Filtern
Erscheinungsjahr
Dokumenttyp
- Vortrag (142)
- Zeitschriftenartikel (77)
- Beitrag zu einem Tagungsband (35)
- Posterpräsentation (12)
- Buchkapitel (6)
- Forschungsbericht (4)
- Sonstiges (3)
- Beitrag zu einem Sammelband (2)
- Forschungsdatensatz (2)
- Monografie (1)
Sprache
- Deutsch (151)
- Englisch (134)
- Mehrsprachig (1)
Schlagworte
- Prozessanalytik (59)
- Online NMR Spectroscopy (36)
- Process Analytical Technology (36)
- Industrie 4.0 (34)
- Prozessindustrie (34)
- CONSENS (29)
- Process analytical technology (24)
- Online NMR spectroscopy (21)
- Reaction Monitoring (17)
- Digitalisierung (16)
Organisationseinheit der BAM
- 1 Analytische Chemie; Referenzmaterialien (119)
- 1.4 Prozessanalytik (118)
- 8 Zerstörungsfreie Prüfung (14)
- 8.6 Faseroptische Sensorik (10)
- 6 Materialchemie (5)
- 8.0 Abteilungsleitung und andere (5)
- 6.3 Strukturanalytik (4)
- 8.1 Sensorik, mess- und prüftechnische Verfahren (4)
- 3 Gefahrgutumschließungen; Energiespeicher (3)
- 1.3 Instrumentelle Analytik (2)
Paper des Monats
- ja (1)
Eingeladener Vortrag
- nein (142)
qNMR provides the most universally applicable form of direct concentration or purity determination without need for reference materials of impurities or the calculation of response factors but only exhibiting suitable NMR properties.
The workshop presents basic terms of statistics and uncertainty analysis, which are the basis for qNMR spectroscopy and data analysis such as, e.g., standard deviations, linear regression, significance tests, etc. and gives typical examples of applications in qNMR spectroscopy.
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).
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.
Cryoadsorption on the inner surface of porous materials is a promising solution for safe, fast, and reversible hydrogen storage. Within the class of highly porous metal−organic frameworks, zeolitic imidazolate frameworks (ZIFs) show high thermal, chemical, and mechanical stability. In this study, we selected ZIF-8 synthesized mechanochemically by twin-screw extrusion as powder and pellets. The hydrogen storage capacity at 77 K and up to 100 bar has been analyzed in two laboratories applying three different measurement setups showing a high reproducibility. Pelletizing ZIF-8 increases the packing density close to the corresponding value for a single crystal without loss of porosity, resulting in an improved volumetric hydrogen storage capacity close to the upper limit for a single crystal. The high volumetric uptake combined with a low and constant heat of adsorption provides ca. 31 g of usable hydrogen per liter of pellet assuming a temperature−pressure swing adsorption process between 77 K − 100 bar and 117 K − 5 bar. Cycling experiments do not indicate any degradation in storage capacity. The excellent stability during preparation, handling, and operation of ZIF-8 pellets demonstrates its potential as a robust adsorbent material for technical application in pilot- and full-scale adsorption vessel prototypes.
Calibration-Free Chemical Process and Quality Control Units as Enablers for Modular Production
(2021)
Modular chemical production is a tangible translation of the digital transformation of the process industry for specialty chemicals. In particular, it enables the speeding-up of process development and thus a quicker time to market by flexibly connecting and orchestrating standardised physical modules and bringing them to life (i.e., parameterising them) with digitally accumulated process knowledge.
We focus on the specific challenges of chemical process and quality control, which in its current form is not well suited for modular production and provide possible approaches and examples of the change towards direct analytical methods, analytical model transfer or machine-supported processes.
Within the Collaborative Research Center InPROMPT a novel process concept for the hydroformylation of long-chained olefins is studied in a mini-plant, using a rhodium complex as catalyst in the presence of syngas. Recently, the hydroformylation in micro¬emulsions, which allows for the efficient recycling of the expensive rhodium catalyst, was found to be feasible. However, the high sensitivity of this multi-phase system with regard to changes in temperature and composition demands a continuous observation of the reaction to achieve a reliable and economic plant operation. For that purpose, we tested the potential of both online NMR and Raman spectroscopy for process control. The lab-scale experiments were supported by off-line GC-analysis as a reference method.
A fiber optic coupled probe of a process Raman spectrometer was directly integrated into the reactor. 25 mixtures with varying concentrations of olefin (1-dodecene), product (n-tridecanal), water, n-dodecane, and technical surfactant (Marlipal 24/70) were prepared according to a D-optimal design. Online NMR spectroscopy was implemented by using a flow probe equipped with 1/16” PFA tubing serving as a flow cell. This was hyphenated to the reactor within a thermostated bypass to maintain process conditions in the transfer lines.
Partial least squares regression (PLSR) models were established based on the initial spectra after activation of the reaction with syngas for the prediction of unknown concentrations of 1-dodecene and n-tridecanal over the course of the reaction in the lab-scale system. The obtained Raman spectra do not only contain information on the chemical composition but are further affected by the emulsion properties of the mixtures, which depend on the phase state and the type of micelles. Based on the spectral signature of both Raman and NMR spectra, it could be deduced that especially in reaction mixtures with high 1-dodecene content the formation of isomers as a competitive reaction was dominating. Similar trends were also observed during some of the process runs in the mini-plant. The multivariate calibration allowed for the estimation of reactants and products of the hydroformylation reaction in both laboratory setup and mini-plant.
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 Unionʼs Research Project CONSENS (Integrated Control and Sensing)
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.
Forderungen nach höherer Geschwindigkeit und die Komplexität der Fragen geben der analytischen Chemie neue Impulse. So nutzt die Prozessanalytik gepulste Ramanspektrometer, die Elemente Arsen und Quecksilber dominieren die Forschung in der Speziesanalytik, und die Omics-Techniken entwickeln sich zu Multi-Omics-Ansätzen. Ambiente MS-Techniken benötigen keine aufwendige Probenvorbereitung, multidimensionale Trenntechniken werden verstärkt in der Routine eingesetzt, und Chip-basierte Trennungen fallen durch Schnelligkeit auf. Molekülspektroskopie und Massenspektrometrie dominieren die bildgebenden Verfahren, und die Lateralauflösung der ToF-Sekundärionenmassenspektrometrie hat sich bei Oberflächenuntersuchungen verbessert.
Online NMR spectroscopy is an excellent tool to study complex reacting multicomponent mixtures and gain process insight and understanding. For online studies under process conditions, flow NMR probes can be used in a wide range of temperature and pressure. This paper compiles the most important aspects towards quantitative process NMR spectroscopy in complex multicomponent mixtures and provides examples. After NMR spectroscopy is introduced as an online method and for technical samples without sample preparation in deuterated solvents, influences of the residence time distribution, pre-magnetization, and cell design are discussed. NMR acquisition and processing parameters as well as data preparation methods are presented and the most practical data analysis strategies are introduced.
Benchtop nuclear magnetic resonance spectroscopy currently develops to an important analytical tool for both quality control and process monitoring. In contrast to high resolution online NMR (HR-NMR), benchtop NMR can be operated under rough environmental conditions. A continuous re-circulating stream of reaction mixture from the reaction vessel to the NMR spectrometer enables a non-invasive, volume integrating online analyses of reactants and products. Here we investigated the esterification of 2,2,2-trifluoroethanol with acetic acid to 2,2,2-trifluoroethyl acetate both by 1H HR-NMR (500 MHz) and 1H and 19F MR NMR (43 MHz and 40 MHz) as a model system. The parallel online measurement was realized by splitting the flow, which allowed the adjustment of quantitative and independent flow rates, both in the benchtop NMR probe as well as in the HR-NMR probe, in ad-dition to a fast bypass line back to the reactor.
Prozess-Sensoren 4.0 vereinfachen ihre Einbindung über Plug and Play, obwohl sie komplexer werden. Sie bieten Selbstdiagnose, Selbstkalibrierung und erleichterte Parametrierung. Über die Konnektivität ermöglichen die Prozess-Sensoren den Austausch ihrer Informationen als Cyber-physische Systeme mit anderen Prozess-Sensoren und im Netzwerk.
Der Wandel von der aktuellen Automation zum smarten Sensor ist im vollen Gange. Automatisierungstechnik und Informations- und Kommunikationstechnik (IKT) verschmelzen zunehmend. Eine Topologie für smarte Sensoren, die das Zusammenwirken mit daten- und modellbasierte Steuerungen bis hin zur Softsensorik beschreibt gibt es heute jedoch noch nicht. Wir müssen jetzt schnell die Weichen für eine smarte und sichere Kommunikationsarchitektur stellen, um zu einer störungsfreien Kommunikation aller Komponenten auf Basis eines einheitlichen Protokolls zu kommen. Unnötiges Schnickschnack ist nicht erwünscht. Wenn die Prozessindustrie dieses nicht definiert, tun es andere.
Für die weitere Entwicklung von der Ist‐Situation zu einer Industrie-4.0-Welt in der Prozessindustrie werden mehrere Szenarien diskutiert. Diese reichen vom erleichterten Abruf sensorbezogener Daten über zusätzliche Kommunikationskanäle zwischen Sensor und mobilen Endgeräten über vollständig bidirektionale Kommunikation bis hin zur Einbindung der Cloud und des Internets in virtualisierte Umgebungen.
Um zu einer störungsfreien Kommunikation aller Komponenten untereinander zu kommen, muss mindestens ein einheitliches Protokoll her, das alle sprechen und verstehen. Der derzeit greifbarste offengelegte Standard, der moderne Kommunikationsanforderungen erfüllt, ist OPC Unified Architecture (OPC-UA). Viele halten das Sortieren der Kommunikationsstandards für eines der wesentlichen Errungenschaften von Industrie 4.0.
Der Vortrag greift die Anforderungen der Technologie-Roadmap „Prozess-Sensoren 4.0“ auf und zeigt Möglichkeiten zu ihrer Realisierung am Beispiel eines Online-NMR-Analysators, der im Rahmen des EU-Projekts „CONSENS“ (www.consens-spire.eu) entwickelt wurde.
Aktuelle und zukünftige öffentliche Förderung von Industrie 4.0-Projekten sind eine gute Investition. Wegen der hohen Komplexität und Interdisziplinarität gelingt die Umsetzung nur gemeinsam zwischen Anwendern aus der Prozessindustrie, Software- und Geräteherstellern sowie Forschungsgruppen. Anwender sind gefragt, diese neue Technologie durch eine beschleunigte Validierung und Akzeptanz umzusetzen. Sie erhalten die einzigartige Chance, ihre Prozesse und Anlagen wettbewerbsfähig zu halten. Kooperativ betriebenen F&E-Zentren und gemeinsam anerkannten Applikationslaboren kommt dafür eine hohe Bedeutung zu.
The departure from the current automation landscape to next generation automation concepts for the process industry has already begun. Smart functions of sensors will simplify their use and enable plug-and-play integration, even though they may appear to be more complex at first sight. This is particularly important for concepts like self-diagnostics, self-calibration and self-configuration/parameterization. Intelligent field devices as parts of digital field networks, Inter-net Protocol (IP)-based connectivity and web interfaces, as well as advanced data analysis soft-ware will provide the basis for future projects like Industrie 4.0, Factory of the Future, or Industrial Internet of Things (IIoT). The talk summarizes the currently discussed general requirements for process sensors 4.0 and introduces an online NMR sensor as example. This sensor was developed to provide integrated control and sensing for sustainable operation of flexible intensified processes (CONSENS) funded by the European Union’s Horizon 2020 research and innovation programme.
Monitoring chemical reactions is the key to 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 cutting the calibration and validation needs to an minimum and thus exhibiting short set-up times. Compact NMR instruments make NMR spectroscopy accessible in industrial and harsh environments for advanced process monitoring and control.
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. Data analysis techniques are available but currently mostly used for off-line data analysis to detect the causes of variations in the product quality.
This is addressed within the EU’s Research Project CONSENS by the development and integration of a smart NMR module for process monitoring. The presented NMR module is provided in a mobile explosion proof housing and involves a compact spectrometer together with an acquisition unit and a programmable logic controller for automated data preparation (phasing, baseline correction), and evaluation. Such “smart sensors” provide the basis for the future project “Industrie 4.0”, and Industrial Internet of Things (IIoT), along with current requirements to process control, model based control, or soft sensing. The module transforms the acquired online spectra of various technically relevant reactions to either conventional 4‒20 mA signals as well as WiFi based OPC-UA communication protocols, which enables NMR-based advanced process control and funny discussions with plant managers along with automation and safety engineers.
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.
Der Vortrag stellt einige aktuelle Herausforderungen für die Prozessanalytik und mögliche Antworten vor.
Gepulste Raman-Spektrometer akkumulieren das Raman-Signal mit Hilfe schneller optischer Schalter im Picosekunden-Bereich, bevor langlebigere Fluoreszenzanregung entsteht. Damit lassen sich stark fluoreszierende Materialien untersuchen, die bislang nicht zugänglich sind. Eine weitere interessante Entwicklung ist etwa die Shifted excitation Raman difference spectroscopy (SERDS) die besonders für biologische Anwendungen interessant ist.
Flexible, modulare Produktionsanlagen stellen einen vielversprechenden Ansatz für die kontinuierliche Produktion von Fein- und Spezialchemikalien dar. In einem EU-Projekt wurde die Feldintegration eines Online-NMR-Sensormoduls als smartes Modul für die Prozesskontrolle vorangebracht. Dieses Modul basiert auf einem kommerziell erhältlichen Niederfeld-NMR-Spektrometer, welche zurzeit für die Anwendung im Laborbereich erhältlich ist. Für die Feldintegration wurde ein ATEX-zertifiziertes, explosionsgeschütztes Gehäuse entwickelt sowie Automationsschemen für den unbeaufsichtigten Betrieb und für die kalibrierfreie spektrale Datenauswertung erstellt.
Eine sehr gut anwendbare analytische Messtechnik zur Kontrolle der elementaren Zusammensetzung von verschiedensten Materialien ist die laserinduzierte Plasmaspektroskopie (LIPS, engl. LIBS - Laser-induced Breakdown Spectroscopy). Bei der LIBS wird ein kurz gepulster Laser auf eine Probe fokussiert, um ein Leuchtplasma zu erzeugen. Das dabei erzeugte Atomemissionsspektrum ermöglicht eine qualitative und quantitative Analyse der Zusammensetzung der Probe bezüglich praktisch aller Elemente des Periodensystems. In einem aktuellen Projekt wird diese Methode neben anderen zur Online-Analyse von Ackerböden für die ortsspezifischer Bewirtschaftung (Precision Agriculture) weiterentwickelt und bewertet.
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.
The talk introduces a smart online NMR sensor module provided in an explosion proof housing as example. This sensor was developed for an intensified industrial process (pharmaceutical lithiation reaction step) funded by the EU’s Horizon 2020 research and innovation programme (www.consens-spire.eu). 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.
The talk also generally covers current aspects of high-field and low-field online NMR spectroscopy for reaction monitoring and process control giving an overview from direct dissolution studies of API cocrystals to studies of emulsions for a hydroformylation.
qNMR provides the most universally applicable form of direct concentration or purity determination without need for reference materials of impurities or the calculation of response factors but only exhibiting suitable NMR properties. The workshop presents basic terms of statistics and uncertainty analysis, which are the basis for qNMR spectroscopy and data analysis such as, e.g., standard deviations, linear regression, significance tests, etc. and gives typical examples of applications in qNMR spectroscopy.
Um in einem veränderten Umfeld erfolgreich bestehen zu können, müssen Chemie-unternehmen neue Pfade beschreiten. Dazu gehört insbesondere das Potential digi-taler Technologien. Mit flexiblen, modularen chemischen Vielzweck-Produktionsanlagen lassen sich häufig wechselnde Produkte mit kürzeren Vorlauf- und Stillstandzeiten zwischen den Kampagnen und dennoch hoher Qualität realisie-ren. Intensivierte, kontinuierliche Produktionsanlagen erlauben auch den Umgang mit schwierig zu handhabenden Substanzen.
Grundvoraussetzung für solche Konzepte ist eine hochautomatisierte "chemische" Prozesskontrolle zusammen mit Echtzeit-Qualitätskotrolle, die "chemische" Informati-onen über den Prozess bereitstellt. In einem Anwendungsbeispiel wurde eine phar-mazeutische Lithiierungsreaktion aus einer modularen Pilot-Anlage betrachtet und dabei die Vorzüge eines vollautomatischen NMR-Sensors untersucht. Dazu wurde ein kommerziell erhältliches Benchtop-NMR-Spektrometer mit Permanentmagnet auf die industriellen Anforderungen, wie Explosionsschutz, Feldkommunikation und voll-automatischer, robuster Datenauswertung angepasst. Der NMR-Sensor konnte schließlich erfolgreich im vollautomatischen Betrieb nach fortschrittlichen Regelkon-zepten und für die Echtzeitoptimierung der Anlage getestet werden. Die NMR-Spektroskopie erwies sich als hervorragende Online-Methode und konnte zusammen mit einer modularen Datenauswertung sehr flexibel genutzt werden. Die Methode konnte überdies als zuverlässige Referenzmethode zur Kalibrierung konventioneller Online-Analytik eingesetzt werden.
Zukünftig werden voll integrierte und intelligent vernetzte "smarte" Sensoren und Pro-zesse eine kontinuierliche Produktion von Chemikalien und Pharmazeutika mit ver-tretbaren Qualitätskosten möglich machen.
There is a need within the NMR community to progress forward in exploring new facets in which we can use analytical techniques to advance our understanding of various systems. One aspect the NMR community hasn’t fully encompassed is the validation process, which also involves setting reference standards, establishing a common language that directly relates to NMR, communication relating to validation, and much more.
This workshopcontribution starts with an overview on international metrology for qNMR spectroscopy. Since NMR is completely described by mathematical equations, the measurement unceartainty can directly be dreived from formula. Examples are presented. These are differentiated between type A and B evaluations. Finally the Expanded Unceartainty is defined. Since the user needs a risk-based unceartainty assessment, different "leagues" for routine, advanced, and high level needs are proposed to make clear, that no all sources of uncertainty have to be taken in considerention at practical levels.
Bei Industry 4.0 dreht sich alles um Interkonnektivität, sensorgestützte Prozesssteuerung und datengesteuerte Systeme. Prozessanalysentechnik (PAT) wie die Online-Kernresonanzspektroskopie (NMR) gewinnt zunehmend an Bedeutung, da sie zur Automatisierung und Digitalisierung in der Produktion beiträgt. Eine klassische Auswertung von Prozessdaten und deren Umsetzung in Wissen ist jedoch bisher in vielen Fällen aufgrund der unzureichend großen verfügbaren Datensätze nicht möglich oder nicht wirtschaftlich. Bei der Entwicklung eines automatisierten Verfahrens für die Prozesskontrolle stehen manchmal nur die Basisdaten einer begrenzten Anzahl von Batch-Versuchen aus typischen Produkt- und Prozessentwicklungskampagnen zur Verfügung. Diese Datensätze sind jedoch nicht groß genug, um maschinengestützte Verfahren zu trainieren.
Um diese Einschränkung zu überwinden, wurde ein neues Verfahren entwickelt, das eine physikalisch motivierte Multiplikation der verfügbaren Referenzdaten erlaubt, um einen ausreichend großen Datensatz für das Training von maschinellen Lernalgorithmen zu erhalten. Das zugrundeliegende Beispiel einer chemischen Synthese wurde spektroskopisch verfolgt und mit der neuen Methode sowie mit einem physikalisch basierten Modell analysiert, wobei sowohl eine anwendungsrelevante Niederfeld-NMR als auch eine Hochfeld-NMR-Spektroskopie als Referenzmethode verwendet wurde.
Künstliche neuronale Netze (ANNs) haben das Potenzial, bereits aus relativ begrenzten Eingabedaten wertvolle Prozessinformationen abzuleiten. Um jedoch die Konzentration unter komplexen Bedingungen (viele Edukte und weite Konzentrationsbereiche) vorherzusagen, sind größere ANNs und damit ein größerer Trainingsdatensatz erforderlich. Wir zeigen, dass ein mäßig komplexes Problem mit vier Edukten unter Verwendung von ANNs in Kombination mit der vorgestellten PAT-Methode (Niederfeld-NMR-Spektroskopie) und mit dem vorgeschlagenen Ansatz zur Erzeugung aussagekräftiger Trainingsdaten bewältigt werden kann.