Filtern
Dokumenttyp
- Vortrag (5)
- Posterpräsentation (5)
- Zeitschriftenartikel (4)
- Beitrag zu einem Tagungsband (2)
- Forschungsdatensatz (1)
Sprache
- Englisch (17) (entfernen)
Schlagworte
- Online NMR spectroscopy (17) (entfernen)
Organisationseinheit der BAM
Eingeladener Vortrag
- nein (5)
Industry 4.0 is all about interconnectivity, sensor-enhanced process control, and data-driven systems. Process analytical technology (PAT) such as online nuclear magnetic resonance (NMR) spectroscopy is gaining in importance, as it increasingly contributes to automation and digitalization in production. In many cases up to now, however, a classical evaluation of process data and their transformation into knowledge is not possible or not economical due to the insufficiently large datasets available. When developing an automated method applicable in process control, sometimes only the basic data of a limited number of batch tests from typical product and process development campaigns are available. However, these datasets are not large enough for training machine-supported procedures. In this work, to overcome this limitation, a new procedure was developed, which allows physically motivated multiplication of the available reference data in order to obtain a sufficiently large dataset for training machine learning algorithms. The underlying example chemical synthesis was measured and analyzed with both application-relevant low-field NMR and high-field NMR spectroscopy as reference method. Artificial neural networks (ANNs) have the potential to infer valuable process information already from relatively limited input data. However, in order to predict the concentration at complex conditions (many reactants and wide concentration ranges), larger ANNs and, therefore, a larger Training dataset are required. We demonstrate that a moderately complex problem with four reactants can be addressed using ANNs in combination with the presented PAT method (low-field NMR) and with the proposed approach to generate meaningful training data.
Data set of low-field NMR spectra of continuous synthesis of nitro-4’-methyldiphenylamine (MNDPA). 1H spectra (43 MHz) were recorded as single scans.
Two different approaches for the generation of artificial neural networks training data for the prediction of reactant concentrations were used: (i) Training data based on combinations of measured pure component spectra and (ii) Training data based on a spectral model.
Synthetic low-field NMR spectra
First 4 columns in MAT-files represent component areas of each reactant within the synthetic mixture spectrum.
Xi (“pure component spectra dataset”)
Xii (“spectral model dataset”)
Experimental low-field NMR spectra from MNDPA-Synthesis
This data set represents low-field NMR-spectra recorded during continuous synthesis of nitro-4’-methyldiphenylamine (MNDPA). Reference values from high-field NMR results are included.
Monitoring specific chemical properties is the key to chemical process control. Today, mainly optical online methods are applied, which require time- and cost-intensive calibration effort. NMR spectroscopy, with its advantage being a direct comparison method without need for calibration, has a high potential for closed-loop process control while exhibiting short set-up times. Compact NMR instruments make NMR spectroscopy accessible in industrial and rough environments for process monitoring and advanced process control strategies.
We present a fully automated data analysis approach which is completely based on physically motivated spectral models as first principles information (Indirect Hard Modelling – IHM) and applied it to a given pharmaceutical lithiation reaction in the framework of the European Union’s Horizon 2020 project CONSENS. Online low-field NMR (LF NMR) data was analysed by IHM with low calibration effort, compared to a multivariate PLS-R (Partial Least Squares Regression) approach, and both validated using online high-field NMR (HF NMR) spectroscopy.
Already Analyzing or Still Calibrating? - Demonstration of an online NMR analyzer in pilot scale
(2018)
The departure from the current automation landscape to next generation automation concepts for the process industry has already begun. Smart functions of sensors simplify their use and enable plug-and-play integration, even though they may appear to be more complex at first sight. Monitoring specific information (i.e., “chemical” such as physico-chemical properties, chemical reactions, etc.) is the key to “chemical” process control. Here we introduce our smart online NMR sensor module provided in an explosion proof housing as example.
Due to NMR spectroscopy as an “absolute analytical comparison method”, independent of the matrix, it runs with extremely short set-up times in combination with “modular” spectral models. Such models can simply be built upon pure component NMR spectra within a few hours (i.e., assignment of the NMR signals to the components) instead of tedious calibrations runs. We present a range of approaches for the automated spectra analysis moving from statistical approach, (i.e., Partial Least Squares Regression) to physically motivated spectral models (i.e., Indirect Hard Modelling and Quantum Mechanical calculations).
Based on concentration measurements of reagents and products by the NMR analyzer a continuous production and direct loop process control were successfully realized for several validation runs in a modular industrial pilot plant and compared to conventional analytical methods (HPLC, near infrared spectroscopy). The NMR analyser was developed for an intensified industrial process funded by the EU’s Horizon 2020 research and innovation programme (“Integrated CONtrol and SENsing”, www.consens-spire.eu).
Monitoring chemical reactions is the key to chemical process control. Today, mainly optical online methods are applied, which require excessive calibration effort. NMR spectroscopy has a high potential for direct loop process control while exhibiting short set-up times. Compact NMR instruments make NMR spectroscopy accessible in industrial and harsh environments for advanced process Monitoring and control, as demonstrated within the European Union’s Horizon 2020 project CONSENS.
We present a range of approaches for the automated spectra analysis moving from conventional multivariate statistical approach, (i.e., Partial Least Squares Regression) to physically motivated spectral models (i.e., Indirect Hard Modelling and Quantum Mechanical calculations). By using the benefits of traditional qNMR experiments data analysis models can meet the demands of the PAT community (Process Analytical Technology) regarding low calibration effort/calibration free methods, fast adaptions for new reactants or derivatives and robust automation schemes.
Monitoring chemical reactions is the key to chemical process control. Today, mainly optical online methods are applied. NMR spectroscopy has a high potential for direct loop process control. Compact NMR instruments based on permanent magnets are robust and inexpensive analysers, which feature advantages like low maintenance, ease of use, and cryogen-free operation. Instruments for online NMR measurements equipped with a flow-through cell, possessing a good signal-to-noise-ratio, robustness, and meeting the requirements for integration into industrial plants (i.e., explosion safety and fully automated data analysis) are currently not available off the rack.
Intensified continuous processes are in focus of current research. Flexible (modular) chemical plants can produce different products using the same equipment with short down-times between campaigns and quick introduction of new products to the market. In continuous flow processes online sensor data and tight closed-loop control of the product quality are mandatory. Otherwise there is a huge risk of producing large amounts of out-of-spec (OOS) products. This is addressed in the European Union’s Research Project CONSENS (www.consens-spire.eu) by development and integration of smart sensor modules for process monitoring and control within such modular plant setups.
The presented NMR module is provided in an explosion proof housing with a module size of 57 x 57 x 85 cm and involves a compact 43.5 MHz NMR spectrometer together with an acquisition unit and a programmable logic controller for automated data preparation (phasing, baseline correction) and evaluation. Indirect Hard Modeling (IHM) was selected for data analysis of the low-field NMR spectra. A set-up for monitoring continuous reactions in a thermostated 1/8” tubular reactor using automated syringe pumps was used to validate the IHM models by using high-field NMR spectroscopy as analytical reference method.
For reaction monitoring and process control using NMR instruments, in particular, after acquisition of the FID the data needs to be corrected in real-time for common effects using fast interfaces and automated methods. When it comes to NMR data evaluation under industrial process conditions, the shape of signals can change drastically due to nonlinear effects. Additionally, the multiplet structure becomes more dominant because of the comparably low-field strengths which results in overlapping of multiple signals. However, the structural and quantitative information is still present but needs to be extracted by applying predictive models.
We present a range of approaches for the automated spectra analysis moving from statistical approach, (i.e., Partial Least Squares Regression) to physically motivated spectral models (i.e., Indirect Hard Modeling). By using the benefits of traditional qNMR experiments data analysis models can meet the demands of the PAT community (Process Analytical Technology) regarding low calibration effort/calibration free methods, fast adaptions for new reactants, or derivatives and robust automation schemes.
Simple calibration concept of an online NMR module demonstrated in a modularised production plant
(2017)
Monitoring specific information (such as physico-chemical properties, chemical re-actions, etc.) is the key to chemical process control. Within the CONSENS Project, the challenge to adapt a commercially available benchtop NMR spectrometer to the full requirements of an automated chemical production environment was tack-led. The developed online NMR module was provided in an explosion proof housing and involves a compact 43.5 MHz NMR spectrometer together with an acquisition unit, a programmable logic controller for automated triggering, flow con-trol, as well as data communication.
First results of an aromatic coupling reaction in lab scale showed a general feasibil-ity according to the signal information in the acquired NMR spectra even though with a considerable overlap. Due to the comparatively low field strength of the NMR spectrometer multivariate methods had to be considered for the prediction of con-centration profiles based on spectral data. Typically, for industrial application of those methods, e.g. Partial Least Squares Regression (PLS-R) as well as Indirect Hard Modeling, large amount of calibration data is demanded, which must be ac-quired in time consuming lab-scale experiments and offline analytics. When it comes to changes in raw materials (e.g., varying functional groups, additional stabi-lizing agents) calibration experiments and data evaluation models are developed again. Here we present an approach of automated data analysis tools for low field NMR spectra with minimal calibration effort. The algorithms are based on Indirect Hard Modeling, whereby each component in each mixture spectra can be rep-resented by several flexible peak functions (pure component models). This means, that only pure component NMR spectra are needed to generate a first evaluation model. The flexibility of peak functions in the spectral model can be adjusted via constraints of peak parameters. The area of any pure component model can either be converted to concentrations based on a one-point calibration on raw material concentration or even neat solvent signals. In several cases it has been shown, the IHM works almost independently of the matrix of the real samples. Such a calibration can be repeated daily in the beginning of each process run with minimal time effort. Moreover, additional pure components can be added to the model or even substitut-ed while keeping the previously adjusted peak function constraints. The proposed method exhibited good agreement of resulting concentration data from low field NMR spectra, when compared to an online high field NMR spectrometer as refer-ence instrument.
Monitoring chemical reactions is the key to chemical process control. Today, mainly optical online methods are applied, which are calibration intensive. NMR spectroscopy has a high potential for direct loop process control while exhibiting short set-up times. Compact NMR instruments make NMR spectroscopy accessible in industrial and harsh environments for advanced process monitoring and control.
Within the European Union’s Research Project CONSENS (Integrated CONtrol and SENsing, www.consens-spire.eu) by development and integration of a smart NMR module for process monitoring was designed and delivers online spectra of various reactions. The presented NMR module is provided in an explosion proof housing of 57 x 57 x 85 cm module size and involves a compact spectrometer together with an acquisition unit and a programmable logic controller for automated data preparation (phasing, baseline correction), and evaluation.
For reaction monitoring and process control using NMR instruments after acquisition of the FID the data needs to be corrected in real-time for common effects using fast interfaces and automated methods.
When it comes to NMR data evaluation under industrial process conditions, the shape of signals can change drastically due to nonlinear effects. Additionally, the multiplet structure becomes more dominant because of the comparably low-field strengths which results in overlapping of multiple signals. However, the structural and quantitative information is still present but needs to be extracted by applying predictive models.
We present a range of approaches for the automated spectra analysis moving from statistical approach, (i.e., Partial Least Squares Regression) to physically motivated spectral models (i.e., Indirect Hard Modelling and Quantum Mechanical calculations). By using the benefits of traditional qNMR experiments data analysis models can meet the demands of the PAT community (Process Analytical Technology) regarding low calibration effort/calibration free methods, fast adaptions for new reactants, or derivatives and robust automation schemes.
Monitoring chemical reactions is the key to chemical process control. Today, mainly optical online methods are applied. NMR spectroscopy has a high potential for direct loop process control. Compact NMR instruments based on permanent magnets are robust and relatively inexpensive analysers, which feature advantages like low cost, low maintenance, ease of use, and cryogen-free operation. Instruments for online NMR measurements equipped with a flow-through cell, possessing a good signal-to-noise-ratio, sufficient robustness, and meeting the requirements for integration into industrial plants (i.e., explosion safety and fully automated data analysis) are currently not available off the rack.
Intensified continuous processes are in focus of current research. Flexible (modular) chemical plants can produce different products using the same equipment with short down-times between campaigns and quick introduction of new products to the market. In continuous flow processes online sensor data and tight closed-loop control of the product quality are mandatory. If these are not available, there is a huge risk of producing large amounts of out-of-spec (OOS) products. This is addressed in the European Union’s Research Project CONSENS (Integrated Control and Sensing [3]) by development and integration of smart sensor modules for process monitoring and control within such modular plant setups.
The presented NMR module is provided in an explosion proof housing of 57 x 57 x 85 cm module size and involves a compact 43.5 MHz NMR spectrometer together with an acquisition unit and a programmable logic controller for automated data preparation (phasing, baseline correction) and evaluation. Indirect Hard Modeling (IHM) was selected for data analysis of the low-field NMR spectra. A set-up for monitoring continuous reactions in a thermostated 1/8” tubular reactor using automated syringe pumps was used to validate the IHM models by using high-field NMR spectroscopy as analytical reference method.