Data Analytics for Process
Optimization
Dr. Markus Roggen
Introduction
Complex Biotech Discovery Ventures
CBDV is a young research venture that seeks to add fundamental scientific
insight to the field of cannabis production.
We seek to support the cannabis industry by establishing a centralized
hub in Vancouver, BC, for collaborative research focused on:
• Process Design
• Process Optimization
• Process Analytics
• Formulation Research
Collaborative Research
CBDV collaborates with academic, industry and private groups around the
globe. Some highlights of those collaborations are:
• University of British Columbia, Vancouver
• Loyalist College, Belleville
• Via Innovations by Dr. Monica Vialpando
• Verdient Science by Dr. Linda Klumpers
• Fritsch Milling
• PerkinElmer
• Mettler Toledo
Fundamental Collaboration
Research Topics
• Chemometrics and data analytics for process control and optimization
• Kinetic studies to understand mechanisms
• In-process analytics for reaction control
• Computational studies to understand mechanisms
• Process development, like crystallization
Fundamental Cannabis Chemistry
Research Topics
• Chemometrics and data analytics for process control and optimization
• Kinetic studies to understand mechanisms
• In-process analytics for reaction control
• Computational studies to understand mechanisms
• Process development, like crystallization
Extraction -> Decarboxylation -> Isolation
Extraction Optimization: Visual Guide
Extraction Optimization: DoE
• Optimize more than extraction vessel
• What about separation columns
• Can we isolate or enrich?
• More factors, more variables!
EC2 C1C3
Extraction Optimization: DoE
SFE optimized for single separator
Cannabinoid Concentration
EC2 C1C3
Extraction Optimization: DoE
SFE optimized for single separator
Cannabinoid Concentration
Extraction Optimization: DoE
SFE optimized for single separator
Cannabinoid Yield
Extraction Optimization: Data Science
• Experiments are costly!
• 21 Extraction Runs
• 4 Factors (temp, pressure) studied
• 105 kg Cannabis used
• Can we utilize previous runs?
• No extra experiments
• >10 Factors tracked
• 0 kg Cannabis wasted
Extraction Data Analytics
• Big data analytics for cannabis extraction
Extraction Data Analytics
• Big data analytics for cannabis extraction
Different
Cultivars
Some grouping
for cultivars
Extraction Data Analytics
• Big data analytics for cannabis extraction
Change axis to
extraction speed
Pump is degrading
Cannabis: PCA
Factors to consider:
• Temperature
• Pressure
• Time
• InputWeight
• Input Concentration
• Input Moisture
• Flow Rate
• Compound Solubility
• Cultivar
• ExtractType
Extractor
Cannabis: Extractor Rating
• Different instruments => different architecture
• Same model => different performance
• Cannot be compared easily
Decarboxylation Observation
Not all Decarboxylations are equal
Don’t Decarboxylate to Long
Problems of excess heating:
• Availability of instruments
• Higher costs of production
• Side reaction and degradation
• Lower yields
0
10
20
30
40
50
60
70
80
0
0.5
1
1.5
2
2.5
3
3.5
0 1000 2000 3000 4000 5000 6000
THC
THC(%)
CBN&d8-THC(%)
d8-THC
CBN
Elapsed Time (Minutes)
Reaction Monitoring of Decarboxylation
Current
• Subjective determination of reaction completeness
• Reactionary approach
• Inconsistent batches
• Lack of quality & process control
vs. Optimal
• Rapid
• Simple
• Accurate
• Small sample volume
In-Process Analytics
Infrared spectroscopy is a useful tool for reaction monitoring.
BG62-64 T0
Name
Sample 023 By Administrator Date Thursday, July 12 2018
Description
1750 8001600 1400 1200 1000
0.23
-0.01
-0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.16
0.18
0.20
0.22
cm-1
A
_ Start THCA
20.87 %
End THCA
1.56 %
Monitoring THCA In-Process
0
5
10
15
20
25
30
35
0 10 20 30 40 50 60 70 80
THCA(%)
Elapsed Time (Minutes)
Monitoring THCA In-Process
0
5
10
15
20
25
30
35
0 10 20 30 40 50 60 70 80
THCA(%)
Elapsed Time (Minutes)
Computational Studies
Steric vs. Electronic: Exploring the Rate Difference in THCA and CBDA
Decarboxylation
https://doi.org/10.26434/chemrxiv.12909887.v1
Computational Studies
https://doi.org/10.26434/chemrxiv.12909887.v1
THC CBD
Computational Studies
https://doi.org/10.26434/chemrxiv.12909887.v1
THC CBD
Computational Studies
Key Findings:
• Rate determining step is the
intermolecular protonation
• Rate difference is due to steric rather
than electronic effects
https://doi.org/10.26434/chemrxiv.12909887.v1
Computational Studies
1: Work by Alex Siegel, presented at Emerald Conference 2020
• Δ9-THC isomerizes to Δ8-THC under heat or acid
• Δ10-THC and Δ6a,10a-THC have also been found1
• What otherTHC isomers are possible?
Post-Processing: Crystallization
• The process of solidifying atoms or molecules into highly
organized structures
• Commonly used in the hemp industry to purify CBD
• Current standard procedures use pentane or petroleum
ether
Post-Processing: Crystallization
Post-Processing: Crystallization
0.22
0.24
0.26
0.28
0.3
0.32
0.34
0.36
0.38
0.4
15 17 19 21 23 25 27 29
Concentration(g/mL)
Temperature (C)
Meta stable zone width for CBD in petroleum
ether
Sol. T (pure)
MSZT (pure)
0.22
0.24
0.26
0.28
0.3
0.32
0.34
0.36
0.38
0.4
15 17 19 21 23 25 27 29
Concentration(g/mL)
Temperature (C)
Meta stable zone width for CBD in petroleum
ether
Sol. T (pure)
MSZT (pure)
Sol. T (curde)
MSZT (crude)
Post-Processing: Crystallization
Collaboration and Help
• Advanced analytical testing
• Beyond compliant testing
Collaboration and Help
• Advanced analytical testing
• Beyond compliant testing
Collaboration and Help
• Advanced analytical testing
• Beyond compliant testing
• Coming soon: particle size, skin permeability & crystallization
Summary
• Track everything, utilize all data
• Be like the Oakland Athletics, be like Billy Beane
• Solve problems from multiple angles
• Strive to understand the fundamentals
• Collaborate, no one can solve it alone
Expertise
CEO: Dr. Markus Roggen
Dr. Roggen has been actively involved in the cannabis industry for over 5 years in executive
positions overseeing production, R&D and process optimization for multiple producers. Dr.
Roggen is also a trusted advisor and mentor for multiple startups, startup accelerators and
organizations.
Co-Founder: Prof. Glenn Sammis
Prof. Sammis is an Associate Professor in the Chemistry Department at the University of British
Columbia. He has built an internationally recognized research group working on the
development of novel synthetic methods for the preparation of natural products and
pharmaceuticals.
CBDV Team
Our team covers a wide range of expertise,
including analytical chemistry, process
chemistry, engineering physics, data science
and statistics.
Dr. Markus Roggen markus@cbdvl.com

Data Analytics for Process Optimization

  • 1.
    Data Analytics forProcess Optimization Dr. Markus Roggen
  • 2.
    Introduction Complex Biotech DiscoveryVentures CBDV is a young research venture that seeks to add fundamental scientific insight to the field of cannabis production. We seek to support the cannabis industry by establishing a centralized hub in Vancouver, BC, for collaborative research focused on: • Process Design • Process Optimization • Process Analytics • Formulation Research
  • 3.
    Collaborative Research CBDV collaborateswith academic, industry and private groups around the globe. Some highlights of those collaborations are: • University of British Columbia, Vancouver • Loyalist College, Belleville • Via Innovations by Dr. Monica Vialpando • Verdient Science by Dr. Linda Klumpers • Fritsch Milling • PerkinElmer • Mettler Toledo Fundamental Collaboration
  • 4.
    Research Topics • Chemometricsand data analytics for process control and optimization • Kinetic studies to understand mechanisms • In-process analytics for reaction control • Computational studies to understand mechanisms • Process development, like crystallization Fundamental Cannabis Chemistry
  • 5.
    Research Topics • Chemometricsand data analytics for process control and optimization • Kinetic studies to understand mechanisms • In-process analytics for reaction control • Computational studies to understand mechanisms • Process development, like crystallization Extraction -> Decarboxylation -> Isolation
  • 6.
  • 7.
    Extraction Optimization: DoE •Optimize more than extraction vessel • What about separation columns • Can we isolate or enrich? • More factors, more variables! EC2 C1C3
  • 8.
    Extraction Optimization: DoE SFEoptimized for single separator Cannabinoid Concentration EC2 C1C3
  • 9.
    Extraction Optimization: DoE SFEoptimized for single separator Cannabinoid Concentration
  • 10.
    Extraction Optimization: DoE SFEoptimized for single separator Cannabinoid Yield
  • 11.
    Extraction Optimization: DataScience • Experiments are costly! • 21 Extraction Runs • 4 Factors (temp, pressure) studied • 105 kg Cannabis used • Can we utilize previous runs? • No extra experiments • >10 Factors tracked • 0 kg Cannabis wasted
  • 12.
    Extraction Data Analytics •Big data analytics for cannabis extraction
  • 13.
    Extraction Data Analytics •Big data analytics for cannabis extraction Different Cultivars Some grouping for cultivars
  • 14.
    Extraction Data Analytics •Big data analytics for cannabis extraction Change axis to extraction speed Pump is degrading
  • 15.
    Cannabis: PCA Factors toconsider: • Temperature • Pressure • Time • InputWeight • Input Concentration • Input Moisture • Flow Rate • Compound Solubility • Cultivar • ExtractType
  • 16.
    Extractor Cannabis: Extractor Rating •Different instruments => different architecture • Same model => different performance • Cannot be compared easily
  • 17.
    Decarboxylation Observation Not allDecarboxylations are equal
  • 18.
    Don’t Decarboxylate toLong Problems of excess heating: • Availability of instruments • Higher costs of production • Side reaction and degradation • Lower yields 0 10 20 30 40 50 60 70 80 0 0.5 1 1.5 2 2.5 3 3.5 0 1000 2000 3000 4000 5000 6000 THC THC(%) CBN&d8-THC(%) d8-THC CBN Elapsed Time (Minutes)
  • 19.
    Reaction Monitoring ofDecarboxylation Current • Subjective determination of reaction completeness • Reactionary approach • Inconsistent batches • Lack of quality & process control vs. Optimal • Rapid • Simple • Accurate • Small sample volume
  • 20.
    In-Process Analytics Infrared spectroscopyis a useful tool for reaction monitoring. BG62-64 T0 Name Sample 023 By Administrator Date Thursday, July 12 2018 Description 1750 8001600 1400 1200 1000 0.23 -0.01 -0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18 0.20 0.22 cm-1 A _ Start THCA 20.87 % End THCA 1.56 %
  • 21.
    Monitoring THCA In-Process 0 5 10 15 20 25 30 35 010 20 30 40 50 60 70 80 THCA(%) Elapsed Time (Minutes)
  • 22.
    Monitoring THCA In-Process 0 5 10 15 20 25 30 35 010 20 30 40 50 60 70 80 THCA(%) Elapsed Time (Minutes)
  • 23.
    Computational Studies Steric vs.Electronic: Exploring the Rate Difference in THCA and CBDA Decarboxylation https://doi.org/10.26434/chemrxiv.12909887.v1
  • 24.
  • 25.
  • 26.
    Computational Studies Key Findings: •Rate determining step is the intermolecular protonation • Rate difference is due to steric rather than electronic effects https://doi.org/10.26434/chemrxiv.12909887.v1
  • 27.
    Computational Studies 1: Workby Alex Siegel, presented at Emerald Conference 2020 • Δ9-THC isomerizes to Δ8-THC under heat or acid • Δ10-THC and Δ6a,10a-THC have also been found1 • What otherTHC isomers are possible?
  • 28.
    Post-Processing: Crystallization • Theprocess of solidifying atoms or molecules into highly organized structures • Commonly used in the hemp industry to purify CBD • Current standard procedures use pentane or petroleum ether
  • 29.
  • 30.
    Post-Processing: Crystallization 0.22 0.24 0.26 0.28 0.3 0.32 0.34 0.36 0.38 0.4 15 1719 21 23 25 27 29 Concentration(g/mL) Temperature (C) Meta stable zone width for CBD in petroleum ether Sol. T (pure) MSZT (pure)
  • 31.
    0.22 0.24 0.26 0.28 0.3 0.32 0.34 0.36 0.38 0.4 15 17 1921 23 25 27 29 Concentration(g/mL) Temperature (C) Meta stable zone width for CBD in petroleum ether Sol. T (pure) MSZT (pure) Sol. T (curde) MSZT (crude) Post-Processing: Crystallization
  • 32.
    Collaboration and Help •Advanced analytical testing • Beyond compliant testing
  • 33.
    Collaboration and Help •Advanced analytical testing • Beyond compliant testing
  • 34.
    Collaboration and Help •Advanced analytical testing • Beyond compliant testing • Coming soon: particle size, skin permeability & crystallization
  • 35.
    Summary • Track everything,utilize all data • Be like the Oakland Athletics, be like Billy Beane • Solve problems from multiple angles • Strive to understand the fundamentals • Collaborate, no one can solve it alone
  • 36.
    Expertise CEO: Dr. MarkusRoggen Dr. Roggen has been actively involved in the cannabis industry for over 5 years in executive positions overseeing production, R&D and process optimization for multiple producers. Dr. Roggen is also a trusted advisor and mentor for multiple startups, startup accelerators and organizations. Co-Founder: Prof. Glenn Sammis Prof. Sammis is an Associate Professor in the Chemistry Department at the University of British Columbia. He has built an internationally recognized research group working on the development of novel synthetic methods for the preparation of natural products and pharmaceuticals. CBDV Team Our team covers a wide range of expertise, including analytical chemistry, process chemistry, engineering physics, data science and statistics.
  • 37.
    Dr. Markus Roggenmarkus@cbdvl.com