Liquid chromatography (LC) is a cornerstone of analytical separations, but comparing the retention times (RTs) across different LC methods is challenging because of variations in experimental parameters such as column type and solvent gradient. Nevertheless, RTs are powerful metrics in tandem mass spectrometry (MS2) that can reduce false positive rates for metabolite annotation, differentiate isobaric species, and improve peptide identification. Here, we present Graphormer-RT, a novel graph transformer that performs the first single-model method-independent prediction of RTs. We use the RepoRT data set, which contains 142,688 reverse phase (RP) RTs (from 191 methods) and 4,373 HILIC RTs (from 49 methods). Our best RP model (trained and tested on 191 methods) achieved a test set mean average error (MAE) of 29.3 ± 0.6 s, comparable performance to the state-of-the-art model which was only trained on a single LC method. Our best-performing HILIC model achieved a test MAE = 42.4 ± 2.9 s. We expect that Graphormer-RT can be used as an LC "foundation model", where transfer learning can reduce the amount of training data needed for highly accurate "specialist" models applied to method-specific RP and HILIC tasks. These frameworks could enable the machine optimization of automated LC workflows, improved filtration of candidate structures using predicted RTs, and the in silico annotation of unknown analytes in LC-MS2 measurements.

Fentanyl and its analogues are potent opioids that pose a significant threat to society. Over the last several years, considerable focus has been on the concerning trend of increasing fentanyl usage among drug users. Fentanyl analogues are mainly synthesized to evade analytical detection or increase their potency; thus, very low concentrations are sufficient to achieve a therapeutic effect. In an effort to help combat the synthetic opioid epidemic, developing targeted mass spectrometric methods for quantifying fentanyl and its analogues at ultralow concentrations is incredibly important. Most methods used to analyze fentanyl and its analogues from whole blood require manual sample preparation protocols (solid-phase extraction or liquid-liquid extraction), followed by chromatographic separation and mass spectrometric detection. The main disadvantages of these methods are the tedious sample preparation workflows, resulting in lengthy analysis times. To mitigate these issues, we present a targeted method capable of analyzing 96 samples containing fentanyl, several fentanyl analogues, and a common fentanyl (analogue) precursor simultaneously in 2.4 min per sample. This is possible by using a high-throughput solid phase microextraction workflow on the Concept96 autosampler followed by manual coupling of solid-phase microextraction fibers to the microfluidic open interface for tandem mass spectrometry analysis. Our quantitative method is capable of extremely sensitive analysis, with limits of quantification ranging from 0.002 to 0.031 ng mL-1 and linearity ranging from 0.010 to 25.0 ng mL-1. The method shows very good reproducibility (1-18%), accuracy (81-100%) of calibration and validation points, and good interday reproducibility (6-15%).

Binders are critical components used in the preparation of a range of extraction devices, including solid-phase microextraction (SPME) devices. While the main role of a binder is to affix the sorbent particles to the selected support, it is critical to select the optimal binder to ensure that it does not negatively impact the coating's particle sorption capability. This work presents the first comprehensive investigation of the interactions between binders and solid sorbent particles as these interactions can significantly impact the performance of the coating. Specifically, the findings presented herein provide a better understanding of the extraction mechanisms of composite coatings and new rules for predicting the particle adhesion forces and binder distribution in the coating. The influence of binder chemistry on coating performance is investigated by examining a selection of the most used binders, namely, polydimethylsiloxane (PDMS), polyacrylonitrile (PAN), poly(vinylidene difluoride) (PVDF), polytetrafluoroethylene amorphous fluoroplastics (PTFE AF 2400), and polybenzimidazole (PBI). The solid particles (e.g., hydrophilic-lipophilic balanced (HLB) and C18) used in this work were selected for their ability to provide optimal extraction coverage for a broad range of analytes. The results show that PDMS does not change the properties of the solid particles and that the binder occupies a negligible volume due to shrinking after polymerization, resulting in the solid particles making up most of the coating volume. Hence, the coating sorption characteristics correspond closely to the properties of the selected solid particles. On the other hand, the results also showed that PTFE AF 2400 can interact with the active surface of the sorbent, leading to the deactivation of the sorbent particles. Therefore, the extraction performance and permeability coefficients decrease as the size of the penetrant increases, indicating a rigid porous structure. The results of this study can aid in the optimization of SPME devices as they provide reference values that can be used to determine the optimal binder and the sorbent affinity for the targeted compounds. Finally, the present work also provides the broader scientific community with a strategy for investigating the properties of sorbent particle/binder structures and defines the characteristics of a good coating/membrane by analyzing all parameters such as kinetics, thermodynamic equilibria, and morphology.

There is great demand for analytical methods capable of providing high-throughput and rapid screening, especially for anti-doping and clinical point-of-care applications. In this work, automated microfluidic open interface-mass spectrometry (MOI-MS) was used for coupling with high-throughput, automated solid-phase microextraction (SPME) to achieve this objective. The design of the MOI-MS interface provides a continuous and stable electrospray fluid flow to the MS without introducing any bubble, a feature that we exploit to introduce the concept of multi-segment injection for the determination of multiple samples in a single MS run. By eliminating the need to start a new MS run between sample assays, the developed approach provides significantly simplified protocols controlled by programmed software and increased reproducibility. Furthermore, the biocompatible SPME device, which utilizes coating consisting of hydrophilic-lipophilic balanced particles embedded in a polyacrylonitrile (PAN) binder, can be directly used for biological sample analysis, as the PAN acts as both a binder and a matrix-compatible barrier, thus enabling the enrichment of small molecules while eliminating interferences associated with the presence of interfering macromolecules. The above design was employed to develop a fast, quantitative method capable of analyzing drugs of abuse in saliva samples in as little as 75 s per sample. The findings indicate that the developed method provides good analytical performance, with limits of detection ranging between 0.05 and 5 ng/mL for analysis of 16 drugs of abuse, good calibration linear correlation coefficients (R2 ≥ 0.9957), accuracy between 81 and 120%, and excellent precision (RSD% < 13%). Finally, a proof-of-concept experiment was performed to demonstrate the method's suitability for real-time analysis in anti-doping applications.

Mass spectrometry analysis can be performed by introducing samples directly to mass spectrometry, allowing the increase of the analysis throughput; however, some disadvantages of direct-to-mass spectrometry analysis include susceptibility to matrix effects and risk of instrument contamination from inadequate sample preparation. Solid-phase microextraction is one of the most suitable sample preparation methods for direct-to-mass spectrometry analysis, as it offers matrix-compatible coatings which ensure analyte enrichment with minimal or no interference from matrix. One of the ways solid-phase microextraction can be coupled directly to mass spectrometry is via a microfluidic open interface. This manuscript reports improvements made to the initial microfluidic open interface design, where the system components have been simplified to mostly commercially available materials. In addition, the analysis of samples has been automated by implementing software that fully controls the analysis workflow, where the washing procedure is optimized to completely reduce the carryover. Herein, the extraction and desorption time profiles from thin and thick SPME devices was studied where the overall workflow consisted of high-throughput sample preparation of 1.3 min per 96 samples and <1 min per sample instrumental analysis.

Solid-phase microextraction (SPME)-direct mass spectrometry (MS) has proven to be an efficient tool for the rapid screening and quantitation of target compounds at trace levels. However, it is challenging to perform screening using both positive and negative modes in one analytical run without compromising scanning speed and detection sensitivity. To take advantage of the special geometry of a coated blade spray (CBS) blade, which consists of two flat sides coated with the same SPME coating, we developed a CBS-MS method that enables desorption and ionization to be performed in positive ionization mode on one side of a coated blade and negative ionization mode on the other side of the same blade. By simply flipping the blade 180°, MS analysis in both ionization modes on different sides can be completed in 40 s. Combining this approach with an automated Concept 96-blade-based SPME system allowed analysis for one sample in positive and negative modes to be completed in less than 1 min. The workflow was optimized by using a biocompatible polyacrylonitrile as an undercoating layer and a binder of polyacrylonitrile/hydrophilic-lipophilic balance (HLB) particles, which enabled the rapid analysis of 20 drugs of abuse in saliva samples in both positive and negative modes. The proposed method provided low limits of quantification (between 0.005 and 10 ng/mL), with calibration linear correlation coefficients ⩾ 0.9925, accuracy between 72% and 126%, and relative precision < 15% for three validation points.

Selecting the optimal binder and the sorbent affinity for selected compounds can cause the composite to behave either as an efficient extraction coating, as a permeable membrane, or as an impermeable barrier. If the compound partitions onto the sorbent with high preference, it becomes stationary and the composite behaves as an impermeable barrier, while appropriately optimized affinity will result in effective permeation. To understand this phenomenon, we utilize solid-phase microextraction to characterize the mass transfer attributes of different separation composites. Our results indicate that for strong sorbents, the extraction rate is primarily controlled by the diffusion in the extraction phase rather than the sample matrix, even if it is relatively thin. Low analyte diffusion is caused by the retarding force generated by the partitioning of analytes into the sorbent, as migration through the composite is driven by the unbound form of the compound in the binder. One of the main contributions of this work is that an understanding of the extraction composite parameters that control mass transfer during extraction enables better optimization of binder/sorbent extraction phase composition for a given application. Another contribution of this work shows how a heterogeneous coating model can be simplified into a homogeneous coating model. The developed models enable an enhanced understanding of mass transfer kinetics, and they provide insight into how to optimize the extraction phase parameters for a given method involving sorbent particles in polymeric media, including membranes and paints, in addition to extraction coatings.