Filtern
Dokumenttyp
- Posterpräsentation (13) (entfernen)
Referierte Publikation
- nein (13)
Schlagworte
- Bioprocess (1)
- Chemometrics (1)
- Chemometrie (1)
- Mineralwolle (1)
- NIR (1)
- Nominal properties (1)
- Online NMR Spectroscopy (1)
- Qualitative RM (1)
- Raman spectroscopy (1)
Organisationseinheit der BAM
Qualitative reference materials (RM) cover a wide range of the overall RM market. Proficiency testing providers attract up to a thousand of participants in PT schemes purely oriented on qualitative results. The RM used for these kinds of PT are poorly regulated, nevertheless with a more and more general acceptance of accreditation in the field of RM production and PT provision, there is an ever increasing interest in assessing producers and providers according to rules already well accepted in the field of quantitative analysis.
The basic governing document, ISO 17034:2016, is written in a form that, at least for the overwhelming majority of requirements, may be applied to both qualitative and quantitative RM. However, problems remain. In particular, the expression of uncertainty of a purely qualitative result is still unresolved, and under discussion, the latter now lasting already dozens of years.
Some handles would be needed. In the poster, existing approaches and some pragmatic, new ways to tackle the problem are displayed and discussed.
Modular plants using intensified continuous processes represent an appealing concept to produce pharmaceuticals. It can improve quality, safety, sustainability, and profitability compared to batch processes, and it enables plug-and-produce reconfiguration for fast product changes. To facilitate this flexibility by real-time quality control, we developed a solution that can be adapted quickly to new processes and includes a compact Nuclear Magnetic Resonance (NMR) spectrometer for online quality monitoring as well as a new model-based control approach. The NMR sensor is a benchtop device enhanced to the requirements of automated chemical production including ro-bust evaluation of sensor data.
Here, we present alternatives for the quantitative determination of the analytes using modular, physically motivated models. These models can be adapted to new substances solely by the use of their corresponding pure component spectra, which can either be derived from experimental spectra as well as from quantum mechanical models or NMR predictors. Modular means that spec-tral models can simply be exchanged together with alternate reagents and products. Beyond that, we comprehensively calibrated an NIR spectrometer based on online NMR process data for the first time within an industrial plant. The integrated solution was developed for a metal organic reac-tion running on a commercial-scale modular pilot plant and it was tested under industrial conditions.
Spectroscopic methods possess important advantages for the observation of bioprocesses: they are robust, gain many data within a short time scale, and allow for the measurement of multiple relevant information at the same time. Next to Near-infrared spectroscopy, Raman spectroscopy is increasingly applied as a powerful online tool to monitor bioprocesses. However, Raman spectra may be obscured by straylight, cosmic rays and fluorescence. To achieve the best performance, optimal conditions for spectral quality and stability should be pursued. This comprises not only the choice of highly-sensitive equipment, but also the way of implementation into the process. Here we compared the potential of introducing a sterilized fiber optical Raman probe, and a flow-through cell into the bypass of a stirred tank reactor to monitor an autotrophic cultivation of Ralstonia eutropha where gasses are the energy and carbon sources. Raman spectra were continuously recorded over seven days and 22 samples were taken for the reference analytics of biomass, optical density, and protein. In addition, data for of phosphate and ammonium were available. The quantitative Evaluation of Raman spectra by Partial Least Squares Regression (PLSR) yielded comparable results for both approaches, however, there was a general trend of lower prediction errors with the immersion probe.
The better performance of the non-contact optic may be either related to a more representative interaction of the measurement optics with the probe or simply to the absence of probe-fouling, which was observed with the immersion optic in the interstice between window and metal frame.
Künstliche Mineralfasern (KMF) werden weltweit in vielfältiger Weise als Dämmstoffe verwendet. Jährlich fallen in Deutschland ungefähr 20 Millionen Tonnen KMF an, die nach dem Abriss auf Deponien entsorgt werden müssen. Grundsätzlich besteht jedoch die Möglichkeit der Rückführung dieser Materialien in den Produktionsprozess. Dafür ist es notwendig, anfallende KMF sowohl nach Material (Glaswolle, Steinwolle) als auch nach Alter zu sortieren.
Auf der Suche nach einer Methode, die es grundsätzlich gestattet eine schnelle Untersuchung vor Ort auf einer Baustelle durchzuführen, wurde die Nah-Infrarot (NIR) Spektroskopie unter Verwendung eines Prozessspektrometers mit Reflexionssonde getestet. Anhand von Laboruntersuchungen an 70 KMF Proben, die entweder betriebsneu waren oder bundesweit auf Baustellen gesammelt wurden, konnte eine NIR-basierte Methode zur Unterscheidung in Glas- und Steinwolle entwickelt werden. Mittels Partial Least Square Diskriminanzanalyse (PLS-DA) der NIR Spektren wurde basierend auf 40 zufällig ausgewählten Proben ein chemometrisches Modell kalibriert, das erfolgreich zur Klassifizierung der verbleibenden 30 Proben eingesetzt werden konnte. Unabhängig von Farbe, Dichte oder Alter der untersuchten KMF Proben war es möglich, eine 100% korrekte Zuordnung des Materials zu erzielen. Die nötigen Referenzwerte wurden unter Nutzung eines Röntgenfluoreszenzspektroskopischen Verfahrens über den Elementgehalt nach VDI 3492 ermittelt.
Insgesamt ist der experimentelle Aufwand für die Durchführung und Auswertung der NIR Messungen gering und manifestiert sich vor allem in der Kalibrierphase, die in diesem Fall die Untersuchung von 8-fach aliquotierten Probestücken der 70 KMF umfasste. Nur bei deutlich feuchtem Material muss ein zusätzlicher Trocknungs-schritt berücksichtigt werden. Das vorgestellte Verfahren wurde nachfolgend zu einem Patent angemeldet. Untersuchungen zu einer weiteren Unterscheidung der KMF nach Alter sind geplant.