-
SHOCK TUBE STUDY OF NITROGEN-CONTAINING
FUELS
A DISSERTATION
SUBMITTED TO THE DEPARTMENT OF MECHANICAL ENGINEERING
AND THE COMMITTEE ON GRADUATE STUDIES
OF STANFORD UNIVERSITY
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
Sijie Li
June 2014
-
http://creativecommons.org/licenses/by-nc/3.0/us/
This dissertation is online at:
http://purl.stanford.edu/nw843zs8610
© 2014 by Sijie Li. All Rights Reserved.
Re-distributed by Stanford University under license with the
author.
This work is licensed under a Creative Commons
Attribution-Noncommercial 3.0 United States License.
ii
http://creativecommons.org/licenses/by-nc/3.0/us/http://creativecommons.org/licenses/by-nc/3.0/us/http://purl.stanford.edu/nw843zs8610
-
I certify that I have read this dissertation and that, in my
opinion, it is fully adequatein scope and quality as a dissertation
for the degree of Doctor of Philosophy.
Ronald Hanson, Primary Adviser
I certify that I have read this dissertation and that, in my
opinion, it is fully adequatein scope and quality as a dissertation
for the degree of Doctor of Philosophy.
Craig Bowman
I certify that I have read this dissertation and that, in my
opinion, it is fully adequatein scope and quality as a dissertation
for the degree of Doctor of Philosophy.
David Davidson
I certify that I have read this dissertation and that, in my
opinion, it is fully adequatein scope and quality as a dissertation
for the degree of Doctor of Philosophy.
Hai Wang
Approved for the Stanford University Committee on Graduate
Studies.
Patricia J. Gumport, Vice Provost for Graduate Education
This signature page was generated electronically upon submission
of this dissertation in electronic format. An original signed hard
copy of the signature page is on file inUniversity Archives.
iii
-
iv
-
v
Abstract
The combustion chemistry of nitrogen-containing fuels is
important in the study
of bio-derived fuels and nitrogen-based propellants. However,
little high-quality shock
tube kinetics data exists for these systems. The primary
objective of the research
presented in this dissertation is to augment the experimental
database and to improve
understanding of the chemical kinetics for four
nitrogen-containing fuels: morpholine,
dimethylamine, ethylamine and monomethylhydrazine.
Morpholine (C4H9NO, 1-oxa-4-aza-cyclohexane) is a good
representative
candidate of a nitrogen-containing fuel because of its cyclic
structure and wide
industrial applications. Morpholine ignition delay times were
measured behind reflected
shock waves. A morpholine mechanism was developed based on this
shock tube study
and previous works in the literature. The simulations from this
morpholine mechanism
were in good agreement with the current morpholine experiments
as well as previous
morpholine flame data. Refinement of this morpholine mechanism
required
improvements in the sub-mechanisms of two major intermediate
species dimethylamine
and ethylamine, as discussed in a progressive manner in this
dissertation.
The overall rate constants of hydroxyl radicals (OH) with
dimethylamine
(DMA: CH3NHCH3) and ethylamine (EA: CH3CH2NH2) were measured
behind
reflected shock waves using UV laser absorption of OH radicals
near 306.7 nm. The
overall rate constants were determined by fitting the measured
OH time-histories with
the computed profiles using the detailed dimethylamine and
ethylamine sub-
mechanisms contained in the morpholine mechanism. Variational
transition state theory
was used to compute the H-abstraction rates by OH for
dimethylamine and ethylamine.
The calculated reaction rate constants are in good agreement
with the experiment. The
calculated reaction rate constants were used to update the
morpholine mechanism for
simulations in the following sections.
-
vi
Dimethylamine (DMA) ignition delay times and OH time-histories
were
investigated behind reflected shock waves. The dimethylamine
ignition delay time
measurements were carried out in 4% oxygen/argon. OH
time-histories were measured
in stoichiometric mixtures of 500 ppm DMA/O2/argon. The
morpholine mechanism was
then updated by adding the DMA unimolecular decomposition
channel: DMA =
CH3NH + CH3. With this modification, the simulation results are
in excellent agreement
with both the dimethylamine ignition delay times and OH
time-history data.
Ethylamine (CH3CH2NH2) pyrolysis and oxidation were studied
behind
reflected shock waves. For ethylamine pyrolysis, NH2
time-histories were measured in
2000 ppm ethylamine/argon mixtures. For ethylamine oxidation,
ignition delay times,
NH2 and OH time-histories were measured in ethylamine/O2/argon
mixtures. By fitting
the simulations to the early time-histories of NH2 and OH, the
rate constants for the two
major ethylamine decomposition pathways in the morpholine
mechanism were updated
for better agreement with the experiment. In addition,
recommendations from recent
theoretical studies of ethylamine radical reactions were
implemented. With these
modifications, the final updated morpholine mechanism provides
significantly
improved agreement with the species time-history measurements
and the ignition delay
times of ethylamine.
The morpholine mechanism, after implementing the aformentioned
updates
based on the dimethylamine and ethylamine data, was compared
with the morpholine
ignition delay time data again. It was shown that those
modifications improve the
agreements of the mechanism with the morpholine data.
Amine groups are common structural features for rocket
propellants as well, and
using the same approach as above, the pyrolysis of an important
propellant
monomethylhydrazine (MMH) was studied using NH2 time-histories
in MMH/argon
mixtures. The MMH pyrolysis mechanism developed by Sun et al.
(2009), with the
updates by Cook et al. (2011), was used to compare with the
experiment. The rate
constant of the reaction: MMH = CH3N.H + NH2 was determined
based on early time of
-
vii
the NH2 time-histories. Pressure dependence of this reaction was
observed at 0.3-5 atm.
The measured reaction rate constants follow a pressure
dependence trend close to the
theoretical results by Zhang et al. (2011) based on transition
state theory master
equation analysis. Using the high and low-pressure limit
expressions by Zhang et al., a
new Troe’s expression in the fall-off region was proposed based
on the current
experimental data. Utilizing the later times of the NH2
time-histories, a new reaction
rate expression was recommended for the reaction: NHNH2 + H =
NH2 + NH2.
-
ix
Acknowledgments
First, I would like to thank my advisor, Prof. Ronald K. Hanson,
for his
generous support during my PhD process. Without his constant
encouragement and
suggestions, I doubt I would ever reach this point. I also want
to thank Dr. David F.
Davidson for offering generous advice and guidance throughout my
time in Stanford,
not only about research but also about life in general. I am
also thankful to Prof.
Bowman and Prof. Wang for serving on my reading committee.
Every time when things are not going well in my life as a PhD
candidate, I often
look into the thesis database of our group. I was not looking
for solutions, but went
directly to the thesis acknowledgements. I want to see what
previous students had in
mind, did they experience similar disappointment, fake hope,
frustration, excitement,
happiness...? Whom did they want to say thanks to? How did it
feel like when they
finally started to write their thesis? And more hauntingly, I
always asked myself
whether I will ever be able to write my own PhD acknowledgement?
What I will say?
I forgot in which thesis acknowledgement I read this line "I
don’t like the idea of
listing people’s name in the acknowledgement, because every list
starts somewhere and
has an end". I’m grateful to all the previous and current
members I met in the Hanson
group. I still remember all the homework we finished together,
the data we collected as
a team and the free time we spent together. I’m also grateful to
all my friends; you made
my life in Stanford more colorful. Lastly, I want to say thank
you to my family and
Ting. Thank you for having faith in me, even when I am lacking
belief myself. Thank
you for reminding me that people with steady, instead of fast
pace, finish the Marathon.
Thank you for encouraging me to try and never be afraid of
losing.
-
xi
Table of Contents
Abstract
................................................................................................................................v
Acknowledgments..............................................................................................................
ix
Table of Contents
...............................................................................................................
xi
List of Tables
..................................................................................................................
xvii
List of Figures
..................................................................................................................
xix
Chapter 1. Introduction
..................................................................................................1
1.1. Background and Motivation
........................................................................1
1.2. Overview of Thesis
......................................................................................3
Chapter 2. Experimental
Method...................................................................................5
2.1. Shock Tube Facility
.....................................................................................5
2.1.1. Stanford High Pressure Shock Tube
(HPST)...................................6
2.1.2. Stanford Kinetic Shock Tube
(KST)................................................7
2.1.3. Stanford NASA Shock Tube (NASA)
.............................................7
2.2. Laser Diagnostic
..........................................................................................8
2.2.1. Overview
..........................................................................................8
-
xii
2.2.2. IR Diagnostic of Fuel
......................................................................
9
2.2.3. NH2 Diagnostic
................................................................................
9
2.2.4. OH Diagnostic
...............................................................................
10
Chapter 3. Shock Tube and Modeling Study of Morpholine
...................................... 13
3.1. Introduction
...............................................................................................
13
3.2. Experimental Details
.................................................................................
13
3.3. Results and Discussion
..............................................................................
15
3.3.1. Morpholine Ignition Delay Times
................................................. 15
3.4. Model Development and Simulations
....................................................... 19
3.5. Summary
....................................................................................................
24
Chapter 4. Reactions of OH with Dimethylamine and Ethylamine
............................ 27
4.1. Introduction
...............................................................................................
27
4.2. Experimental Setup
...................................................................................
29
4.3. Kinetic Measurements
...............................................................................
30
4.3.1. Dimethylamine (DMA) + OH
....................................................... 31
4.3.2. Ethylamine (EA) + OH
..................................................................
36
4.4. Theoretical Study
.......................................................................................
42
4.5. Summary
....................................................................................................
46
Chapter 5. Dimethylamine Oxidation
.........................................................................
49
5.1. Introduction
...............................................................................................
49
-
xiii
5.2. Experimental Setup
....................................................................................50
5.3. Results and Discussion
..............................................................................50
5.3.1. Dimethylamine Ignition Delay Times
...........................................50
5.3.2. OH
Time-Histories.........................................................................54
5.3.3. Update to the Morpholine Mechanism
..........................................56
5.4. Summary
....................................................................................................63
Chapter 6. Ethylamine Pyrolysis and Oxidation
.........................................................65
6.1. Introduction
................................................................................................65
6.2. Experimental Methods
...............................................................................65
6.3. Experimental Results
.................................................................................66
6.3.1. Ethylamine Pyrolysis
.....................................................................66
6.3.2. Ethylamine Oxidation
....................................................................67
6.4. Update to the Morpholine Mechanism
......................................................71
6.5. Summary
....................................................................................................80
Chapter 7. Revisiting the Morpholine Data
.................................................................81
7.1. Introduction
................................................................................................81
7.2. Morpholine Ignition Delay Times
.............................................................82
7.3. Sensitivity Analysis
...................................................................................85
7.4. Summary
....................................................................................................86
Chapter 8. MMH Pyrolysis
..........................................................................................89
-
xiv
8.1. Introduction
...............................................................................................
89
8.2. Experimental Method
................................................................................
91
8.3. Results and Discussion
..............................................................................
92
8.4. Summary
..................................................................................................
106
Chapter 9. Summary and Future Work
.....................................................................
107
9.1. Summary of Results
................................................................................
107
9.1.1. Morpholine Oxidation
.................................................................
107
9.1.2. Dimethylamine and Ethylamine Combustion
............................. 108
9.1.3. MMH Pyrolysis
...........................................................................
110
9.2. Recommendations for Future Work
........................................................ 111
9.2.1. Dimethylamine and Ethylamine Pyrolysis
.................................. 111
9.2.2. Dimethylamine and Ethylamine Oxidation
................................. 112
9.2.3. Morpholine Pyrolysis and Oxidation
........................................... 113
9.3. Conclusion
...............................................................................................
113
Appendix A. Morpholine Oxidation Set
.........................................................................
115
Appendix B. Morpholine Pyrolysis Set
...........................................................................
119
Appendix C. Thermochemistry for Morpholine Species
................................................ 121
Appendix D. Computational Methods
.............................................................................
127
D.1. CHEMKIN Simulation
................................................................................
127
D.2. Multiwell Calculation
..................................................................................
128
-
xv
D.3. Gaussian
Calculation....................................................................................130
Bibliography
....................................................................................................................133
-
xvi
-
xvii
List of Tables
Table 3.1. Shock tube ignition delay times. Test gas
mixture:
morpholine/O2/argon........................................................................................15
Table 4.1. Reactions describing DMA and EA + OH experiments
...................................31
Table 4.2. Measured rate constants for DMA + OH = Products
.......................................35
Table 4.3. Measured rate constants for EA + OH = Products
...........................................40
Table 4.4. Summary of the zero Kelvin electronic energies and
rotational data
used for dimethylamine and ethylamine + OH VTST calculations
.................43
Table 4.5. The vibrational frequencies computed at the
BH&HLYP/6-
311++G(2d,2p) level of theory
........................................................................44
Table 4.6. VTST reaction rate constants of individual channels
for DMA and EA
+ OH.
...............................................................................................................45
Table 5.1. Ignition delay time data for dimethylamine
......................................................53
Table 5.2. Summary of rate recommendations to the dimethylamine
sub-
mechanism of the morpholine mechanism.
.....................................................59
Table 6.1. Updated reaction rate constants to ethylamine
sub-mechanism .......................72
-
xviii
Table 7.1. Modifications to the morpholine mechanism [54],
recommended in
Chapters 4-6
.....................................................................................................
81
Table 8.1. Measured reaction rate constants for the N-N bond
scission of MMH ............ 98
-
xix
List of Figures
Figure 2.1. Shock tube schematic.
.......................................................................................6
Figure 2.2. He-Ne laser diagnostic of fuel.
..........................................................................9
Figure 2.3. Ring dye laser diagnostic of NH2, using the
fundamental output of a
Spectra Physics 380 ring dye
laser...................................................................10
Figure 2.4. Ring dye laser diagnostic of OH, using the frequency
doubled output
of a Spectra Physics 380 ring dye laser.
..........................................................11
Figure 3.1. Sample pressure trace for morpholine ignition delay
time
measurements.
..................................................................................................15
Figure 3.2. Ignition delay time measurements in morpholine/air
mixtures near 15
atm and with equivalence ratios of 0.5, 1 and 2. Dashed lines:
linear
fit to data.
.........................................................................................................17
Figure 3.3. Ignition delay time measurements in morpholine/air Φ
= 1 mixtures
near 15 and 25 atm. Dashed lines: linear fit to data.
........................................18
Figure 3.4. Ignition delay time measurements in stoichiometric
morpholine/air
mixtures and morpholine/4% O2/argon mixtures around 15 atm.
Dashed lines: linear fit to data.
........................................................................19
Figure 3.5. Comparison of model predictions to morpholine/air
ignition delay
time measurements around 15 atm and under different
equivalence
-
xx
ratios. Dashed lines: Simulation results using the mechanism in
[8].
Solid lines: Simulation results using the morpholine mechanism
[54]. .......... 22
Figure 3.6. Comparison of model predictions to morpholine/air
ignition delay
time measurements for stoichiometric mixtures around 15 and 25
atm
respectively. Dashed lines: Simulation results using the
previous
mechanism [8]. Solid lines: Simulation results using the
morpholine
mechanism [54].
..............................................................................................
23
Figure 3.7. Comparison of model predictions to morpholine
ignition delay time
measurements around 15 atm for morpholine/4% O2/argon and
morpholine/air mixtures. Dashed lines: Simulation results using
the
previous mechanism [8]. Solid lines: Simulation results using
the
morpholine mechanism [54].
...........................................................................
24
Figure 4.1. Sensitivity analysis of OH using the dimethylamine
sub-mechanism
[25] with TBHP chemistry set, in the mixture of 320 ppm
DMA/Ar
with 22 ppm TBHP and 140 ppm H2O, at 1176 K and 0.9 atm.
..................... 33
Figure 4.2. Sample OH trace in 320 ppm DMA/Ar with 22 ppm TBHP
and 134
ppm H2O, at 1176 K and 0.9 atm.
...................................................................
33
Figure 4.3. Error analysis for measured k4.1 in 320 ppm DMA/Ar
with 22 ppm
TBHP and 134 ppm H2O, at 1176 K and 0.9 atm.
.......................................... 34
Figure 4.4. Measured overall reaction rate for k4.1: DMA+OH =
Products, in
comparison with the estimation by Lucassen et al. [25].
................................ 36
Figure 4.5. Sensitivity analysis of OH using the ethylamine
sub-mechanism [25]
with inclusion of TBHP chemistry, in a mixture of 470 ppm
EA,50
ppm TBHP, 190 ppm H2O, in Ar at 1067 K and 0.83 atm.
............................. 38
-
xxi
Figure 4.6. Sample OH trace in in a mixture of 470 ppm EA, 50
ppm TBHP, 190
ppm H2O, in Ar at 1067 K and 0.83 atm.
........................................................39
Figure 4.7. Error analysis for measured k4.2 in 470 ppm EA/Ar
with 50 ppm
TBHP and 190 ppm H2O, at 1067 K and 0.83 atm.
.........................................40
Figure 4.8. Measured overall reaction rates for EA + OH =
Products, in
comparison with the estimation by Lucassen et al. [25].
.................................42
Figure 4.9. Comparison of the measured reaction rates and
theoretical study
results for DMA + OH and EA + OH.
.............................................................46
Figure 5.1. Sample pressure traces for dimethylamine ignition
delay times. ....................51
Figure 5.2. Ignition delay times in stoichiometric DMA/4%
O2/argon mixtures,
with P ~ 0.9, 1.5 and 2.8 atm, measurements and simulation
results
using the morpholine mechanism described in Chapter 3 [54].
.......................52
Figure 5.3. Ignition delay times in stoichiometric DMA/4%
O2/argon mixtures at
P ~ 1.5 atm, with Φ = 0.5, 1, and 2, measurements and
simulation
results using the morpholine mechanism described in Chapter 3
[54]. ...........53
Figure 5.4. OH time-histories in stoichiometric mixture of 500
ppm
DMA/O2/argon at 1417 K and 2.2 atm, with simulation results
using
the morpholine mechanism described in Chapter 3 [54].
................................55
Figure 5.5. OH time-histories in stoichiometric mixture of 500
ppm
DMA/O2/argon at 1504 K and 2.1 atm, with simulation results
using
the morpholine mechanism described in Chapter 3 [54].
................................56
Figure 5.6. Sensitivity analysis of temperature, at initial
condition of 1300 K and
1.5 atm in a stoichiometric mixture of DMA/4% O2/argon, using
the
morpholine mechanism described in Chapter 3 [54].
......................................57
-
xxii
Figure 5.7. Ignition delay times in stoichiometric mixtures of
DMA/4%
O2/argon, at P ~ 0.9, 1.5 and 2.8 atm, measurements and
simulations
using the morpholine mechanism [54] with the modifications
in
Table 5.2.
.........................................................................................................
60
Figure 5.8. Ignition delay times in stoichiometric mixtures of
DMA/4%
O2/argon at P ~ 1.5 atm, with Φ = 0.5, 1, and 2, measurements
and
simulation results using the morpholine mechanism [54] with
the
modifications in Table 5.2.
..............................................................................
60
Figure 5.9. Measured dimethylamine ignition delay times in 4%
O2/argon,
scaled to Φ = 1 and P = 1.5 atm, in comparison with the
simulations
using the morpholine mechanism [54] with and without the
changes
recommended in Table 5.2.
.............................................................................
61
Figure 5.10. Comparison of the simulated OH time-histories,
using the
morpholine mechanism [54] with and without the modifications
in
Table 5.2, to the experiment in stoichiometric mixture of 500
ppm
DMA/O2/argon at 1417 K and 2.2 atm.
........................................................... 62
Figure 5.11. Comparison of the simulated OH time-histories,
using the
morpholine mechanism [54] with and without the modifications
in
Table 5.2, to the experiment in stoichiometric mixture of 500
ppm
DMA/O2/argon at 1504 K and 2.1 atm.
........................................................... 63
Figure 6.1. NH2 time-histories in 2000 ppm ethylamine/Ar
mixtures,
measurements and simulation results using the morpholine
mechanism [54] reported in Chapter 3.
........................................................... 67
Figure 6.2. Ethylamine ignition delay time measurements near
0.85, 1.35 and 2
atm in stoichiometric mixture of ethylamine/4% O2/Ar, and
simulations based on the morpholine mechanism [54] in Chapter 3.
.............. 68
-
xxiii
Figure 6.3. Ethylamine ignition delay time measurements near
1.35 atm, with Φ
= 0.75, 1 and 1.25, and simulation results using the
morpholine
mechanism [54] presented in Chapter 3.
..........................................................69
Figure 6.4. NH2 time-histories in 2000 ppm ethylamine/0.8% O2/Ar
mixtures;
simulations are based on the morpholine mechanism presented
in
Chapter 3.
.........................................................................................................70
Figure 6.5. OH time-histories in 500 ppm ethylamine/0.2% O2/Ar
mixtures,
measurements and simulations using the morpholine mechanism
[54]
presented in Chapter 3.
....................................................................................71
Figure 6.6. NH2 sensitivity analysis using the morpholine
mechanism [54], with
updates shown in Table 6.1, at 1428 K, 1.2 atm in 2000 ppm
ethylamine/Ar mixtures.
..................................................................................74
Figure 6.7. Measured NH2 time-histories in 2000 ppm
ethylamine/Ar mixtures;
simulation results are based on the morpholine mechanism [54]
with
updates shown in Table 6.1.
.............................................................................75
Figure 6.8. Measurements of ethylamine ignition delay times near
0.85, 1.35 and
2 atm in stoichiometric mixture of ethylamine/4% O2/Ar;
simulation
results utilize the morpholine mechanism [54] with updates in
Table
6.1.....................................................................................................................76
Figure 6.9. Measured ethylamine ignition delay times near 1.35
atm, with Φ =
0.75, 1 and 1.25; simulation results utilize the morpholine
mechanism [54] with updates in Table 6.1.
.....................................................76
Figure 6.10. NH2 sensitivity analysis using the morpholine
mechanism [54] with
updates in Table 6.1, at 1441 K, 2.1 atm in 2000 ppm
ethylamine/0.8% O2/Ar mixtures.
....................................................................77
-
xxiv
Figure 6.11. NH2 time-histories in 2000 ppm ethylamine/0.8%
O2/Ar mixtures:
measurements (solid lines) and simulation results based on
the
morpholine mechanism [54] with updates in Table 6.1
(dash-dotted
lines).
...............................................................................................................
78
Figure 6.12. OH sensitivity analysis at 1399 K, 1.9 atm in 500
ppm
ethylamine/0.2% O2/Ar mixtures, using the morpholine
mechanism
[54] with updates in Table 6.1.
........................................................................
79
Figure 6.13. OH time-histories in 500 ppm ethylamine/0.2% O2/Ar
mixtures:
measurements (solid lines) and simulation results using the
morpholine mechanism [54] with updates in Table 6.1
(dash-dotted
lines).
...............................................................................................................
80
Figure 7.1. Comparisons of model predictions with the ignition
delay time data
in morpholine/air mixtures near 15 atm and under different
equivalence ratios. Solid lines: simulation results using
the
morpholine mechanism in Chapter 3 [54]. Dash-dotted lines:
simulation results using the morpholine mechanism with
modifications in Chapter 4-6.
..........................................................................
83
Figure 7.2. Comparisons of model predictions to ignition delay
time data in
stoichiometric morpholine/air mixtures near 15 and 25 atm
respectively. Solid lines: simulation results using the
morpholine
mechanism in Chapter 3 [54]. Dash-dotted lines: simulation
results
using the morpholine mechanism with modifications in Chapter
4-6. ............ 84
Figure 7.3. Comparisons of model predictions to ignition delay
time data near 15
atm in morpholine/4% O2/argon and morpholine/air mixtures.
Solid
lines: simulation results using the morpholine mechanism in
Chapter
3 [54]. Dash-dotted lines: simulation results using the
morpholine
mechanism with modifications in Chapter 4-6.
............................................... 85
-
xxv
Figure 7.4. . Sensitivity analysis of temperature using the
morpholine
mechanism [54] with modifications in Chapter 4-6, in a
stoichiometric mixture of morpholine/air under initial condition
of
1000 K and 15 atm.
..........................................................................................86
Figure 8.1. Representative NH2 time-history measurement at 1217
K and 0.34
atm in 350 ppm MMH/argon, in comparison with simulation
results
using the Cook et al. mechanism [34].
.............................................................93
Figure 8.2. NH2 sensitivity analysis at 1217 K and 0.34 atm in
350 ppm
MMH/argon, using the Cook et al. mechanism [34] with the
constant
energy and volume assumptions.
.....................................................................94
Figure 8.3. NH2 sensitivity analysis at 1163 K and 5.2 atm in
170 ppm
MMH/argon, using the Cook et al. mechanism [34] with the
constant
energy and volume assumptions.
.....................................................................94
Figure 8.4. Representative NH2 time-history measurement at 1217
K and 0.34
atm in 350 ppm MMH/argon, with an error bar ±15% due to the
uncertainty of NH2 cross section, in comparison with the Cook et
al.
mechanism with the best-fit k8.1a (dotted line), and the Cook et
al.
mechanism with the best-fit k8.1a and with k8.2 increased by a
factor of
3 (dash-dotted line).
.........................................................................................96
Figure 8.5. NH2 time-history measurement near 0.3 atm, in
comparison with the
simulation results using the modified Cook et al. mechanism
[34]
with the updated k8.1a and k8.2 (dash-dotted line).
............................................97
Figure 8.6. NH2 time-history measurement near 1 atm, in
comparison with the
simulation results using the modified Cook et al. mechanism
[34]
with the updated k8.1a and k8.2 (dash-dotted line).
............................................97
-
xxvi
Figure 8.7. NH2 time-history measurement near 5 atm, in
comparison with the
simulation results using the modified Cook et al. mechanism
[34]
with the updated k8.1a and k8.2 (dash-dotted line).
............................................ 98
Figure 8.8. Measured MMH N-N bond scission reaction rate
constants k8.1a, in
first-order reaction form. Data points: current study. Dashed
line:
reaction rate constant expression for k8.1a near 2.5 atm by Cook
et al.
[34].
...............................................................................................................
100
Figure 8.9. Uncertainty analysis for k8.1a at 1217 K and 0.34
atm in 350 ppm
MMH/argon.
..................................................................................................
102
Figure 8.10. Representative MMH N-N bond scission reaction rate
constants
k8.1a, in comparison with the previous studies.
.............................................. 104
Figure 8.11. Pressure dependence of the measured k8.1a at
representative
temperatures, in comparison with the theoretical study by Zhang
et al
[94].
...............................................................................................................
105
Figure 9.1. Representative species time-histories for ethylamine
oxidation in
stoichiometric mixture with 4% O2/argon, at 1500 K and 1
atm:
simulations using the modified morpholine mechanism.
.............................. 112
-
1
Chapter 1. Introduction
1.1. Background and Motivation
Biofuels, as additives and alternatives to petroleum-based
transportation fuels,
are of increasing interest in national strategic fuel planning.
These fuels may have more
nitrogen-containing compounds than petroleum-based fuels,
especially in those biofuels
derived from biomass [1–4]. This is because nitrogen atoms bound
in biomass, in the
form of proteins and free amino acids for example, may stay as
nitrogen-containing
compounds in the derived fuels. Thus, understanding the
combustion chemistry of
nitrogen-containing fuels is becoming of great importance;
however, little shock tube
kinetics data exists for these systems.. Significant new
insights into fuel-nitrogen
chemistry can be gained by applying shock tube/laser absorption
methods to study the
kinetics of nitrogen-containing fuels.
Previous works have shown that the structure of
nitrogen-containing compounds
and the pyrolysis conditions affect the decomposition pathways
of biomass and
biofuels, with mechanisms involving complicated oxygenated and
nitrogenated ring
structures [1,4]. Because of the complicated structure of real
biomass and biofuels, in-
depth studies of simpler model biofuels are needed first to gain
insight into fuel-
nitrogen chemistry. Morpholine (C4H9NO,
1-oxa-4-aza-cyclohexane), which is a six-
membered ring both oxygenated and nitrogenated, is an excellent
representative
candidate for these studies [5–8]. In previous studies,
molecular-beam mass
spectrometry (MBMS) has been used to identify intermediates and
products for
morpholine flames stabilized on a flat low-pressure burner at 40
mbar [6]. Cavity ring-
down spectroscopy (CRDS) was also employed to detect the
profiles of intermediate
-
2
species including CH2, CH and NH2 [7] in morpholine flames under
the same condition
as in [6]. Combining photoionization (PI) and electron
ionization (EI) MBMS, the
mole-fraction profiles of major and intermediate species in a
morpholine flame were
further determined in [8]. In combination with the MBMS
experimental work, a
morpholine combustion mechanism was developed using analogies to
cyclohexane
combustion [8]. That mechanism for morpholine captures relevant
features of the
morpholine flame quite well. However, the morpholine combustion
database needs to
be augmented with shock tube data, and the morpholine mechanism
developed in [8]
needs to be validated and updated.
The combustion of morpholine as a 6-membered cyclic amine starts
with ring
opening and pyrolysis process to form smaller aliphatic amine
compounds, in particular,
dimethylamine and ethylamine radicals. Accurate dimethylamine
and ethylamine sub-
mechanisms are thus needed if the morpholine mechanism is to be
refined.
Dimethylamine and ethylamine are among the most abundant amines
found in the
atmosphere, with sources found in agricultural and industrial
processes such as fish and
meat production [1–3,9,10]. In industry applications,
dimethylamine and ethylamine are
the base structures for various substances used for crop and
wood protection, paints, and
finishes [11], as well as in amine-based fuel additives.
Only a few early works on the reactions of aliphatic amines
related to
atmospheric chemistry are available in the literature [12–16].
Atkinson et al. [12] and
Slagle et al. [17] examined the kinetics of the reactions of
oxygen atoms with amines,
using a photoionization technique. Atkinson and coworkers also
investigated the
reactions of methylamine with OH over the temperature range of
299-426 K [13]. A
similar study was then carried out to measure the rate constants
of dimethylamine,
ethylamine and trimethylamine with OH over the temperature range
of 298-426 K [14].
Carl et al. studied the reaction rate constants of aliphatic
amines with OH at 295 K,
including those for methylamine, dimethylamine, ethylamine and
trimethylamine [15].
Galano and Alvarez-Idaboy analyzed the different reaction
channels of methylamine,
dimethylamine and ethylamine with OH, using variational
transition-state theory [16].
-
3
Even fewer studies of aliphatic amines have been carried out at
combustion
temperatures, and most of these were early studies of the
effects of aliphatic amines on
hydrocarbon ignition as combustion inhibitors [18–23]. Votsmeier
et al. [24] studied
methylamine thermal decomposition in a shock tube, employing NH2
concentration
time-history measurements. Recently, Lucassen et al. studied the
laminar premixed
flames of dimethylamine and ethylamine under one-dimensional
low-pressure
conditions [25]. In that work, a detailed combustion model was
developed to analyze
the major pathways in the two flames, which successfully
captured many trends
observed in the flame experiments [25]. More detailed
investigations of dimethylamine
and ethylamine are required for both their own research values
and better understanding
of morpholine combustion.
Amine groups are common structural features for rocket
propellants as well. For
example, the important rocket propellant monomethylhydrazine
(MMH) contains both
primary and secondary amino groups (in the form of a hydrazine
group). The
combinations of MMH with certain oxidizers such as nitrogen
tetroxide (N2O4, NTO)
are hypergolic and can ignite spontaneously [26–28]. Hypergolic
propellants play vital
roles in orbital maneuvering and reaction control systems in
aerospace industries. While
MMH satisfies the flight performance requirements, it presents
ground safety hazards
because of its toxic, corrosive and carcinogenic properties,
which make it challenging to
study MMH experimentally. Safety precaution is very important
for MMH experiments.
As researchers have worked to develop detailed MMH pyrolysis and
oxidation
mechanisms for MMH with a variety of oxidizers, accurate
reaction rate constants for
the MMH thermal decomposition reactions have become increasingly
important.
1.2. Overview of Thesis
The dissertations is divided into nine chapters:
-
4
Chapter 2 describes the shock tube facilities used for the study
of the thesis, and
the laser diagnostic techniques utilized in this work for
species time-history
measurements.
Chapter 3 presents the ignition delay time study during
morpholine oxidation in
a high-pressure shock tube, and a theoretical study based on the
experiment to develop a
morpholine mechanism.
Chapter 4 describes the overall rate constant measurements of
hydroxyl radicals
(OH) with dimethylamine (DMA: CH3NHCH3) and ethylamine (EA:
CH3CH3NH2).
Accompanied with the experimental study, a variational
transition state theory study is
also included in this chapter with the potential energy surface,
geometries, frequencies
and electronic energies at CCSD(T)/6-311++G(2d,2p) level of
theory in the literature.
Chapter 5 discusses the oxidation study of dimethylamine,
including ignition
delay time and OH time-history measurements. Based on the
experimental data, the
mechanism discussed in Chapter 3 was further updated.
Chapter 6 presents the ethylamine pyrolysis and oxidation study
behind reflected
shock waves. With the current experimental data and the
recommendations from recent
studies of ethylamine reactions, final modifications to the
morpholine mechanism were
recommended.
Chapter 7 revisits the morpholine ignition delay time data to
show the effects of
the modifications, recommended in Chapter 4-6, on predicting
morpholine ignition
delay times.
Chapter 8 describes the pressure dependence study of the
important MMH
decomposition reaction CH3NHNH2 = CH3N.H + NH2, making use of
NH2 time-history
measurement during MMH pyrolysis process.
Chapter 9 provides a summary of the thesis and proposes several
future works.
-
5
Chapter 2. Experimental Method
Three different shock tubes were used for this thesis work. This
chapter provides
an overview of a shock tube facility in general and presents
some details for the three
shock tubes used for the thesis work, respectively. Three
different laser absorption
diagnotic methods were used to monitor different species during
the combustion
processes of the fuels covered in this thesis. This chapter
first discusses the fundamental
theory for laser absorption measurement in general and then
provides a more detailed
description of each diagnotic method.
2.1. Shock Tube Facility
A shock tube is a test facility close to an ideal
zero-dimensional reactor with
uniform temperature and pressure that can be readily generated
by shock heating. A
diaphragm is used to separate the tube into the driver and
driven sections. When the
diaphragm breaks a shock wave will form, travel down the tube,
reach the end wall of
the shock tube and then reflect back. Optical diagnostics can be
implemented in the
heated test gas behind the incident or the reflected shock
waves. With accurate
measurement of the incident shock speed, the test conditions
behind the shock waves
can be determined accurately using the ideal shock jump
relations [29]. A schematic for
shock tube operation is shown in Figure 2.1. Three different
shock tubes in Stanford
were used for this work, and are described in the following
sections.
-
6
Figure 2.1. Shock tube schematic.
2.1.1. Stanford High Pressure Shock Tube (HPST)
Morpholine ignition delay times in morpholine/air and
morpholine/oxygen/argon mixtures were measured behind reflected
shock waves using
the Stanford high-pressure shock tube (HPST). This shock tube
has a stainless steel
driven section of 5 m length with a 5 cm inner diameter and a
driver section that is 3 m
long with an inner diameter of 7.5 cm. Shock tube driver inserts
were used to achieve
uniform test conditions at lower temperatures where facility
effects at long test times
(dP/dt and dT/dt) are most significant [30]. In the current
study, a test time about 2.5 ms
was achieved with uniform test conditions using helium driver
gas. The incident shock
speed, which is critical to the accurate determination of
reflected shock pressure and
temperature, was determined using five piezoelectric pressure
transducers that were
spaced at approximately 30 cm intervals over the last 2 m of the
shock tube. The driven
section was heated to 86 oC to mitigate condensation of fuel on
the wall. Temperatures
and pressures in the post-shock region were determined from the
incident shock speed
at the end wall using standard normal shock relations. Ignition
pressure was monitored
-
7
using a piezoelectric pressure transducer (Kistler Model 603B1)
located 1 cm from the
end wall. Other details concerning this shock tube can be found
in [31].
2.1.2. Stanford Kinetic Shock Tube (KST)
Dimethylamine and ethylamine + OH, and dimethylamine oxidation
were
studied behind reflected shock waves in a shock tube that has a
3.35 m driver section
and an 8.54 m driven section, both with an inner diameter of
14.13 cm. Shock tube
driver inserts were used to achieve uniform test conditions. In
the current study, a test
time about 2 ms was achieved with uniform test conditions using
helium driver gas.
The incident shock speeds were measured using five piezoelectric
pressure transducers
near the driven section endwall. Between experiments, the shock
tube was routinely
evacuated to ~5 µtorr to ensure purity of the test mixtures.
More details concerning this
shock tube are included in [32,33].
2.1.3. Stanford NASA Shock Tube (NASA)
Ethylamine pyrolysis and oxidation, and MMH pyrolysis were
studied behind
reflected shock waves in a shock tube with a 3.7 m driver
section and a 10 m driven
section, both with an inner diameter of 15.24 cm. The incident
shock speeds were
measured using five piezoelectric pressure transducers over the
last 1.5 meters of the
shock tube and linearly extrapolated to the endwall. The
ignition pressures were
monitored using a piezoelectric pressure transducer (Kistler
Model 603B1) located 2 cm
from the end wall; laser absorption measurements were conducted
at the same axial
location. More details concerning this shock tube can be found
in [34].
-
8
2.2. Laser Diagnostic
2.2.1. Overview
In this thesis, the primary laser diagnostic method utilized is
the fixed-
wavelength direct absorption technique, which is a powerful tool
for chemical kinetics
study. One advantage of a laser absorption diagnostic is that it
enables rapid real-time
measurement at kHz-MHz rates, which can be used to determine the
time evolution of
important species in combustion processes. Besides, laser
absorption diagnostics are
non-intrusive and do not perturb the chemical kinetics
processes. In this work, species
time-history measurements for important combustion compounds
were used to provide
valuable kinetic information for mechanism validation and
modification.
Species concentration can be inferred from laser absorption
measurement via the
Beer-Lambert law shown in equation 2.1:
( ) (
) (Eq 2.1),
where α is the absorbance, T is the transmission, I is the
transmitted laser intensity with
absorption through the test region, I0 is the laser intensity
without absorption, P is the
pressure, x is the mole fraction of the absorbing species, k is
the absorption coefficient
of the target species, and L is the laser pathlength in the test
region.
With the measured absorbance and pressure, and the known
absorption
coefficient and laser pathlength, the mole fraction of the
absorbing species can be
derived as:
x = α/PkL (Eq 2.2)
More details on the specific laser absorption diagnostic methods
used for the
current work can be found in the following sections.
-
9
2.2.2. IR Diagnostic of Fuel
For morpholine ignition delay time measurements, initial fuel
concentrations
were monitored using the 3.39 m emission of a Spectral Physics
model 124B He-Ne
laser. This fuel diagnostic relies on the strong absorption band
near 3.39 m due to the
C-H stretch vibration. Common mode rejection was used to reduce
laser intensity noise.
The experimental setup for this fuel diagnostic is shown in
Figure 2.2. In support of this
work, the absorption coefficient of morpholine at 3.39 m and 86
oC was also measured
using an FTIR instrument. Details on the FTIR measurement
technique can be found in
[35,36].
Figure 2.2. He-Ne laser diagnostic of fuel.
2.2.3. NH2 Diagnostic
NH2 was measured using the output of a narrow-linewidth ring dye
laser near
597.4 nm. This NH2 laser absorption diagnostic employed the
overlapping ÃA1 ←
X2B1(090 ← 000)∑PQ1,N7 doublet lines, which was previously
characterized in our
laboratory [24,34,37]. Visible light near 597.4 nm was generated
by pumping
-
10
Rhodamine 6G dye in a Spectra Physics 380 laser cavity with the
5 W, continuous
wave, output of a Coherent Verdi laser at 532 nm. Using a
common-mode rejection
detection setup, a minimum NH2 detection sensitivity of 5 ppm
could be achieved for
most conditions studied in this work. A schematic of the NH2
diagnostic is shown in
Figure 2.3, and more details of the NH2 laser diagnostic setup
are described in [34].
Figure 2.3. Ring dye laser diagnostic of NH2, using the
fundamental output of a Spectra
Physics 380 ring dye laser.
2.2.4. OH Diagnostic
OH was measured near 306.7 nm using the frequency-doubled output
of the
same ring dye laser system as for the NH2 diagnostic. The chosen
wavelength was the
peak of the well-characterized R1(5) absorption line in the OH
A-X(0,0) band [32].
Visible light near 613.4 nm generated in the Spectra Physics 380
laser cavity was
intracavity frequency-doubled using a temperature-tuned AD*A
nonlinear crystal to
generate ~1 mW of UV light near 306.7 nm. Using a common-mode
rejection detection
setup, a minimum OH detection sensitivity of 0.5 ppm could be
easily achieved. A
schematic of the OH diagnostic is shown in Figure 2.4, and more
details of the laser
diagnostic setup are can be found in [32].
-
11
Figure 2.4. Ring dye laser diagnostic of OH, using the frequency
doubled output of a
Spectra Physics 380 ring dye laser.
-
13
Chapter 3. Shock Tube and Modeling Study of
Morpholine
3.1. Introduction
As is introduced in Chapter 1, morpholine (C4H9NO,
1-oxa-4-aza-cyclohexane)
is an excellent representative candidate to study oxygenated and
nitrogen-containing
biofuel because of its unique structure and wide industrial
applications [5–8]. The
morpholine combustion experimental database needs to be
augmented, and we are
aware of no shock tube data that have been published for
morpholine ignition delay
times before this study. In this chapter, morpholine ignition
delay times measured
behind reflected shock waves were provided. A morpholine
combustion mechanism,
developed based on a previous morpholine flame study [8] and the
current shock tube
data, was used for comparison with the experiment.
3.2. Experimental Details
Morpholine ignition delay times in morpholine/air and
morpholine/oxygen/argon mixtures were measured behind reflected
shock waves using
the Stanford high-pressure shock tube (HPST). More details about
this shock tube have
been introduced in section 2.1.1.
Prior to each experiment, morpholine mixtures were
manometrically prepared in
a 12.8 L, magnetically stirred stainless steel mixing tank. To
avoid condensation of the
fuel, the mixing tank and mixing assembly were heated to
approximately 86 °C. Liquid
morpholine was added into the mixing tank using a gas-tight
syringe. A sufficient time
-
14
period (about 15 min) was provided to ensure the full
evaporation of fuel liquid inside
the tank; then the oxidizer and bath gas were added. The test
mixtures were stirred
using a magnetically driven vane assembly for at least 1 hour
before actual shock tube
experiments. At 86 °C the vapor pressure of morpholine is around
25 kPa [38], while
the partial pressures of morpholine inside the tank, mixing
assembly, and the shock tube
driven section never went above 3 kPa during the experiments, so
that morpholine
remained in the vapor phase throughout the experimental
process.
The infrared diagnostic of fuel at 3.39 mm, together with FTIR
measurements of
morpholine absorption cross section as described in section
2.2.2, was used to confirm
the initial morpholine concentration in the shock tube test
section.
Ignition delay times were determined by extrapolating, back to
the baseline
pressure, the steep increase in pressure concurrent with
ignition. A sample pressure
trace for ignition delay time determination can be found in
Figure 3.1. Shock tube driver
inserts were used for all the experiments to reduce the
non-ideal pressure variation
caused by viscous effects, and to achieve test time with small
pressure variation behind
the reflected shock waves [30].
0 500 1000 1500 2000
0.0
0.5
1.0
1.5
Pre
ssure
[A
. U
.]
Time [s]
ign
= 1070 s
Morpholine/Air
925 K, 15.8 atm
= 1
-
15
Figure 3.1. Sample pressure trace for morpholine ignition delay
time
measurements.
3.3. Results and Discussion
3.3.1. Morpholine Ignition Delay Times
Morpholine ignition delay times were measured during morpholine
oxidation
experiments under different conditions behind the reflected
shock waves and are shown
in Figure 3.2-3.4 and listed in Table 3.1.
Table 3.1. Shock tube ignition delay times. Test gas mixture:
morpholine/O2/argon.
T5 1000/T5 P5 Φ XO2 IDT
[K] [1/K] [atm] [µs]
910 1.099 14.71 1 0.21 1350
915 1.093 15.76 1 0.21 1130
925 1.081 15.83 1 0.21 1070
938 1.066 14.66 1 0.21 1000
966 1.035 14.79 1 0.21 636
1009 0.991 14.04 1 0.21 374
1062 0.942 13.77 1 0.21 197
1097 0.912 13.02 1 0.21 115
921 1.086 13.79 0.5 0.21 1970
983 1.017 13.94 0.5 0.21 1030
1023 0.978 13.37 0.5 0.21 547
1110 0.901 13.19 0.5 0.21 192
1168 0.856 12.53 0.5 0.21 97
875 1.143 16.45 2 0.21 969
901 1.110 16.60 2 0.21 775
935 1.070 16.29 2 0.21 562
965 1.036 15.25 2 0.21 402
994 1.006 16.13 2 0.21 223
-
16
1041 0.961 15.08 2 0.21 136
1027 0.974 16.50 1 0.04 1150
1042 0.960 16.64 1 0.04 845
1091 0.917 16.30 1 0.04 430
1139 0.878 15.94 1 0.04 244
1197 0.835 15.56 1 0.04 145
932 1.073 25.21 1 0.21 625
955 1.047 29.18 1 0.21 389
986 1.014 26.70 1 0.21 279
994 1.006 27.02 1 0.21 241
1046 0.956 27.05 1 0.21 117
1074 0.931 26.36 1 0.21 82
A shock tube can reproduce close, but not identical, pressures
from shock
experiment to shock experiment. For a uniform graphic
presentation of the results, a
pressure scaling of all the data in a similar pressure regime is
needed. Many previous
studies have observed that ignition delay times have pressure
dependence close to P-1
.
Since the actual pressures are close to the reported pressure
for one set of data on an
ignition delay time plot, this simple power law dependence is
used for Figure 3.2-3.4. A
more accurate pressure dependence will be established based on
regression analysis of
the ignition delay time data over the entire pressure range of
the current study. In Figure
3.2, ignition delay times in morpholine/air mixtures at
pressures near 15 atm are shown
for different equivalence ratios. Synthetic air with 21% O2 and
79% N2 was used. The
stoichiometric case was defined with the following reaction:
C4H9NO + 5.75(O2 + 3.76N2) = 4CO2 + 4.5H2O + 22.12N2 (Eq
3.1)
As can be seen from Figure 3.2, in morpholine/air mixtures, when
other
conditions are held the same, auto-ignition occurs faster with
increasing equivalence
ratio. The data are characterized by small scatter, and within
the current temperature
range, ignition delay times vary monotonically with temperature.
In Figure 3.2, the
slopes of the ignition delay time data are similar at different
equivalence ratios, thus the
-
17
activation energy of morpholine/air ignition is not sensitive to
equivalence ratio. At the
current relatively low-temperature and high-pressure conditions,
fuel-rich mixtures are
fastest to ignite due to the major chain-branching reactions
emanating from the fuel.
0.9 1.0 1.1
100
1000
= 0.5
= 1
= 2
Ignitio
n D
ela
y T
ime [s]
1000/T [K-1]
Morpholine/Air
Scaled to 15 atm with P-1
1111K 1000K 909K
Figure 3.2. Ignition delay time measurements in morpholine/air
mixtures near 15 atm
and with equivalence ratios of 0.5, 1 and 2. Dashed lines:
linear fit to data.
The ignition delay times in morpholine/air mixtures are shown in
Figure 3.3 for
an equivalence ratio of 1, and pressures around 15 and 25 atm.
The ignition delay times
near 25 atm are shorter than those near 15 atm, as expected.
-
18
0.9 1.0 1.1
100
1000
1111K 1000K 909K
P = 15 atm
P = 25 atm
Ignitio
n D
ela
y T
ime [s]
1000/T [K-1]
Morpholine/Air = 1
Scaled with P-1
Figure 3.3. Ignition delay time measurements in morpholine/air Φ
= 1 mixtures near 15
and 25 atm. Dashed lines: linear fit to data.
To study the effects of oxidizer concentration, ignition delay
times were
measured in stoichiometric morpholine/4% O2/argon mixtures as
well, and compared
with morpholine/air mixtures in Figure 3.4. The ignition delay
time clearly decreases
with an increasing oxygen concentration, due in part to the
decreasing dilution.
-
19
0.8 0.9 1.0 1.1
100
1000
1250K 1111K 1000K 909K
Morpholine/4% O2/Ar
Morpholine/Air
Ignitio
n D
ela
y T
ime [s]
1000/T [K-1]
= 1
Scaled to 15 atm with P-1
Figure 3.4. Ignition delay time measurements in stoichiometric
morpholine/air mixtures
and morpholine/4% O2/argon mixtures around 15 atm. Dashed lines:
linear fit to data.
A regression analysis was carried out based on all the
experimental data reported
in this section, and the following scaling relation was found
for morpholine ignition
delay time:
τ = 1.7×10-3
Φ-0.8
P-0.9
XO2-0.84
exp(13400/T) [s] (Eq 3.2)
over the temperature range of 866-1197 K, pressures 15-25 atm
and equivalence ratios
0.5-2.
3.4. Model Development and Simulations
A mechanism for morpholine flame chemistry has been previously
presented,
constructed using simple analogies with cyclohexane combustion
[8]; however, this
mechanism was tested only against low-pressure flat-flame data,
and the conditions of
shock tube oxidation were not considered. A new mechanism was
thus proposed for
morpholine oxidation (see Appendix A), based on a previous
cyclohexane oxidation
-
20
study [39] (see Appendix A) and the current data. Additionally,
the H/C/O chemistry
was updated to reflect recent works on acetylene [40] and
tetrahydropyran (THP) [41].
Improvements were also made to the base nitrogen-chemistry set,
with rate constants
drawn from several sources [25,28,42–47].
Rate coefficients for morpholine will be different from those
for cyclohexane or
other 6-membered ring species, but the transition-state
structures have useful
similarities. There are six saturated heavy atoms (C, N, O) in
the morpholine ring, so it
is proposed that, similar to cyclohexane, morpholine oxidation
in the shock tube occurs
by O2 addition to a radical site. The key difference is that
while each hydrogen on a
cyclohexane ring is symmetrically equivalent, morpholine has
three different sites from
which an H-atom may be abstracted: the carbon ortho to the ether
oxygen, the meta
carbon, and the para amine nitrogen. Thus, three distinct
hydrogen-abstraction routes
exist. Once an RO2 morpholine species is formed, as in
cyclohexane, the O2 group can
internally abstract a hydrogen atom from the ring and either
form HO2 + an unsaturated
morpholine-ene cyclic species or one of several morpholine QOOH
species. The
resulting QOOH radicals can then undergo β-scission by
ring-breaking, forming linear
and branched, unsaturated radicals. The linear and branched
radicals can undergo
further β-scission reactions until small products with 2 to 3
heavy atoms are formed.
Reaction rate coefficients for the oxidation reactions for
morpholine were derived based
on analogous reactions from the cyclohexane [39] model for
oxidation.
Accounting for the thermal decomposition of the ring requires
assumptions
about the product channels and rate constants. There are no
previous morpholine
pyrolysis data in the literature, and the products and rate
constants are not settled in the
literature for morpholine or similar ring species that might
provide analogies. However,
for cyclohexane, Tsang [48] detected only 1-hexene as an
experimental product.
In this work, the ring decomposition was assumed to proceed
analogously to the
mechanisms proposed for 1,4-dioxane [49] and cyclohexane [50],
modifying Arrhenius
pre-exponential factors for the proper reaction-path degeneracy
(RPD). Three
-
21
decomposition mechanisms were drawn from theoretical studies: 1)
homolytic cleavage
of the ring into three 2-heavy-atom species, 2) opening of the
ring into a diradical
species which can then internally abstract a hydrogen in a
6-centered transition state,
then decomposing into two 3-heavy-atom radicals, and 3)
1,4-hydrogen shifts,
transforming the ring into 6-heavy-atom linear species with a
single π bond at the end,
similar to the 1-hexene from cyclohexane observed by Tsang. Rate
constants from
Altarawneh and Dlugogorski [51] were also tested where possible.
They had applied
G3MP2B3 and RRKM calculations to investigate these pathways,
concluding that the
fastest channel in the present temperature range was to
CH2=CH-O-CH2-CH2-NH.
Their rate constants could not all be used, as their second
fastest channel was homolytic
scission of the C-N bond to an unphysical product,
•CH2-O-CH2-CH2-CH2-N•.
Thermochemistry for the 28 species introduced in [8], as well as
five further
species from the new pyrolysis mechanism and 41 additional
species from the oxidation
mechanism, were calculated theoretically using the
complete-basis-set method CBS-
QB3 [52] (see Appendix C). Geometry and frequency calculations
were completed with
Gaussian09 software [53] using the tight convergence criteria
and rigid-rotor/harmonic-
oscillator approximations. Thermochemistry was estimated for an
additional nine
radical species for the oxidation set based on their non-radical
analogues. The resulting
mechanism is 290 species and 2130 reactions. This mechanism was
also published in
[54], with the morpholine oxidation reaction set, pyrolysis
reaction set and
thermochemistry included in Appendix A-C. This mechanism will be
referred to as the
morpholine mechanism in the following sections, with reference
to [54]. Shock tube
simulations were performed using a closed homogenous reactor
model using
CHEMKIN Pro [55] assuming constant volume and constant internal
energy conditions.
The predictions, using the current morpholine mechanism
developed based on
ignition delay time data, for the morpholine oxidation system
(dashed lines) are
compared to the shock tube data in Figure 3.5-3.7. As can be
seen in Figure 3.5, for the
equivalence ratio dependence of the ignition delay time, the
newly proposed morpholine
mechanism captures the same trend as the experiment. Also shown
in the figure are the
-
22
simulation results using the mechanism of [8], represented by
dash lines. It is evident
that the current model matches much better with the experimental
data than the model
from [8] because of the addition of the O2 addition chemistry.
The current mechanism
is quite good at the equivalence ratio 0.5, but overpredicts the
ignition delay times as
richness increases.
0.8 0.9 1.0 1.1 1.2
100
1000
10000
= 0.5
= 1
= 2
Ignitio
n D
ela
y T
ime [s]
1000/T [K-1]
Morpholine/Air
Scaled to 15 atm with P-1
1250K 1111K 1000K 909K 833K
Figure 3.5. Comparison of model predictions to morpholine/air
ignition delay time
measurements around 15 atm and under different equivalence
ratios. Dashed lines:
Simulation results using the mechanism in [8]. Solid lines:
Simulation results using the
morpholine mechanism [54].
A sensitivity analysis for morpholine concentration was
performed for the
stoichiometric case at P=15 atm. At lower temperatures (~800K),
the model has a
heightened sensitivity to morpholine unimolecular decomposition
to three-heavy-atom
products. However, as temperature increases (1000K and higher),
the unimolecular
decomposition reactions become less important. Instead, the
model predictions for
morpholine become more sensitive to radical chemistry,
especially to the orthomorphyl
→ CH2CH2NHCH2CHO and the metamorphyl → CH2CH2OCH2CHNH
beta-scission
reactions.
-
23
Pressure effects are shown in Figure 3.6 at 15 and 25 atm, using
both the current
mechanism (solid lines) and the mechanism published in [8]
(dashed lines). Significant
improvement is evident in the modeling results using the
morpholine mechanism [54].
The influence of oxidizer concentration on ignition delay time
can be seen in
Figure 3.7. The mechanism used in [8] overpredicts ignition
delay times by a factor of
5, whereas the new model performs well. It overpredicts the
ignition delay times only
by about 50% in stoichiometric morpholine mixtures, using air
(21% O2 in N2) or 4%
O2 in Ar.
0.8 0.9 1.0 1.1 1.2
100
1000
10000
1250K 1111K 1000K 909K 833K
P = 15 atm
P = 25 atm
Ignitio
n D
ela
y T
ime [s]
1000/T [K-1]
Morpholine/Air = 1
Scaled with P-1
Figure 3.6. Comparison of model predictions to morpholine/air
ignition delay time
measurements for stoichiometric mixtures around 15 and 25 atm
respectively. Dashed
lines: Simulation results using the previous mechanism [8].
Solid lines: Simulation
results using the morpholine mechanism [54].
-
24
0.8 0.9 1.0 1.1 1.2
100
1000
10000
1250K 1111K 1000K 909K 833K
Morpholine/4% O2/Ar
Morpholine/Air
Ignitio
n D
ela
y T
ime [s]
1000/T [K-1]
= 1
Scaled to 15 atm with P-1
Figure 3.7. Comparison of model predictions to morpholine
ignition delay time
measurements around 15 atm for morpholine/4% O2/argon and
morpholine/air mixtures.
Dashed lines: Simulation results using the previous mechanism
[8]. Solid lines:
Simulation results using the morpholine mechanism [54].
3.5. Summary
Morpholine ignition delay times were measured in the Stanford
high-pressure
shock tube, covering temperatures from 866 to 1197 K,
equivalence ratios of 0.5, 1 and
2, two pressure groups near 15 and 25 atm, and two oxygen
concentration values of 4%
O2 in Ar and synthetic air with 21% O2. The current shock tube
work extends the
morpholine combustion experimental database and a new morpholine
mechanism was
generated using the current data.
The morpholine mechanism developed for low-pressure flames in
[8]
significantly over-predicts the ignition delay times under all
conditions. The
simulations, using the morpholine mechanism proposed in this
chapter, are much closer
matches with the morpholine ignition delay times than those from
the previous
-
25
mechanism developed in [8], and can successfully capture the
equivalence ratio
dependence near 15 atm.
Combustion of morpholine as a 6-membered cyclic amine may start
with ring
opening and pyrolysis process to form smaller aliphatic amine
compounds, in particular,
dimethylamine and ethylamine radicals. Further refinement of the
morpholine
mechanism requires improvements in the sub-mechanisms of
dimethylamine and
ethylamine. In the following chapters, shock tube studies of
dimethylamine and
ethylamine combustion are presented, to improve understanding of
the reaction kinetics
of those two important aliphatic amines, and also to improve the
morpholine
mechanism.
-
27
Chapter 4. Reactions of OH with Dimethylamine and
Ethylamine
4.1. Introduction
The efforts to update the dimethylamine and ethylamine
sub-mechanisms of the
morpholine mechanism begin with the direct reaction rate
measurements of OH with
dimethylamine and ethylamine. The reactions of aliphatic amines
are relevant to both
atmospheric chemistry and biofuel combustion processes. In the
context of atmospheric
chemistry, aliphatic amines are potential precursors of HCN and
stratospheric NOx [56–
58]. Additionally, the degradation of dimethylamine within the
environment can lead to
carcinogenic nitrosamines [59]. In the context of combustion,
the amine group is a
common functional group in bio-derived fuels [5–7,25,54]. The
hydrogen abstraction
reactions by OH radical are important steps in the combustion of
amines, thus
dimethylamine and ethylamine + OH reactions are of great
research value.
Experimental and computational studies of the reactions between
aliphatic
amines and OH are scarce. Atkinson et al. investigated the
reactions of methylamine
(MA: CH3NH2, CAS: 74-89-5) with OH over the temperature range of
299-426 K using
a flash photolysis-resonance fluorescence technique [13]. The
same method was used to
measure the rate constants for the reactions of dimethylamine,
ethylamine and
trimethylamine with OH over the temperature range of 298-426 K
[14]. Carl et al.
studied the reaction rate constants of aliphatic amines with OH
at 295 K, including
those for methylamine, dimethylamine, ethylamine and
trimethylamine, using the
sequential two-photon dissociation of NO2 in the presence of H2
as a source of OH [15].
Galano and Alvarez-Idaboy analyzed the different reaction
channels of methylamine,
-
28
dimethylamine and ethylamine with OH, using variational
transition-state theory [16].
Geometry optimization and frequency calculations were performed
at the
BH&HLYP/6-311++G(2d,2p) level of theory, with electronic
energy values improved
by single-point calculations at the CCSD(T) level of theory and
using the same basis
set. The overall reaction rate constants and the branching
ratios for reactions of amines
with OH were reported within the temperature range 290-310 K
[16]. Recently,
Lucassen et al. studied the laminar premixed flames of
dimethylamine and ethylamine
under one-dimensional low-pressure conditions [25]. A detailed
combustion model was
developed to analyze the major pathways in those two flames,
which successfully
reproduced many trends observed in the flame experiments.
Lucassen et al. estimated
the reaction rates for amine + OH reactions based on previous
work in the literature.
To the best of the author’s knowledge, there is no direct
experimental or
theoretical study of the reactions of aliphatic amines with OH
under combustion
conditions. The present work determines the reaction rate
constants for the overall
reactions of dimethylamine (DMA: CH3NHCH3, CAS: 124-40-3) and
ethylamine (EA:
CH3CH2NH2, CAS: 75-04-7) with OH.
DMA + OH = Products (R4.1)
EA + OH = Products (R4.2)
The OH radical was generated by near-instantaneous pyrolysis of
tert-butyl
hydroperoxide (TBHP, CAS: 75-91-2). The pseudo-first order decay
of OH behind
reflected shock waves was monitored using laser absorption at
306.7 nm, and the
reaction rate constants of amine with OH were inferred from the
measured OH time-
histories. Variational transition state theory was used to
compute the H-abstraction rates
by OH for dimethylamine and ethylamine, with potential energy
surface geometries,
frequencies and electronic energies calculated by Galano and
Alvarez-Idaboy at
CCSD(T)/6-311++G(2d,2p) level of theory [16].
-
29
4.2. Experimental Setup
The Stanford Kinetic Shock Tube as described in section 2.1.2
was used for the
dimethylamine and ethylamine + OH experiments, with the OH decay
time-histories
monitored using the OH diagnostic presented in section 2.2.4.
The chemicals used in the
experiments include 97% ethylamine, anhydrous ≥ 99%
dimethylamine, and a solution
of 70%, by weight, tert-butyl hydroperoxide (TBHP) in water, all
supplied by Sigma-
Aldrich with no further purification. Research grade argon
(99.99%) supplied by
Praxair was employed as the bath gas. All the mixtures were
prepared manometrically
using a double-dilution method in a 12 liter electro-polished
stainless steel tank, and
mixed with a magnetically driven stirring vane for at least one
hour prior to the
experiments. Before each experiment, the shock tube was
passivated to avoid loss of
amine to the shock tube wall. Since each TBHP decomposes
near-instantaneously to
form one OH, the TBHP concentrations were determined based on
the peak OH value
for each experiment. Before the experiment, controlled mixtures
of fuel diluted in argon
were made; the amine concentrations were then confirmed by
sampling a portion of the
mixture, after filling into the shock tube, from near the
endwall to an external multipass
cell with 29.9 m pathlength. The fuel concentration in the
multipass cell was measured
using a Jodon helium-neon laser at 3.39 μm, and Beer’s law was
used to convert the
measured absorption data to the fuel mole fraction. Further
details about this multipass
cell laser diagnostic of fuel are reported in [33,60]. The
absorption cross sections of
dimethylamine and ethylamine from the PNNL database [61] were
used in the Beer’s
law concentration calculation, and the measured fuel
concentrations were consistent
with the manometric values to within ±5%, which gives confidence
in the manometric
values. The manometric fuel concentrations were used for
comparisons with
simulations.
-
30
4.3. Kinetic Measurements
Experiments were performed behind reflected shock waves over the
temperature
range of 901-1368 K and pressures near 1.2 atm. At temperatures
greater than 1000 K,
TBHP dissociates near-instantaneously to form an OH radical and
a tert-butoxy radical.
The tert-butoxy radical, (CH3)3CO, further dissociates into
acetone and a methyl radical.
TBHP also reacts with OH radical to form other products. The
TBHP chemistry set can
be described as follows:
TBHP = (CH3)3CO + OH (R4.3)
(CH3)3CO = CH3COCH3 + CH3 (R4.4)
TBHP + OH = H2O + O2 + tert-C4H9 (R4.5)
TBHP + OH = H2O + HO2 + iso-C4H8 (R4.6).
Further details about the TBHP chemistry set can be found in the
literature
[33,60,62–65].
The rate constants for Reaction 4.3, 4.5, and 4.6 were adopted
from Pang et al.
[33], and the reaction rate constant for Reaction 4.4 was
obtained from Choo and
Benson [66]. The thermodynamic parameters for TBHP and
tert-butoxy radical were
taken from the thermodynamic database by Goos et al. [67].
Methyl radical is formed in
Reaction 4.4 and previous works [33,62] have shown that the
accuracy of the CH3 + OH
rate constant around 1.5 atm is important for accurate
determination of the fuel + OH
reaction rate constant. There are two major channels for CH3 +
OH,
CH3+OH + M = CH3OH + M (R4.7)
CH3 + OH = CH2(s) + H2O (R4.8).
Reaction 4.7 was updated using the results from Srinivasan et
al. [68] at ~0.3-1.1 atm,
and their values agree well with the theoretical study by Jasper
et al. [69] and the
-
31
measured values from Vasudevan et al. at 1.3 atm [62]. The rate
constant for Reaction
4.8 was updated with the value measured by Pang et al. [33] ,
which agrees well with
the values by Srinivasan et al. [68] and Vasudevan et al. [62].
Sangwan et al. recently
measured the reaction rate for Reaction 4.8 over the temperature
range of 294-714 K
[70]. The rate constant for Reaction 4.8 by Pang et al. also
agrees with extrapolation of
the Sangwan et al. results, within the uncertainty limit used
for error analysis in the
following sections.
The rate constants for reactions 4.3-4.8 are provided in Table
4.1. The same set
of reaction rate constants for TBHP chemistry has been used
before by Lam et al.
[65,71]
Table 4.1. Reactions describing DMA and EA + OH experiments
# Reaction Rate Constant [cm
3mol
-1s
-1]
Ref. A n E
4.1 DMA + OH = Products See text This work
4.2 EA + OH = Products See text This work
4.3 TBHP = (CH3)3CO + OH1 3.57 × 10
13 0 3.575 × 10
4 [33]
4.4 (CH3)3CO = CH3COCH3 + CH31 1.26 × 10
14 0 1.530 × 10
4 [66]
4.5 TBHP + OH = H2O + O2 + tert-C4H9 2.30 × 1013
0 5.223 × 103 [33]
4.6 TBHP + OH = H2O + HO2 + iso-C4H8 2.49 × 1013
0 2.655 × 103 [33]
4.7 CH3 + OH + M = CH3OH + M 1.73 × 108 1.41 -3.32 × 10
4 [68]
4.8 CH3 + OH = CH2(s) + H2O 1.65 × 1013
0 0 [33] 1Rate coefficient units for 1
st order reactions: s
-1
The above TBHP chemistry set was implemented into the base
dimethylamine
and ethylamine sub-mechanisms of the morpholine mechanism. The
dimethylamine and
ethylamine sub-mechanisms were originally developed by Lucassen
et al. [25] . The
CHEMKIN PRO [55] package was used to simulate the OH
time-histories, with the
standard constant energy and volume assumptions.
4.3.1. Dimethylamine (DMA) + OH
The reaction of dimethylamine with OH consists of two different
channels:
DMA + OH = CH3NHCH2 + H2O (R4.1a)
-
32
DMA + OH = CH3NCH3 + H2O (R4.1b)
In the dimethylamine sub-mechanism by Lucassen et al. [25], the
total rate
constant of DMA with OH at 295 K measured by Carl et al. [15],
and the branching
ratio by Galano and Alvarez-Idaboy at 295 K [16] was used for
all temperatures, with
k4.1a = 2×1013
cm3mol
-1s
-1 and k4.1b = 1.9 ×10
13 cm
3mol
-1s
-1.
A OH sensitivity analysis was carried out for the overall rate
constant
determination of Reaction 4.1 (k4.1 = k4.1a + k4.1b) in the
mixture of 320 ppm DMA with
22 ppm TBHP (and 140 ppm water) diluted in argon, at 1176 K and
0.9 atm. The OH
sensitivity is defined as SOH = (∂XOH/∂ki)×(ki/XOH), where XOH
is the local OH mole
fraction and ki is the rate constant for reaction i. As
illustrated in Figure 4.1, the
sensitivity analysis shows that Reaction 4.1 is the dominant
reaction with minor
interferences from secondary reactions. The measured OH
time-history under the same
conditions is shown in Figure 4.2. The mechanism with the TBHP
chemistry set in
Table 4.1 was used to simulate the experimental data, and a
best-fit overall rate constant
of k4.1 = 3 × 1013
cm3mol
-1s
-1 was obtained between the experiment and the simulation.
Also shown in Figure 4.2 are the simulations for the
perturbations of ±50% in the best-
fit overall rate constant. Note in this figure that non-kinetic
effects, i.e. laser beam
steering by the shock passage, contribute to the measured
absorption profiles at times
before t=0. The branching ratios for Reaction 4.1 in the
mechanism were kept the same
in the simulations. It is worth noting that the presence of H2O
in the test mixture has no
noticeable influence on the simulated OH profiles.
-
33
0 20 40 60 80-7
-6
-5
-4
-3
-2
-1
0
1
320 ppm DMA / Ar
22 ppm TBHP / 140 ppm H2O
1176 K, 0.9 atm
OH
Se
nsitiv
ity
Time [s]
DMA + OH = Products
TBHP = tert-Butoxy + OH
CH3+OH=CH
2(S)+H
2O
CH3+CH
2=C
2H
4+H
CH3NHCH
2=CH
3+CH
2NH
Figure 4.1. Sensitivity analysis of OH using the dimethylamine
sub-mechanism [25]
with TBHP chemistry set, in the mixture of 320 ppm DMA/Ar with
22 ppm TBHP and
140 ppm H2O, at 1176 K and 0.9 atm.
0 10 20 30 40
10
50
k4.1
1.5
OH
[pp
m]
Time [s]
Experiment
k4.1
= 3.2E13 cm3mol
-1s
-1
(best-fit)
320 ppm DMA/ Ar
22 ppm TBHP/ 134 ppm H2O
1176 K, 0.9 atm
k4.1
0.5
2
Figure 4.2. Sample OH trace in 320 ppm DMA/Ar with 22 ppm TBHP
and 134 ppm
H2O, at 1176 K and 0.9 atm.
Under the current pseudo first-order conditions, OH decays
exponentially and
close to be a straight line in the semi-log plot of Figure
4.2.
-
34
Figure 4.3. Error analysis for measured k4.1 in 320 ppm DMA/Ar
with 22 ppm TBHP
and 134 ppm H2O, at 1176 K and 0.9 atm.
A detailed error analysis was conducted to evaluate the overall
uncertainty of the
measured rate constant for Reaction 4.1 in the mixture of 320
ppm dimethylamine with
22 ppm TBHP and 140 ppm water in argon at 1176 K and 0.9 atm.
The primary sources
of uncertainty for the rate constant determination include 2
uncertainties in (a)
pressure (±1%), (b) temperature (±1%), (c) mixture composition
(±5%), (d) OH cross
section(±3%), (e) fitting data (±7%), (f) the rate constant for
TBHP = tert-Butoxy + OH
(Reaction 4.3, ±30%), (g) the rate constant for CH3 + OH =
CH2(s) + H2O (Reaction
4.8, uncertainty factor used: 2), (h) the rate constant for CH3
+ CH2 = C2H4 + H
(uncertainty factor used: 2), (i) the rate constant for CH3NHCH2
= CH3 + CH2NH
(uncertainty factor used: 2). Figure 4.3 presents the
contributions from each source of
uncertainty, which were obtained by perturbing each uncertainty
source to its error
limits and refitting an overall rate constant for Reaction 4.1.
The uncertainty in Reaction
4.8 is the major contributor to the measured rate constant k4.1,
and no significant
influence on k4.1 determination was observed due to the
uncertainties in other secondary
-
35
reactions. All the uncertainties were assumed to be uncorrelated
and combined in a root-
sum-squared method to yield a total uncertainty of ±26% in the
rate constant k4.1 at
1176 K.
Similar tests were carried out for the reaction of DMA with OH,
over the
temperature range of 925-1307 K, and pressures of 0.89-1.24 atm,
with the measured
overall rate constants summarized in Table 4.2. Different
initial dimethylamine
concentrations were implemented to confirm the pseudo-first
order kinetics in OH.
Table 4.2. Measured rate constants for DMA + OH = Products
T P k4.1 x 10-13
[K] [atm] [cm3mol
-1s
-1]
60 ppm TBHP, 440 ppm DMA, Ar
925 0.99 2.80
995 1.24 2.90
1016 0.93 3.00
1170 1.16 3.15
22 ppm TBHP, 320 ppm DMA, Ar
926 1.16 2.80
1141 1.16 3.15
1176 0.89 3.20
1278 1.17 3.40
1307 1.04 3.50
30 pp