Benchtop NMR paves the way for real-time reaction monitoring

Benchtop NMR paves the way for real-time reaction monitoring

Benchtop NMR paves the way for real-time reaction monitoring

Nuclear magnetic resonance (NMR) spectroscopy is a powerful analytical technique widely used in chemistry, biochemistry and materials science for identification, structural characterization, and quantification of compounds. It relies on the interaction between atomic nuclei and magnetic and radiofrequency fields.

In addition, it provides valuable information about the chemical environment, bonding, and connectivity within a molecule. In each case, NMR offers significant advantages over a variety of other spectroscopic techniques; it is non-destructive, is inherently quantitative and does not require dedicated reference/ calibration samples.[1]

High-field NMR

When considering NMR spectroscopy, most people think of so-called high-field NMR, instruments which use superconducting magnets to generate very strong magnetic fields – typically above nine tesla. This enhances sensitivity and resolution and ensures clean spectra with sharp, well-defined peaks that are easy to interpret.

High-field NMR has become indispensable in fields such as biopharmaceutical development,[2] which requires the detailed analysis of large molecules such as proteins and nucleic acids, as the high magnetic field strengths provide strong resolution and sensitivity.

Researchers have reported, for example, that high resolution, coupled with advances in pulse sequences, isotopic labeling strategies, and the development of competition experiments have supported the study of higher molecular weight protein targets, facilitated higher-throughput, and expanded the range of binding affinities that can be evaluated, improving the utility of NMR for identifying and characterizing potential therapeutics to druggable protein targets.[3]

However, high-field instruments can be cost-prohibitive to purchase, run and maintain, and must be operated by specialist NMR technicians. These instruments are most often installed in a separate facility because they require a large amount of space and require controlled handling (due to the use of liquid helium and nitrogen).

A benchtop revolution

As a result, access to high-field NMR has been limited, particularly for smaller research institutions or laboratories with restricted budgets. The analytical advantages of NMR could benefit a much broader range of scientists working in industry, as well as research, but until recently, this has not been possible.

A ‘first-generation’ of benchtop NMR instruments, which use compact permanent lower-field magnets to generate the required magnetic field, was proposed as a viable alternative to high-field NMR in many instances. However, the homogeneity of the permanent magnets used was not sufficient to produce consistent spectra with adequate resolution. In recent years, technological and manufacturing developments have boosted performance to a point where benchtop NMR instruments now offer a viable alternative to high-field systems.

These improved instruments are driving the development of a growing range of important applications. The smaller size, the elimination of the need for cryogenic liquids, and a significant reduction in sample workup, have all made it possible for research institutions and laboratories to bring NMR spectroscopy directly to the laboratory bench.[4]

Benchtop systems can now be operated without expert input and provide answers to important questions like “have we produced the molecule that we need?” in an instant. Moreover, benchtop NMR is unlocking the technology for use in, for example, reaction monitoring, biomedical,[5] and point-of-care clinical diagnostic scenarios.[6]

Solving solvent suppression

As samples are measured in solution and regular solvents have NMR signals that are orders of magnitude larger than the signal of the compound you wish to analyze, the standard procedure requires samples to be dissolved in deuterated solvents in order to eliminate the solvent signal and allow the peaks of interest to be seen.

In addition, deuterated solvents have a second and very important function in high field spectrometers. The signal of deuterium is used to track and correct the drift of the magnetic field generated by the magnet. By locking the magnetic field, the high field stability required for the measurement is achieved.

In practice, replacing a regular solvent with the deuterated counterpart means a demanding and time- consuming sample work-up has to be performed before each sample can be measured.

In contrast, the lock system of Magritek Spinsolve spectrometers uses a capillary sample that is mounted permanently in the magnet. In this way, the need for deuterated solvents to lock the magnetic field is eliminated and samples extracted directly from a reactor can be analyzed without sample preparation. However, as these samples are dissolved in protonated solvents with large signals, the spectra of the compound under test can be obscured.

To overcome this, solvent suppression methods like PRESAT or WET have been developed – but these only work if the magnets have a high level of homogeneity. To achieve the high suppression efficiency required for these measurements, Magritek has developed the Spinsolve ULTRA systems. The ultra-high field homogeneity achieved, when combined with solvent suppression, makes the Spinsolve ULTRA capable of measuring compounds dissolved at sub-millimolar concentrations in protonated solvents, as if they were dissolved in deuterated equivalents.

Spinsolve and the power of WET suppression

A recent Magritek study[7] demonstrated the enhanced performance achieved when using a Spinsolve 80 ULTRA with the WET (water-suppression enhanced through T1 effects) solvent suppression method. For this study, analytes acetylsalicylic acid, diethyl phthalate and ethyl crotonate were each dissolved in a selection of common organic solvents in their protonated form. Because of the concentration difference between the analytes of interest and the protonated solvent, only the solvent peaks were visible when plotted at full scale (see Figure 1).

Figure 1: Stacked plot of 1D 1H NMR of analytes in common protonated organic solvents (16 scans, 10 s rep., 90 pa).

As an example, Figures 2 and 3 show the spectra for diethyl phthalate dissolved in tetrahydrofuran, with a quartet visible at 4.25ppm and a triplet visible at 1.27ppm. In the standard 1H spectrum (shown in red), the protons of the phthalate are buried beneath the tetrahydrofuran signal. As shown in both the 80-time zoom (Fig. 2) and the 160-time zoom (Fig. 3), the WET solvent suppression method with 13C decoupling (shown in cyan), has attenuated the signals of the tetrahydrofuran to baseline-separate the diethyl phthalate signals. Similar attenuation was achieved with the remaining analytes in dimethyl sulfoxide, ethanol, diisopropyl ether, benzene, and chloroform.

Figure 2: Comparison of a standard 1D 1H NMR spectrum of 20 mMolar diethyl phthalate in THF (red) and the 1D spectrum collected with the WET solvent suppression NMR sequence in the presence of 13C decoupling (cyan).
Figure 3: Zoomed comparison of a regular 1D 1H NMR spectrum of 20 mMolar diethyl phthalate in THF (red) and the spectrum collected with the 1D 1H WET solvent suppression protocol with 13C decoupling (cyan).

This example clearly shows the powerful effect of WET suppression in the presence of carbon decoupling when using the Spinsolve 80 ULTRA. The superior homogeneity of Magritek’s Spinsolve ULTRA magnets boosts the performance of the suppression methods. The result is that the residual signals of the protonated solvents after suppression are comparable to the very small signal of any remaining protons in deuterated solvents used in high field NMR. This removes the need for researchers to carry out laborious sample workups to replace regular solvents with deuterated counterparts, thus simplifying and speeding up the analysis process. It is also advantageous in the monitoring of chemical reactions online or at different stages of synthesis as part of the quality control process.

Real-time quality control

In conjunction with solvent suppression, benchtop NMR is becoming more powerful, and it has proven to be a valuable analytical tool for monitoring reactions in both batch and continuous flow systems.[8] Its real-time, non-destructive nature allows for efficient optimization of reaction conditions, improved process control, and faster development of organic synthesis methods. It can also provide valuable information on reaction kinetics and thermodynamics. By measuring reaction rates and equilibrium constants, researchers can gain a deeper understanding of reaction mechanisms and make informed decisions about reaction conditions and catalyst selection.

Batch reactions

The observation and characterization of new reactions or molecules has, in the past, relied on serendipity, or chance observation, as most organic synthesis is target-oriented.[9] This has seen an increase in methods using a ‘closed-loop’ approach, which automates chemical exploration through trial and error using a decision-making algorithm and online analytics.

The closed-loop framework requires three components: a chemical robot to carry out analysis, a program for data interpretation, and an algorithm that links reaction outcomes to the input and process parameters and suggests appropriate parameters for the next reactions (see Figure 4). To enable autonomous reactivity- first discovery, both the analytical approach and reactivity detection algorithm must be general-purpose and robust; the speed and small size of benchtop NMR systems has facilitated this in situ monitoring, especially when contrasted with expensive and time-consuming analysis required by larger machines.

Figure 4: Closed-loop framework for chemical space exploration. A liquid handling robot performs an experiment and collects NMR and mass spectrometry (MS) spectra. These data are processed to assess reactivity and create a model of the chemical space that is queried to formulate the next experiment to be performed10

In 2021, a research team based at the University of Glasgow presented a robotic chemical discovery system (Chemputer) capable of navigating a chemical space based on a learned general association between molecular structures and reactivity, while incorporating a neural network model that can process data from online analytics and assess reactivity without knowing the identity of the reagents. They recorded the NMR spectra using a Magritek Spinsolve 60.

Figure 5: Five reactions showing high reactivity have been found and characterized. Based on the data acquired by the Spinsolve, the researchers selected five combinations from a pool of reactions and discovered two novel products (see d and e). This suggests that NMR-driven universal reactivity detection can significantly contribute to discovering new reactions in uncharted chemical spaces, as well as enabling autonomous discovery in closed loops. The team has now integrated the Spinsolve as a critical component of its novel “Chemputer.”

Continuous flow

Benchtop NMR spectroscopy also enables online, real-time monitoring of reaction parameters in continuous flow systems, and offers numerous advantages over conventional batch protocols and off-line analysis. In fact, the synergy between continuous flow reactors and benchtop NMR spectroscopy has transformed chemical synthesis and analysis. [11] The combination of benchtop NMR, which offers non-destructive and quantitative insights, and continuous flow reactors – the benefits of which include precise reaction parameter control, increased safety, and more efficiency – enables real-time monitoring and alteration of reaction parameters.

More specifically, by integrating the NMR probe directly into a flow system, researchers can monitor reaction progress and control reaction conditions in real-time, allowing for rapid and automatic optimization of key reaction parameters, such as temperature, flow rate, and reactant concentration. This improves yields and selectivity, streamlines workflows, and enables more efficient reaction scale-up. The combination of continuous flow and online NMR also minimizes waste, both by reducing the amount of solvent required, and by eliminating the need for arduous sample preparation and off-line analysis.[12]

Speedier reaction optimization

Online monitoring by NMR relies on the high performance of the solvent suppression method to be successful. Magritek scientists integrated a Spinsolve 80 ULTRA at the outlet of a flow reactor. In this case, to maximize the hydrogenation of cinnamyl alcohol, they optimized the reaction parameters – pressure, temperature, reacting mixture flow rate and equivalents of H2. They monitored the reaction mixture of cinnamyl alcohol in methanol every 2 minutes, with the WET sequence set to suppress the methanol signals at 3.3 ppm and 4.9 ppm (Fig. 6). In the study mentioned above, they performed the signal acquisition with carbon decoupling to eliminate the carbon satellites from the spectrum. The measurements were performed under continuous flow conditions.

Figure 6: Comparison of the spectra acquired for the product with a standard pulse and acquire sequence (blue) and a WET suppression with carbon decoupling sequence (red). The measurements were performed under continuous flow conditions with a liquid flow rate of 1mL/min.

Figure 6 shows the spectra acquired with a standard pulse and acquire sequence (blue), compared to those acquired by WET suppression with carbon decoupling (red). In the former method, the methanol signals obscure the signals of interest for the analyte. Figure 7 shows a subset of spectra acquired using the WET sequence with carbon decoupling during an optimization run of the double bond reduction of cinnamyl alcohol. This indicates that when pressure and temperature are increased, the signals of the product increase, while the starting materials decrease. In Figure 8, the marked regions from Figure 7 are then plotted against time, with each point in the curves corresponding to a spectrum from the stack plot. It can be seen here that there is excellent agreement between the curves and minimal scatter of the points collected during each steady-state phase, showing the robustness and precision of the quantification method.

Figure 7: Spectra acquired with the Spinsolve RMX software during an optimization run of the double bond reduction of cinnamyl alcohol. The data was collected with a WET sequence set to suppress the signals of methanol at 3.3 and 4.9 ppm. To eliminate the carbon satellites, the signal was acquired in the presence of carbon decoupling.
Figure 8: Screenshot of the concentration plots of the starting material (left) and product (right) as they are displayed in real time on the Spinsolve reaction monitoring software (RMX). The different curves on each plot correspond to the concentration calculated using the signal of different chemical groups available for both starting material and product. The values are in very good agreement, showing how linear the NMR response is.

This study demonstrates the efficacy of combining an online NMR spectrometer with a continuous flow reactor, and its value to chemists. The steady state is achieved in under 15 minutes after the insertion of new reaction conditions, which would allow the analysis of 30 or more different parameter combinations in a day – a significant improvement for the optimization process, and one that saves time and money.

Autonomous reactions: monitoring chemical synthesis

The Spinsolve has also been used successfully in real-time process analytics in the development of an autonomous self-optimizing flow reactor. Three significant papers are discussed here, in chronological order of publication.

The first, a collaboration between researchers in Germany and the UK, published in 2021 [13] describes their work to integrate online NMR into an automated chemical synthesis machine (CSM). As noted above, the team has named their system the ‘Chemputer’. It is capable of small-molecule synthesis using a universal programming language to allow automated analysis and adjustment of reactions on the fly.

For this ‘proof of concept’ study, the Grignard reaction was the chosen pathway due to its relevance in synthesis and challenging analytical conditions involving solids.

Despite online or in-line optical or vibrational spectroscopy techniques being well established in industry and on some automated platforms, NMR was their chosen analytical technique – as opposed to UV/Vis, IR, Raman spectroscopy – as the benchtop NMR system offered the group:

  • Reliable and easily accessible relative quantitative results without prior calibration
  • Matrix-independent linearity between measured signal and species concentrations
  • No compromised results due to solid particles in the process stream, something often seen with other techniques

In summary, the authors concluded that the availability of qualitative and quantitative real-time NMR data enabled the application of simple feedback control. Controlling the synthesis pathway using measured current species ratios instead of hardcoded waiting times was implemented (enabling potentially higher productivity of the synthesis platform) and the measurement and evaluation of NMR spectra was successfully performed during the whole synthesis.

A second research team, the Kappe group at the University of Graz, published a 2022 paper that explored the potential benefits of using spectroscopic methods combined with chemometric modelling in closed- loop applications for the synthesis of the active pharmaceutical ingredient, edaravone.

As the authors highlight, output chemical composition using PAT (process analytical technology) poses a significant analytical challenge, and while chromatographic methods are able to provide a high level of accuracy, they are both time consuming and can only provide a single measurement, which causes errors if the sample is inhomogeneous. In contrast, spectroscopic techniques, when combined with chemometric modelling, provide more accurate quantification in faster measurement times.

Figure 9: Simplified reaction scheme of the two-step reaction—synthesis of edaravone 7 by condensation of hydrazine 4 and ketoester 5, followed by cyclization of imine intermediate 6. Upper and lower bounds of optimizable variables are shown in red.

In the study, the researchers used rapid-flow NMR and Fourier transform infrared (FTIR) alongside chemometric modelling for the timely and efficient analysis of reaction outcomes, applying it to a seven variable two- step optimization problem (imine formation and cyclization – see Figure 9). Their results showed that the Magritek Spinsolve ULTRA allowed for real-time evaluation and optimization of the reaction conditions, demonstrating the effectiveness of the self-optimization approach in finding optimal reaction parameters within a short operational time.

Recently, a third research team, the Noe?l Group at the University of Amsterdam, published a landmark paper in Science [14] that describes how their software and hardware platform, known as RoboChem, iteratively determines optimal conditions for photochemical processes in a scalable, flow-based architecture.

Photocatalysts exploit light to drive a chemical reaction. They are important in a wide range of applications and industries: pharmaceuticals, agrochemicals, and material science, for example, however, challenges persist in optimizing, replicating, and scaling these techniques. Practical considerations such as uneven light absorption and experimental variability, as well as chemical aspects such as poorly understood reaction mechanisms and intricate interactions among various variables, are all important.

The Noe?l Group assembled several key components, including a liquid handler, syringe pumps, a tunable continuous-flow photoreactor, cost-effective Internet of Things devices, and an in-line NMR system (Magritek Spinsolve ULTRA) (Figure 10). By integrating readily available hardware, customized software, and a Bayesian optimization (BO) algorithm, the platform offers a hands-free and safe solution that operates autonomously, exploring a chosen parameter space that includes both discrete and continuous variables.

Figure 10: RoboChem: a benchtop robotic platform for the self-optimization, intensification and scale-up of photocatalytic transformations

The author’s working hypothesis was that the common practice of chemists to develop and define a single set of conditions and apply them widely across experiments could be bettered by conditions derived for individual reactions by RoboChem. Indeed, this was what they found. The Science paper contains the detailed descriptions, methodology and results but in summary, across a diverse set of 19 molecules and encompassing various facets of photocatalysis, such as hydrogen atom transfer photocatalysis, photoredox catalysis, and metallaphotocatalysis, the platform either matched or significantly surpassed yields reported in the literature.

Conclusion

Magritek’s Spinsolve ULTRA technology has significantly contributed to making NMR spectroscopy more accessible, helping scientists and researchers across a wider variety of disciplines achieve faster, more accurate results.

While NMR spectroscopy is a powerful analytical technique for studying the composition, structure, and dynamics of molecules, traditional NMR systems are often large, expensive, and require specialized training to operate. In contrast, the Spinsolve ULTRA, a compact, benchtop NMR spectrometer, brings the benefits of NMR to laboratories that have not previously had access to this technology.

The Spinsolve ULTRA’s small footprint, portability and ease of installation allows it to be easily placed in any laboratory setting without the need for dedicated lab space or costly modifications. This also makes it ideal for use in applications such as reaction monitoring at remote locations, or onsite quality control in industries such as food and beverages or pharmaceuticals. Its high-performance magnets offering unprecedented homogeneity mean it can detect and differentiate between even very small chemical shifts, making it suitable for a wide range of applications, including the analysis of small organic molecules and quantification of impurities in complex samples.

The user-friendly interface of the Spinsolve ULTRA simplifies the operation of the instrument, reducing the need for extensive NMR expertise and making it accessible to a broader range of researchers. This ease of use empowers users to obtain and interpret valuable NMR data promptly and in real time. The initial cost of the Spinsolve ULTRA and the cost of maintenance is significantly lower than that of traditional NMR spectrometers, allowing researchers with lower budgets to incorporate NMR spectroscopy into their scientific workflows, improving the quality and robustness of their experiments.

Magritek’s Spinsolve ULTRA technology has revolutionized the accessibility of NMR spectroscopy in scientific applications and is facilitating the development of batch reaction and continuous flow systems and helping to drive forward autonomous discovery.

By democratizing NMR spectroscopy, Magritek is empowering scientists and researchers to make discoveries and improve processes across a broader range of scientific fields including medicinal chemistry, organic synthesis, polymer characterization and small molecule drug development.

References

  1. Emwas AHM. The strengths and weaknesses of NMR spectroscopy and mass spectrometry with particular focus on metabolomics research. Metabolomics. 2015; 1277. DOI: 10.1007/978-1-4939-2377-9_13
  2. Kiss R et al. What NMR can do in the biopharmaceutical industry. J Pharm Biomed Anal. 2017; 147, 367-377
  3. Skinner AL & Laurence JS. High-field solution NMR spectroscopy as a tool for assessing protein interactions with small molecule ligands. J Pharm Sci. 2008; 97(11), 4670-4695. DOI: 10.1002/jps.21378
  4. Blumich B. A briefcase full of NMR. The Analytical Scientist. 2015. Available at: https://theanalyticalscientist.com/techniques-tools/a- briefcase-full-of-nmr. Last accessed November 5, 2023
  5. Alonso-Moreno P et al. Benchtop NMR-Based Metabolomics: First Steps for Biomedical Application. Metabolites. 2023; 29, 13 )5): 614. DOI: 10.3390/metabo13050614
  6. Percival BC et al. Low-field, benchtop NMR spectroscopy as a potential tool for point-of-care diagnostics of metabolic conditions. High Throughput. 2019; 8(1): 2. DOI: 10.3390/ht8010002
  7. Magritek. Lifting the curtain of undesired solvent signals [application note]. Available at: https://magritek.com/products/benchtop-nmr- spectrometer-spinsolve. Last accessed: November 3, 2023
  8. Knox ST et al. benchtop flow-NMR for rapid online monitoring of RAFT and free radical polymerization in batch and continuous reactors. Polym Chem. 2019; 35(10), 4774-4778. DOI: 10.1039/C9PY00982E
  9. Burke MD & Lalic G. Teaching target-oriented and diversity-oriented organic synthesis at Harvard University. Chem Biol. 2002; 9, 535-541
  10. Caramelli D et al. Discovering new chemistry with an autonomous robotic platform driven by a reactivity-seeking neural network. ACS Cent Sci. 2007; 7, 1821-1830. DOI: 10.1021/acscentsci.1c00435
  11. Giraudeau P & Felpin FX. Flow reactors integrated with in-line monitoring using benchtop NMR spectroscopy. React Chem Eng. 2018; 3, 399-413. DOI: 10.1039/C8RE00083B
  12. Magritek. Coupling a Spinsolve NMR spectrometer to an H-Cube? Pro flow reactor for fast reaction optimization [application note]. Available at: https://magritek.com/products/benchtop-nmr-spectrometer-spinsolve. Last accessed: November 3, 2023
  13. Bornemann-Pfeiffer M et al. Standardization and Control of Grignard Reactions in a Universal Chemical Synthesis Machine using online NMR.
  14. Aidan Slattery et al., Automated self-optimization, intensification, and scale-up of photocatalysis in flow. Science 383, eadj1817(2024). DOI:10.1126/science.adj1817 Angew. Chem. Int. Ed. 2021, 60, 23202–23206. DOI: 10.1002/anie.202106323

Contact

For further information on the Magritek Spinsolve ULTRA instruments, please contact us: [email protected]

要查看或添加评论,请登录

Magritek Ltd的更多文章

社区洞察

其他会员也浏览了