Noninvasive Study of Powder Blending Using NIR Spectrometry and Acoustic Emission

1
Non-invasive Study of Powder Blending Using NIR
Spectrometry and Acoustic Emission
L.J. Bellamy, A. Nordon, D. Littlejohn CPACT,
University of Strathclyde, Glasgow, G1 1XL
Luke Bellamy CPACT Strathclyde
2
Ideal Mixing Profile
Composition Variance
Mixing Time
Luke Bellamy CPACT Strathclyde
3
Experimental Plan

• Binary systems have been mixed to determine the
  effect of the physical properties of the
  particles and mixer parameters on mixing

• Microcrystalline cellulose as bulk material
• Second components chosen to display variation in
  key physical properties
• Mixing monitored with non-invasive techniques

Luke Bellamy CPACT Strathclyde
4
Selected Materials
Luke Bellamy CPACT Strathclyde
5
Laboratory Scale Blending

• Vessel and impeller scaled down from process unit
  used at GlaxoSmithKline, Ware,UK
• Glass upper section allows non-invasive optical
  measurements
• Mixing speed ranges from 0-125 rpm
• Contained in cabinet with extraction

Luke Bellamy CPACT Strathclyde
6
Near Infrared Spectrometry

• Zeiss Corona 45 NIR
• Process instrument designed for non-invasive
  reflectance measurements
• 128 element InGaAs array detector
• Operated in absorbance mode
• Spectra acquired every 0.5 seconds, 10
  co-added scans with 32 ms integration
• Reference was reflective white paper inside the
  mixing vessel

Luke Bellamy CPACT Strathclyde
7
Zeiss Corona 45 NIR
Aspect software used to acquire spectra through
PC link
Luke Bellamy CPACT Strathclyde
8
Passive Acoustic Emission Spectrometry

• Agilent Infiniium oscilloscope
• Broadband transducer 150 750 kHz
• Signal acquired with 2 mHz sample rate and 8 k
  points every 2 seconds
• Data imported into Matlab and signal, s,
  converted to power spectra using

Luke Bellamy CPACT Strathclyde
9
Broadband transducer Linked directly to
oscilloscope
Agilent Infiniium oscilloscope Data acquired in
C-program using GPIB link
Luke Bellamy CPACT Strathclyde
10
Mixing Procedure

• Acoustic transducer coupled to glass
• Mass of avicel added to mixer
• NIR instrument positioned 13 mm from glass
• Second compound added after 300 s through sliding
  window in top of cabinet
• Mixing initially monitored for 7200 s to test
  stability of the mixed powder to segregation
• Mixing generally stopped after 1200 s

Luke Bellamy CPACT Strathclyde
11
Corona must be positioned 13 mm from the vessel
for optimum performance
Transducer coupled with silicone sealant and held
in place with tape
Luke Bellamy CPACT Strathclyde
12
First Derivative Spectra Mixing Avicel and
Aspirin
8956 cm-1
6086 cm-1
Luke Bellamy CPACT Strathclyde
13
First Derivative NIR Absorbance at 8956 cm-1
Averaged mixing profile for mixing avicel and
aspirin
Luke Bellamy CPACT Strathclyde
14
NIR Mixing Profiles at 8956 cm-1
Aspirin concentration varied at 75 g avicel fill
and 50 rpm
Luke Bellamy CPACT Strathclyde
15
Monitor Mixing Using RSD
Mixing 10 g aspirin into 75 g avicel at 50 rpm
Monitor RSD of NIR signal at 8956 cm-1
Luke Bellamy CPACT Strathclyde
16
Summed Passive Acoustic Power Spectra
75 g avicel fill with 10 g aspirin at 50 rpm
Luke Bellamy CPACT Strathclyde
17
Acoustic Mixing Profile
75 g avicel fill with 10 g aspirin at 50 rpm
18
Mixing Experiments

• Vessel properties investigated
• - Fill level (65, 75, 85 and 90 g avicel)
• - Impeller speed (25, 50, 75,100 and 125 rpm)
• Particle properties investigated
• - Shape (spherical, needles, granular, cubic)
• - Concentration (5, 10, 20 and 25 g aspirin)
• - Particle size (lt251, 251-500 and 500-853 µm)
• - Density (1.1 2.2 g/cm-3)

Luke Bellamy CPACT Strathclyde
19
Effect of Density and Shape in NIR Profile
Povidone 30 Spherical 1.1 g cm-3
Povidone 90 Platelets 1.1 g cm-3
Potassium chloride cubic 2.2 g cm-3
Aspartame Needles 1.35 g cm-3
Luke Bellamy CPACT Strathclyde
20
Effect of Particle Size - Citric Acid (static)
Citric acid NIR spectra measured through glass
beaker
Luke Bellamy CPACT Strathclyde
21
Effect of Particle Size - Citric acid (mixing)
75 g avicel fill with 7.5 g citric acid at 50
rpm, NIR
Luke Bellamy CPACT Strathclyde
22
Effect of Size at Different Concentrations NIR
Monitored at 6814 cm-1
75 g avicel fill, three size ranges, 50 rpm
Luke Bellamy CPACT Strathclyde
23
Effect of Size at Different Concentrations
Acoustics
75 g avicel fill, three size ranges, 50 rpm
Luke Bellamy CPACT Strathclyde
24
Effect of Impeller Speed and Shape NIR
75 g avicel fill, 10g addition
25
Selected Materials
Luke Bellamy CPACT Strathclyde
26
Effect of Impeller Speed Acoustics
75 g avicel fill, 10g addition
27
Conclusions

• Both NIR and acoustic spectrometry can be used to
  monitor mixing processes non-invasively
• NIR and passive acoustics are sensitive to
  particle size and concentration
• Some regions of NIR spectra more sensitive to
  particle size variations
• Shape and density seem to give variations in NIR
  mixing profile and mix at different rates
• Impeller speed variations may lead to composition
  changes from NIR data

Luke Bellamy CPACT Strathclyde
28
Future Work

• Mix multi-component systems
• Tumble blender
• Use pre-amps and filters to optimise acoustic
  signal detection
• Fluorescence spectrometry will be investigated in
  near future

Luke Bellamy CPACT Strathclyde
29
Acknowledgements

• Dr David Rudd and Dr Paul Frake, GlaxoSmithKline,
  Ware, UK
• Centre for Ultrasonic Engineering (CUE)
• CPACT and University of Strathclyde

Luke Bellamy CPACT Strathclyde