Transcript Bewerbungsgespraech bei Henkel
PAT-enabling Hard- and Software
Andrey Bogomolov J&M GmbH, Germany
Outline
• • • • Good news • in 2007 one more person started to earn his leaving from doing chemometrics J&M: the company Process monitoring • • J&M special equipment for PAT two industrial cases PAT philosophy • • PAT chemometrics software concept components making up a PAT solution
J&M: Introduction
• • About J&M GmbH • • • • since 1987 spectroscopic equipment and components specifically, UV-Vis/NIR, DAD, fiber optics, microspectroscopy positioned as “fast and precise” recent years trend: spectroscopic equipment for PAT What kind of company?
• • • • • strongly PAT-oriented pro-chemometrics active, growing innovative open-minded and collaboration-friendly
J&M Products
TIDAS II NIR
(1100 – 2100 nm) Integrated in-situ
Dissolution Test System
8-channel parallel
Chromatography System NOVA
fully-integrated Process Analyser
TIDAS
fully-integrated spectrometer for pharmaceutical applications
PAT and beyond…
• • • Process Analytical Technology (PAT) • • • is an initiative of FDA to create a regulatory framework for innovative pharmaceutical development and quality assurance innovative methods [ of data analysis=chemometrics ] are allowed not only for research, but to provide the legal evidence ( e.g.
of product quality) for FDA the impact goes beyond the regulated environment, since any production is more or less bound with quality requirements The basic idea behind PAT is on-line process monitoring • • • • product quality assurance process control process understanding process optimisation PAT and Chemometrics • • chemometrics and spectroscopy are important components of PAT multivarite statistical process control (MSPC) replaces classical [ univariate ] statistics in the industry
PAT players and driving forces quality& safety price
producer
profit, image
regulative organisation
quality& safety
consumer
LightHouse Probe
TM
(LHP)
• • • Joint patent with GEA Pharma Systems Features and applications: • for on-line measurement of diffuse reflectance NIR-spectra • e.g.
granulation, drying, blending, coating processes • • • through 7 sapphire windows self-cleaning technology multiple applications moisture content, composition, homogeneity particle size monitiring process end-point detection etc.
Advantages (compared to surface-mounted probes): • reduced window fouling (cleaning) • • • enhanced contributing sample weight better sampling: 3D-effect => better performance
Case 1. Moisture monitoring
• • • Granulation and drying • • • • granulation is routinely carried out in the pharmaceutical production fluid-bed granulator drying is the last step proceeds to a sertain residual moisture content ( e.g.
2%) Moisture monitoring • • process end-point detection Loss on Drying (LOD) the direct off (at-) line technique PAT Classics • NIR monitoring of H 2 O is well established in-line technique
0 .4
0 .3
0 .2
0 .1
0 -0 .1
1 2 0 0 1 4 0 0 1 6 0 0 Wavelength (nm) 1 8 0 0 2 0 0 0
Case 1. Moisture monitoring (2)
r = 0.993
RMSECV = 0.58
a = 2 2 5 2 0 1 5 1 0 5 0 0 r = 0.996
RMSECV = 0.17
a = 1 (!) 1 0 2 0 3 0 4 0 5 0 time (min) 8 6 6 0 7 0 8 0 9 0 4 heating was interrupted 2 0 4 0 5 0 6 0 7 0 8 0
Case 2. Pellet Coating
• • • • • • Polymer coating of pellets for capsule filling PrecesionCoater TM • • patented in 1998 well-controlled particle flow due to the Swirl Accelerator Process conditions • constant heating; product temperature is controlled • coating is followed by drying stage Materials • pellet size up to 1100 µm • layer thickness up to 30-40 µm Critical parameters • layer thickness • moisture content Is it possible to predict the dissolution profile?!
Case 2. Pellet Coating (2)
1 0 -1 3 SNV-corrected spectra 2 -2 1 2 0 0 1 4 0 0 1 6 0 0 Wavelength (nm) 1 8 0 0 RMSECV = 4.05
a = 3 2 0 0 0 1 1 8 0 1 1 6 0 1 1 4 0 1 1 2 0 1 1 0 0 0 3 .5
3 moisture 2 .5
2 1 .5
1 0 .5
0 0 1 2 2 0 5 0 particle size 1 2 0 0 5 0 RMSEP = 0.12
a = 2 1 0 0 1 5 0 2 0 0 2 5 0 3 0 0 1 0 0 1 5 0 time (min) 2 0 0 predicted deviation calibration sample 2 5 0 3 0 0
ProDicht Project
• • • On-going J&M project funded till 2010 • • • • Reutlingen University (Prof. R. Kessler) University of Applied Sciences, Offenburg (Prof. B. Spangenberg) Light-Technical Institute, Karlsruhe Laser technique by Optimags Spectroscopic probe for liquids • transmission and reflection modes • • • tunable pathlength supports UV-Vis, NIR, IR, Raman, Imaging, … ability to combine multiple techniques Analysis in condense/ dispersive media • milk, creams, grease • fat content
Additional wash tube Mirror Probe window Wash tubes Movable Tip
Technically, PAT is…
• • •
PAT = Hardware + Software
PAT hardware • sensors • • analytical [ spectroscopic ] instrumentation computers PAT software • instrument control software (data acquisition) • data processing and modelling (chemometric packages) • • • operating software (process visualisation, alarms, end-point detection) information/knowledge management software (LIMS)
PAT software !?
connect all hardware and aquire the data synchronise the data acquisition build and connect models predict, observe and operate process document the process
Data handling software
• Levels of data handling • data acquisition • • • instrument control, visualisation FFT, averaging data [ pre ]processing baseline correction, smoothing, derivation, subtraction, normalisation, peak picking, integration, transforms etc.
[ advanced ] data analysis modelling decision making (chemometrics) information management LIMS, ERP, …
manage analyse
x=!
process acquire
Towards the PAT Software
• •
Concept elements
• process as the main object • PAT solution as an output “universal process modelling language”?
• • • • modular [ open ] toolkit PAT platform expert (model building) and operating (routine application) level should be clearly separated new principles of data handling delivering the document on-line “web-compatible” remote information sources web-based solutions
Software becomes a dominating part of the data analytical system
PAT Solution
• PAT company • • • • software development linked to hardware vendors scientific expetise
independent Technical companies hardware science Consulting companies Software companies Academia
Welcome to Collaboration
• • • Scientific expertise • experimental design • • • • sampling image analysis new algorithms uncertainty estimation Practical applications • industrial process control • environmental monitoring • • quality control … Joint development • PAT hardware • PAT software • integrated systems
The end
Thank you for attention!