Cyberinfrastructure-enabled Molecular Products Design and - - PowerPoint PPT Presentation

SLIDE 1

1

Cyberinfrastructure-enabled Molecular Products Design and Engineering: Challenges and Opportunities

Venkat Venkatasubramanian

Laboratory for Intelligent Process Systems Purdue University West Lafayette, IN, USA

SLIDE 2

Venkat Venkatasubramanian, Keynote Lecture, European Congress in Chem. Engg, Copenhagen, Sept 2007

2

LIPS

Outline

Molecular Products Design

Data, Information and Knowledge Modeling

Challenges

Cyberinfrastructure

Ontological Informatics

Industrial Case Studies

Lubrizol: Fuel Additives Design Caterpillar: Rubber Products Design ExxonMobil: Catalyst Design Eli Lilly: Seromycin Formulation for MDR-TB

Summary

SLIDE 3

Venkat Venkatasubramanian, Plenary Lecture, Foundations of Computer-Aided Process Design, Princeton, July 2004.

Acknowledgements

• Prof. J. F. Caruthers and Prof. W. N. Delgass
• Dr. K. Chan (Bear Stearns), Dr. A. Sundaram (ExxonMobil), Dr. P.

Ghosh (ExxonMobil), Dr. S. Katare (Dow), Dr. A. Sirkar (Intel), Dr.

• P. Patkar (ZS), Dr. S-H. Hsu (Purdue), S. Syal (Purdue), B. K.

Balachandra (Purdue)

Funding Agencies

• NSF
• DOE
• Indiana 21st Century Research and Technology Fund
• GE, Bangalore, India
• Lubrizol
• Caterpillar
• ExxonMobil
• Equistar
• Cummins

Materials Design

SLIDE 4

Venkat Venkatasubramanian, Keynote Lecture, European Congress in Chem. Engg, Copenhagen, Sept 2007

4

LIPS

Acknowledgements

• Dr. C. Zhao (Bayer)
• Dr. A. Jain (United)
• Dr. S-H. Hsu (Purdue)
• L. Hailemariam (Purdue)
• P. Akkisetty (Purdue)
• P. Sureshbabu (Purdue)
• I. Hamdan (Purdue)
• M. Chaffee (Purdue)
• A. Shukla (Purdue)

Funding Agencies: NSF ERC Indiana 21st Century R&T Fund Eli Lilly Pfizer Abbott

• Dr. Prabir Basu
• Dr. Girish Joglekar
• Prof. Ken Morris
• Prof. G. V. Reklaitis

Information Technology@Purdue

Pharmaceutical Engineering

SLIDE 5

Venkat Venkatasubramanian, Plenary Lecture, Foundations of Computer-Aided Process Design, Princeton, July 2004.

LUBRIZOL: Fuel Additive Design

• EPA requirement: Minimize

intake-valve deposits (IVD)

• Approach: Fuel Additives
• Performance measure
• BMW Test for IVD
• Stipulated to be less than 100 mg
• ver a 10,000 mile road test
• Expensive and time-consuming

testing

• Around $8000 for a single datum
• Problem: Design fuel-additives

that meet desired I VD performance levels

Intake Valve and Manifold

SLIDE 6

Venkat Venkatasubramanian, Plenary Lecture, Foundations of Computer-Aided Process Design, Princeton, July 2004.

About 1000 Rubber Parts in Failure Critical Functions Tires, Treads, Hoses, Shock Absorbers, O- rings, Gaskets, Mounts … Reliability and Warranty Problems

CATERPILLAR

SLIDE 7

Venkat Venkatasubramanian, Plenary Lecture, Foundations of Computer-Aided Process Design, Princeton, July 2004. Quantum Chemistry

˜ T = 2ρR ∂ ψ ∂I 1 + I 1 ∂ ψ ∂I 2 ⎧ ⎨ ⎩ ⎫ ⎬ ⎭ I − ∂ ψ ∂I 2 C + I 3 ∂ ψ ∂I 3 C−1 ⎡ ⎣ ⎢ ⎤ ⎦ ⎥

• ∂ C

∂C

Constitutive Model 3D Finite Element Part Model Spring-Dashpot Approximation with Chemical Aging

A

x + S 8 k1

⎯ → ⎯ ⎯ A

x+8

Ax + R

k2

⎯ → ⎯ ⎯ Bx B

x k3

⎯ → ⎯ ⎯ Bx

*

etc . Time To r q u e

Cure Curve

Kinetics Polymer Network Mechanical Response

Mixing + Heat T ransfer + Mold Filling

Part Manufacturing

Overcure Undercure S C N S S S S S S S S Zn

New Molecules Material Formulation Manufac- turing Process Part Shape Design Sub Assembly Design Full Product Design & Operation

Design & Operation of Complex Sub-Assembly

Engine

Over 1000 rubber parts in failure critical functions

Multi-Scale Modeling

Quantum Mechanics Kinetics Heat Transfer FEA Mechanics FEA Part Design CAD /CAM

Caterpillar’s Multi-Scale Modeling Challenge: Design of Formulated Rubber Parts

SLIDE 8

Venkat Venkatasubramanian, Plenary Lecture, Foundations of Computer-Aided Process Design, Princeton, July 2004.

ExxonMobil’s Grand Challenge: Catalyst Design via Combinatorial Chemistry

Fundamental Chemistry Reaction kinetics Performance measures

time Rate/Selectivity

C

k6

A B F

k1

H D R M E G N

k3 k

4

k2 k7 k5

Reaction network Electronic/Structural descriptors

d band center Electronegativity

. . .

Quantum calculations Catalyst Structures

Synthesis Characterization

SLIDE 9

10/ 11/ 2005 9

Pharmaceutical Product Development and Engineering

• Two Products

Drugs Documents

Preliminary product recipe Drug is converted into Particles Drug Synthesis Raw Chemicals Formulation Process Development Preliminary process recipe Scale/up -Tech Transfer Adjusted process recipe Manufacturing A Delayed Product

SLIDE 10

Venkat Venkatasubramanian, Plenary Lecture, Foundations of Computer-Aided Process Design, Princeton, July 2004.

Molecular Products Design

Given a set of desired performance specifications, how does

• ne rationally and efficiently identify the optimal product

structures and formulations?

$200+ Billion/yr industry: Major Opportunity for ChEs

• Areas of Application
• Engineering Materials

Fuel and Oil Additives Polymer Composites Rubber Compounds Catalysts Solvents, Paints and Varnishes…..

• Pharmaceuticals

Drug Design

• Agricultural Chemicals

Pesticides Insecticides

Zeolite!!

SLIDE 11

Venkat Venkatasubramanian, Plenary Lecture, Foundations of Computer-Aided Process Design, Princeton, July 2004.

Different Products, But Common Challenges

Lots of Data Uncertain/Noisy Data Complex chemistry, lots of species Nonlinear systems and processes Incomplete, Uncertain, Mechanisms Combinatorially large search spaces Need Multi-scale Models Hundreds of Differential and Algebraic Equations to

formulate and solve

Laborious and time consuming model development

process

Limited human expertise

SLIDE 12

Venkat Venkatasubramanian, Plenary Lecture, Foundations of Computer-Aided Process Design, Princeton, July 2004.

Traditional Approach

Revise Design No

Designer

Hypothetical Molecule Synthesize Evaluate Design Objectives Yes Solution

Meets Design Objectives ?

Drawback: Protracted and Expensive Design Cycle Need a Rational, Automated Approach: Discovery Informatics

EUREKA!!!

SLIDE 13

Venkat Venkatasubramanian, Keynote Lecture, European Congress in Chem. Engg, Copenhagen, Sept 2007

13

LIPS

Need a New Paradigm

Cyberinfrastructure Methods and

Tools

Ontology-based Discovery

Informatics

SLIDE 14

Venkat Venkatasubramanian, Plenary Lecture, Chemical Process Control 7, Canada, Jan 2006

14

LIPS

Data, Information, and Knowledge

Data

What’s going on?

Raw, does not inform much

Information

How are the variables related?

Correlations and relationships

Knowledge

Why are they related?

Develop Mechanistic understanding First-Principles Based Math Models Heuristics

Tower of Babel of Data, Information and Models

SLIDE 15

Venkat Venkatasubramanian, Plenary Lecture, Pharmaceutical Sciences World Congress, Amsterdam, April 2007

15

LIPS

What is an Ontology?

Originated in Philosophy

Study of Existence Not to be confused with Epistemology: Theory of Knowledge

and Knowing

Computer Science and Artificial Intelligence

A formal, explicit specification of a shared conceptualization Knowledge representation in AI

Logic, Semantic Networks, Frames, Objects, Ontology

Web-driven development

Semantic Richness: Description Logic (DL) provides

foundation for formal reasoning

Easily create, share and reuse knowledge Semantic Search

SLIDE 16

Venkat Venkatasubramanian, Plenary Lecture, Pharmaceutical Sciences World Congress, Amsterdam, Apr 2007

16

LIPS

Based on Set Theory and First-order Logic

An ontology is a 5-tuple ) , , , ,

• c

A rel H R C ( := Ο consisting of

• Two disjoint sets C and R whose elements are called concept identifiers and

relation identifiers, respectively

• A concept hierarchy

C

H :

C

H is a directed, transitive relation C C H C × ⊆ which is called concept hierarchy or taxonomy. ) , (

2 1 C

C H C means that

1

C is a subconcept of

2

C .

• A function

C C R rel × → : that relates concepts non-taxonomically. The function dom: C R → with )) ( ( : ) (

1

R rel R dom Π = gives the domain of R, and range: C R → with )) ( ( : ) (

2

R rel R range Π = give its range. For ) , ( ) (

2 1 C

C R rel = we also write ) , (

2 1 C

C R .

• A set of axioms
• A , expressed in an appropriate logic language, e.g. first order

logic

SLIDE 17

17

Ontology: A Simple Example

Excipient Properties Material API Is a Is a has Concept Instance Property Roles Composition Range has has Flow aid Filler Lubricant Minimum Maximum Average Is an instance of has Contact angle Moisture content Is a Material Composition has

Web Ontology Language: OWL

SLIDE 18

Venkat Venkatasubramanian, Plenary Lecture, Foundations of Computer-Aided Process Design, Princeton, July 2004.

Catalyst Design Problem – The Grand Challenge

Fundamental Chemistry Reaction kinetics Performance measures

time Rate/Selectivity

C

k6

A B F

k1

H D R M E G N

k3 k

4

k2 k7 k5

Reaction network Electronic/Structural descriptors

d band center Electronegativity

. . .

Quantum calculations Catalyst Structures

Synthesis Characterization

SLIDE 19

Venkat Venkatasubramanian, Plenary Lecture, Foundations of Computer-Aided Process Design, Princeton, July 2004.

1 A° 100 A° 10 μm 1 cm 10-15 s 10-12 s 10-6 s 1 s 102 s

Quantum Mechanics

DFT

Atomistic Simulation

• Molec. Dyn.

Statistical Mechanics

TST

Continuum Models

Transport Models Elect. Struct. Thermo. Props. Molec. Struct. Rate Consts. Overall Rxn Rates Conc. Profiles

Microscopic Mesoscopic Macroscopic

Modeling at different Length and Time Scales

SLIDE 20

Venkat Venkatasubramanian, Plenary Lecture, Foundations of Computer-Aided Process Design, Princeton, July 2004.

Paraffin Aromatization

Identify a catalyst formulation for light

paraffin aromatization that is superior to Ga/H-ZSM-5 in terms of:

Higher Benzene, Toluene, Xylene

(B/T/X) selectivity

Higher Hydrogen selectivity

Microkinetic model development for the

kinetic description of the system

SLIDE 21

Venkat Venkatasubramanian, Plenary Lecture, Foundations of Computer-Aided Process Design, Princeton, July 2004.

Previous Work

Microkinetic Analysis (Dumesic, 1993)

First unified view of reaction engineering on catalytic surfaces Incomplete forward problem Lacks clear view of the design perspective (inverse problem)

Empirical Catalyst Design (Baerns, 2000)

First attempt to “design” catalysts Completely based on “guided” experiments – time consuming and expensive No fundamental understanding of the system (forward problem) Ertl, Rasmussen, Lauterbach: Surface Science Studies Steve Jaffe: Composition based modeling of large systems Jens Norskov: Computation based studies Symyx, Novodynamics, Lauterbach: Combinatorial HTE

SLIDE 22

Venkat Venkatasubramanian, Plenary Lecture, Foundations of Computer-Aided Process Design, Princeton, July 2004.

Reaction Network – Propane Aromatization on HZSM-5

• 31 gas phase species + 29 surface species + 271 reaction steps
• Model with 31 ODEs, 29 algebraic equations
• 13 parameters with up to 10 orders of magnitude bounds on each

Hydride transfer

Cx-y

=

β-scission/Oligomerization

Cx

+

Cy

+

H2 Cx-y

D e h y d r

• g

e n a t i

• n

P r

• t
• l

y s i s

Cx

=

adsorption/desorption adsorption/desorption

Cy

=

H+ CxH+ Cx

i n i t i a t i

• n

Paraffin

Dehydro- cyclization

Aromatics

Cx Paraffin Cx

= Olefin

Px

=

Poly-olefin Cx

=c Cyclic-olefin

Active site

H+

SLIDE 23

Venkat Venkatasubramanian, Plenary Lecture, Foundations of Computer-Aided Process Design, Princeton, July 2004.

Ontological Informatics: Modeling Super Highway

• DAE System: 30 to 1000 species and 10 to 50 parameters
• Parameter search space is highly nonlinear with numerous local minima

GA based pseudo global parameter estimator Fast and efficient coverage of the parameter space

• Optimization: Traditional least-squares and features-based objectives

Chemistry Compiler

Reaction Description Language Plus (RDL+)

Equation Generator

Automatic generation of differential algebraic equations (DAEs)

Parameter Optimizer

Solution of DAEs Least squares Features

Advanced Statistical Analyzer

Sensitivity Analysis Uncertainty Analysis Error Propagation Reactions Grouping

Chemistry Rules

time Rate/Selectivity

Performance Curves Data

J.M. Caruthers, J.A. Lauterbach, K.T. Thomson, V. Venkatasubramanian, C.M. Snively, A. Bhan, S. Katare and G. Oskarsdottir, “Catalyst Design: Knowledge Extraction from High Throughput Experimentation”, In “Understanding Catalysis from a Fundamental Perspective: Past, Present, and Future”, A. Bell, M. Che and W.N. Delgass, Eds., Invited Paper for the 40th Anniversary Issue of the Journal of Catalysis, 2003. Katare, S., Caruthers, J. M., Delgass, W. N., Venkatasubramanian, V., “An Intelligent System for Reaction Kinetic Modeling and Catalyst Design”, Ind. Eng. Chem. Res. and Dev., 2004.

SLIDE 24

Model Refinement Procedure

Crude models Initial data

Model screening

Infer new knowledge about models (mechanism) and data

Model discrimination Model enhancement

Enhanced models Richer data set

Planning new experiments

Add/Delete models Add/Delete data

Model analysis Data analysis

SLIDE 25

Chemistry Rules for Propane Aromatization on HZSM-5

Chemistry Rules Representative Chemical Reactions

• 1. Alkane adsorption
• 2. Alkane desorption
• 3. Carbonium ion protolysis
• 4. Carbonium ion dehydrogenation
• 5. Olefin adsorption
• 6. Olefin desorption
• 7. Aromatization
• 8. Beta-Scission
• 9. Hydride Transfer
• 10. Oligomerization

CH4 + + H2 + + + +

SLIDE 26

Venkat Venkatasubramanian, Plenary Lecture, Foundations of Computer-Aided Process Design, Princeton, July 2004.

Modeling Super Highway: Reaction Modeling Suite (RMS)

Reaction Network Reaction Description Language Plus English Language Rules

C

k6

A B F

k1

H D R M E G N

k3 k

4

k2 k7 k5

Mathematical Equations

1 / /

5 4 1 1

= + + − + = − =

C B A A B A A

B k D k C k dt dC C k dt dC θ θ θ

100’s of DAE’s

. . .

Model Generator

Beta Scission

Label-site c1+ (find positive carbon) Label-site c2 (find neutral-carbon attached-to c1+) Label-site c3 (find neutral-carbon attached-to c2) Forbid (primary c3) Forbid (less-than (size-of reactant) 9) Disconnect c2 c3) Increase-order-of (find bond connecting c1+ c2) Add-charge c3 Subtract-charge c1+

Beta Scission

Label-site c1+ (find positive carbon) Require (c1+ primary and product) set-k k1 Label-site c2+ (find positive carbon) Require (c2+ secondary and product) set-k 20*k1 Label-site c3+ (find positive carbon) Require (c2+ tertiary and product) set-k 60*k1

Chemistry

• 8. Beta Scission

transforms a carbenium ion into a smaller carbenium ion and an olefin

Grouping

• 8. a. Formation of a secondary carbenium ion

is 20 times faster than a primary carbenium ion

• b. Formation of a tertiary carbenium ion

is 60 times faster than a primary carbenium ion

. . . . . . . . . . . .

SLIDE 27

Venkat Venkatasubramanian, Plenary Lecture, Foundations of Computer-Aided Process Design, Princeton, July 2004.

Results Over prediction of C2s and under prediction of aromatics

Space time x 104 [hr] Wt %

C2H6 C3H6 C3H8 Aromatics

Data*

* Lukyanov et. al., Ind. Eng. Chem. Res. , 34, 516-523, 1995

Model

SLIDE 28

Venkat Venkatasubramanian, Plenary Lecture, Foundations of Computer-Aided Process Design, Princeton, July 2004.

Results – Model Refinement

Refined model adds alkylation step to convert lighter alkanes to higher ones

Space time x 104 (hr) Wt %

C2H6 C3H6 C3H8 Aromatics

Data

Original Model Refined Model

SLIDE 29

GA Based Pseudo Global Parameter Estimator

S.no. Name Esposito & Floudas (2000) Global optimizer GA based procedure υ Objective CPU s reported min max υ Objective Time taken

(CPU s)

Scaled CPU s Time taken *671/211 (% Saving)

1 First order irreversible chain reaction 5.0035 1.0000 1.18584x10-6 2.92 20.05 5.0122 0.9976 9.25418x10-6 0.24 0.76 (74) 2a First order reversible chain reaction 4.0000 2.0000 40.013 20.007 1.8897x10-7 568.44 6164.3 3.9813 1.9759 39.787 19.887 8.1313x10-6 0.45 1.43 (100) 2b Same as 2a but with error in data 4.021 2.052 39.45 19.62 1.586x10-3 272.91 1899.82 4.001 2.027 39.22 19.50 1.589x10-3 0.5 1.59 (99) 3 Catalytic cracking of gas

• il

12.214 7.9798 2.2216 2.65567x10-3 79.77 1185 12.246 7.9614 2.2351 2.68017x10-3 0.38 1.21 (98) 4 Bellman’s problem 4.5704x10-6 2.7845x10-4 22.03094 36.16 12222 4.5815x10-6 2.7899x10-4 22.2885 3.54 11.26 (69) 5 Methanol-to- hydrocarbon process 5.1981 1.2112 0.10652 1361.8 19125 5.2212 1.2320 1.6363x10-31 4.0059x10-12 0.004665 0.10586 2.08 6.61 (100) 6 Lotka-Volterra Problem 3.2434 0.9209 1.24924x10-3 367.26 9689.67 3.1434 0.9583 2.39784x10-3 0.56 1.78 (100)

Our zeolite model 31 ODEs, 29 Algebraic equations 13 parameters 9 concentration curves Performance comparison on test problems – worst case: 3 ODEs/5 parameters Reasonable comparison to experimental data 2 hours on a Sun 400 MHz solaris machine

SLIDE 30

Venkat Venkatasubramanian, Plenary Lecture, Foundations of Computer-Aided Process Design, Princeton, July 2004.

Ontological Informatics for Catalyst Design: Summary

Chem istry Rules High Throughput Experim ental Data

Reaction Know ledge Base Kinetic Modeling Toolkit ( Reaction Modeling Suites)

Chem istry Rules High Throughput Experim ental Data

Reaction Know ledge Base Kinetic Modeling Toolkit ( Reaction Modeling Suites) Represents reaction

chemistries explicitly

Customizable for different

catalyst chemistries

Results Real-time evaluation of

100s of reaction pathways and 1000s of DAEs

Saved months of model

development effort

Improved mechanistic

understanding and first principles models

Molecule Ontology Reaction Mechanism Ontology Reaction Ontology Chemistry Knowledge Base Molecule Classification Reaction Classification Reaction Network Generator Correctness and Consistency Checking Proposed Mechanism Graphical User Interface

Chemistry Ontology

Application Ontology

Molecule Ontology Reaction Mechanism Ontology Reaction Ontology Molecule Ontology Reaction Mechanism Ontology Reaction Ontology Chemistry Knowledge Base Molecule Classification Reaction Classification Reaction Network Generator Correctness and Consistency Checking Proposed Mechanism Graphical User Interface

Chemistry Ontology

Application Ontology

Application

SLIDE 31

Proposed Framework for Knowledge Extraction from HTE – Kinetic Modeling

Reaction Modeling Suite

Automatic processing of data

Reaction Modeling Suite

Automatic processing of data

Model Refinement

Rules Features mapping Suggestions for location of new rules

New HTE for Model Discrimination via Experimentation

Experiments in new composition regimes Measure critical variables identified

High Throughput Experiments

time Rate/Selectivity

Performance Curves

FTIR GC

. . . Chemistry Rules

Reactions Lumping

Chemistry Rules

Reactions Lumping

Feature Extractor/ Analyzer Feature Extractor/ Analyzer

SLIDE 32

Inverse Model Genetic Algorithms

Selection Recombination

Catalyst Library

HTE

Target Catalyst

Model Revision

Compare Performance

Catalyst Design Challenge

Target Catalyst Performance

time Rate/Selectivity

Forward Model Hybrid Model

Physical Model Statistics/Neural-Nets

A + S A-S k1 C + R D k2 A-S +D k3

time Rate/Selectivity

Kinetics AI/Systems Tools Catalyst Performance F

Pseudo Global Optimizer

Pseudo English Rule Compiler

Statistical Analyzer

Feature Extractor

Catalyst

{k}

Reaction Modeling Suite

SLIDE 33

Venkat Venkatasubramanian, Plenary Lecture, Foundations of Computer-Aided Process Design, Princeton, July 2004.

LUBRIZOL: Fuel Additive Design

• EPA requirement: Minimize

intake-valve deposits (IVD)

• Approach: Fuel Additives
• Performance measure
• BMW Test for IVD
• Stipulated to be less than 100 mg
• ver a 10,000 mile road test
• Expensive and time-consuming

testing

• Around $8000 for a single datum
• Problem: Design fuel-additives

that meet desired I VD performance levels

Intake Valve and Manifold

SLIDE 34

Fuel + Oil (Liquid + Vapor Mixture) Additive Component that keeps it in solution Additive Component that binds to deposit “precursors” Fuel-Additive Molecules High-Boiling Deposit Forming Precursors

Fuel-Additives : A Functional Description

SLIDE 35

Component “length” directly indicative of stability of additive Breakage of this bond removes “dirt” carrying capacity totally Chemical nature of this component (polar/non-polar) controls “dirt” removing capacity Breaking of these bonds control “length”

The first-principle model tracks the structural distribution of fuel-additive with time due to reactive degradation

First First-

• Principles Model for Additive Degradation

Principles Model for Additive Degradation

SLIDE 36

Performance Pattern Recognition Neural Networks Regression Ab Initio Methods Group Contribution Topological Techniques Molecular Mechanics Material/Molecule Formulation Primary Model Structural Descriptors Secondary Model

General Framework of Predictor Models General Framework of Predictor Models

SLIDE 37

Modeling Degradation Modeling Degradation

Dynamic in nature

Modeled as first-order irreversible reactions for additive

degradation

Rate constants in proportion to rates of pure thermal

degradation

Distribution of molecular species differing in structure

Identified by effective "length" Population balance model tracks distribution with time

SLIDE 38

From Stability to IVD From Stability to IVD

Define "amount of active additive"

The fraction of the additive species that remains active (i.e. intact viable structure) in the fuel at any point of time

A dynamic quantity decreasing with time

Depends on additive distribution

Solubility correlated to IVD via regression

Linear models Neural-Network models Input descriptors are the amounts of active additive at

different times

SLIDE 39

If A goes higher, B should go higher ..... If C goes higher, D should go lower ..... Expert Rules/ Heuristics Fundamental Physics/Chemistry ) v ( t ρ

• ∇

− = ∂ ρ ∂ ) ( P ) v ( v t ) v ( τ

• ∇

+ ∇ − = ρ ∇

• +

∂ ρ ∂

How to integrate diverse scientific knowledge/information into a single unified knowledge architecture ?

Forward Problem Approaches

Empirically speaking, a polynomial expansion fits my data............ Data-driven/empirical models

SLIDE 40

Venkat Venkatasubramanian, Plenary Lecture, Foundations of Computer-Aided Process Design, Princeton, July 2004.

Modeling Philosophy

All models are wrong….

… but some are useful!

• George Box

(U. Wisconsin)

Hybrid Models

Combine First-principles and

Data-driven models

SLIDE 41

Hybrid Model for IVD Prediction Hybrid Model for IVD Prediction

Time Amount of Active Additive

τ τ

2 τ i

τ

N

Additive A Additive B

Intake-Valve Deposit

SLIDE 42

Computer Computer-

• Aided Design of Fuel

Aided Design of Fuel-

• Additives

Additives

Hybrid Model Additive Performance Intake Valve Deposit Additive Structure Fuel Property Engine Conditions Genetic Algorithms

. .

Selection Recombination

Physical Model Statistical/Neural-Net Correlation

SLIDE 43

Previous Approaches to Inverse Problem

Previous Methodologies Random Search Heuristic Enumeration Math Programming Knowledge-Based Systems Graph Reconstruction Disadvantages Combinatorial Complexity

Nonlinear Search Spaces Local Minima Traps

Difficulties in Knowledge Acquisition Difficulties in using high-level chemical/bio-chemical

knowledge

SLIDE 44

Overview of Genetic Algorithms Overview of Genetic Algorithms

Definition

Genetic Algorithms are stochastic, evolutionary search procedures based on Darwinian model of natural selection

Essential Components

Genetic Operators

Crossover Mutation

Reproductive Plan

Fitness Proportionate Selection

SLIDE 45

Genetic Algorithms for Product Design Genetic Algorithms for Product Design

Global Search

Diversity of solutions High potential for novelty Global Optima

Development is de-coupled from forward problem

Robust to non-linearity

Population based search

Ability to provide several near-optimal solutions

Captures transparently the rich chemistry of the design

problem

SLIDE 46

Overview of Genetic Algorithms Overview of Genetic Algorithms

Select Operator Molecular Designs Initial Population Calculate Fitness Select Parent(s) Mutation Crossover Apply Operator Create New Population

SLIDE 47

Genetic Algorithms (GA) Genetic Algorithms (GA)

GAs are stochastic evolutionary search procedures based on the Darwinian model of natural selection

I nitial Population (random) Fitness Calculn, Parent Selectn Operators New Population “ Survival of the fittest ”

Evolution

SLIDE 48

• Single-point Crossover

Genetic Operators: Crossover Genetic Operators: Crossover

]n

O C O C O

[

O H H C H H C O H H C H H C CH3 CH3 C O C O

[ ]n

O O C O C O]n

[

CH3 CH3 C O C O O

]n [

Parent # 1 Parent # 2 Offspring # 1 Offspring # 2

SLIDE 49

• Main-chain and Side-chain Mutation

Mainchain Mutation

C H H C H H

n

C H H

Parent: Offspring:

C H H

n

C H H C H H C F H

n

Parent:

C H H C H

n

Offspring:

Sidechain Mutation

Replace F by Replace CH 2 by

Mutation Operators Mutation Operators

SLIDE 50

Fitness Function for GA-based CAMD

α = 0.0001 α = 0.01 α = 0.001 α = 0.0005

Fitness (F) = exp - α Pi – P Pi,max – Pi,min

2

Fractional Error (FE) = Pi – P P

SLIDE 51

Polymer Design Case Study

• Base Groups
• Target Properties
• Density

Glass Transition Temperature Thermal Expansion Coefficient Specific Heat Capacity Bulk Modulus 15 Side-chain Groups

3

• C2H5
• nC H

3 7

• iC H

3 7

• Cl
• Br
• OH
• H

tC 4H 9

• OCH

3

• CN
• CH
• F

3

• C-

O O CH - 3

• C-

O CH -

• C-

O NH-

• NH

X X X

17 Main-chain Groups

>C<

• S-
• SO2-
• O-
• NH-

O-C- O

• C-

O

• C-

O

• O

O-

• C-

O O

• C-

O

• C-

O NH-

• O-C-

O NH-

SLIDE 52

Target Properties

Target Polymer Density (gm/cm3)

2.96x10-4 1152.67 5.18x109

TP2:

1377.82 2.51x109

TP3:

1135.10 5.40x109 1073.96 5.39x109

TP5:

995.95 5.31x109

TP6:

1016.55 6.12x109

TP7:

1.34 1.18 1.21 1.19 1.28 1.25 1.06 350.75 225.24 420.83 406.83 472.00 421.12 322.55 1455.90 3.85x109

C O O C O O C H H C H H n C H H C F F C H H C H CH 3 n C O O O CH 3 C CH 3 n CH S O 3 C CH 3 O n O n SO 2 O O O C O n C H H C H H C H H C H H C H H C O n N H

Glass Transition (K) Thermal Expansion 1/K Heat Capacity J/kg K Bulk Modulus K N/m3

TP1:

2.81x10-4 2.90x10-4 2.90x10-4 2.90x10-4 2.98x10-4 2.89x10-4

TP4:

SLIDE 53

Polymer Design Case Study: Results

Legend: Success Rate; Average Conv. Generation; Number with fitness >0.99

Target Polymer

C O O C O O C

H H

C

H H

n C

H H

C

F F

C

H H

C

H

CH 3 n C O O O CH 3 C CH 3 n S O CH 3

C CH 3 O n O n SO 2 O O O C O n C H H C H H C H H C H H C H H C O n N H C O O C O O C H H

C

H H

n CH

3

CH

3

O CH

3

O n

60% 240 213 48% 522 6 12% 193 74 0%

• 589

48% 232 142 32% 632 168 100% 64 198 88% 109 161 4% 868 70 12% 184 282 36% 411 6 0%

• 163

0%

• 861

32% 400 175 8% 548 199 100% 61 217 68% 210 162 8% 382 144

Random MC, SC Random MC, SC Feasibility Constraints

SLIDE 54

Near Optimal Designs

Polymer Design Overall Error Fitness

Target Polymer: #4

3

S O

n

C C H 3 C H O

0% 1.0 Case 1: Standard GA 0.74% 0.995

n

H H C H O C O C H

n 2

C H 5 O C C N O O O C O

1.18% 0.991 Case 2: Modified GA, Stability

H O H O

0.25% 0.999

n

C H 3 C H C O O C

1.10% 0.991 Case 3: Modified GA, Stability & Complexity

O C H S C O C

2 5

H

n

O C O C H H

n

0.21% 0.999

H C

3

O O

n

C C H 3 O C O C

0.83% 0.995

SLIDE 55

Advantages of GAs for Product Design

Global Search

Diversity of solutions High potential for novelty Global Optima

Development is de-coupled from forward problem

Robust to non-linearity

Population based search

Ability to provide several near-optimal solutions

Captures transparently the rich chemistry of the design

problem

SLIDE 56

• Powerful generic method with some drawbacks

No convergence guarantees Performance sensitive to parameters Performance dependent on search space structure

Drawbacks of GAs

SLIDE 57

Evolutionary Design of Fuel Evolutionary Design of Fuel-

• Additives

Additives

• Knowledge Based

Initial Population

• Random

No

Termination Criteria

• Last Generation ?
• Fitness = 1.0 ?

New Generation Probabilistic Select Operator Select Parent(s) Fitness Proportionate Apply Operator Crossover Mutation Elitist Policy

Retain k best in population

Calculate Fitness (F)

Structure ---> Hybrid Model Fitness <---- Additive Performance

Head Linker Branch Tail Yes

Stop Start

IVDdesired< IVDlimit; Maximize Solubility Objective

SLIDE 58

Branch Crossover

L J Site 2 Branch A

Parent-1

N K

Parent-2

L N K Branch A moved to site-2 to ensure feasibility of offspring-1

Offspring-1

J

Offspring-2

Genetic Operators: Branch Crossover Genetic Operators: Branch Crossover

SLIDE 59

Run Rank/Identifier

Fitness

Predicted IVD (PLS-NN Model) Structural Description

1, I-1 0.997

11.4 mg Novel Structure. Infrequently used linker.

2, I-2 0.996

11.5 mg Novel Structure. Same tails as best structure, different heads and linkers I

6, I-6 0.993

12.0 mg Variant of structure found in the BMW

• database. Same head and linkers, different

tails

1, II-1 0.999

10.1 mg Novel Structure. Different from I-1. Infrequently used linker component.

2, II-2 0.989

12.6 mg Slight variant of additive structure found in BMW and HONDA databases. Different tails but same head and linker II

4, II-4 0.983

13.2 mg Minor variation of structure II-2 above. Slight modification of the head

1, III-1 1.00

8.9 mg Novel Structure. Different from I-1 and II-1. Commonly used components

2, III-2 0.994

11.9 mg Variant of III-1. One linker and tail modified. III

3, III-3 0.993

12.1 mg Variant of structure II-2 above. Slight modification of head. A linker and tail inserted.

Evolutionary Design of Fuel Additives: Results Evolutionary Design of Fuel Additives: Results

Objective: Determine a structure with IVD < 10 mg Dosage: 50 PTB; Population Size: 25; Generations: 25 Objective: Determine a structure with IVD < 10 mg Dosage: 50 PTB; Population Size: 25; Generations: 25

SLIDE 60

School of Chemical Engineering, Purdue University

Tires, Treads, Hoses, Shock Absorbers, O- rings, Gaskets, Mounts …….. Rubber Parts in Service

SLIDE 61

School of Chemical Engineering, Purdue University

Rubber Formulation

Accelerator Activator Rubber

Retarder

Filler Antidegradants Oils Sulfur

Mixing (~ hrs)

Part

Curing (~ hrs)

Thermal History

Final Vulcanized Part

Chemical Environment Deformational Environment Thermal Environment

Application (~ yrs)

Given the Desired Property/Performance of Interest, what is the Optimal Rubber Formulation? Problem Definition

Rubber Formulation

Accelerator Activator Rubber Retarder Filler Antidegradants Oils Sulfur

Which materials to include? How much of them to include? What should be the processing conditions?

SLIDE 62

School of Chemical Engineering, Purdue University

What are Sulfur-Vulcanized Elastomers?

Rubber Formulation

Accelerator Activator Rubber Retarder Filler Antidegradants Oils Sulfur

Accelerated Sulfur Vulcanization

S S S S

Vulcanization

+ S8

SLIDE 63

School of Chemical Engineering, Purdue University AIChE Annual Meeting, Reno 2001

OBJECTIVE OF THIS WORK

• Understand the Mechanistic Details of Accelerated-Sulfur-

Vulcanization.

• Provide a Mathematical Description of the Kinetics of Accelerated-

Sulfur Vulcanization consistent with Mechanistic Chemistry Rubber Formulation

Accelerator Activator Retarder Rubber Filler Sulfur Oils Antidegradants

N C S S N O

NR ZnO/Stearic-Acid Benzothiazole Sulfenamide S8

SLIDE 64

School of Chemical Engineering, Purdue University AIChE Annual Meeting, Reno 2001

COMPLEXITY OF VULCANIZATION REACTIONS

W.Scheele (1956) 1 Perhaps no-where in chemistry is there encountered a field which even in its literature alone shows so many uncertainties and (possibly only apparent) contradictions as that of the vulcanization of rubber. L.Bateman (1963) 2 Whilst it has long been appreciated, albeit intuitively, that sulfur vulcanization is a very complex chemical process, the actual complexity as revealed in the studies described above is probably far in excess of what has ever been envisaged.

• 1W. Scheele, O. Lorenz and W. Dummer, Rubber Chemistry and Technology, 29, 1 (1956)

2 L. Bateman, C.G. Moore, M. Porter and B. Saville, “The Chemistry and Physics of Rubber-Like

Substances”, L Bateman, Ed., Maclaren & Sons Ltd., London, 1963, pp 449.

P.J. Nieuwenhuizen (1997) 3 It must be considered remarkable that despite all the efforts and progress in the field of vulcanization during the past decade, one has to conclude that the statements of Scheele and Bateman, made 30 to 40 years ago, are still true to a great extent.

3 P.J. Nieuwenhuizen, Rubber Reviews, Rubber Chemistry and Technology, 29, 70 (1997)

SLIDE 65

School of Chemical Engineering, Purdue University AIChE Annual Meeting, Reno 2001

SCHEMATIC OF A VULCANIZATION CURVE

Temporal Evolution of Crosslinks

Time

Concentration of Crosslinks, ν

Curing Phase Reversion Phase Induction Phase

A S S

S S S S S S

Time Scale ~ Years Time Scale ~ Hrs

SLIDE 66

School of Chemical Engineering, Purdue University

CAMD Sub-Problems

Structure Space Property

INVERSE PROBLEM (Given Properties, Predict Structure/Formulation)

Properties/ Performance ( P1, P2, ……..PN )

FORWARD PROBLEM (Given Structure, Predict Properties/Performance)

Material Structure/ Formulation Neural Network Hybrid Models Fundamental Models

SLIDE 67

School of Chemical Engineering, Purdue University

THE OVERALL FORWARD MODEL

Sub Assembly Design Full Product Design & Operation

Design & Operation of Complex Sub-Assembly

Engine

Over 1000 rubber parts in failure critical functions

Mixing + Heat Transfer + Mold Filling

Part Manufacturing

Overcure Undercure

Manufac- turing Process Material Formulation

A

x + S 8 k1

⎯ → ⎯ ⎯ A

x+8

Ax + R

k2

⎯ → ⎯ ⎯ Bx B

x k3

⎯ → ⎯ ⎯ Bx

*

etc . Time Torque Cure Curve

Kinetics Polymer Network Mechanical Response

˜ T = 2ρR ∂ ψ ∂I 1 +I 1 ∂ ψ ∂I 2 ⎧ ⎨ ⎩ ⎫ ⎬ ⎭ I − ∂ ψ ∂I 2 C + I 3 ∂ ψ ∂I 3 C−1 ⎡ ⎣ ⎢ ⎤ ⎦ ⎥ •

• ∂ C

∂C

Constitutive Model 3D Finite Element Part Model Spring-Dashpot Approximation with Chemical Aging

Part Shape Design

Quantum Chemistry

S C N S S S S S S S S Zn

New Molecules

SLIDE 68

School of Chemical Engineering, Purdue University

Accelerator Activator + Active Accelerator Complex S8 Active Sulfurating Agent Rubber Bound Intermediate (Crosslink Precursor) Rubber Bound persulfenyl radical Polysulfidic crosslinks Aged Network Desulfuration Degradation N C S S Sx S C S N

Ax Bx

C C CH Sx S C S N

Bx

• C

C CH Sx*

Vux

C C CH Sx C C C

N C S S N O

(ZnO + Stearic Acid) 2-morpholinothiobenzothiazole

Schematic of a Vulcanization Curve

SLIDE 69

School of Chemical Engineering, Purdue University

Accelerator Activator + Active Accelerator Complex S8 Active Sulfurating Agent

N C S

Bt = BtS-NR2 BtSH + R2NH BtS-NR2 + BtSH BtSSBt + R2NH BtS-SBt + S8 BtS-S-SBt + S7 BtS-S2-SBt + S6 ••• BtS-Sx-SBt + BtS-Sy-SBt BtS-Sz-SBt + BtS-Sw-SBt

N C S S Sx S C S N

N C S S N O

N C S S Sx S N C S Zn

L L

Without Activator With Activator

Ax = BtS-Sx-SBt

Accelerator Chemistry

BtS-SBt + S8 BtS-S8-SBt

SLIDE 70

School of Chemical Engineering, Purdue University

CROSSLINKING CHEMISTRY (I)

Rubber Bound Intermediate (Crosslink Precursor) Active Sulfurating Agent N C S S Sx S C S N CH HC C CH

3

H S N C S S N C S x S S

CH C CH S S x N C S S

+

N

C

S S H N

C

S SH

(R.34) C C CH Sx S C S N

Ax Bx Rubber

Crosslinking Chemistry (I)

SLIDE 71

School of Chemical Engineering, Purdue University

Rubber, RH BtS• S8 S8 S7 RS-Sy+1

• RS-Sy+8
• S8

S8 S7 Bt Sz+1

• S8

S8

(Regeneration of Sulfurating Species)

Bt Sz+8

• + Bt-SH

+ Bt-Sz

• Crosslink Precursor

RSx –SBt, RSx-Zn-SBt Crosslink Precursor RSx –SBt, RSx-Zn-SBt

Crosslinks

RSSy -R

Crosslinks

RSSy -R Rubber, RH Sulfurating Species BtS-Sx-SBt, BtS-Zn-Sx-SBt Sulfurating Species BtS-Sx-SBt, BtS-Zn-Sx-SBt

Persulfenyl Radical

RS-Sy

• Persulfenyl Radical

RS-Sy

• Rubber, RH

Bt Sz+2

• Rubber, RH

RS-Sy Cyclic Sulfide BtSZnSBt Main- Chain Modification, Conjugated Dienes/Trienes, Inactive thiols RSSy-1R + BtSZnSSBt (Desulfuration) (Degradation) ZnO BtSZnSBt BtS-ZnSx-SBt (Scorch Delay) CDB + Pthalimide (Retarder Action) S8

…

S7 S8

BtS-SBt

BtSNR2 BtSSSBt BtSSxSBt BtSS8SBt

BtS-SBt

BtSSxSBt + BtSSySBt

BtS-SBt

BtSSxSBt + BtSSySBt + CTP

• R2NH

Lumped S8 Pickup Sequential S8 Pickup Both

Mechanisms BtSH

820 Reactions, 107 Species

SLIDE 72

School of Chemical Engineering, Purdue University

Equations for MBTS and other sulfurating species

[ ]

{ }

{ }

{ }

{ }

{ }

{ }

{ }

{ }

{ }

{ }

{ }

⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎦ ⎤ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎣ ⎡ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + − + − − − − − − =

∑ ∑ ∑

= ∗ = ∗ = 2 7 8 6 9 5 10 4 11 3 12 2 13 1 7 6 8 5 9 4 10 3 11 2 12 1 2 6 7 5 8 4 9 3 10 2 11 1 6 5 7 4 8 3 9 2 10 1 2 5 6 4 7 3 8 2 9 1 5 4 6 3 7 2 8 1 2 4 5 3 6 2 7 1 4 3 5 2 6 1 2 3 4 2 5 1 3 2 4 1 2 2 3 1 2 1 2 1 A

• A

16 2 r r DESULF 2 E

• E

16 1 r r BST

• A

R

• A

8 S

• A

14 2 r r A

• A

MBT

• MBS

MBS

] A [ 5 . ] A ][ A [ ] A ][ [A ] A ][ [A ] ][A [A ] ][A [A ] ][A [A 2 ] A ][ A [ ] A ][ [A ] A ][ [A ] ][A [A ] ][A [A ] ][A [A 2 ] A [ 5 . ] A ][ [A ] A ][ [A ] ][A [A ] ][A [A ] ][A [A 2 ] A ][ A [ ] A ][ [A ] ][A [A ] ][A [A ] ][A [A 2 ] A [ 5 . ] A ][ [A ] ][A [A ] ][A [A ] ][A [A 2 ] A ][ [A ] ][A [A ] ][A [A ] ][A [A 2 ] 0.5[A ] ][A [A ] ][A [A ] ][A [A 2 ] ][A [A ] ][A [A ] ][A [A 2 ] [A 5 . ] ][A [A ] ][A [A 2 ] ][A [A ] ][A [A 2 ] [A 5 . ] ][A [A 2 ] ][A [A 2 ] [A k Vu ] [A k ] [E k .5 B ] [A k ] [A k ] [S ] [A k A 1) (r ] [A k [MBS] [MBT] k [MBS] k A dt d ] [E ] [E k A ] [A 2k Vu ] [A k A ] [A 2k ] [S ] [A k ] [A 2k ] [A dt d

1 E

• E

13 1 r r 1 A

• A

16 2 r r DESULF 14 2 r r A

• A

8 1 S

• A

1 R

• A

1 ∗ ∗ = = =

+ − + + − − =

∑ ∑ ∑

POPULATION BALANCE EQUATIONS

SLIDE 73

School of Chemical Engineering, Purdue University

POPULATION BALANCE EQUATIONS

[ ]

⎥ ⎦ ⎤ ⎢ ⎣ ⎡ + − − − =

∑

= 16 1 i i R1 1 4 A 8 1 A3 1 C1 1

Vu ] [A k ] A [ ] MBTS [ k 2 ] S [ ] [A k ] [A k 2 A dt d

[ ] [ ][ ]

] [A k A 1) (i ] [A k ] [S ] [A k MBS MBT k A dt d

C1 14 2 i i A4 8 A3 A2

− ⎥ ⎦ ⎤ ⎢ ⎣ ⎡ − − − =

∑

=

⎥ ⎦ ⎤ ⎢ ⎣ ⎡ − ⎥ ⎦ ⎤ ⎢ ⎣ ⎡ −

∑ ∑

= = 16 1 i i 1 R 16 1 i i C4

Vu ] A [ k * B ] [A k

[ ]

⎩ ⎨ ⎧ > ≤ ≤ + + − ⎩ ⎨ ⎧ > ≤ ≤ − − = 7 i ] S [ 6 i 2 , ] A [ k ] A [ ] [A k 1) (i 7 i , 6 i 2 ], S [ ] A [ k ] [A k 2 A dt d

8 i 3 A i A4 8 i 3 A i C1 i

Equations for MBTS and other sulfurating species Equations for Crosslink Precursors

[ ]

) *] [E k ] [MBTS k ( B dt d

1 C7 C1

+ =

[ ]

*] [E k ] B [ k ) 1 i ( ] [A k 2 *] B ][ A [ k B dt d

1 i C7 i 2 C i C1 i A4 i +

+ − − + − =

SLIDE 74

School of Chemical Engineering, Purdue University

[ ] [ ]

j conc and i time at Point Data al Experiment Constants Rate c dt d t s k k k k

j i P q i m j j i j i j i

= = ≥ = = ⎥ ⎥ ⎦ ⎤ ⎢ ⎢ ⎣ ⎡ −

∑∑

= = exp , 3 2 1 2 1 1 exp , exp , , k

k k ) k , , ( . . ...... , , k , ) k ( min ν ν φ ν ν ν ν

m time points, q concentrations

RATE-CONSTANT DETERMINATION

FACTS

Total of 820 different reactions 107 Coupled Ordinary Nonlinear Differential Equations 9 Optimizable Rate Constants

SLIDE 75

School of Chemical Engineering, Purdue University

B2

• Example Writing Population Balance Equations

kB

C C HC S S C S N S S* C S N + C C CH S2* + C C CH S* * C S N S S

kA

B2 E• E2

• Potential Breaking Sites

B•

[ ] ( )[ ]

[ ]

[ ]

[ ]

[ ]

[ ]

[ ]

[ ]

[ ]

2 B 2 2 B 2 A 2 A 2 B B B A A A 2 B A 2

B k E dt d , B k B dt d B k E dt d , B k B dt d RT E exp k k , RT E exp k k , B k k B dt d = = = = ⎟ ⎠ ⎞ ⎜ ⎝ ⎛− = ⎟ ⎠ ⎞ ⎜ ⎝ ⎛− = + − =

SLIDE 76

School of Chemical Engineering, Purdue University

Equations for MBTS and other sulfurating species

Population Balance Equations

[ ]

y 13 1 x x 13 1 y x A

• A

16 2 r r DESULF 2 E

• E

16 1 r r BST

• A

R

• A

8 1 y y S

• A

14 2 r r A

• A

MBT MBS-

A A k Vu ] [A k ] [E k 2 1 B ] [A k ] [A k ] [S ] [A k A 1) (r ] [A k [MBS] [MBT] k A dt d

∑∑ ∑ ∑ ∑ ∑

= − = = ∗ = ∗ = =

+ − + − − − − − = ] [E ] [E k A ] [A k 2 Vu ] [A k A ] [A 2k ] [S ) ] [A

• ]

[A ( k ] [A 2k ] [A dt d

1 E

• E

13 1 r r 1 A

• A

16 2 r r DESULF 14 2 r r A

• A

8 1 y y 1 S

• A

1 R

• A

1 ∗ ∗ = = = =

+ − + + − − =

∑ ∑ ∑ ∑

] A ][ [A k ] [E ] [E k A ] [A k 2 ] ][A [A k 1) (x A ] [A 2k ] [S ) ] [A

• ]

[A ( k ] [A 2k ] [A dt d

r

• x

1

• x

1 r r A

• A

x E

• E

x 14 1 r r x A

• A

x A

• A

14 1 x r r A

• A

8 1 y y 1

• x

x S

• A

x R

• A

x

∑ ∑ ∑ ∑

= ∗ ∗ − = + = =

+ + − − − + − − =

x = 0 x = 1 2 ≤ x ≤ 13

SLIDE 77

School of Chemical Engineering, Purdue University

Final Set of Population Balance Equations

( )

( ) ( ) ( )

( ) ( )

( )⎥

⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎦ ⎤ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎣ ⎡ = ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎦ ⎤ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎣ ⎡

∗ ∗ ∗ ∗ ∗

E Bx Bx Ax k k k k f Ax k f S A k f Vu Bx k k f Bx Bx Ax k k k f Bx Bx Ax k k k f Ax k k k k f E MBT S Vu Bx Bx Ax

x x

, , , , , , , , , , . . . , , , . . . , , , , , . . . , , , , , . . . , , , , . . . . . . . . . . . . . . . . . .

52 51 4 2 7 1 6 8 01 5 6 3 4 4 3 2 3 4 2 1 2 4 1 02 01 1 8

d dt

Initial Conditions:

Ax(>0)(0) = Bx(0) = B*x(0) = Vux(0) = MBT(0) = E(0) = 0 A0(0) = [MBS]0 S8(0) = [Sulfur]0

Most simple kinetic model that explains all the important features of sulfur vulcanization. Total of 820 different reactions considered with 107 coupled ordinary nonlinear differential equations and 9 optimizable rate-constants.

SLIDE 78

School of Chemical Engineering, Purdue University

Population Balance Model - Predictions

Open Symbols = Experimental Data Solid Line = Model Predictions Open Symbols = Experimental Data Solid Line = Model Predictions

T = 330 F T = 330 F

Effect of Accelerator and Sulfur

(1.0, 4.0) (0.5, 4.0) (0.5, 2.0)

+

(1.0, 2.0) (0.75, 2.0)

x

Legend

SLIDE 79

School of Chemical Engineering, Purdue University

Population Balance Model – Different Temperatures

T = 310 F T = 310 F T = 298 F T = 298 F

Open Symbols = Experimental Data Solid Line = Model Predictions Open Symbols = Experimental Data Solid Line = Model Predictions

SLIDE 80

School of Chemical Engineering, Purdue University Quantum Chemistry

˜ T = 2ρR ∂ ψ ∂I 1 + I 1 ∂ ψ ∂I 2 ⎧ ⎨ ⎩ ⎫ ⎬ ⎭ I − ∂ ψ ∂I 2 C + I 3 ∂ ψ ∂I 3 C−1 ⎡ ⎣ ⎢ ⎤ ⎦ ⎥

• ∂ C

∂C

Constitutive Model 3D Finite Element Part Model Spring-Dashpot Approximation with Chemical Aging

A

x + S 8 k1

⎯ → ⎯ ⎯ Ax+8 Ax + R

k2

⎯ → ⎯ ⎯ Bx Bx

k3

⎯ → ⎯ ⎯ Bx

*

etc . Time To r q u e

Cure Curve

Kinetics Polymer Network Mechanical Response

Mixing + Heat T ransfer + Mold Filling

Part Manufacturing

Overcure Undercure S C N S S S S S S S S Zn

New Molecules Material Formulation Manufac- turing Process Part Shape Design Sub Assembly Design Full Product Design & Operation

Design & Operation of Complex Sub-Assembly

Engine

Over 1000 rubber parts in failure critical functions

THE CATERPILLAR GRAND CHALLENGE

SLIDE 81

School of Chemical Engineering, Purdue University

Spatial Profile of the State-of-Cure for the Thermal History in the Mold

T h e r m a l P r

• f

i l e

85-87 89-91 91-93 93-95 95-97 > 99 97-99 87-89

C u r e P r

• f

i l e

Actual Part Design

SLIDE 82

School of Chemical Engineering, Purdue University

Result Summary Vulcanization

Vulcanization Chemistry

Mathematical Quantification Insights into Actual Part Design Mechanistic Evaluation

• Evaluated all possible

mechanisms critically for accelerated sulfur vulcanization

• Identified the simplest set of

reactions that describe all the important aspects of cure.

• Developed Population Balance

Model based on mechanistic details to describe the different aspects of cure.

• Single set of parameters that

predicts cure details for all formulations at all temperatures.

• Model extended for retarders,

filled systems and other accelerator and elastomer classes.

• Incorporation of Population

Balance model with a finite element code to predict spatial cure profile.

• Identification of undercured

and overcured regions visually.

SLIDE 83

School of Chemical Engineering, Purdue University Quantum Chemistry

˜ T = 2ρR ∂ ψ ∂I 1 + I 1 ∂ ψ ∂I 2 ⎧ ⎨ ⎩ ⎫ ⎬ ⎭ I − ∂ ψ ∂I 2 C + I 3 ∂ ψ ∂I 3 C−1 ⎡ ⎣ ⎢ ⎤ ⎦ ⎥

• ∂ C

∂C

Constitutive Model 3D Finite Element Part Model Spring-Dashpot Approximation with Chemical Aging

A

x + S 8 k1

⎯ → ⎯ ⎯ Ax+8 Ax + R

k2

⎯ → ⎯ ⎯ Bx Bx

k3

⎯ → ⎯ ⎯ Bx

*

etc . Time To r q u e

Cure Curve

Kinetics Polymer Network Mechanical Response

Mixing + Heat T ransfer + Mold Filling

Part Manufacturing

Overcure Undercure S C N S S S S S S S S Zn

New Molecules Material Formulation Manufac- turing Process Part Shape Design Sub Assembly Design Full Product Design & Operation

Design & Operation of Complex Sub-Assembly

Engine

Over 1000 rubber parts in failure critical functions

THE OVERALL FORWARD MODEL

SLIDE 84

School of Chemical Engineering, Purdue University

CAMD Framework

The solution of INVERSE PROBLEM involves searching for the optimal rubber

formulation that has the desired macroscopic performance.

Prediction

P1 P2 P3 ... Pn

Structure or Formulation Product Performance Design

Inverse Problem

Prediction

P1 P2 P3 ... Pn

Structure or Formulation Design

Inverse Problem Forward Problem

Rubber Accelerator Sulfur Activator Filler Retarder Cure-State Modulus Stress-Strain Stiffness

SLIDE 85

School of Chemical Engineering, Purdue University

Genetic Algorithms (GA)

GAs are stochastic evolutionary search procedures based on the Darwinian model of natural selection

Fitness Calculn, Parent Selectn Fitness

⎥ ⎥ ⎦ ⎤ ⎢ ⎢ ⎣ ⎡ ⎟ ⎟ ⎠ ⎞ ⎜ ⎜ ⎝ ⎛ − − − =

∑

2

min i, max i, des i, i

P P P P (F) Fitness

α exp

I nitial Population (random) Formulation = [1 4 0.1 30 330]

SLIDE 86

School of Chemical Engineering, Purdue University

Fitness Calculn, Parent Selectn Fitness “ Survival of the fittest ” New Population

Evolution

Genetic Algorithms (GA, contd)

Operators

SLIDE 87

School of Chemical Engineering, Purdue University

Formulation Representation Binary Representation

Rubber Formulation

Accelerator Activator Rubber Retarder Filler Antidegradants Oils Sulfur

Rubber Formulation

Accelerator Activator Accelerator Activator Rubber Retarder Rubber Retarder Filler Filler Antidegradants Oils Sulfur Antidegradants Oils Sulfur Sulfur

1 1

• •

1 1 1

Phr of Different Ingredients

SLIDE 88

School of Chemical Engineering, Purdue University

Genetic Operators

Parent 1 Parent 2 Offspring 1 Offspring 2

1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

Formulation Fitness

Crossover

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

Mutation

SLIDE 89

School of Chemical Engineering, Purdue University

Tmax (time to reach max cure) = 16 min Vumax (crosslink @ Tmax) = 71.5 mol/m3 Stress (100% Elongation) = 0.92 MPa Stress (200% Elongation) = 1.65 MPa

Optimal Formulations Desired Property

Inverse Problem (Results)

• Binary Representation
• Fitness Proportionate Selection
• Crossover Probability = 0.8, Mutation Probability = 0.2
• Population Size = 40, Number of Generations = 40, Elitism = 10%

[1.75 1.25 0.05 0 315] [0.50 4.00 0.10 0 310] [1.75 1.50 0.05 0 310] [0.75 2.75 0.10 0 315] [3.00 0.75 0.10 0 315] [2.75 1.00 0.25 0 330] 0.9999 0.9999 0.9996 0.9994 0.9991 0.9990 15 16 17 13 17.5 10.5 69.92 71.57 75.75 67.88 65.34 70.46 0.91 0.92 0.97 0.88 0.85 0.93 1.63 1.65 1.75 1.58 1.52 1.67

Formulation Fitness Tmax Vumax σ100 σ200 Example 2

SLIDE 90

School of Chemical Engineering, Purdue University

[1.75 1.25 0.05 0 315] [0.50 4.00 0.10 0 310] [1.75 1.50 0.05 0 310] [0.75 2.75 0.10 0 315] [3.00 0.75 0.10 0 315] [2.75 1.00 0.25 0 330] 0.9999 0.9999 0.9996 0.9994 0.9991 0.9990 15 16 17 13 17.5 10.5 69.92 71.57 75.75 67.88 65.34 70.46 0.91 0.92 0.97 0.88 0.85 0.93 1.63 1.65 1.75 1.58 1.52 1.67

Formulation Fitness Tmax Vumax σ100 σ200

10 20 30 40 50 60 10 20 30 40 50 60 70 80 1 2 3

Time, min Concentration of Crosslinks, mol/m3 New Formulations

BEST THREE FORMULATIONS

Solution Found in 24th generation. i.e.only 960 out

• f a total of

131072 formulations were searched ~ 0.73 %

SLIDE 91

School of Chemical Engineering, Purdue University

Interactive Software Used At Caterpillar on a Daily Basis

SLIDE 92

10/ 11/ 2005 1

Pharmaceutical Product Development and Engineering

• Two Products

Drugs Documents

Preliminary product recipe Drug is converted into Particles Drug Synthesis Raw Chemicals Formulation Process Development Preliminary process recipe Scale/up -Tech Transfer Adjusted process recipe Manufacturing A Delayed Product

SLIDE 93

Venkat Venkatasubramanian, Keynote Lecture, European Congress in Chem. Engg, Copenhagen, Sept 2007

2

LIPS

• Prescription drug recalls
• 176 in 1998
• 248 in 2001
• 354 in 2002
• Schering-Plough Corp. recalled 59

million asthma inhalers in 1999 and 2000 -- unknown number were shipped empty.

• Semiconductor Manufacture

6 σ

• Chemicals

5 to 5.5 σ

• Pharma

2.5 σ

SLIDE 94

Venkat Venkatasubramanian, Keynote Lecture, European Congress in Chem. Engg, Copenhagen, Sept 2007

3

LIPS

Pharma Industry’s Scale-up Challenge

engineering pharmaceutical

SLIDE 95

4

Purdue Ontology for Pharmaceutical Engineering (POPE)

Information D e c i s i

• n

t r e e s a n d h e u r i s t i c s M a t h e m a t i c a l m

• d

e l s

ModLAB OntoMODEL

SLIDE 96

Venkat Venkatasubramanian, Keynote Lecture, European Congress in Chem. Engg, Copenhagen, Sept 2007

5

LIPS

Ontologies Developed So Far in Our Project

Equipment ontology

Standards referenced: STEP, AP231, FIATECH

General recipe ontology

RecipeElement, UnitProcedure, Operation etc. Standards referenced: ISA S88, S95, OntoCAPE

Process safety ontology

Deviation, Cause, Consequence etc.

Material ontology for pharmaceutical product development

Flow property, Angle_of_Fall, Carrs_Index etc.

Reaction mechanism ontology

Molecule, Atom, Bond, Reaction etc. Referenced: Chemistry Development Kit

Model ontology Guideline ontology

Referenced: GuideLine Interchange Format

SLIDE 97

Venkat Venkatasubramanian, PSWC, April 2007, Amsterdam

6

LIPS

Material Property Ontology in Protégé

SLIDE 98

Venkat Venkatasubramanian, PSWC, April 2007, Amsterdam

7

LIPS

Experiment Ontology

Experiment Data Files Experimental Methods Details

SLIDE 99

Venkat Venkatasubramanian, Keynote Lecture, European Congress in Chem. Engg, Copenhagen, Sept 2007

8

LIPS

Guideline Ontology for Dosage Form Formulation

Follows GLIF (Guideline Interchange Format), A standard ontology for clinical guidelines

SLIDE 100

9

Model Ontology

SLIDE 101

Venkat Venkatasubramanian, Plenary Lecture, Foundations of Computer-Aided Process Design, Princeton, July 2004.

SuperModel Model Container Model Container Model Model

hasModelContainers hasModel hasModel hasModelIndex hasModelIndex

Equation Partial Differential Integral Algebraic DAE Function Evaluation Assumptions Universal Constants Dependent Var Independent Var Model Parameters Variable Link Variable Variable Simple Variable Array Variable SizeOfList Variable Link

hasInitialGuess hasVar PointsToValue PointsToValue hasVar hasEqn hasAssumptions hasUnivConstants hasDepVar hasIndepVar hasModelParms hasVar hasSymbol hasML

Text Input

SLIDE 102

Venkat Venkatasubramanian, Keynote Lecture, European Congress in Chem. Engg, Copenhagen, Sept 2007

11

LIPS

Integrated Model-based Decision Support

Recipe Process Information Repository Material Form in XForms Recipe Network Used by PHASuite Used by Batches/Batch+

SLIDE 103

Venkat Venkatasubramanian, Keynote Lecture, European Congress in Chem. Engg, Copenhagen, Sept 2007

12

LIPS

Integrating with Mathematical Knowledge

User created model instance Model Ontology

Equations in MathML

• Variables linked through URI

Process Description (instances of Process Ontology)

Automatically generated Mathematica statements

Interface

Mathematica Notebook

Access ontology instances

Mathematica Kernel Rich Client Web Applications

SLIDE 104

1 3

Literature Experiments Experiment Report Data

Intelligent Systems for Preformulation

Formulation Process Monitoring and Control

Tools for Organic Solids Mfg.

Optimization, Simulation, Safety Analysis, Scheduling

Multiscale models for synthesis mehtods Modeling and Analysis Crystalline Structure Unit Operation models

DEM/CFD/FEM

Process modeling

First Principles Process Model

Organic composites:

Physical, chemical, and powder properties

Equipment:

blender, tabletting press, roller compactor, fluid bed

Process:

Crystallization, Granulation, Drying, Size Reduction etc.

Knowledge relationships Systematic functionalization Control methodology for spatial structure of organic composites Material synthesis method development Determine structure-function- performance relationships

Instruments

Cyberinfrastructure for Real-time Decision Support for Design, Control, and Optimization

Dissolution Problems

Spatial Structure Process Monitoring and Control

Tools for Design and Manufacturing

Modeling User Ontological Informatics Infrastructure Material Science Experiments

Process

DEM Experiment Report Physical Properties Powder Properties

eLabNotebook

Blender Literature Experiments Experiment Report Data

Intelligent Systems for Preformulation

Formulation Process Monitoring and Control

Tools for Organic Solids Mfg.

Optimization, Simulation, Safety Analysis, Scheduling

Multiscale models for synthesis mehtods Modeling and Analysis Crystalline Structure Unit Operation models

DEM/CFD/FEM

Process modeling

First Principles Process Model

Organic composites:

Physical, chemical, and powder properties

Equipment:

blender, tabletting press, roller compactor, fluid bed

Process:

Crystallization, Granulation, Drying, Size Reduction etc.

Knowledge relationships Systematic functionalization Control methodology for spatial structure of organic composites Material synthesis method development Determine structure-function- performance relationships

Instruments

Tools for Design and Manufacturing

Modeling User Ontological Informatics Infrastructure Material Science Experiments

SLIDE 105

Venkat Venkatasubramanian, Keynote Lecture, European Congress in Chem. Engg, Copenhagen, Sept 2007

14

LIPS

Summary

Reviewed Modeling and Informatics Challenges in Molecular

Products Design and Engineering

Need for Cyberinfrastructure Concepts, Methods, and Tools Ontological approach for information and knowledge

modeling

Beyond ERP, SAP, Oracle, Expert Systems etc. This is not just programming! Nor can it be done by CS folks

alone! POPE: Purdue Ontology for Pharmaceutical Engineering

Conceptual foundation for next the generation, integrated,

decision support tools environment

This is only a beginning….Miles to go before we sleep…. Huge opportunities for Process/Product Systems

Engineers

SLIDE 106

15

Thank You for Your Attention!

Any Questions?