly as possible, most cases of interferences may at least be minimized. We had the best results in IMS detection when we made sure that the IMS cell did not have to “digest” too much substance signal. However, it is correct that the ionization process is not selective when similar polar substances enter the ion source. This is an issue that may happen in GC–MS as well, yet the problem in IMS is the formation of so-called 

, which are basically adducts on from two different molecules. This not only complicates signal identification because these hetero dimers will feature different drift times to the pure substance ions but also often leads to nonlinear effects in detection.

The use of dopants is still challenging in routine handling, as the constant dosage is a key aspect. We used a specifically produced calibration gas with 100 ppm nitric oxide (NO) in nitrogen (N2 ) to safely dose the NO into the drift cell, which can be done using a separate electronic pneumatic controller (EPC) valve. This may demonstrate that our prototype setup is still a couple of steps away from routine.

While we used the dopant to demonstrate the use in the separation of charged NO adducts of two terpenes—and this is a remarkable way of using specific ion chemistry as a separation principle—my assumption is that the use of a dopant might be even more useful when applied to nontargeted screening (NTS) workflows. The dopant adds a large number of additional, substance-group-specific signals to the already complex fingerprint, which then can be analyzed by machine learning-based workflows. Since different substances will react differently to the dopant, this may add more sample-specific variance to the data, which is what we are heading for.

So, in short, dopant addition may not be the best choice to overcome an inferior chromatography, but rather a good choice for complex matrices to add selectivity by specific signals to the NTS workflow.

We have not tested any other dopants other than NO yet, but I would assume that ammonia would also work—albeit in a different manner; the choice is limited to gases with a sufficiently high proton affinity, as other than the buffer gas nitrogen, the dopant itself is part of the ion itself. An important consideration obviously would be toxicology, as the IMS cells use relatively high flow rates.

Q. Food fraud is a major concern in modern food markets, with the geographical origin of products coming under intense scrutiny. Your paper (1) contains a very useful table with studies utilizing GC–IMS to combat food fraud and follows that up with a NTS workflow. What advice can you give researchers looking to utilize GC–IMS for nontargeted screening for the first time?

The good thing about GC–IMS is that handling wise, it is very similar to GC–MS. So, if you are already using GC–MS for NTS workflows, you can stick closely to what you have done so far—with one exemption. Usually, there is no need for enrichment steps, such as SPME or comparable, simply because GC–IMS is usually sensitive enough for polar to medium polar compounds to be measured directly from the headspace. An important key to the successful use of GC–IMS is to not oversaturate your cell with high analyte concentrations. This is the Achilles’ heel of drift tube ion mobility spectrometry (DTIMS) systems, as upon doing so you’ll leave the path of linear detector response, and that makes things more complicated because you´ll have to deal with nonlinear data analysis.

Q. Your paper also addresses the chemometric approaches that can be paired with GC–IMS. However, the sheer number of techniques can be quite intimidating. How did you initially approach data processing and model building when carrying out this research? And what advice can you offer to those seeking to utilize the same techniques?

 To be honest, we made every error that a data scientist could make on spectral data, which was obviously frustrating in the beginning. Yet we learned a lot about our data, its potential, and also about its limits. There are a plethora of methods, this is true, but there are a couple of rules of thumb that I would give a beginner. First of all, keep your approach as simple as possible and use principal component analysis (PCA) to get an overview. PCA is often underestimated, but helps so much to get a quick glance of your data and to get an initial idea of what is in there—or what is not. A second rule is a very general one in data science that says “garbage in = garbage out”—if your data is not good because, for instance, your analysis did not work as it should have, don´t use it—it is usually a waste of time. A massively important rule is validation and a critical view on your data analysis: if you are doing supervised learning, such as partial least squares regression (PLS-R) or partial least squares-discriminant analysis (PLS-DA), which would also be my recommendation for a start, do not chose the model that calibrates the best, but the one that predicts the best—so know your figures of merit.

Q. Supervised and unsupervised data analysis and machine learning techniques—are there benefits and downsides to each or can they be used in tandem to ensure accurate conclusions from GC–IMS generated data?

The concepts of both approaches are very different, so in short, yes, I would always use both unsupervised exploratory and supervised techniques together. As already said, exploratory analyses tell you so much about your data and also about your analyses, such as systematic errors, retention time or drift time shifts, or getting first ideas about potentially relevant features. Furthermore, PCA can help you judge whether the model will be able to cover the data space in a sufficient manner and also whether the data contains potential outliers by using the T2/Q-plot. In my opinion, the latter is one of the most important figures you should generate.

Supervised learning does a brilliant job when you already have an idea on your sample data and your focus is to either classify or to quantitate latent information. However, it does require much more care when building and validating a model. Furthermore, one has to keep in mind that often class-based models, as it is often done in food authentication, might not be the best choice. Classes should be made up of samples of the same distribution in order for the classification algorithm to properly do its job. But what happens if the suspicious sample that you are analyzing is not part of the distribution you calibrated your model with? Usually the results don´t make sense or the validation fails. In these cases, class‑free algorithms, such as SIMCA, may be more reasonable.

Q. In another recent paper (2), GC–IMS was used for the volatilomic profiling of citrus juices in an authentication study. What challenges arose in the application of the technique for such a study? And what were the conclusions of the paper?

 We encountered more issues in working with the sample data than we anticipated beforehand (and also those anticipated in the very good discussions we had with our reviewers). One of the most profound ones was that it is absolutely vital to get hold of truly authentic material, which is really a challenge in food authentication in general.

A second issue was a conceptual one. When we began the study, the initial idea was to pair GC–IMS and GC–MS in one machine run to obtain complementary data that could be used for data fusion, which is the term for “merging” the data on a data level to obtain more information than from the individual techniques. IMS and EI-MS seemed like a good pairing, but it turned out that much of the data was not at all complementary, but rather relatively similar in its output. We therefore did not end up with that perfect moment that proved data fusion to be the ideal tool for every question, but it rather showed that it did make a good job for some of the analytical questions—and for some it did not. One of the conclusions of the paper was that data fusion of analytical data is a brilliant strategy, but it is no “fire and forget” technology and still requires optimization for routine use.

Q. Can GC–IMS also be applied for targeted screening?

 Yes, absolutely. GC–IMS works very well for polar and medium polar compounds—and with notably less sensitivity for apolar compounds in a target analysis type of strategy. An example could be short organic acids in fermentations. A user could even go both ways: a targeted approach for specific marker compounds such as linalool oxides in grapefruit juices, while at the same time using an NTS approach for finding unusual patterns. Sensitivity does not change in IMS whether you focus on individual compounds or on the whole pattern. Yet, in some cases, it might make sense to stay with GC–MS instead. An example of this would be if you´re in need of mass‑to‑charge ratio information because of regulations.

We have several projects running: one large one deals with deep learning in authenticity and quality analysis of citrus oils, juices, and coffee; a second one focuses on data augmentation of GC–IMS data by deep learning, to compensate for too low sample numbers; and the third one is machine/deep learning-based bioprocess analysis, out of which we are going to release a Python‑based open-source toolbox for GC–IMS data. This is hopefully going to help other GC–IMS users, who might be less experienced with programming than we are.

R. Brendel, S. Schwolow, S. Rohn, and P. Weller, 

Philipp Weller is a full professor at the Institute of Instrumental Analytics and Bioanalytics at the Mannheim University of Applied Sciences. Furthermore, he also heads the competency center for chemometrics and material analysis CHARISMA (www.charisma.hs-mannheim.de). He holds a Ph.D. in food chemistry (2004) from the University of Hohenheim. His research is mainly focused on data‑driven omics approaches based on the multimodal analysis of complex food matrices in the field of quality and authenticity analysis. His research group develops their own specialized machine learning/deep learning toolboxes for high‑dimensional data and data fusion.

The Column spoke to Philipp Weller from the Institute of Instrumental Analytics and Bioanalytics at the Mannheim University of Applied Sciences, Germany, about the benefits of nontargeted screening approaches in food safety and environmental analysis using GC–IMS, and the best way to approach data processing and model building with the generated data.

New Adventures in GC–IMS for Nontargeted Screening of Food and Environmental Samples

Chemical synthesis of nucleic acids can result in a number of by-products, such as shorter length components and protecting groups. Therefore, adequate separation and purification of the target compounds is a crucial step in the production process. This article introduces an analytical method for the separation of oligonucleotides of different chain length by ion‑exchange chromatography (IEC) using a bio-inert ultrahigh-pressure liquid chromatography (UHPLC) system. The effect of changes in mobile phase pH due to CO2 absorption from ambient air during storage was also investigated.

Risa Suzuki received her Ph.D. at the University of Tokyo, Japan, in 2015, focusing on protein engineering and construction of a universal screening system for elucidating enzyme functions involved in drug metabolism. From 2015 to 2017 she worked as a postdoctoral research associate at Nagoya University and Tokyo Institute of Technology, working on a JST funded ALCA project on microbe reaction engineering. In 2017, she was employed as a project assistant professor in the School of Medicine at the University of Tokyo. Her research interests and expertise encompass the developing and manufacturing of biopharmaceuticals and crystal structure analysis of proteins. She now works for Shimadzu Corporation in Kyoto, Japan, as a member of the Healthcare Solution Unit in the Solutions COE.


Chihiro Hosoi graduated with a master’s of engineering from the Tokyo University of Agriculture and Technology, Japan, in 2019, focusing on interaction between nucleic acids and proteins and developing a biosensing system. In 2017, she studied graphene-based biosensors at the University of California, Riverside, USA. She now works for Shimadzu Corporation in Kanagawa, Japan, as a member of the Healthcare Solution Unit in the Solutions COE.

Chemical synthesis of nucleic acids can result in a number of by-products, such as shorter length components and protecting groups. Therefore, adequate separation and purification of the target compounds is a crucial step in the production process. This article introduces an analytical method for the separation of oligonucleotides of different chain length by ion‑exchange chromatography (IEC) using a bio-inert ultrahigh-pressure liquid chromatography (UHPLC) system. The effect of changes in mobile phase pH due to CO

 absorption from ambient air during storage was also investigated.

Antisense oligonucleotides are a variety of nucleic acid drugs, synthetic molecules with a chemical backbone structure composed of deoxyribonucleic acid (DNA) or ribonucleic acid (RNA) bases. Therapeutic oligonucleotides have become more prominent over recent years because they can be designed to accurately target a specific disease gene or its encoded messenger RNAs, providing the next generation of therapeutics. While having the targeted specificity of antibody therapeutics, they can also be synthesized chemically like “low-molecular‑weight” drugs, which reduces manufacturing costs compared to the biological process of protein-based biomedicines. Therapeutic oligonucleotides combine some of the best properties of “low‑molecular-weight” therapies and biopharmaceuticals.

Chemical synthesis of nucleic acid results in a number of by-products, such as shorter length components and protecting groups (1). Therefore, adequate separation and purification of the target compound is a crucial step in the production process.

This article presents an analytical method for the separation of oligonucleotides of different chain length by ion-exchange chromatography (IEC), assuming that shorter length components are impurities derived from the synthesis process. To avoid any adsorption of the metal-coordinating compounds to the wetted surface of the instrument, a bio-inert ultrahigh-pressure liquid chromatography (UHPLC) system was used. The effect of mobile phase pH on the separation was also investigated.

The sequence of the oligonucleotide to be analyzed is shown in Table 1. Target oligonucleotides in 20 mer and four sequences that were deleted from n-1 to n-4 on the 3’ terminus of the target were prepared as impurities derived from the synthesis. All of them were unmodified single-stranded DNA and were synthesized by a solid-phase synthesis (high performance liquid chromatography [HPLC] purification). These five sequences were diluted to 5 μmol/L with water, and an oligonucleotide mixture was prepared.

In ion-exchange chromatography, analytes are separated and eluted using a mobile phase of varying salt concentration or pH. In this analysis, an aqueous solution of sodium hydroxide (A) was used in a gradient with a solution of sodium hydroxide containing sodium perchlorate (B), to gradually increase elution strength for the analytes of interest. A detailed description of analytical conditions is listed below.

 System: Nexera XS inert (Shimadzu); column: 100 mm × 4.6 mm, 5-μm Shim-pack Bio IEX Q-NP (Shimadzu); mobile phase A: 10 mmol/L NaOH, mobile phase B: 10 mmol/L NaOH containing 1 mol/L NaClO

; time program: 25%B–32.5%B (0–15 min) " 100%B (15–20 min) " 25%B (20–25 min); flow rate: 0.8 mL/min; oven temperature: 30 °C; injection volume: 4 µL; detection: UV 260 nm.

Using these IEC conditions, separation occurred based on the difference in negative charge of the analytes, which was equivalent to the number of phosphate groups in the oligonucleotide. Consequently, shorter oligonucleotides eluted first. Figure 1 shows a chromatogram of a five-sequence oligonucleotide mixture; all compounds of interest were separated by their length.

Repeatability of the method was determined based on retention time and peak area from six successive analyses of the 16–20 mer oligonucleotide mixture shown in Table 1. With variation of ≤ 0.5% in retention time and peak area for all analytes, good repeatability of the assaywas established. Detailed results can be found in Table 2.

Sample Analysis: Oligonucleotides in the Presence of By-Products: 

A mixture of five oligonucleotides, at 5 μmol/L concentration each, was prepared (four of them were HPLC-purified while one was only desalted) and IEC results were compared to those of the mixture of all HPLC-purified nucleotides (Figure 2). The target oligonucleotides were completely separated from impurities such as free protecting groups and shorter length oligonucleotides.

Effect of Change in Mobile Phase pH on Retention:

 As described above, in the IEC separation of an oligonucleotide mix, the analytes were separated and eluted using a salt concentration gradient. In an aqueous mobile phase (pH ≥ 7), absorption of carbon dioxide from ambient air can change the effective concentration of a basic eluent due to the formation of carbonic acid (2). In IEC, where separation is based on the analytes’ differences in charge, even a slight change in the pH can have a significant impact on retention and hence the analytical results. It is therefore crucial to prevent possible changes in mobile phases pH, to ensure reliable, stable analytical performance.

In this study, results obtained using mobile phase stored in a bottle closed with a standard screw cap were compared to those achieved with a mobile phase from a bottle with a filter cap that suppresses evaporation of the solvent. Analyses were performed using freshly prepared mobile phase and after two days of storage, both using the conditions described earlier. A comparison of chromatograms obtained are shown in Figure 3. Figure 4 depicts the percentage of change in mobile phase pH during the two days of storage under the respective conditions.

As can be seen in Figure 3, when using a standard mobile phase bottle cap 3), the retention time increased by an average of 8% as the pH changed compared to that on the day of preparation of the mobile phase 1); when using a cap with a filter 2), the retention time did not change significantly. The use of a cap with a filter suppressed the change in the pH of the mobile phase, resulting in stable results, even after two days of mobile phase preparation. Also, the change in pH was measurable in the bottles with a standard cap, with a larger impact seen in mobile phase B containing sodium perchlorate, as shown in Figure 4.

A simple gradient method for the analysis of oligonucleotides by ion-exchange chromatography using a bioinert UHPLC system has been presented. Good separation of the oligonucleotides under investigation was achieved, both from one another and from impurities such as protecting groups and shorter length oligonucleotides, which are common by‑products from chemical synthesis. It was shown that even small changes in eluent pH did affect the analytes’ retention time. Therefore, it is important to carefully control the mobile phase pH to obtain reproducible, reliable results.

C. Hosoi and R. Suzuki, Shimadzu Corporation Application News No. 01-00227-EN (Shimadzu Corporation, February 2022).

Ion Chromatography, Third, completely revised and enlarged edition

graduated with a diploma in chemical engineering from the Technical University NTA in Isny, Germany, in 2004 and as a master of science in pharmaceutical analysis from the University of Strathclyde in Glasgow, UK, in 2005. She worked until 2006 as a consultant in HPLC method development and validation in an analytical laboratory of the FAO/IAEA in Vienna, Austria. She gained her doctorate for research in pharmaceutical sciences at the University of Strathclyde in 2010 and was employed as an HPLC specialist in the R&D department at Hichrom Ltd in Reading, UK, from 2009. Since 2013, she has worked as a HPLC product specialist and since 2015 as HPLC Product Manager in the analytical business unit of Shimadzu Europa in Duisburg, Germany

received her Ph.D. at the University of Tokyo, Japan, in 2015, focusing on protein engineering and construction of a universal screening system for elucidating enzyme functions involved in drug metabolism. From 2015 to 2017 she worked as a postdoctoral research associate at Nagoya University and Tokyo Institute of Technology, working on a JST funded ALCA project on microbe reaction engineering. In 2017, she was employed as a project assistant professor in the School of Medicine at the University of Tokyo. Her research interests and expertise encompass the developing and manufacturing of biopharmaceuticals and crystal structure analysis of proteins. She now works for Shimadzu Corporation in Kyoto, Japan, as a member of the Healthcare Solution Unit in the Solutions COE.

graduated with a master’s of engineering from the Tokyo University of Agriculture and Technology, Japan, in 2019, focusing on interaction between nucleic acids and proteins and developing a biosensing system. In 2017, she studied graphene-based biosensors at the University of California, Riverside, USA. She now works for Shimadzu Corporation in Kanagawa, Japan, as a member of the Healthcare Solution Unit in the Solutions COE.

An analytical method for the separation of oligonucleotides of different chain length by IEC using a bio-inert UHPLC system is presented.

IEC Analysis of Oligonucleotides and Evaluation of the Effect of Changes in Mobile Phase pH on Separation

This month we interview Alexandre Goyon, Senior Scientist in the Small Molecule Pharmaceutical Sciences Organization of Genentech, about his work focused on the online digestion and analysis of RNA molecules using immobilized RNase cartridges attached to a mDLC–MS system, and why the accurate sequencing of sgRNA is important.

This month we interview Alexandre Goyon, Senior Scientist in the
Small Molecule Pharmaceutical Sciences Organization of Genentech, about his work focused on the online digestion and analysis of RNA molecules using immobilized RNase cartridges attached to a multidimensional liquid chromatography–mass spectrometry (mD‑LC–MS) system, and why the accurate sequencing of single guide RNA (sgRNA) is important.

Q. When did you first encounter chromatography and what attracted you to the subject?

My first experience with chromatography was during an internship in the laboratory of Professor Jessop (Kingston, Canada). The Jessop group is specialized in green chemistry. I analyzed samples using gas chromatography (GC) and liquid chromatography (LC) to conclude my experiments, and I knew that analytical chemistry was an area I would like to explore more. After completing my master’s degree, I joined the Nestle Research Center (Lausanne, Switzerland) where I developed LC–tandem mass spectrometry (MS/MS) methods to quantify food contaminants in various matrices.

Q. Can you tell us more about your Ph.D. thesis?

I performed my Ph.D. thesis at the University of Geneva under the supervision of Professor Jean-Luc Veuthey, Dr Davy Guillarme, and Dr Szabolcs Fekete. It was the first Ph.D. thesis in the laboratory dedicated to the analysis of therapeutic proteins. I have developed non-denaturing size-exclusion chromatography (SEC), cation-exchange chromatography (CEC),capillary zone electrophoresis (CZE), and hydrophobic interaction (HIC) methods to characterize monoclonal antibodies (mAbs) and antibody-drug conjugates (ADCs). I had unique opportunities to perform internships in pharmaceutical companies—Novartis (Basel, Switzerland) and Genentech (South San Francisco, USA)—where I developed innovative multidimensional (mD)-LC–MS methods to streamline existing analytical workflows and facilitate the characterization of complex drug modalities.

Q. What chromatographic techniques have you worked with?

I have worked with reversed-phase LC, hydrophilic interaction liquid chromatography (HILIC), SEC, CEC, anion exchange chromatography (AEC), HIC, and mixed mode chromatography (MMC) for the analysis of diverse molecules, including small pharmaceutical molecules, therapeutic antibodies, protein-polymer conjugates, oligonucleotides, and nucleic acids.

Q. You were involved in research that fully sequenced CRISPR/Cas9 single guide RNA (sgRNA) using parallel ribonuclease digestions and HILIC−high-resolution MS analysis (1). Why is accurate sequencing of sgRNA important?

With the CRISPR-Cas technique, a single guide RNA (sgRNA) directs the ribonucleoprotein (RNP) complex to a specific target. The CRISPR-Cas technique has attracted significant interest in the field of cell therapy. For example, T cells can be reengineered using this approach. It is crucial to verify the sequence of the sgRNA is correct to limit off-target effects and ensure the safety of the patients. Regulatory agencies expect investigational new drug (IND) applicants to sequence sgRNAs.

Q. What is novel about the method you developed? What were the main challenges from an analytical perspective and how did you overcome them?

A sequencing method should enable the verification of the correct nucleoside order. Ideally, the sequencing method should also confirm the presence and correct position of the chemical modifications, which is not feasible using next generation sequencing (NGS) approaches due to the loss of information during the library preparation.

To my knowledge, this parallel RNase digestion and HILIC–MS/MS approach was the first to confirm the correct order of the nucleosides and the presence of the chemical modifications. Ideally, the sequencing approach should provide 100% sequence coverage. However, the RNase digestion of 100-mer sgRNAs produces non-unique fragments, which reduces the sequence coverage. The challenge was overcome by using multiple RNases in parallel. Our approach required the development of robust i) digestion procedures, ii) chromatographic separation, and iii) MS detection. It was the first time that an RNase digestion mixture was analyzed by HILIC, which provided both truly MS-compatible conditions in the absence of ion-pairing agents and selectivity to discriminate sgRNA sequences. We found that RNase could contaminate HILIC columns and interfere with the analysis of RNase digestion mixtures.

Q. Is HILIC becoming more commonly used in biopharmaceutical research? If so, why, and in what areas?

In my experience, HILIC is becoming more and more utilized and in particular for the analysis of oligonucleotides. Advances in this field are fast, and I am confident future research works will unleash the potential of HILIC for the characterization of new drug modalities. The pharmaceutical landscape has become incredibly diverse, with the development of complex drug modalities beyond the Lipinski’s rule of five. In parallel, the coupling of MS instrument to LC separation has become essential to identify impurities. In many cases, HILIC is an ideal mode, preventing the use of ion-pairing agents needed for the reversed-phase LC analysis of polar molecules.

Q. You also recently published a paper presenting, for the first time, the online digestion and analysis of an RNA molecule using immobilized RNase cartridges attached to a two‑dimensional (2D)-LC–MS system and three-dimensional (3D)-LC–MS system (2). How did this project arise?

I am convinced that the automated digestion and analysis of large oligonucleotides offer unique benefits for their characterization. Bottom-up MS approaches enable i) the sequencing of large oligonucleotides, and ii) the identification/quantification of impurities at single residue level.

Conventional offline, in-solution digestions require a large amount of sample and intensive laboratory work, which may cause errors. Conversely, the automated digestion and analysis of large oligonucleotides via the use of immobilized RNase cartridge and HILIC analysis streamline the process. The advantage of an online procedure was even greater since three RNases were used for the parallel offline, in-solution digestions of sgRNAs.

A 3D-LC–MS approach enabling the online fractionation and bottom-up analysis of fractions was also developed to understand the cause of a peak tailing observed on the ion-pairing (IP) reversed-phase LC profiles of sgRNAs.

Q. What were your main findings and how does your approach benefit the analyst in practice?

The same sequence coverage was found using the online approach in comparison to the offline one. A 10-fold reduction in digestion time and 20-fold reduction in sample amount was achieved using the online 2D-LC–MS approach in comparison to the offline approach. Due to the limited amount of sgRNA available, the characterization of impurity peaks observed at the single nucleotide level could only be performed when using a 3D-LC method, enabling the automated fraction collection and digestion.

In practice, the analyst can save time and focus on other activities, such as data treatment, when using the online approach, since the digestion and analysis are automated.

Q. Is 2D-LC–MS becoming used more commonly for routine analysis in the pharma/biopharma sector? If so, in what application areas? Is 3D-LC likely to be used more widely?

In the field of pharmaceutical small molecules, 2D-LC–MS is used for the characterization of impurity peaks and/or screening of method conditions, such as mobile phases and LC columns. In the field of large molecules, 2D-LC–MS is also used for the desalting of fractions collected from one-dimensional (1D) analysis for the identification of impurities.

In addition, mD-LC–MS systems enable bottom-up MS characterization and fraction collection to be integrated. We have developed several of these methods to characterize fractions collected from reference purity methods—in the past for therapeutic antibodies and now for large oligonucleotides.

Q. Are chromatographers overcautious about implementing multidimensional methods?

Multidimensional methods may often sound complicated and only ever applied to specific projects. As a general rule, the simpler, the better. However, with the increasing complexity of drug modalities and delivery systems, some challenges cannot be solved using 1D-LC methods.

There is indeed an important challenge when going beyond two dimensions, which is mainly due to limited software solutions providing a user-friendly environment. The broader adoption of mD-LC–MS methods will depend on close collaboration between instrument vendors, pharmaceutical companies, and academic laboratories.

Q. What projects are you working on next?

I am hoping to see the implementation of mD-LC–MS methods to help understand manufacturing and purification processes. In particular, an ideal method would isolate pharmaceutical molecules from a biological matrix prior to their comprehensive impurity characterization. Such a streamlined method would provide multiple quality attributes, including titer and purity, and help guide the process.

A. Goyon, B. Scott, K. Kurita, C.M. Crittenden, D. Shaw, A. Lin, P. Yehl, and K. Zhang,

A. Goyon, B. Scott, K. Kurita, C. Maschinot, K. Meyer, P. Yehl, and K. Zhang, 

is Senior Scientist in the Small Molecule Pharmaceutical Sciences Organization of Genentech. He received a Ph.D. degree in pharmaceutical analytical chemistry in 2019 from the University of Geneva, Switzerland. His team supports early- andlate-stage research programs. They develop innovative multidimensional LC–MS methods and complementary approaches to support new drug modalities and drug delivery systems. He has published 36 papers and book chapters, including 23 as first author.

This month we interview Alexandre Goyon, Senior Scientist in the Small Molecule Pharmaceutical Sciences Organization of Genentech, about his work focused on the online digestion and analysis of RNA molecules using immobilized RNase cartridges attached to a mDLC–MS system, and why the accurate sequencing of sgRNA is important.

Rising Stars of Separation Science: Alexandre Goyon

Siloxane contamination can interfere with low-level analysis of target compounds whether they are discrete peaks or manifested as high background bleed. They are one of the most common types of “ghost peaks” in gas chromatography (GC) and can come from a variety of sources such as inlet septa, columns, valves, solvents, and vial cap seals. While both column bleed and other sources of contamination contain siloxanes, the composition is different. Examining the mass spectrum and structure provides clues to the source of the contamination. This article analyzes both septa bleed and column bleed fragments and reviews best practices in reducing siloxane “ghost peaks”, in addition to understanding the differences in composition.

Siloxanes or polysiloxanes are molecules representing [-Si-O-Si-]n, where the Si has two other organic groups attached. For gas chromatography (GC) stationary phases, those side groups could be methyl (“1”‑type column), phenyl (“5”‑type column), combinations such as phenyl/cyano (“624”-type column), or a variety of other functional groups. These long siloxane chains can fold back upon themselves as temperature increases and depolymerize, yielding cyclosiloxanes, predominately trimers (D3), tetramers (D4), and pentamers (D5) (1). It is important to understand these molecules because they can impact a gas chromatogram in several ways.

Oxygen and moisture can greatly accelerate linear siloxane breakdown, which is observed as bleed in GC. Siloxane compounds can originate from carrier gas line. To reduce this risk it is recommended that purification filters are used on carrier gas lines to reduce column bleed, minimize surface activity, and extend column lifetime. This was demonstrated by Linx Waclaski, who performed an experiment attaching different ball valves behind the GC system without purification filters, and set the oven at 35 °C to collect the off‑gassing from the ball valve lubricant. These silicone-based lubricants make opening and closing the valve easier, prevent leaks, and extend valve lifetime. There was a significant amount of contamination observed, which was eliminated by the addition of a triple filter (hydrocarbon, moisture, and oxygen) (2).

Septum bleed from vial caps is less common but can be difficult to isolate since chromatographically it looks identical to inlet septa bleed. Solvents such as methylene chloride can leach siloxanes from septa, especially if the vial has been penetrated more than once. Amanda Rigdon demonstrated this phenomenon by injecting derivatized amphetamines five consecutive times using the same vial and cap septa. In this case, the septum bleed peaks were almost as large as the analyte peaks (3). In another study, repeated use of the same vial cap and vial with a 50-µL insert resulted in endrin and dichlorodiphenyltrichloroethane (DDT) breakdown, in addition to contamination of the inlet (4). Most vial cap septa are lined with a protective layer of polytetrafluorethylene (PTFE) to prevent direct contact with solvent vapour. Without a protective layer, vial cap septa bleed can be detected after 24 h at ambient temperature (3).

The septum on the top of the injection port of the GC system requires care when handling since the septum can adsorb phthalates, finger oils, perfumes, makeup, skin creams, and hand soap residue. GC inlet septum bleed occurs when volatile degradation products enter the column and is seen as a late eluting series of evenly spaced peaks in a programmed GC analysis. During continuous instrument use, change the septum daily to avoid coring and to stop small pieces of septum from entering into the GC liner (5).

Further advice includes the selection of the correct syringe needle to help minimize coring and septa fragments. One popular choice is the tapered cone style 23s–26s needle. Furthermore, understanding the thermal profiles of injection ports for setting inlet temperatures will help to reduce inlet septa bleed. The temperature at the septum is affected by the heating element and the overall inlet design, which varies significantly among manufacturers. For instance, on most GC systems, an inlet set point of 250 °C results in a septa temperature of 100 °C lower. Compare this to some manufacturers of older instruments where the inlet set point is similar to the septa temperature (6,7,8).

If GC column bleed and septa bleed are both made up of siloxanes, what makes them different? At first glance septa bleed has a base peak of 

 73, while column bleed has a base peak of 

 207. But keep in mind both column bleed and septa bleed are made up of cyclosiloxanes. These differences are investigated below.

The first part of this experiment was the extraction and concentration of both septa bleed and column bleed. A 30 m × 0.25 mm, 1.0-µm polydimethylsiloxane stationary phase column (“1-type”) (Restek) was installed into the GC inlet, with the detector end of the column protruding from the GC oven into 1 mL of methylene chloride. By operating the column at its maximum temperature (340 °C), bleed could be captured and concentrated in the methylene chloride, which was placed in a beaker of ice water to minimize evaporative loss. Bleed was collected for 6 h, with methylene chloride being added to the sample as necessary to maintain a volume of approximately 1 mL. Bleed from GC inlet septa was collected by submerging approximately 5 g of septa in 5 mL of methylene chloride for 6 h. Both septum and column bleed were concentrated 10-fold by evaporation at 60 °C. Samples were analyzed individually and combined for a single analysis to evaluate bleed profiles (Figure 1).

The GC column chosen to examine both septa and column was a “1-type” 30 m × 0.25 mm, 0.25-µm (Restek) installed into a GC–MS system (Agilent 7890 GC, 5975 GC–MS) with a split/splitless inlet and a 10:1 split, flow 2 mL/min, and a GC program of 40 °C (hold 1 min) 15 °C per min to 330 °C (hold for 20 min). The MS scan rate was 

 45 to 550 with a 3-min solvent delay. Compounds associated with bleed profiles were identified using a mass spectra database from the National Institute of Standards and Technology (NIST).

Polydimethylsiloxanes (PDMS), the stationary phase in many GC columns, are long‑chain polymers that slowly decompose into short-chain cyclic molecules. Column bleed had two prominent ions, 

 281, which are the base peaks for hexamethylcyclotrisiloxane (D3) and octamethylcyclotetrasiloxane (D4), respectively. Since bleed occurred throughout the chromatogram (Figure 2) and was a mix of these compounds that were not focused, it was seen as background. To confirm the specific compounds in the bleed profile, it was necessary to collect the column bleed and analyze it. Septa bleed concentrated at the head of the column was seen as chromatographic peaks of cyclosiloxanes, but they were in fact much larger molecules. There were over a dozen of these compounds, such as eicosamethylcylcosiloxane (D10), which has a less stable structure and produces a base peak of 

 73 (Figure 2). Figure 3 presents the spectra of D3 and D4 separately and examines the chromatogram bleed spectrum at 19.00 min, confirming that the bleed is a combination of both compounds. While column bleed also contains D5, D6, and D7, the concentrations of these compounds were less compared to D3 and D4. Since the scan rate was limited to 550 amu, compounds with spectral fragments higher than this scan rate were not able to be correctly identified; however, the fragmentation pattern suggested that they were larger cyclosiloxanes. Figure 4 demonstrates how septum-related bleed compounds were still eluting after 15 min at 330 °C. If the column was not baked out at this temperature, the septum compounds would carryover into the next analysis and would be observed as a bump in the baseline or an increase in background.

One overlooked area for septa bleed is the vial cap septa, and, while less common, it is characterized as sharp repetitive peaks even after several blanks have been analyzed (3). With vial inserts, the issues can be greatly exaggerated and can also degrade target compounds; an example of this would be endrin or DDT (4). Coring, whether from the sample vial or the injection port septum, will result in an increased number of siloxanes that will manifest themselves as “ghost peaks” (5). Whether it is valve bleed, septum vial bleed, or GC inlet septum bleed, there will be late eluting compounds, with 

 73 for many of them. Identifying the source of siloxanes makes troubleshooting much easier. While it’s true that both septa bleed and column bleed are the same—that is, they are both made up of cyclosiloxanes—the molecular weights are quite different, resulting in different spectra.

Special thanks to Jaap de Zeeuw and Erica Pack (Restek) for their technical advice and review.

D. Allan, S.C. Radzinski, M.A. Tapsak, and J.J. Liggat, 

, 553–562 (2016) DOI: 10.1007/s12633-014-9247-6

www.restek.com/en/chromablography/chromablography/are-my-shut-off-valves-contaminated

How to Diagnose GC Septum Bleed Contamination Sources: Could it be Your Vial Cap?

www.restek.com/en/technical-literature-library/articles/how-to-diagnose-gc-septum-bleed-contamination-sources-could-it-be-your-vial-cap

Silicone Autosampler Vial Septa Cause Endrin Breakdown and Sample Contamination

www.restek.com/en/chromablography/chromablography/silicone-autosampler-vial-septa-cause-endrin-breakdown-and-sample-contamination

www.restek.com/globalassets/pdfs/literature/gnar2671-unv.pdf

Thermo Trace 1310 Inlet Temperature Profile vs Agilent 7890 for Split/Splitless Injectors 

www.restek.com/en/chromablography/chromablography/thermo-trace-1310-inlet-temperature-profile-vs-agilent-7890-for-splitsplitless-injectors

 It’s a Matter of Degrees, but Do Degrees Really Matter? 

https://www.restek.com/globalassets/pdfs/literature/gnar1937-unv.pdf

 https://pdf2.chromtech.net.au/580134_ADV07-2_22.pdf

has managed a team of chemists in Restek’s innovations laboratory since 2004. Before taking the reins of the laboratory, he spent seven years as an environmental chemist and was critical to the development of Restek’s current line of volatile GC columns. Prior to joining Restek, he operated a variety of gas chromatographic detectors conducting method development and sample analysis. Chris holds a B.S. in environmental science from Saint Michael’s College, USA.

Identifying the source of siloxanes makes troubleshooting that much easier.

Featured Content

Advances in Small Molecule Analysis

New Adventures in GC–IMS for Nontargeted Screening of Food and Environmental Samples

Essentials of LC Troubleshooting, Part III

Sample Preparation Strategies

SFC Coupled with Drift Time Ion Mobility QTOF Mass Spectrometry