[email protected]
richardhwest
rwest
Reaction Mechanism Generator:
Toward High-Throughput Transition State Calculations,
and Interpreting Existing Kinetic Models
Richard H. West
4 August 2014
.edu/comocheng
1
Presented at the 2nd International Workshop on Flame Chemistry in San Francisco
Automatic reaction mechanism generation
needs methods to:
1. Represent molecules (and identify duplicates)
The common theme to the two projects in the talk is Reaction Mechanism Generator software. First, a brief introduction to it.
Automatic reaction mechanism generation
needs methods to:
2. Create reactions (and then new species)
CH3
+
Automatic reaction mechanism generation
needs methods to:
3. Choose which reactions to include (and which to leave out)
You’ll notice a lot of slide designs are top-heavy. This is because the room layout meant most people couldn’t see the bottom of the screen!
Automatic reaction mechanism generation
needs methods to:
4. Estimate thermo and kinetic parameters
(quickly!)
Molecules are represented as 2D graphs
CH3CH2.
=
H
H
H
C
C*
H
H
Reaction families propose all possible
reactions with given species
• Template
• Recipe
• Rules
for recognizing reactive sites
for changing the bonding at the site
for estimating the rate
There are ~40 reaction families such as as Hydrogen abstraction,
unimolecular homolysis, radical addition to a double bond…
All possible reactions are found, then core is
expanded by following the fastest pathways.
H
B
A
E
D
C
F
B
A
E
D
C
F
G
Model can be started from “seed mechanism”
(if you have this in RMG format)
H
F
G
Getting it in “RMG format” is the hard part, and will be addressed in the second half of the talk.
Rate estimates are based on the local
structure of the reacting sites.
O
H
O
• Hydrogen abstraction: XH + Y. → X. + YH
• Rate depends on X and Y.
Rate estimation rules are organized in a tree
Part of the tree for X
The most generic expression goes at the top of the tree. You use the most precise node that you can, then fall up until you find data.
Part of the tree for Y
New “RMG-Py” designed to be more extensible and developer-friendly.
• See poster on Wednesday
We presented two posters describing this work at the 35th International Symposium on Combustion,
occurring the week following this presentation.
New “RMG-Py” designed to be more extensible and developer-friendly.
• See poster on Wednesday
We presented two posters describing this work at the 35th International Symposium on Combustion,
occurring the week following this presentation.
New “RMG-Py” designed to be more extensible and developer-friendly.
• See poster on Wednesday
We presented two posters describing this work at the 35th International Symposium on Combustion,
occurring the week following this presentation.
Toward High-Throughput
Transition State Calculations
Pierre L. Bhoorasingh
Acknowledgment is made to the Donors of
the American Chemical Society Petroleum
Research Fund for support of this research.
.edu/comocheng
14
Now that you understand RMG, here’s the first of the two projects.
Ideal tree:
lots of data
Ideally we climb the tree, find the matching node for our reaction, and pluck the leaf containing the rate expression.
Typical tree:
sparse data
But what if there are no data? How can we make it?
Transition State calculations can provide
missing data
Transition State calculations can provide
missing data
Heat release (red) and progress variable (purple) of a turbulent lean methane-air Bunsen flame simulated using S3D. DOE
At my first Combustion Symposium (in 2006) I was impressed by the computer resources used by the CFD modelers
16,000,000
CPU-hours
Heat release (red) and progress variable (purple) of a turbulent lean methane-air Bunsen flame simulated using S3D. DOE
At my first Combustion Symposium (in 2006) I was impressed by the computer resources used by the CFD modelers
Assoc. Lib Director, ORNL Computing
This is one of the large federal computers they use. Imagine what we could do if combustion chemists had 16M CPU hours!…
We’d need 16,000 graduate students to set up the calculations!
We’d need 16,000 graduate students to set up the calculations!
Automatic TS searches remain an important
energy research goal
“An accurate description of the often
intricate mechanisms of large-molecule
reactions requires a characterization of all
relevant transition states... Development
of automatic means to search for
chemically relevant configurations is the
computational-kinetics equivalent of
improved electronic structure methods.”
- Basic Research Needs for Clean and
Efficient Combustion of 21st Century
Transportation Fuels. US Dept of Energy (2006)
I skipped this slide, because the previous talk had made the motivation clear.
Automatic TS searches remain an important
energy research goal
!
First Annual Conference of the
Combustion Energy Frontier Research
Center (CEFRC)
September 23-24, 2010
Princeton
“...transformation from by-hand
calculations of single reactions
to automated calculations of
millions of reactions would be a
game-changer for the field of
chemistry, and would be a good
‘Grand Challenge’ target...”
- Combustion Energy Frontier
Research Center (2010)
I skipped this slide too.
Can you predict TS geometries
from molecular groups alone?
!
(this would be great)
Can you predict TS geometries
from molecular groups alone?
Radical
Molecule
Length of bond being broken, at TS for Hydrogen abstraction
Can you predict TS geometries
from molecular groups alone?
!"!#$
!"!%%
!"!&'
!"!($
Length of bond being broken, !")()
!"*+$
!"*!#
at TS for!")'&
Hydrogen
abstraction
!")('
!")$%
!")'+
!"*+&
!"*%%
!"*&)
!"*&$
in Å with M06-2X/6-31+G(d,p)
Can you predict TS geometries
!"!#$
!"#$%
!"#$#
!"#$%
from molecular
groups alone?
!"!#$
!"!%%
!"!&'
!"!($
!")()
!")'&
!"*+$
!"*!#
!")('
!")$%
!")'+
!"*+&
!"*%%
!"*&)
!"*&$
in Å with M06-2X/6-31+G(d,p)
Guess which number goes in the blank?
You can predict TS geometries
from molecular groups alone!
!"!#$
!"!%%
!"!&'
!"!($
!")()
!")'&
!"*+$
!"*!#
!")('
!")$%
!"#$%
!"*%%
!")'+
!"*+&
!"*&)
!"*&$
in Å with M06-2X/6-31+G(d,p)
So you can predict distances! But… (1) I gave you 15 numbers and you gave be one; and (2) you only gave me a distance, not a 3D geometry.
Add the two numbers to get any distance
within 1 pm
0.525
0.657
0.607
0.622
0.658 0.666
0.680
0.691
0.525
0.657
0.607
0.622
0.658 0.666
0.680
0.691
in Å with M06-2X/6-31+G(d,p)
First challenge: getting more data out than you put in. Now, I give you 8 numbers and you can give me 16! Arrange these groups in a tree as mentioned before, and job done.
Use distance geometry to position atoms
RMG
⇌
Molecule
Connectivity
3D
Structure
Second challenge: getting from a distance to a 3D structure. We use the distance-geometry algorithms in RDKit.
Use distance geometry to position atoms
RMG
⇌
Embed
in 3D
Molecule
Connectivity
Atoms List
Atoms List
Upper limits
Lower limits
Generate bounds
matrix
Second challenge: getting from a distance to a 3D structure. We use the distance-geometry algorithms in RDKit.
Use distance geometry to position atoms
RMG
⇌
Embed in 3D
Atoms List
Atoms List
Molecule
Connectivity
Atoms List
Atoms List
Upper limits
Lower limits
Edit
bounds matrix
Generate bounds
matrix
Second challenge: getting from a distance to a 3D structure. We use the distance-geometry algorithms in RDKit.
Estimate geometry directly via
group additive distance estimates
Generate
Edit Bounds
Bounds
Matrix!
close to TS
Matrix
Generate
Edit Bounds
Bounds
Matrix!
Reactants
close to TS
Matrix
Products
Reaction
from RMG
Generate
Bounds
Matrix
Embed
Matrix in
3D Embed
Matrix in
3D
Edit Bounds
Matrix!
for TS
Doubleended
Search
Optimize TS
geometry
IRC!
Calculation
Embed
Matrix in
3D
• Database arranged in tree structure as for kinetics
• Trained on successfully optimized transition states
• Direct guess much faster than double ended search
• Success depends on training data
Test the group-additive method on H-abstraction reactions in Diisopropyl Ketone
Allen, J. W. et al. Com
bust. Flame 161, 711
–724 (2014).
Combustion and
Flame 161 (2014)
711–724
Contents lists available
at ScienceDirect
Combustion and
Flame
journal home
page: www.e
lsevie
r.com/locate/
c o m b u s t fl a m
e
A coordinated inves
tigation of the comb
ustion chemistry
ketone, a prototype
of diisopropyl
for biofuels produ
ced by endophyti
Joshua W.
a
c fungi
Allen , Adam M.
Scheer b, Connie
John D. Savee b,
W. Gao a, Shame
David L. Osborn b
l S. Merchant a, Subith
, Changyoul Lee d
Ravi X. Fernandes d,f
S. Vasu c, Oliver Welz b
, Stijn Vranckx d,
, William H. Green a,⇑
,
Zhandong Wang e
, Masood Z. Hadi g,1
, Fei Qi e,
Department of Chemical
, Craig A. Taatjes b,⇑
Engineering, Massachus
Combustion Research
etts Institute of Technology
Facility,
a
b
Sandia National
, Cambridge, MA
Mechanical and
Laboratories Livermore,
02139, USA
Aerospace Engineerin
CA 94551, USA
g, University of Central
Physico-Chemical
Florida, Orlando,
Fundamentals of
FL 32708, USA
Combustion, RWTH
National Synchrotro
Aachen University,
n Radiation Laboratory
52056 Aachen, Germany
, University of Science
Physikalisch-Technisc
he Bundesanstalt,
and Technology of
Bundesallee 100,
China, Hefei, Anhui
Biomass Conversion
D38116 Braunschw
230029, PR China
Technologies, Sandia
eig, Germany
National Laboratorie
s, Livermore, CA
94551, USA
c
d
e
f
g
a r t i c l e
i n f o
Article history:
Received 25 July
2013
Received in revised
form 18 September
2013
Accepted 18 October
2013
Available online
20 November 2013
Keywords:
Diisopropyl ketone
Automatic mechanism
generation
Ignition delay
Pyrolysis
Combustion
Detailed kinetics
modeling
1. Introduction
a b s t r a c t
Several classes
of endophytic fungi
have been recently
range of ketones
identified that convert
and other oxygenat
ed molecules, which
cellulosic biomass
oxidation chemistry
to a
are potentially viable
is not yet well
understood. In
as biofuels, but
describing the pyrolysis
this work, we present
whose
and oxidation of
a predictive kinetics
using the Reaction
diisopropyl ketone
model
Mechanism Generato
(DIPK) that was
generated automati
r (RMG) software
against three experime
cally
package. The model
nts that cover a
predictions are evaluated
range of temperat
(1) Synchrotron
photoionization
ures, pressures,
and oxygen concentra
mass spectrometry
800–1340 K at 30
(PIMS) measurem
tions:
Torr and 760 Torr;
ents of pyrolysis
(2) Synchrotron
oxidation from 550
in
the
PIMS
range
measurem
K to 700 K at 8 Torr;
ents of laser photolyti
and (3) Rapid-com
delay between 591
c Cl-initiated
pression machine
K and 720 K near
measurements of
10 bar. Improvem
in the areas of
ignition
ents made to the
hydrogen abstractio
model paramete
n from the initial
chemistry, are discussed
rs, particularly
DIPK molecule
. Our ability to
and low-temperature
automatically generate
its parameters without
peroxy
this model and
fitting to the experime
systematically improve
chemical kinetics
ntal results demonstr
paradigm.
ates the usefulnes
s of the predictive
! 2013 The Combusti
on Institute. Published
by Elsevier Inc.
All rights reserved.
Liquid petroleu
m-based fuels account
world energy consump
for more than
30% of
tion, the majority
occurring in the
tation sector, where
transporliquid fuels dominat
e because of their
ergy density and
high enease of storage
and distribution
insecurity of oil
[1]. However,
supply and negative
impacts of climate
are motivating
increased developm
change
ent of alternative
[2]. Lignocellulosic
liquid fuels
biofuels are a promisin
to conventional
g renewable alternati
petroleum fuels
ve
that could lower
emissions. Due
greenhouse gas
to the recalcitr
ance of lignocell
new means for
ulosic materials,
breaking down
biomass into fermenta
are a central area
ble sugars
of biofuels research
[3]. However, the
products
⇑
that are most efficient
ly produced from
well-characterize
biomass may not
d combustion properti
have
es. Furthermore,
clean, efficient combust
advanced
ion strategies, particula
on compression
rly those that rely
ignition, are often
very sensitive to
[4,5]. To address
fuel chemistry
these issues compreh
ensively, coordina
towards biofuel-e
ted efforts
ngine co-development
are required.
Figure 1 depicts
the Sandia collabor
ative biofuel developm
strategy. Syntheti
c biologists working
ent
in close collabor
combustion research
ation with
ers develop fundame
the production and
ntal mechanisms
combustion of potentia
for
engine trials then
l biofuels. Ignition
provide feasibilit
and
y tests for fuels
and yield recomm
and mixtures
endations for the
bioengineering scale-up
cific metabolic
pathways. This
of specoupling of fundame
plied combustion
ntal and apchemistry and syntheti
identify and investiga
c biology is a strategy
to
te the most promisin
through mutual
g fuel compou
feedback. Moreove
nds
r, the developm
tion models will
ent of combusprovide the predictiv
e capability needed
tual efficient utilizatio
for evenn of the new
biofuel stream.
A similar
Corresponding authors.
E-mail addresses:
[email protected]
Taatjes).
(W.H. Green),
[email protected]
1
(C.A.
Present address:
NASA Ames Research
Center, Moffett Field,
CA 94035, USA.
0010-2180/$ - see
front matter ! 2013
The Combustion
http://dx.doi.org/10.1
Institute. Published
016/j.combustflame.2
by Elsevier Inc.
013.10.019
All rights reserved.
These data were hot off the press. We tried finding every H-abstraction reaction in this large published kinetic model.
Parity plots of Estimated against Optimized
X-H distances at the transition state
Estimated Distance XH (Å)
1.6
Parity
0.7
0.7
1.6
Optimized Distance XH (Å)
We’ll see how well we predict the distances as we re-train the groups on more and more data.
Trained on 44 reactions,
estimates not so great…
Estimated Distance XH (Å)
1.6
Training Data
44
Parity
0.7
0.7
1.6
Optimized Distance XH (Å)
Trained on 148 reactions,
estimates improve…
Estimated Distance XH (Å)
1.6
Training Data
44
148
Parity
0.7
0.7
1.6
Optimized Distance XH (Å)
Trained on 230 reactions,
estimates improve…
Estimated Distance XH (Å)
1.6
Training Data
44
148
230
Parity
0.7
0.7
1.6
Optimized Distance XH (Å)
Trained on 767 reactions,
estimates are very good.
Estimated Distance XH (Å)
1.6
Training Data
44
148
230
767
Parity
0.7
0.7
1.6
Optimized Distance XH (Å)
• See Pierre
Bhoorasingh's
poster on
Wednesday
For more details, talk to Pierre (who was up all night generating those results)
Interpreting Existing Kinetic Models
Grant No. 1403171
.edu/comocheng
37
Now for the second project, that is also built on RMG-Py.
Proceedings
of the
Proceedings of
the Combustion
Institute 31 (200
7) 125–140
Combustion
Institute
www.elsevier.com
Transforming da
ta into knowledg
e—Pr
Informatics for
combustion chem ocess
istry
Michael Frenkla
ch
/locate/proci
*
Department of Mec
hanical Enginee
ring, University
of California, and
Division, Lawrenc
Environmental Ene
e Berkeley Nationa
rgy Technologies
l Laboratory, Ber
keley, CA 94720,
USA
Abstract
The present fron
tier of combus
namely, chemic
al kinetics models tion chemistry is the develop
ties. While the
usual factors like capable of accurate numerical ment of predictive reaction
models,
pre
individual measur
deficient knowle
dge of reaction dictions with quantifiable unc
ements and/or
ertainsistency of accum
the
pat
ulating data and oretical calculations impede pro hways and insufficient accura
new paradigm.
cy of
gress, the key obs
proliferating rea
It relies on thre
ctio
tacl
n
e is the inconmechanisms. Pro
e major compon
of scientific too
cess
Info
ls for analysis and
ents:
rma
munity in the dat
processing of the proper organization of scientifi tics introduces a
a collection and
se
c
dat
dat
a,
a,
and
ava
ilability
engage
tific method by
analysis. The pro
considering the
per infrastructure ment of the entire scientific com
entire content
scientific consist
will enable a new
of information
ency of the dat
form of scienava
a,
available data,
evaluating model rigorously assessing data unc ilable, assessing and assuring
mutual
ertainty, identify
the highest pos
pre
dict
abi
lity
,
sible impact, rea
sug
ing problems wit
ching community gesting new experimental and
knowledge.
h the
the
consensus, and
! 2006 The Com
merging the asse oretical work with
mbled data into
bustion Institute.
new
Published by Els
evier Inc. All righ
Keywords: Kinetic
ts reserved.
s; Modeling; Opt
n; Consistency;
Informatics; PrIM Symposium in 2006, Again, we go back imiz
to atio
my
first Combustion
e
where I was introduced to the idea of Process Informatics / Data Collaboration.
1. Introduction
1.1. Progress in
combustion dep
reaction kinetic
ends on reliable
s
Research is mo
explain phenom tivated by the human desire to
ena surrounding
us and the passion
for new discove
ries
tion such activiti . Driven by individual satisfac
es
of human experie aim at improving the quality
nce. One can gen
eralize the stimulus for research
as developing the
ability of making
predictions. Indeed
, predicting proper
ties of a mate*
Fax: +1 510 643
5599.
E-mail address:
[email protected]
ey.e
rial before it is syn
thesized, predict
of a device before
ing performanc
e
it
nitude of a natura is built, or predicting the magl disaster ahead
obvious benefit
of time all have
s to society.
In the area of com
elements: the fasc bustion, we have all the above
times, its broad ination with fire from ancient
application to num
of everyday life
erous aspects
, and the definite
predictions (see
need for reliabl
e
Fig. 1). In recent
decades the com
bustion researc
h was largely driv
en by the desired
increase in energy
efficiency, reduct
formation, and
ion in pollutant
material synthe
sis.
tying the researc
h to practical app The need for
lica
ther emphasized
in the Hottel add tions was furSymposium by
Professor Glassm ress of the 28th
an [1].
du
1540-7489/$ - see
front matter !
doi:10.1016/j.proci
2006 The Combus
.2006.08.121
tion Institute. Pub
lished by Elsevier
Inc. All rights rese
rved.
“the key obstacle is the inconsistency of accumulating
data and proliferating reaction mechanisms.
Process Informatics introduces a new paradigm. It relies
on three major components: • proper organization of scientific data,
• availability of scientific tools for analysis and
processing of these data,
• and engagement of the entire scientific community
in the data collection and analysis.”
M. Frenklach, Proc. Combust. Inst. 31 (2006)
I assumed the entire scientific community would go home and properly organize their scientific data…
Model Complexity is Increasing
Number of species in model
10000
2-methylalkanes/Sarathy
Methyloleate/Naik
Heptamethylnonane/Westbrook
1000
100
Methydecanoate/Herbinet
n-Octane/Ji
n-Heptane/Curran
n-Heptane/Mehl
n-Heptane/Babushok
Methylhexanoate/Westbrook
Ethylbenzene/Therrien
Neopentane/Wang
1-butanol/Grana
n-Butane
Furan/Yasunaga
USC-Mech
Marinov
Methylcrotonate/Bennadji
C3/Appel
Ethane
Isobutene/Dias
Propane/Qin
Acetaldehyde/Leplat
C2/Karbach GRI-Mech
Methane/Leeds
Ethanol/Leplat
Ethylene/Egolfopoulos
10
1995
H2/Marinov
H2/Connaire
H2/Klippenstein
2000
2005
2010
Year of publication
So is everything published since 2006 in PrIMe format? No. It’s mostly in CHEMKIN format. (This chart is very incomplete random sampling)
And again, here is a photo of a large federal computer. This time at NASA
This is the computer model Gordon and McBride used when they devised the NASA polynomial format
for thermochemistry (now used in CHEMKIN). It was designed to fit on 80-column punch-cards.
This is the computer model Gordon and McBride used when they devised the NASA polynomial format
for thermochemistry (now used in CHEMKIN). It was designed to fit on 80-column punch-cards.
APPENDIX D
LISTING)
THERMO DATA (FORMAT AND
The
order
in the follow-
given
are
appendix
in this
cards
data
of the input
and format
ing table"
Card
Format
Contents
Card
column
order
1 to6
3A4
THERMO
ranges
Temperature
for
T
highest
name
Species
Date
of species
Phase
or
(S,
or G for
in equations
Coefficients
(90)
80
1 to75
5(E15.8)
(90) to (92)
80
1 to75
a T for
(a6,
to (92)
and a 3
and a 1, a2,
interval,
temperature
upper
for lower)
in equations
Coefficients
a, i for
lower
cards
Repeat
(90)
(a4,
to (92)
(Indicates
data)
end of thermodynamic
I20
8O
3A4
ito3
each
1 to 4 in cc 80 for
numbered
species
1 to 60
4(E15.8)
a6,
a5,
interval)
temperature
4
Integer
80
15
3
Integer
END
45
46 to 65
2
Integer
(Final
25 to 44
I15
ai(i = 1 to 5) in equations
interval)
temperature
Coefficients
(for upper
19 to 24
2F10.3
range
t
Integer
1 to 12
2A3
A1
liquid,
solid,
respectively)
gas,
Temperature
(a)
L,
3A4
4(A2, F3. O)
and formula
symbols
Atomic
1 to 30
3F10.3
of coefficients:
2 sets
T, and
common
T.
lowest
card)
aGaseous
any
species,
to their
and condensed
species
However.
order.
species
must
their
the
be adjacent.
sets
temperature
must
sets
species
for
If there
be either
are
one
only
with
two or
condensed
condensed
more
more
than
in increasing
phases
two condensed
can be in
phase
of the
phases
same
of a
order
according
ORIGINAL
PAGE
or decreasing
intervals.
IS
OF. POOR QUALITY
168
nq
This is the computer model Gordon and McBride used when they devised the NASA polynomial format
for thermochemistry (now used in CHEMKIN). It was designed to fit on 80-column punch-cards.
NASA (Chemkin) format is very dense –
not much room for species identifiers
Species Name Chemical Formula Parameters for H(T), S(T)
Space constraints led to “creative” naming schemes
C3KET21
MVOX
IIC4H7Q2-T
C3H5-A
C4H8O1-3
C6H101OOH5-4
CH3COCH2O2H
C3H5-S
TC4H8O2H-I
Notice hydroperoxypropan-2-one has two different names, often in the same mechanism file!
Space constraints led to “creative” naming schemes
C3KET21
MVOX
IIC4H7Q2-T
C3H5-A
C4H8O1-3
C6H101OOH5-4
CH3COCH2O2H
C3H5-S
TC4H8O2H-I
• C3KET21 is generated from alkyl peroxy radical
isomerization pathway
• CH3COCH2O2H is generated from low-temperature
oxidation of acetone
Notice hydroperoxypropan-2-one has two different names, often in the same mechanism file!
Human-Computer team can
identify species more quickly
proposed
matches
confirmed
matches
Human
Computer
Review evidence,
confirm matches.
confirmed
matches
Generate reactions,
check for equivalents,
propose matches
proposed
matches
• Our new tool uses RMG to generate reactions to
compare with the target model
• A human reviews the evidence and confirms matches.
Identify small molecules first
• Identify species with only one possible structure
• CO2
• H2O
• C3H8
• Then species with "borrowed" thermochemistry
• Then boot-strap based on how these react
The first thing the tool will do is to check previously imported models for matching thermochemistry
blocks. For example, it could tell the user that the LLNL model for n-Heptane [29] has a species with
the same parameters that has already been identified as but-2-enoyl (
), and that LLNL
called it “SC3H5CHO”. This would probably be enough evidence for the user to confirm the match.
Identifying
‘sc3h5co’ from its reactions
If the species cannot be found in a previously imported model, the reactions give additional clues.
The seven reactions containing sc3h5co in the methyl butanoate model are:
R1
sc3h5cho + o2 ⇌ sc3h5co + ho2
+
⇌
R2
sc3h5cho + oh ⇌ sc3h5co + h2o
+
⇌
R3
sc3h5cho + o ⇌ sc3h5co + oh
+
R4
sc3h5cho + ch3 ⇌ sc3h5co + ch4
+
R5
sc3h5cho + h ⇌ sc3h5co + h2
+
R6
sc3h5cho + ho2 ⇌ sc3h5co + h2o2
R7
sc3h5co ⇌ c3h5-s + co
The first six are all hydrogen abstractions
from but-2-enal (assume for now that this
species has already been identified), which
has four types of hydrogen atom, implying
sc3h5co could be one of four possible radicals. The seventh reaction is the decomposition into propenyl and carbon monoxide,
also limiting sc3h5co to four possible species. The Venn diagram in Fig. 4 shows that
only one species satisfies all seven reactions:
but-2-enoyl. The tool will present the user
with this data, as well as a comparison of
+
sc3h5co ⇌
+ sc3h5co
+ sc3h5co
⇌
+ sc3h5co
⇌
+ sc3h5co
⇌
+ sc3h5co
⇌
+ sc3h5co
+
Fig. 4. Venn diagram showing identification of
“sc3h5co” from reactions R1 through R7.
The first 6 reactions are all H-abstractions giving 4 possibilities for sc3h5co. 7
The 7th reaction is beta-scission,
also giving 4 possibilities.
The first thing the tool will do is to check previously imported models for matching thermochemistry
blocks. For example, it could tell the user that the LLNL model for n-Heptane [29] has a species with
the same parameters that has already been identified as but-2-enoyl (
), and that LLNL
called it “SC3H5CHO”. This would probably be enough evidence for the user to confirm the match.
Identifying
‘sc3h5co’ from its reactions
If the species cannot be found in a previously imported model, the reactions give additional clues.
The seven reactions containing sc3h5co in the methyl butanoate model are:
R1
sc3h5cho + o2 ⇌ sc3h5co + ho2
+
⇌
R2
sc3h5cho + oh ⇌ sc3h5co + h2o
+
⇌
R3
sc3h5cho + o ⇌ sc3h5co + oh
+
R4
sc3h5cho + ch3 ⇌ sc3h5co + ch4
+
R5
sc3h5cho + h ⇌ sc3h5co + h2
+
R6
sc3h5cho + ho2 ⇌ sc3h5co + h2o2
R7
sc3h5co ⇌ c3h5-s + co
The first six are all hydrogen abstractions
from but-2-enal (assume for now that this
species has already been identified), which
has four types of hydrogen atom, implying
sc3h5co could be one of four possible radicals. The seventh reaction is the decomposition into propenyl and carbon monoxide,
also limiting sc3h5co to four possible species. The Venn diagram in Fig. 4 shows that
only one species satisfies all seven reactions:
but-2-enoyl. The tool will present the user
with this data, as well as a comparison of
+
sc3h5co ⇌
+ sc3h5co
+ sc3h5co
⇌
+ sc3h5co
⇌
+ sc3h5co
⇌
+ sc3h5co
⇌
+ sc3h5co
+
Fig. 4. Venn diagram showing identification of
“sc3h5co” from reactions R1 through R7.
The first 6 reactions are all H-abstractions giving 4 possibilities for sc3h5co. 7
The 7th reaction is beta-scission,
also giving 4 possibilities.
The first thing the tool will do is to check previously imported models for matching thermochemistry
blocks. For example, it could tell the user that the LLNL model for n-Heptane [29] has a species with
the same parameters that has already been identified as but-2-enoyl (
), and that LLNL
called it “SC3H5CHO”. This would probably be enough evidence for the user to confirm the match.
Identifying
‘sc3h5co’ from its reactions
If the species cannot be found in a previously imported model, the reactions give additional clues.
The seven reactions containing sc3h5co in the methyl butanoate model are:
R1
sc3h5cho + o2 ⇌ sc3h5co + ho2
+
⇌
R2
sc3h5cho + oh ⇌ sc3h5co + h2o
+
⇌
R3
sc3h5cho + o ⇌ sc3h5co + oh
+
R4
sc3h5cho + ch3 ⇌ sc3h5co + ch4
+
R5
sc3h5cho + h ⇌ sc3h5co + h2
+
R6
sc3h5cho + ho2 ⇌ sc3h5co + h2o2
R7
sc3h5co ⇌ c3h5-s + co
The first six are all hydrogen abstractions
from but-2-enal (assume for now that this
species has already been identified), which
has four types of hydrogen atom, implying
sc3h5co could be one of four possible radicals. The seventh reaction is the decomposition into propenyl and carbon monoxide,
also limiting sc3h5co to four possible species. The Venn diagram in Fig. 4 shows that
only one species satisfies all seven reactions:
but-2-enoyl. The tool will present the user
with this data, as well as a comparison of
+
sc3h5co ⇌
+ sc3h5co
+ sc3h5co
⇌
+ sc3h5co
⇌
+ sc3h5co
⇌
+ sc3h5co
⇌
+ sc3h5co
+
Fig. 4. Venn diagram showing identification of
“sc3h5co” from reactions R1 through R7.
The first 6 reactions are all H-abstractions giving 4 possibilities for sc3h5co. 7
The 7th reaction is beta-scission,
also giving 4 possibilities.
The first thing the tool will do is to check previously imported models for matching thermochemistry
blocks. For example, it could tell the user that the LLNL model for n-Heptane [29] has a species with
the same parameters that has already been identified as but-2-enoyl (
), and that LLNL
called it “SC3H5CHO”. This would probably be enough evidence for the user to confirm the match.
Identifying
‘sc3h5o’ from its reactions
If the species cannot be found in a previously imported model, the reactions give additional clues.
The seven reactions containing sc3h5co in the methyl butanoate model are:
R1
sc3h5cho + o2 ⇌ sc3h5co + ho2
+
⇌
R2
sc3h5cho + oh ⇌ sc3h5co + h2o
+
⇌
R3
sc3h5cho + o ⇌ sc3h5co + oh
+
R4
sc3h5cho + ch3 ⇌ sc3h5co + ch4
+
R5
sc3h5cho + h ⇌ sc3h5co + h2
+
R6
sc3h5cho + ho2 ⇌ sc3h5co + h2o2
R7
sc3h5co ⇌ c3h5-s + co
The first six are all hydrogen abstractions
from but-2-enal (assume R7
for now that this
species has already been identified), which
has four types of hydrogen atom, implying
sc3h5co could be one of four possible radicals. The seventh reaction is the decomposition into propenyl and carbon monoxide,
also limiting sc3h5co to four possible species. The Venn diagram in Fig. 4 shows that
only one species satisfies all seven reactions:
but-2-enoyl. The tool will present the user
with this data, as well as a comparison of
+
sc3h5co ⇌
+ sc3h5co
+ sc3h5co
⇌
+ sc3h5co
⇌
+ sc3h5co
⇌
+ sc3h5co
⇌
+ sc3h5co
+
R1-R6
Fig. 4. Venn diagram showing identification of
“sc3h5co” from reactions R1 through R7.
By considering all seven reactions, we can deduce the most likely isomer.
7
‘sc3h5o’ is but-2-enoyl
Interact with a Web UI
• Computers make mistakes, and humans can see
patterns
• All matches must be approved by an operator using a
web user interface
Two undergraduates this summer imported models from recent volumes of the journals Combustion & Flame and Proceedings of the Combustion Institute. More details on our poster.
Two undergraduates this summer imported models from recent volumes of the journals Combustion & Flame and Proceedings of the Combustion Institute. More details on our poster.
Identified 8,299 species from 58 models
• Malformed chemkin files
• Incorrect glossary entries
• 100 kJ/mol disagreements in enthalpies of formation
• 30 orders of magnitude disagreements in rates
• Only about 10% pressure-dependent
We frequently hear we can get 1 kJ/mol errors in energy and 30%
30 in rate constants, yet our supplementary material differ by 100 kJ/mol and 10 respectively! These discrepancies are usually undetected.
Some recent developments in RMG
• Developer-friendly "RMG-Py" being released
• New features (Nitrogen!)
• Web tools
• Group-Additive Transition State Estimates
• Fast estimates of TS geometries
• Gets better at guessing the more it guesses
• Mechanism Importer tool
• Facilitates identification of species in chemkin files
• Started to collect and curate data
Final Summary!
Richard H. West
4 August 2014
[email protected]
richardhwest
rwest
.edu/comocheng
54
And of course, http://www.slideshare.net/richardhwest