characterization of etbe + cl and isooctane + cl
TRANSCRIPT
The University of San FranciscoUSF Scholarship: a digital repository @ Gleeson Library |Geschke Center
Master's Theses Theses, Dissertations, Capstones and Projects
Summer 8-1-2016
Characterization of ETBE + Cl and Isooctane + ClCombustion Products using SynchrotronPhotoionizationrong yaouniversity of San Francisco, [email protected]
Follow this and additional works at: https://repository.usfca.edu/thes
Part of the Physical Chemistry Commons
This Thesis is brought to you for free and open access by the Theses, Dissertations, Capstones and Projects at USF Scholarship: a digital repository @Gleeson Library | Geschke Center. It has been accepted for inclusion in Master's Theses by an authorized administrator of USF Scholarship: a digitalrepository @ Gleeson Library | Geschke Center. For more information, please contact [email protected].
Recommended Citationyao, rong, "Characterization of ETBE + Cl and Isooctane + Cl Combustion Products using Synchrotron Photoionization" (2016).Master's Theses. 188.https://repository.usfca.edu/thes/188
Characterization of ETBE + Cl and Isooctane + Cl
Combustion Products using Synchrotron Photoionization
To the Department of Chemistry
of the University of San Francisco
in partial fulfillment of the degree of
master of Science in Chemistry
A thesis written by Rong Yao
Bachelor of Clinical Pharmacology
Anhui Medical University, China
05/28/2016
Characterization of ETBE + Cl and Isooctane + Cl
Combustion Products using Synchrotron Photoionization
Thesis written by Rong Yao
This thesis is written under the guidance of Faculty Advisory Committee, and approved by all its
members, has been accepted in partial fulfillment of the requirements for the degree of
Master of Science in Chemistry
Thesis Committee
Giovanni Meloni, Ph.D. Date
Associate Professor, Research Director
William Melaugh, Ph.D. Date
Professor
Ryan West, Ph.D. Date
Assistant Professor
Marcelo Camperi, Ph.D. Date
Dean, College of Art and Science
Acknowledgments
This thesis would not have been possible without the support and help from many people in
many different ways. Those people to whom I owe a great deal of thanks I would like to address
my appreciation:
First and the foremost, I would like to express my gratitude to my advisor, Dr. Giovanni
Meloni. During my two-year graduate study here, Dr. Meloni is providing me continuous
invaluable supports, supervisions, encouragements and guidance for research. He shows me what
a scientist should be. His irrepressible enthusiasm and constant striving to be the best keep
pushing us to learn more. I am very honored to having the chance working with Dr. Meloni.
Thank you for your attention until the completion of my dissertation. It is meaningful for me.
At the beginning of my master career, Joseph Czekner and Ken Smith had helped and
trained me. With great patience, they supplied me excellent training on both experiment and
theory that servers as important foundations for me to become independent and successful in my
later research. What they had done are really appreciated. I also want to thank all the other group
members in Dr. Meloni’s group, including Brittany Bryan, Magaly Wooten, Audrey Smith, Giel
Muller, for providing such a friendly and supportive environment.
Furthermore I would like to thank the Chemistry Department at USF. Gratitude goes to Dr.
West and Dr. Melaugh for their generous help on my thesis. I am also grateful to the department
supporting staff including Chad and Angela. Special thanks go to the program assistant Deidre.
Last, but not least, I want to thank my family and my husband, Yuntao Xu for their endless
support and encouragement. They are always here when I need them. I could not have been what
I am today without them.
Abstract
The low temperature (298 – 700 K) oxidation characterizations of a series of fuel
additives (ETBE and isooctane) were studied at advanced light source in Lawrence Berkley
National lab, using a multiplexed chemical kinetics photoionization mass spectrometer with
tunable synchrotron radiation. Using Cl atoms as initiators in presence of oxygen,
photoionization data were collected. Data analysis was performed via characterization of the
reaction species photoionization spectra and kinetic traces, from which the products and also the
intermediates were identified. Relevant reaction dynamics on potential energy surfaces were
calculated, reaction enthalpies of each individual step were presented using the CBS-QB3
composite model, and reaction pathways were proposed.
v
Table of Contents
Chapter 1 Introduction
1.1 Energy Insufficient …………………………………………………………………………….....1
1.2 Environmental Crisis …………………………………………………………………………......1
1.3 Combustion …………………………………………………………………………………….....1
1.4 Homogeneous Charge Compression Ignition (HCCI) Engine .......................................................2
1.5 Fuel Additives………………………………………………………………………………….. ..4
1.6 Reference ………………………………………………………………………………………....5
Chapter 2 Experimental Set-Up
2.1 Introduction……………………………………………………………………………………..6
2.2 Excimer Laser…………………………………………………………………………………...6
2.3 Slow-Flow Reaction Tube…………………………………………………………………........8
2.4 Vacuum Systems.……………………………………………………………………………….8
2.5 Photoionization Source…………………………………………………………………………10
2.6 Beamline…………………………………………...……...........................................................12
2.7 Microchannel Plates (MCP) Detector………………………………………………………….14
2.8 Time-of-Flight Mass Spectrometer…………………………………………………………….16
2.9 Reference……………………………………………………………………………………….17
Chapter 3 Theory and Methodology
3.1 Introduction……………………………………………………………………………………..19
3.2 Radical…………………………………………………………………………………………..19
3.3 Ionization and dissociation Energy……………………………….……………………………..20
3.4 Photoionization Cross Section……………………………………………………………….….21
3.5 Frank-Condon Principle………………………………………………………………………....23
3.6 Photoionization Spectra………………………………………………………………………....25
3.7 Electronic Structure Calculations…………………………………………………………….….26
3.8 Data Analysis……………………………………………………………………………………29
vi
3.9 Reference………………………………………………………………………………………..31
Chapter 4 Synchrotron Photoionization Investigation of the Oxidation of Ethy Tert-Butyl Ether
4.1 Introduction……………………………………………………………………………………..34
4.2 Experimental Section…………………………………………………………………………...35
4.3 Computation Method…………………………………………………………………………...36
4.4 Results…………………………………………………………………………………………..36
4.5 Postulated Mechanism…………………………………………………………………………..42
4.6 Branching Ratios………………………………………………………………………………..47
4.7 Conclusions……………………………………………………………………………………..49
4.8 Reference……………………………………………………………………………………….49
Chapter 5 Characterization of Isooctane Oxidation at Low Temperature via Synchrotron
Photoionization Mass Spectrometry
5.1 Introduction………………………………………………………………………………………53
5.2 Experimental Section…………………………………………………………………………….54
5.3 Computation Methods……………………………………………………………………………54
5.4 Results……………………………………………………………………………………………55
5.5 Postulated Mechanism……………………………………………………………………………60
5.6 Conclusions……………………………………………………………………………………….66
5.7 Reference…………………………………………………………………………………………67
Chapter 1.Introduction
1.1 Energy Insufficient The rising population, heavy industrialization, and motorization have led to a steep rise of the energy demand. The major energy sources in global scale are petroleum (oil), natural gas, coal, nuclear, and renewable. According to US Energy Information Administration (EIA),1 the total primary energy consumption will grow from 95 quadrillion Btu ,which is the abbreviation of British thermal unit, a traditional energy unit defined as the energy to cool or heat one pound of water by one degree and used in the power, steam generation, heating and air conditioning industries,2 in 2012 to 106 quadrillion Btu in 2040, an increase of 12% in the Annual Energy Outlook 2014.3 The domestic production of natural gas and crude oil will grow annually of 0.8 million barrels per day (MMbbl/d), through 2016 in total it will be 9.5 MMbbl/d to meet the residential and commercial needs.1 Since the enormous consumption of coal and gasoline and the increasing need for energy and the limited reserve, seeking new kinds of replaceable energy and improving the combustion efficiency of engines is gaining public and scientific attention.
1.2 Environmental Crisis The environmental crisis represented by greenhouse effect, climate change, air pollution, ozone depletion, and severe water contamination, requires an immediate solution. It is expected that the concentration of CO2 in the atmosphere will be around 480 parts per million (ppm) in 2050, it will continue to rise to around 585 ppm in 2100 with the growing demand for fossil fuel.4 It has been shown that the total green house gas emission has increased by 65% to 222,988 Gg, or gigagram that equals to 106Kg,5 in 2000 from the 1994 level.6 In January 2013 Beijing experienced a prolonged and severe smog, which contained 40 times the amount of hazardous particles than the safe level defined by the World Health Organization.7 These hazardous particles mostly came from coal combustion and vehicles exhaust emissions. For these reasons, there are many studies concerning the energy output optimization and lowering of pollutant emissions. For example, new biodiesel, produced by enzymatic transesterification of cottonseed oil with methanol, showed great performance in greenhouse gas emission reduction: CO emission gets reduced by 33.3% and CO2 gets reduced by 8.4%.8
1.3 Combustion Combustion is basically the exothermic oxidation chemical reactions of fuels. Its general equation is:
Fuel + Air CO2 + H2O + Heat However, this only occurs in simple and complete combustion. But for most of the fuels, such as diesel, oil, coal, or wood, incomplete combustion occurs, which involves numerous individual free radicals and intermediates. As shown in Fig. 1.1, several different intermediates emerged during the low pressure oxidation of n-butanol.
1
Figure 1.1. Reaction pathway for n-butanol in low pressure premixed flames.9 These different intermediate reaction pathways will give rise to unwanted pollutants. In this thesis, we specifically focus on reactions in which the fuel and oxidizers are premixed before combustion at lower temperature (<900 K).
1.4 Homogeneous Charge Compression Ignition ( HCCI ) Engine The quality of combustion can be improved by advanced combustion engines. Homogeneous Charge Compression Ignition or HCCI is one example of new generation advanced engines, which has no spark plug and relies on compression ignition of premixed fuels autoignition. HCCI engine basically has two benefits. It lowers NOx emission10 and improves combustion efficiency. Via a homogeneous charge in the chamber at high compression ratio through multiple ignition points the combustion is more homogeneous and, therefore, more efficient. At the same time, the maximum temperature of the charge is kept low by means of super-lean mixtures and/or inert dilution. With such conditions, very low NOx are produced and relatively high efficiency cycles can be designed.11 Figure 1.2 shows the schematics of traditional spark and Diesel engines compared to a HCCI engine.
2
Figure 1.2. General layout of traditional spark induced diesel and petroleum engines compared to a homogeneous charge compression ignition (HCCI) engine.11
The challenges of HCCI engines are ignition timing, combustion duration,12 and load range of fuel.13 Fuels like DME14 and n-butanol12 can be operated in such engines, where low temperature auto-ignition occurs. Therefore, the low temperature oxidation of fuels is crucial to improve the HCCI systems and combustion efficiency, and lower pollutants emission. Reactivity Controlled Compression Ignition (RCCI) is another example of advanced engines, which allows increased combustion efficiency and lower possible emissions. RCCI is a dual fuel, light-duty, and low-temperature combustion technique that operates using fuel premixing like methane/gasoline blend or methanol/diesel and obtains diesel-like efficiency or better with ultra-low nitrogen oxide and soot emissions.15 Figure 1.3 shows the schematic of a RCCI engine, in which gasoline and diesel are mixed. RCCI uses in-cylinder fuel blending with at least two fuels of different reactivity and multiple injections to control fuel reactivity and optimize combustion phasing duration and magnitude. First, the low reactivity fuel (in this case gasoline) is injected into the cylinder through the port, and is mixed with air and recycled exhaust gases to create the premixed fuel. Then, the higher reactivity fuel (in this case diesel) is injected directly. Meanwhile, the mixed fuel ignites.16 RCCI is actually a variant of HCCI, which still does not have the spark ignition system but relies on the compression of mixed fuel.
3
Figure 1.3. General layout of a reactivity controlled compression ignition (RCCI) engine.16
There are still many questions to be answered for the development of RCCI, such as: (a) which low and high reactivity fuel are chosen?; (b) What is the most optimized ratio of their relative amounts?; (c) What is the perfect ignition timing?; (d) How is the controlled temperature chosen?17 All these questions need us to explore and investigate the low temperature oxidation of fuels.
1.5 Fuel Additives Additives to petroleum fuels have been proven to make the fuels more stable with higher
combustion efficiency and lower noxious emissions. Specific fuel additives are used depending on the internal combustion engine. For the petrol engines, typical fuel additives are oxidation inhibitors or antioxidants, which
prevent oxidation, polymerization, and polycondensation processes in petrol. Phenol and amines derivatives are currently used the most. Specifically, the main function of amines additives is to control combustion chamber deposit, which cause an increase in the requirement of octane number18 that is a value used to indicate the resistance of a fuel to knock. Its value is based on a scale in which isooctane is 100 and heptane is 0.19 Knocking is the sharp sound caused by combustion of part of compressed air-fuel mixture in the cylinder of an internal engine. Knocking causes overheating of the spark-plug points, erodes the combustion chamber, and decreases combustion efficiency.20 Unsaturated hydrocarbons (mainly alkenes and aromatics) and oxygenates (such as methyl tert-butyl ether and ethanol) comprise another group of very important fuel additives. Unsaturated hydrocarbons can polymerize and form gums, which are injected into the combustion chamber to prevent engine damages. The unsaturated hydrocarbons also can increase the amount of dissolved oxygen to improve the combustion efficiency.20 For the diesel engines, additives are used to increase lubricity, stability, and combustion cetane
number, which refers to the combustion quality of diesel fuel and represents the time delay between the start of injection process and the point where the fuel ignites. The cetane number is determined by the % amount of cetane in isocetane21. The diesel engines will produce more NOx pollutants than gasoline engine since the different burning processes inside the cylinders.22 For example, ethylene glycol monoacetate, an oxygenated additive derived from biomass, has been showed that it improves the combustion performance and lower the NOx pollutants emission.23 In general, additives are relatively unstable compounds, which will be converted to radicals to initiate and promote the combustion of diesel fuel.24 Currently, the most used gasoline additive in the market is ethanol. The oxygenated additives
employed in gasoline engines are mainly alcohols, such as methanol, ethanol isopropyl alcohol, n-
4
butanol, and ethers like methyl tert-butyl ether (MTBE), tert-amyl methyl ether (TAME), ethyl tert-butyl ether (ETBE), and tert-amyl ethyl ether (TAEE). The other group of gasoline additives is antiknock agents like toluene, ferrocene, isooctane, and triptane.25 The antiknock additives interrupt the chain reaction of auto ignition by lowering the temperature, pressure and increasing the auto-ignition rate.26 The diesel additives currently used are 2-ethylhexyl nitrate, isopropyl nitrite, tetrathylene glycol
dinitrate, and di(tert-butyl)peroxide, which are the primary cetane improvers.
1.6 References
(1) (EIA), U. E. I. A. 2014. (2) http://en.wikipedia.org/wiki/British_thermal_unit. (3) http://www.eia.gov/forecasts/aeo/er/pdf/0383er%282014%29.pdf. (4) Dietz, S.; Stern, N. Environmental Econmomics & Policy 2009, 3, 138. (5) http://www.convertunits.com/info/Gg. (6) Leong1, Y. P.; Mustapa, S. I.; Hashim, A. H. International Journal of Climate Change: Impacts & Responses. 2011, 2, 9. (7) Xu, B. Council on Foreign Relations 2014. (8) Chattopadhyay, S. Applied Energy May 2013, 105, 319. (9) Hansen, N.; Harper, M. R.; Green, W. H. Phys. Chem. Chem. Phys. 2011, 13, 20262. (10) Andrae, J. C. G. Energy Fuels 2013, 27, 7098. (11) Hairuddin, A. A.; Yusaf, T.; Wandel, A. P. Renewable Sustainable Energy Rev. 2014, 32, 739. (12) Liu, M.-b.; He, B.-q.; Yuan, J.; Zhao, H. Neiranji Xuebao 2013, 31, 324. (13) Kobayashi, K.; Sako, T.; Sakaguchi, Y.; Morimoto, S.; Kanematsu, S.; Suzuki, K.; Nakazono, T.; Ohtsubo, H.; International Gas Union: 2011; Vol. 3, p 2066. (14) Jang, J.; Lee, Y.; Cho, C.; Woo, Y.; Bae, C. Fuel 2013, 113, 617. (15) Curran, S. J.; Hanson, R. M.; Wagner, R. M. International Journal of Engine Research 2012, 13, 216. (16) Curran, S.; Gao, Z.; Szybist, J.; Wagner, R. 2014 CRC Workshop on Advanced Fuels and Engine Efficiency 2014. (17) Wagner, R.; Curran, S.; Hanson, R.; Barone, T.; Briggs, T.; Kokjohn, S.; Reitz, R. Directions in Engine-efficiency and emissions research conference 2011. (18) Rang Heino; Juri, K. Estonian Academy of Sciences:Chemistry Sep 1, 2003. (19) Hosoya, H. Croatica Chemica ACTA 2002, 2, 433. (20) A.Groysman Corrosion in Systems for storage and transporation of petroleum products and biofuel identification,monitoring and solutions. Springer Science 2014. (21) Bamgboye, A. I.; Hansen, A. C. Int.Agrophysics 2008, 22, 21. (22) C.-Y. Lin; Huang, J.-C. Ocean Engineering Sep. 2003, 30, 1699. (23) Diana Hernández; Jhon J. Fernández; Fanor Mondragón; López, D. Fuel Feb, 2012, 92, 130. (24) Jimmie C. Oxley ; James L. Smith ; Evan Rogers ; Ye, W. Energy& Fuels 2001, 5, 1194. (25) http://en.wikipedia.org/wiki/List_of_gasoline_additives. (26) E. A. Dem'yanenko; A. V. Sachivko; V. P. Tverdokhlebov; P. S. Deineko; A. M. Bakaleinik; V. M. Manaenkov; V. E. Emel'yanov; Onoichenko, S. N. Chemistry and Technology of Fuels and Oils 1993, 29, 267.
5
6
Chapter 2．Experimental Set-Up
2.1 Introduction
This chapter describes the main experimental instruments we used for our research. The experiments
are conducted at the Chemical Dynamics Beamline 9.0.2 of the Advanced Light Sources (ALS) at the
Lawrence Berkeley National Laboratory. Radical precursor and reactant diluted with helium gas flow into
the top of the reactor tube with a flow velocity of about 400 cm s-1
at a total flow of 100 standard cubic
centimeter per minute (sccm) and a total pressure of 4 Torr at room temperature and 7 Torr at 550 K.
Their flows are metered by calibrated individual mass flow controllers. Reactions are initiated by
formation of free radicals from a precursor using an unfocused excimer laser pulse (possible wavelengths
include 193, 248, or 351 nm depending on the photolytic precursor), which propagates collinearly down
the reaction tube, ensuring spatial homogeneity of radicals on a relatively short time scale. The quasi-
continuous tunable vacuum ultraviolet radiation combined with the time of flight (TOF) mass
spectrometer (MS) is used to probe our systems. The overall schematic is presented below (Figure 2.1).
The important components will be discussed further.
Figure 2.1. Cross section view of reactor tube and multiplexed photoionization time of flight mass
spectrometer using the tunable VUV radiation of the Advanced Light Source.1
2.2 Excimer Laser
The experiments are initiated with the reaction of the target reactant molecule with a radical
species, which is generated by the photolysis or photodissociation of precursors with the unfocused
excimer laser pulses. The excimer (excited dimer) laser breaks specific chemical bonds and generates
7
radical precursors at precise wavelengths. For example, Cl radical is produced by Cl2 photolysis at 351
nm, and OH radicals are generated by H2O2 photolysis at 248 nm.
Figure 2.2 shows the components of a state-of-the art modular excimer laser. Optical pulse
multiplier is one of its important components. It receives the laser output beam and produces a larger
number of pulses. Usually a noble gas (argon, krypton, or xenon) and a reactive gas (an halogen such as
fluorine or chlorine) are used as the lasing medium. For a KrF laser, the lasing medium is about 0.1%
fluorine, 1.3% krypton, and the rest helium as buffer gas. For an ArF laser, the laser medium is about
0.1% fluorine, 3% argon and the rest helium.2 The laser gas is introduced in the laser cavity at a pressure
between 3 and 3.5 atm using the tangential blower located in the laser discharge chamber. The natural
bandwidth of the excimer laser is narrowed by line narrowing modules that comprise of more optical units
used for the purpose of selecting a center wavelength and, therefore, narrowing the bandwidth. These
components are designed for quick replacement as individual units to minimize downtime to the laser
when maintenance is performed. A commutation module, a compression module, and a high voltage
power supply module serve as a pulse power system and provide electrical pulses to the electrodes located
in the discharge chamber. An interface module and a control module include an elliptical mirror and a
diffraction grating, which is a laser beam dispersion and extraction unit. It receives the laser beam
extracted through the laser oscillator and is configured to selectively extract a specific wavelength. A
stabilization module and an auto-shutter module serve as a prism and a selective transmission surface,
respectively, which transmit the laser beam with the selected wavelength and reflect other lines of the
laser beam. The interface and control modules also include a diffraction grating cooling system for
cooling the grating and preventing the gas from diffusing toward the front surface of the diffraction
grating. If the gas reaches the front surface of the diffraction grating by diffusing, the grating which is
very sensitive will malfunction due to contamination.3
Figure 2.2. The schematic of KrF excimer laser2.
Lasers work on the basic principle of population inversion, which occurs when the number of
atoms or molecules in an excited state is more than the number existing in lower energy states.4
Specifically, population inversion in excimer lasers is achieved by an electric current from the high-
voltage electrode that produces molecules in an excited state. This electric discharge will excite Kr to an
excited state, or kick an electron off Kr creating Kr ions. Meanwhile, some electrons collide with F2
molecules breaking their bond and ionizing the individual atoms. At this point all these excited-state Kr
atoms, F ions, and Kr ions in the laser discharge region can combine in several ways to create excimer
(i.e., excited dimers) molecules, which can only exist in an excited state. These excimer molecules decay
8
to their ground state by emitting a photon and give rise to the laser light in the ultraviolet range. Their
ground state is inherently unstable and dissociates back into two unbound atoms. 5,6
The unfocused excimer laser in our experiments is operated at a repetition rate of 4 Hz, with the
fluence of 10-60 mJ cm-2
and pulsewidth of 20 ns.
2.3 Slow-Flow Reaction Tube
The reaction is carried out in a 62 cm long side-sample quartz flow reactor tube with the inner
diameter of 1.05 cm and the outer diameter of 1.27 cm. The reactions are investigated under a selected
temperature, which can be varied between 298 and 700 K, and pressure between 4 and 10 Torr. In order
to assure uniform initial radical density, the total pressure is kept low so that gaseous diffusion in the
reactor smoothes any inhomogeneity on a short time-scale. The temperature inside the cell is measured by
a thermocouple utilizing a closed-loop feedback circuit controlling the gas temperature over the range
300−1050 K. The reactor tube is thermally insulated by a 18 m thick nichrome tape in order to minimize
temperature gradients. The heating tape is insulated with square-weave, yttria-stabilized zirconia cloth
(ZYW-15, Zircar Zirconia, Inc.) and encapsulated by a gold-plated copper sheath in order to keep
temperature homogeneity.1 The pressure in the reactor is maintained at a specific chosen value by gas
removal using a Roots pump controlled by a closed-loop feedback throttle valve. The total gas flow rate
in the reactor tube is maintained at 400 cm s-1
at each pressure and temperature in order to have a fresh
sample at each laser shot. A 400−650 m pinhole in the side of the reactor tube, which is 35 cm
downstream from the opening, allows gases to effuse as a sampled beam into the source chamber. The
remainder of the main chamber (source chamber) is evacuated using a 3200 Ls-1
oil-free turbo molecular
pump. Reaction gases are measured by individual calibrated mass flow controllers and fed down the tube.
2.4 Vacuum Systems
The reactions are carried out under differential vacuum environment. The pressure in the
beamline before our experimental apparatus is extremely low (on the order of 10-9
Torr) to minimize the
photon absorption. There are three different vacuum pumps, which are designed to be free of hydrocarbon
contamination: roots pumps, turbomolecular pumps, and backing scroll pumps.7
The roots pumps evacuate the reactor tube. Their working mechanism consists of two
symmetrically counter-rotating rotors separated by a thin gap (0.05−0.25 mm) rotating in opposite
directions.8 Figure 2.3 (a) shows the picture of a roots pump and (b) shows its internal diagram. Its
vacuum capacity can vary from 21 to 8,333 L/s (75 to 30,000 m3/h) with a total operating pressure range
from 10 to 10-3
mbar.
(a) (b)
Figure 2.3. (a) General view and (b) diagram of a roots pump.8
9
The turbo molecular pumps, as shown in Figure 2.4, consist of a rotor, a stator, and turbine blades,
which are located around the circumferences of the stator and the rotor.9 The space between the stator and
the rotor disks is on the order of millimeters or less. The high-speed rotating blades transfer momentum to
the gas molecules, with a component towards the pump outlet. Concentration, arrival rate, and
directionality are determined by the flow rates and chamber pressure.10
Kinetic gas theory shows us the
parameters affecting the gas average molecular speed c in the following equation:
M
TRc
*
**8
(2.1)
where M is the molar mass of the pumped gas, R is the universal ideal gas constant, and T is the
temperature of the source chamber.9
Three turbo molecular pumps are employed in our experiments. A 3200L/s turbomolecular pump
is used to evacuate the reaction chamber. A 600 L/s turbomolecular pump is employed to evacuate the
detector region and a 1600 L/s turbomolecular pump is used to pump the ionization region.1
Figure 2.4. Cross section of a turbo molecular pump.9
10
The scroll pumps are used to pump the entire apparatus before the turbo pumps start. Their
operating pressure range may vary from 1,000 to 10-2
mbar with a pumping speed of 5.6 to 14 L/s (20 to
50 m3/h).
7 The scroll pump consists of two mated plates, each with a spiral machined into it, as shown in
Figure 2.5 (a).11
The two plates are interweaved, and one remains fixed while the other is driven in an
orbital motion, so that one spiral sways involuntarily within the second one.12
A schematic of how scroll
pumps work is given in Figure 2.5 (b).13
The motion of the two plates traps volumes of gas and
compresses them along the spiral in chambers that are constantly becoming smaller until the gas is
expelled to the atmosphere.
(a) (b)
Figure 2.5. (a) Picture of the spiraled plate of a scroll pump,11
and (b) its general schematic.13
2.5 Photoionization Source
A specific energy range is needed to ionize the reaction species. In our experiments, 3rd
generation synchrotron radiation from the Chemical Dynamics Beamline of the Advanced Light Source
(ALS) at Lawrence Berkeley National Laboratory in the UV−VUV region (7.2−25 eV) is employed to
provide this ionizing energy. The details of synchrotron radiation will be discussed further below.
Synchrotron radiation is generated when charged particles (at the ALS electrons are used) move
at relativistic speeds in a curved trajectory. The ALS is a third generation facility, in which the storage
ring with a lower emittance and long straight sections for insertion devices permits achieving even higher
brightness than other light sources and with it a considerable degree of spatial coherence.14
The ALS
produces high brightness (1021
photo cm-2
sr-1
), tunable (7.2–25 eV), quasi-continuous, and medium
resolution vacuum ultraviolet (VUV) radiation. The ALS is composed of an electron generator, linear
accelerator, booster ring, storage ring, and the beamlines as shown in Figure 2.6.
11
Figure 2.6. Layout of the ALS at the Lawrence Berkeley National Laboratory.15
The most outstanding feature of synchrotron radiation is the high brightness (i.e., photon flux per
unit area). The brightness of the photon beam from the ALS is a million times more intense than that
produced by X-ray tubes and can be several orders of magnitude larger than any conventional laboratory
source. The other remarkable feature of synchrotron radiation is its tunability. The radiation can be
continuously changed using appropriate monochromators on each beamline. The energy range of the
beamline we use is 7.2−25 eV with an energy resolution of approximately 25 meV.
At the ALS first the electrons are produced in the electron gun by thermionic emission from a
heated barium aluminate cathode and accelerated towards a doughnut-shape anode in pulses at the rate of
500 MHz. The Linac or linear accelerator, shown in Figure 2.7, rapidly speeds up the generated electrons
by means of electromagnetic fields leading most of the electrons to the buncher where powerful
microwave radiation accelerates the pulsed electrons and compresses them into bunches.16
After the linear
accelerator, electrons are injected into the booster ring, a circle structure, which boost the electrons to
99.99994% the speed of light. At this point the electrons are ready to be transferred to the storage ring.
12
Figure 2.7. The overview of ALS 50 MeV Linac.16
Once the electrons reach an optimum energy, they are injected into a roughly circular storage ring under
vacuum, where the electrons circle millions of times per second. In the storage ring, bending magnets,
undulators, and wigglers are used to make and deliver the synchrotron radiation. Bending magnets
provide broad spectrum, low power radiation at the turns in the storage ring. Undulators and wigglers are
periodic magnets, repeated as N and S poles, placed in the straight section of the storage ring, as shown in
Figure 2.8. These periodic arrays of alternating magnets force the electrons through a series of oscillations
around their mainly straight trajectory, “wiggle” electrons back and forth to form a narrow beam of light
that is a hundred times brighter than conventional X-ray sources. Since wigglers have less magnets than
undulators, undulators provide bright, laser like radiation, while wigglers provide broad spectrum high
photon flux radiation.17
The tuning of the undulator is achieved by varying the value of the so-called
“magnetic deflection parameter” which is proportional to the strength of the magnetic field (B0) and the
spacing (λμ) between every adjacent two magnets with same polarity (as shown in Fig. 2.8).18
Three
undulators are used at the ALS: two U5.0 that have 89 periods and each period is 5 cm, and one U8.0,
which has 55 periods and each period is 8 cm.19
Figure 2.8. The periodic magnets of an undulator.17
Undulator radiation generates high power, tunable X-ray sources, but is not inherently 100%
spectrally and spatially coherent. Radiation from an undulator is quasi monochromatic. Increased spectral
coherence is achieved with a coupled monochromator.17
2.6 Beamline
When the electromagnetic radiation leaves the storage ring, gas filters and monochromators are
used to narrow the light in a specific energy range and deliver into different beamlines. As shown in
13
Figure 2.9, there are 43 beamlines at the ALS, each optimized for a particular field of research according
to its energy range and characteristics.
Figure 2.9. The ALS beamline diagram.20
The beamline used in our experiments is 9.0.2, also known as Chemical Dynamics, which
operates in the energy range from 7.8 to 25 eV. Its schematic is displayed in Figure 2.10. 30 Torr of argon,
krypton, or neon are filled into the gas filter and used as an absorption cell to remove radiation with
higher harmonics than the ionization energy of gases Ar (15.76 eV21,22
), Kr (14.00 eV21,23
), and Ne (21.56
eV21,24
). Then, a 3 m or 6.5 m monochromator is used at the 9.0.2 beamline depending on the specific
utilized end-station. The monochramator is an optical device that uses a diffraction grating to filter a
broad, polychromatic band of light to a specific wavelength. For our experiments, a 3 m monochromator
is employed to narrow the wavelength of the radiation into the 10-50 meV resolution.1
14
Figure 2.10. Schematic layout of Beamline 9.0.2.
The monochromator selects a specific wavelength value based on Braggs’ law (equation 2.2). An
incident beam that scatters on a diffraction grating is reflected according to the following equation,
sin2dn (2.2)
where d is the distance between layers in a crystal, θ is the incident angle, λ is the wavelength of the
incident beam of light, and n is order of the reflection. Essentially, the grating works as an optical filter
allowing only a specific wavelength to be reflected from the incident radiation. The reflection occurs
when Bragg’s law is satisfied. By changing the angle θ different wavelengths can be selected.
2.7 Microchannel Plates (MCP) Detector
Electron multipliers have been used as detectors in mass spectrometers for over forty years. One
of the most used types is the continuous dynode multiplier that typically is made of glass. Some of them
consist of coated ceramic materials or a combination of glass and ceramic. Most of these detectors are of
the single channel type and are called channeltron.25
An extreme reduction of the size of a linear
channeltron tube to some micrometers in diameter can be achieved by a single linear channel electron
multiplier,26
as shown in Figure 2.11 (a). As the incident beam of radiation strikes on the channel wall,
electrons are emitted, which strike the channel wall again, and secondary electrons are generated. The
number of secondary electrons generated is dependent on the velocity of the initial particle and the
electron work function of the surface of impact. The length and diameter of the tube (number of the
collisions) will affect the generation of the secondary electrons as well. These secondary electrons are
accelerated by an electric field that is created by a voltage applied across both ends of the electron
multiplier. The electrons follow a parabolic trajectory and are delivered to the rear plate, where the signal
is measured as a current. However, the gain for this linear channel electron multiplier is relatively limited.
One limitation is due to possible interactions of the electrons with the residual gas located in the tube.
Cations, which come from the collision of an accelerated electron with a neutral gas molecule, attract the
electron, thus affecting the electron gain. Usually the electron gain of a single linear channel electron
multiplier is about 103-10
4.27
15
A microchannel plate (MCP) consists of millions of ultra-thin conductive glass capillaries from 4
to 25 μm in diameter and 0.20 to 1.0 mm in length fused together and sliced in the shape of a thin plate, as
shown in Figure 2.11 (b). Each capillary or channel functions as an independent secondary electron
multiplier and they form together a two-dimensional secondary electron multiplier.27
For the purpose of
increasing the electron gain, two MCPs are often sandwiched together so that the small angles oppose
each other (Chevron configuration) to obtain gains of 106-10
7, which is used in our experiment. A
stacking of three MCP’s (Z-stack configuration) will be up to gains of 108. These three types of MCP are
showed in Figure 2.12.
(a)
(b)
Figure 2.11. (a) Schematic of a single linear channel multiplier,26
and (b) an MCP detector from
R.M.Jordan Co.28
16
Figure 2.12. Multichannel plates arrays: Single, Chevron (double or V-stack) and z-stack configuration.
2.8 Time-of-Flight Mass Spectrometer
The time-of- flight (TOF) mass spectrometer (MS) is used to detect and identify the species in our
experiments. TOF-MS operate on the simple principle that ions of different mass-to-charge (m/z) ratios,
but equal kinetic energy, when projected into an electric field free region, will separate according to their
m/z ratios.29
Since charge is nominally unity, the separation occurs as a function of mass, which is
described in equation 2.3
Ekin = ezU = 1/2 mv2 (2.3)
Then, the velocity can be expressed as
m
ezUv
2 (2.4)
If the ions travel a known distance s to a detector, then the TOF is given by the expression 2.5
t
s
m
ezUv
2 (2.5)
where e is the electronic charge, z the ion charge, U the accelerating voltage, m the ion mass, and t the
flight time. In this case, with the same Ekin the flight time is proportional to the square root of the ion m/z
ratio. The ions with smaller m/z ratio will reach the detector before the ions with larger m/z ratios. By
measuring t, with the known ion charge, the ion mass is determined.
17
Figure 2.12. Schematic overview of orthogonal time-of-flight mass spectrometer.30
The time-of-flight mass spectrometer utilized in our experiments has an orthogonal
configuration,24, 29
which is shown in Figure 2.12. The ions produced at low pressure expand continuously
into a vacuum chamber by means of a 3-step differential pumping system. Several ion optical lenses
collimate the continuous ion beam into an extraction region that consists of several electrodes. The ions
enter a field free extraction region. Ions of all masses, which have the same velocity, cross this region
without experiencing any electrical force. When an electrical pulse or bias is supplied to the back
electrode, the ions in the extraction region are accelerated in a direction perpendicular to the original axis
of the beam in the vacuum chamber. Repeating this pulse periodically, ion packets are generated at the
repetition rate of the pulse. The ions are further accelerated by a constant electrical field into a field- free
region (flight tube) and then are mass analyzed according to the equation 2.5.
2.9 References
(1) Osborn, D. L.; Zou, P.; Johnsen, H.; Hayden, C. C.; Taatjes, C. A.; Knyazev, V. D.;
North, S. W.; Peterka, D. S.; Ahmed, M.; Leone, S. R. Rev. Sci. Instrum. 2008, 79, 104103/1.
(2) Richard G. Morton; Partlo, W. N. US Patents 2000, US 6067311 A.
(3) Kim, J.-S. US Patents:US 7426229 B2 2006.
(4) Koerber, K.; Kulik, C.; Springer GmbH: 2005, p 33.
(5) McNaught, A. D.; Wilkinson, A. IUPAC. Compendium of Chemical Terminology, 2nd ed
by Blackwell Scientific Publications, Oxford 1997.
(6) Basting, D.; Marowsky, G. Excimer Laser Technology. Springer: 2005.
(7) Salikeev, S.; Burmistrov, A.; Bronshtein, M.; Fomina, M. Vak. Forsch. Prax. 2014, 26,
40.
(8) Burgmann, W.; Goehler, K. Metallurgist (N. Y., NY, U. S.) 2013, 57, 516.
(9) Hermann Adam ; Alfred Bolz; Hermann Boy; Heinz Dohmen; Karl Gogol; Wolfgang
Jorisch; Walter Mönning; Hans-Jürgen Mundinger; Hans-Dieter Otten; Willi Scheer; Helmut Seige;
Wolfgang Schwarz; Klaus Stepputat; Dieter Urban; Heinz-Josef Wirtzfeld; Zenker, H.-J. Fundamentals
of vacuum technology 1998.
(10) R. Y. Jou; S. C. Tzeng; Liou, a. J. H. International Journal of Rotating Machinery 2004,
10, 1.
18
(11) Yau, P. M.; Patil, A. N. R&D (Cahners) 2000, 42, 80.
(12) Anon R&D (Cahners) 1997, 39, 55.
(13) Lafferty, J. M., ed Foundation of Vacuum Science and Technology 1998, Wiley.
(14) Albert Thompson; David Attwood; Eric Gulllikson; Malcolm Howells; Kim, K.-J.; Kirz,
J.; Kortright, J.; Winick, H.; Lindau, I.; Liu, Y.; Pianetta, P.; Robinson, A.; Scofield, J.; Underwood, J.;
Williams, G. X-ray data booklet 2009, Center for A-ray optics and Advanced Light Source.
(15) Berhardt, K.-H. Vacuum Technology 2008, Braunfels, Germany: Pfeiffer Vacuum
GmbH. .
(16) Selph, F.; Massoletti, D.; IEEE: 1991; Vol. 5, p 2978.
(17) Rosfjord, K. M.; Liu, Y.; Attwood, D. T. IEEE J. Sel. Top. Quantum Electron. 2004, 10,
1405.
(18) Hofmann, A. The Physics of Synchrotron Radiation; Cambridge Univ. Press, 2004.
(19) Mossessian, D. A.; Heimann, P. A. Rev. Sci. Instrum. 1995, 66, 5153.
(20) http://www-als.lbl.gov/index.php/beamlines/beamlines-directory.html.
(21) Lide, D. R. Ionization potentials of atoms and atomic ions in Handbook of Chem and
Phys 1992, 10, 211.
(22) Weitzel, K. M.; Mahnert, J.; Penno, M. Chem. Phys.Lett. 1994, 224, 371.
(23) Wetzel, R. C.; Baiocchi, F. A.; Hayes, T. R.; Freund, R. S. Phys Rev.A 1987, 35, 559.
(24) Kelly, R. L. J. Phys. Chem. Ref. Data, Suppl. 1987, 16, 1.
(25) applications, C. e. m. h. f. m. s. Photonis.
(26) Gross, J. B. Mass Spectrometry. A textbook from Springer. Berlin 2004.
(27) Hamamatsu Technical Information.
(28) http://www.rmjordan.com/mcpdetector.html.
(29) Li, G.; Hieftje, G. M. Time-of-flight mass spectrometer US Patent 5614711A 1997.
(30) Christine Rivera; Andrew Saint; Balsanek, W. Spectroscopy 2006.
Chapter 3． Theory and Methodology 3.1 Introduction
The aim of our research is to characterize intermediates and final products of low temperature oxidation of biofuels using Synchrotron Photoionization Mass Spectrometry, investigate the kinetics properties of the reaction, and probe the possible reaction pathways. From the generation of photons to the kinetics calculations, many concepts and analysis methods are involved, such as radical, photoionization, absolute cross section, Frank-Condon Principle, photoionization spectrum, and electronic structure calculations, just to mention a few.
In this chapter, these concepts and analysis methods will be introduced. The data acquisition and analysis programs will also be briefly discussed.
3.2 Radical
A radical is a neutral reactive intermediate with unpaired electrons. Radicals are usually formed from breaking a covalent bond by means of heat or electromagnetic radiation. Radicals are generally highly reactive and, therefore, unstable with a very short lifetime. Nevertheless, they are key reactive intermediates in both combustion and atmospheric chemistry.1
There are several ways to form radicals: photolysis, photoionization, photodissociation, electron impact, and reaction with some chemical agents.2 In our experiments, the formation of radicals is crucial to initiate reactions. Cl* and OH* radicals are the most used by photolysis of Cl2 and H2O2, respectively, through 355 nm and 248 nm lasers.3,4
(3.1)
(3.2)
(3.3)
(3.4)
(3.5)
As an example, the Cl initiated reaction with methane is presented in the scheme above. Step 3.1 is
19
the initiation step, ultraviolet light provides sufficient energy to split the chlorine molecule bond and leave two chlorine atoms with an unpaired electron. Steps 3.2 and 3.3 are the propagation steps, where reactive intermediates react with stable molecules to generate other reactive intermediates. Cl* radical abstracts a hydrogen from CH4 to form a methyl radical CH3*; CH3* abstracts a Cl atom from Cl2 to form CH3Cl and another Cl*. Steps 3.4 and 3.5 are referred to as termination steps, in which usually two reactive intermediates combine into one stable molecule to terminate the reaction.5 3.3 Ionization and dissociation energy
Mass spectrometry is a very powerful technique for identifying unknowns, and studying molecular structures. The identification of species is based on the ions separation based on their mass-to-charge ratio (m/z). There are several ways to generate ions: electron ionization, chemical ionization, negative-ion chemical ionization, field desorption, particle bombardment, electrospray ionization, and photoionization.6 However, photoionization is one of the classic methods and is also the method we utilized in our experiments, which will be introduced further below.
Photoionization is based on the well-known photoelectric effect. It occurs when electrons are removed from an atom or molecular species by absorption of photons of visible or ultraviolet light. More commonly, the species we are dealing with are ionized in the vacuum ultraviolet (VUV) range Of 7.2– 25 eV.7
Figure 3.1. Possible processes under electron ionization conditions.2 The ionization process can produce different types of ions or fragments as shown in Figure 3.1
depending on the photon energy the atoms or molecule absorb. It can produce a molecular ion, which will have the same molecular weight and elemental composition of the starting analyte, while the
20
amount of energy required to remove an electron from isolated molecular or atoms in the gaseous state is its ionization energy:
AB + υh AB+ + e- (3.6) Computationally (see later), the ionization energy of the species AB [IE(AB)] is equal to IE(AB) = E(AB+) – E(AB) = ∆fHo(AB+) – ∆fHo(AB) (3.7)
where E is the total electronic zero-point corrected energy, and ∆fHo is the standard enthalpy of formation. If both the neutral and cation are in their ground state this quantity is called adiabatic ionization energy (AIE). Ionization can also produce a fragment ion, which corresponds to a smaller piece of the analyte molecule, while bond dissociation energy is involved.2 In this case the energy at which the fragment or daughter ion appears is called the appearance energy (AE). The bond dissociation energy is defined as the standard enthalpy Do change of the following fission:
A B A + B (3.8)
Do is usually derived by the thermochemical equation 3.9 Do(A−B) = ∆fHo(A) + ∆fHo(B) – ∆fHo(AB) (3.9) The relationship between ionization energy (IE) and dissociation energy (Do) is illustrated in the
Figure 3.2 showing the formation of AB+ and AB in a cycle, then using different expressions we can get the connection between Do and IE.
Figure 3.2. Formation Cycle for reactions of AB. From equations 3.7 and 3.9, we have IE(AB) = E(AB+) – E(AB) = ∆fHo(AB+) – ∆fHo(AB) IE(B) = E(B+) – E(B) = ∆fHo(B+) – ∆fHo(B) Do(A−B) = ∆fHo(A) + ∆fHo(B) – ∆fHo(AB) Do(A−B+) = ∆fHo(A) + ∆fHo(B+) – ∆fHo(AB+)
Then, we can obtain the relationship of Do and IE as shown in equation 3.10.
AB+
(∆fH
o
(AB+
))
AB (∆fH
o
(AB))
A (∆fH
o
(A)) B+
(∆fH
o
(B+
))
A (∆fH
o
(A)) B (∆fH
o
(B))
＋
＋
Ionization IE(AB)
Ionization IE(B)
Dissociation
Dissociation
Do
(A−B)
21
Do(AB) = Do(AB+) + IE(AB) - IE(B) ( 3.10)
3.4 Photoionization Cross Section Photoionization cross section is the effective area in which there is the greatest probability to
ionize the species at a given energy. It describes an area through which the electron must travel in order to effectively escape the neutral, therefore, photoionization cross section is often given in units of megabarns (Mb = 10-18 cm2).8
When an atom or molecule gets excited by photoabsorption, the total photoabsorption cross section (σabs
T) is given by:
σabsT = σn + σion
T (3.11)
where σn is cross section for neutral processes and σionT is the total photoionization cross section9.
While the relation between σionT and σabs
T can also be defined as: σion
T = Γ σabsT (3.12)
where Γ is the total photoionization efficiency or quantum yield10 given by:
=Γ 𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁 𝑜𝑜𝑜𝑜 𝑖𝑖𝑜𝑜𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝑁𝑁𝑒𝑒𝑁𝑁𝑖𝑖𝑒𝑒𝑒𝑒/𝑒𝑒𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁 𝑜𝑜𝑜𝑜 𝑝𝑝ℎ𝑜𝑜𝑒𝑒𝑜𝑜𝑖𝑖𝑒𝑒 𝑎𝑎𝑁𝑁𝑒𝑒𝑜𝑜𝑁𝑁𝑁𝑁𝑁𝑁𝑎𝑎/𝑒𝑒
(3.13)
The photoionization cross section at a certain frequency υ is expressed as
υ
πσ υddf
mce
p 2
2
, = (3.14)
or
2
23
,'3
8 Mifhcge
i
pυυπσ υ = (3.15)
where f is the optical oscillator strength, 'υh is the photon energy passed the ionization threshold and gi is the statistical weight of the initial discrete state. The matrix element is defined in terms of continuous wave functions as follows:
υµψψ dfiMif if ∑∫∑∑= *2 2 (3.16)
However, since it is difficult to get accurate electronic wave functions, it is difficult to calculate photoionization cross sections for complicated atoms, not even mention the lager molecules like ethyl tert-butyl ether (ETBE) and isooctane. Recently, Anna Krylov and her coworkers proposed that with a modified central potential model that accounts for the nonspherical charge distribution of the core, using approximate treatments of the photoelectron wave function, plane and Coulomb waves, the accurate molecular photoionization cross sections could be computed.11 Meanwhile, the semi-empirical model by Bobeldijk12 is utilized to estimate the absolute photoionization cross section.
22
A molecule’s photoionization cross section is found through the summation of the cross sections of its component atoms or groups of atoms, and this model reproduces the molecular ionization cross sections within 20%.12 Bobeldijk also mentioned that this approximation tends to breakdown to the vicinity of the ionization energy threshold.12
The cross section is important to determine the product concentrations and calculate the branching ratio and branching fraction of a specific mass product and its isomers. Usually absolute or total photoionization cross sections are obtained by taking the photoionization spectrum of a target species at a known concentration with a calibration species, which has a known photoionization cross section at a specific photon energy. The concentration of a species Cp is related to the photoionization cross section (σ ) by the following relationship:
Sp(E) = kσ p(E)δ pCp (3.17) where Sp(E) is the ion signal of the species p at a specific energy E in the photoionzation spectra, k the instrument constant, and δ p the mass-dependent response of the species p, which in the current investigation is approximately equal to the mass (m) of the observed species to the power of 0.67.13 When the σ p(E) is calculated, the branching fraction, BF, can be obtained by relating the ion signals of the product P and the reactant R at the specific photon energy E as
𝐵𝐵𝐵𝐵𝑃𝑃 = 𝐶𝐶𝑃𝑃𝐶𝐶𝑅𝑅
= ⎣⎢⎢⎢⎡𝑆𝑆𝑃𝑃(𝐸𝐸)
𝜎𝜎𝑃𝑃(𝐸𝐸)𝛿𝛿𝑃𝑃�
⎦⎥⎥⎥⎤
⎣⎢⎢⎢⎡𝑆𝑆𝑅𝑅(𝐸𝐸)
𝜎𝜎𝑅𝑅(𝐸𝐸)𝛿𝛿𝑅𝑅�
⎦⎥⎥⎥⎤
= 𝑆𝑆𝑃𝑃(𝐸𝐸)𝜎𝜎𝑅𝑅(𝐸𝐸)𝛿𝛿𝑅𝑅𝑆𝑆𝑅𝑅(𝐸𝐸)𝜎𝜎𝑃𝑃(𝐸𝐸)𝛿𝛿𝑃𝑃
�𝑚𝑚𝑅𝑅 𝑚𝑚𝑃𝑃� �0.67
(3.18)
The ion signal, at the specific photon energy E, is chosen at a plateau or constant ion signal of the photoionization spectrum, which is done to minimize error due to experimental signal fluctuations. A plot of photon energy against photoionization cross-section is called absolute photoionization spectrum or absolute PI curve, which, if unknown, can be obtained by comparing the species pure sample experimental PI spectrum with a known concentration to the absolute photoionization cross-section of a reference species, such as propene in this work. The absolute cross-section of propene used as a reference was recorded by Person and Nicole.14 3.5 Frank-Condon Principle
A molecule at room temperature normally resides in its ground electronic (S0) and lowest vibrational state ( o=ν ). A ground-state molecule absorbs a photon and undergoes a transition to its excited state (S1). Once a molecule is in this excited vibronic state, it immediately relaxes back down to the lowest vibrational level in the S1 state by releasing energy. After a delay the electron relaxes back down to its ground electronic state S0. But similar to absorption, this relaxation to S0 can occur between the lower vibrational states (ν) and if the transition occurs between the molecular ground state and ionic ground state, the energy required is the minimum and is known as adiabatic ionization energy (AIE).
23
Since nuclei can be considered massive compared to the electron and the ionization occurs in a very short time scale, the position of the nuclei of the target molecule does not change when a vertical electronic transition happens. Frank–Condon (FC) Principle states that when a molecule is ionized by photons or by interaction with energetic electrons, the highest probability configuration of the resulting ionic states will be the one identical to the original neutral molecule. In quantum mechanical terms, the Frank–Condon principle states that the molecule will undergo a transition to the upper vibrational state that most resembles the vibrational wave function of the vibrational ground state of the lower electronic state, as shown in Figure 3.3 (a). The degree of the overlap of the two wave functions is defined by the Frank–Condon factor.15
(a)
(b)
Figure 3.3. (a)Illustration of the transitions from the neutral to the ionic state for a diatomic molecule AB; (b) Schematic distribution of Frank–Condon factors, fFC, for various transitions.2
An electrical dipole transition from the initial vibrational state ( )υ of the ground electronic level
(ε ) to its vibrational state ( 'υ ) of an excited electronic state( 'ε ) occurs when there is a change in the dipole moment. The molecular dipole operator µ is determined by the charge (-e) and location (ri) of the electrons and the charges (+Zje) and locations (Rj) of the nuclei (electron dipole moment µe and nuclear dipole moment µN):
24
∑∑ +−=+=j
jji
iNe RZereµµµ (3.18)
The probability amplitude P for the transition between these two states is given by:
∫== τµψψψµψ dP '*' (3.19)
where ψ and 'ψ are the overall wavefunctions of the initial and final state, respectively. The
overall wavefunction can be expressed approximately as product of the individual vibrational (depending on spatial coordinates of the nuclei), electronic, and spin wavefunctions:
se ψψψψ υ= (3.20)
According to the Born-Oppenheimer approximation, then
∫ ∫ ∫ ∫∫ ∫∫ ∫∫
+=
+=
+=
=
sssNeeessseeen
seNseseese
Nese
sese
ddddded
dd
sde
P
τψψτψµψτψψτψψτψµψτψψ
τψψψµψψψτψψψµψψψ
τψυψψµµψψψ
ψψψµψψψ
υυυυυ
υυυυ
υ
υυ
'*'*'*'*'*'*
'*'*'*'*'*'*
)(***
'''
The first integral after the plus sign is equal to zero because electronic wavefunctions of different states are orthogonal, so
P = ∫∫ ∫ ssseeen dded τψψτψµψτψψ υυ '*'*'* (3.21)
Here the first integral ∫ υυυ τψψ d'* is the Frank-Condon factor, which represents the vibrational
overlap in quantum mechanics, the second integral ∫ eee de τψµψ '* represents the orbital selection rule,
and the third integral ∫ sss dτψψ '* is the spin selection rule quantum expression.16
The quantum mechanical formulation of Frank–Condon principle is the result of a series of approximations, principally the electrical dipole transition assumption and the Born-Oppenheimer approximation. For any given transition, the P value is determined by all of the selection rules; however, spin selection is the largest factor, followed by electronic selection rule. The Franck–Condon factor modulates the relative intensity of the transitions.17
3.6 Photoionization Spectra
In our experiments at the Advanced Light Source, due to the multiplex capability of the apparatus a three-dimensional (3D) data block is obtained, where the ion signal is a function of m/z, reaction time (t), and photon energy (E). Visualization is accomplished by two-dimensional orthogonal slicing of the data by integrating over the desired range of the third dimension. For example, integration over 20-90 ms gives a 2-D plot of intensity as a function of photon energy and mass-to-charge ratio. The
25
photoionization spectrum of the specific mass will be gotten by integrating a photoelectron spectrum at a specific mass. Using the one-dimensional plot of ion signal as a function of synchrotron energy (photoionization energy of target molecule), the isomeric composition at a specific mass-to-charge ratio will be gotten as shown in Figure 3.4.
Figure 3.4. Literature PI curve of formaldehyde superimposed (black solid line) onto the experimental m/z = 30 for the reaction of ETBE with Cl radical at RT.
As shown in Figure 3.4, the ionization threshold is the point at which there is enough kinetic energy for the electron to be able to escape and it is known as adiabatic ionization energy (AIE). The shape of the curve is determined by the overlap of the vibronic wavefunctions of the initial (neutral) and final (cation) electronic states, which are known as Franck-Condon factors. If the vibronic wavefunction overlap between two electronic states is good (very similar geometries of the initial and final states), the PI spectra produced would have a steeper signal increase at the threshold. Otherwise, the PI spectra would present a gentler increase. Comparison of experimental spectra with literature and/or calculated plots is a very important method for the identification of potential products in our experiments.
3.7 Electronic Structure Calculations
Electronic structure calculations are performed routinely for data analysis in our research. The two systems, ethyl tert-butyl ether (ETBE) and isooctane, were studied using the Gaussian 09 program18 by several different levels of theories and basis sets. The CBS-QBS composite method12,19 is employed to obtain optimized molecular structural parameters, such as bond lengths, bond angles, harmonic vibrational frequencies, force constants, ionization energies, and transition state characterization. The Gaussian program is also used to calculate Franck-Condon (FC) factors for the transition between the ground vibronic states of the neutral and cation to generate a photoelectron spectrum, which can be integrated into simulated photoionization spectra by Igor software.
3.7.1 Density-Functional Method
26
Density functional theory (DFT) provides a powerful tool for investigating the properties of the ground state molecule structures according to the functional of the spatially dependent electron density. DFT is constructed based on two Hohenberg-Kohn theorems.20 The first Hohenberg-Kohn theorem states that an electron density depending on three spatial coordinates determines the ground state properties of a many-electron system.21 The second Hohenberg-Kohn theorem gives the definition of an energy functional for the system and confirms that the correct ground state electron density minimizes this energy functional.22 Kohn and Sham adapted DFT into a practical version,23,24 which describes the mathematics of electron densities and their subsequent correlations to molecular energies as shown following:
EDFT[ ρ ]=T[ ρ ]+Ene[ ρ ]+J[ ρ ]+Exc[ ρ ] (3.22)
where E is the total energy, T the electron kinetic energy, Ene the nuclear-electron attraction (Coulombic) energy, J the electron-electron repulsive (Coulombic) energy, and Exc the electron-electron exchange-correlation energy, which is the result of unknown interactions between the electrons and is the most important effect limiting the accuracy of DFT.23,25-28 Each of these terms is a function of the electron density ρ , which is a function of the three positional coordinates (x, y, and z) itself.
The B3LYP method is one of the DFT methods used most in our work. B3LYP is one of the hybrid functionals, which combines the local gradient-corrected correlation-energy functional of Lee-Yang-Parr (LYP) with Hartree-Fock theory.28,29
3.7.2 Complete-Basis-Set method
A basis set in theoretical and computational chemistry is a set of functions, which correspond to linear combinations of electronic wavefunctions and define the space of a range of atoms, to create the molecular orbitals. Generally there are four types of basis sets:
(1) Minimal: in which a single basis function is used for each orbital in a Hartree-Fock calculation for each atom in the molecule. Slater-type orbitals, STO-nG, is the most common minimal basis, where n is an integer and represents the number of Gaussian primitive functions comprising a single basis function.
(2) Double zeta/triple zeta: with the minimal basis sets, all orbitals are approximated to be of the same shape, which is not accurate. The double-zeta basis set treats each atomic orbital as the sum of the two Slater-type orbitals (STOs). One set is closer to the nucleus with larger exponents, the other is further from the nucleus with smaller exponents. The triple zeta basis sets use three Slater-type orbitals instead of two, which results in better accuracy but is more time-consuming.
(3) Split valence: since the inner-shell electrons are not as vital to the calculation, the core orbitals are still represented by a minimal basis, only the valence part of the basis set is calculated using double, triple, quadruple-zeta basis sets. When higher angular momentum functions are added, for example d-functions to carbon atoms or p-functions to hydrogen atoms, angular flexibility will be allowed. Diffuse functions with small exponents added are usually applied to anions.
6-311+G is the basis set used mostly in this thesis, where “6” means that six contracted Guassian basis functions are used to approximate the core atom orbitals, “3” suggests that inner
27
components of valence atom orbitals are expressed with three contracted Guassian functions, “1” indicates outer component is presented by one Gaussian function, “+” implies a highly diffuse function (anions, compounds with lone pairs, and hydrogen-bonded dimers have significant electron density at large distances from nuclei) with a very small orbital exponent.
(4) Correlation-consistent basis set, which are designed to converge systematically using empirical extrapolation techniques.
Complete-Basis-Set (CBS) models involve low-level (Self-consistent field method (SCF)) calculations on large basis sets, mid-sized sets for second-order corrections, and small sets for high level corrections. CBS models are developed to compensate the limits of DFT30 (incomplete treatment of dispersion) and coupled-cluster method (insufficient accuracy or relatively inexpensive for big system), such as CCSD, and to get a more accurate energetic result. CBS-QB3 is the method used most in our work. CBS-QB3 is developed by Petersson and his coworkers19,31-37, and involves the following steps:
I) B3LYP/6-311G(2d,d,p) geometry optimization.
II) B3LYP/6-311+G(2d,d,p) frequencies with a 0.99 scale factor for the ZPE (∆EZPE).
III) Unrestricted Moller-Plesset UMP2/6-311+G(3d2f,2df,2p) energy (EMP2) and CBS extrapolation.
IV) MP4(SDQ)/6-31+G(d(f),p) energy (∆EMP4).
V) CCSD(T)/6-31+G+ energy (∆ECCSD(T)).
Finally, the total energy ECBS-QB3 and free energy GCBS-QB3 are calculated as:
ECBS-QB3=EMP2+∆EMP4+∆ECCSD(T)+∆EZPE+∆ECBS+∆Eemp+∆Eint (3.23)
GCBS-QB3= EMP2+∆EMP4+∆ECCSD(T)+∆Ethermalcorrection+∆ECBS+∆Eemp+∆Eint (3.24)
where∆ECBS is the energy corrected for the basis set truncation error in the second-order energies.31
3.7.3 Transition State Calculation methods
Transition state is defined as the state corresponding to the highest potential energy along the reaction coordinate, which is critical to understand and estimate a given chemical reactions.38 At the transition state, one vibrational normal mode, which aligns with the nuclear motion along the reaction coordinate, should have a negative force constant. This translates in the corresponding vibrational frequency to have an imaginary value. This provides an important criterion for transition states: there is one and only one negative eigenvalue for the Hessian matrix.39
Three-dimensional geometries of transition states now can be visualized explicitly as those of the saddle points located by systematic computation of potential-energy surfaces. Furthermore, the
28
activation barrier can be evaluated accurately. Within the Gaussian 09 program, three ways are used to optimize transition state structures in our work:
I. Using Opt = TS. The input geometry has to be a good initial guess geometry for the transition structure. The input file is a guess structure for the transition state. If the geometry is feasible, the calculation will present the optimized structure with an imaginary frequency proving that it is a first-order saddle point.
II. Using Opt = QST2. Finding the good guess for the right transition state structure may be difficult. QST2 method will help generating the guess geometry. The input file includes two title and molecule specification sections, which are the optimized reactant and product geometries, respectively. The program will calculated a possible transition state between these two.
III. Opt = QST3 is used in more difficult cases. The input file will supply optimized geometries of both reactant and product, and a guessed transition state structure calculated by the TS method.
3.8 Data Analysis.
Our experiments are carried out at the ALS located in the Lawrence Berkeley National Laboratory. The ion intensity, reaction time, photoionization energy, and mass-to-charge ratio of the reaction species (reactants, products, and intermediates) are recorded by a multiplexed time- and energy- resolved mass spectrometer. The ion intensity is plotted against reaction time (t), m/z, and E, then a 3-D “picture” of the reaction will be observed. Whereas 3-D data is not directly used for the analysis due its complexity, 2-D “slices” are employed. The 2-D slices plots are the combinations of photon energy (E), mass-to-charge (m/z), and the time (ms). The upper right corner of Figure 3.5 is a 2-D plot of intensity as a function of photon energy E and mass-to-charge integrated. Other useful 2-D plot is the intensity as a function of time and mass-to-charge obtained by integration over a desired photon energy E range as shown at the bottom right corner of Figure 3.5.
Figure 3.5. Overview of relationship between 3-D data block and 2-D slices.
29
1-D slice photoionization energy vs. intensity for each mass is also known as the photoionization spectrum. The photoionization energy is the energy required to move the outermost electrons from a neutral molecular species. The photoionization spectrum is characteristic for each species. An unknown can be identified based on its m/z and by comparing its PI spectrum to a literature or computed photoionization curve. In Figure 3.6 the red open circles represent the experimental PI spectrum of m/z = 58 unknown, the blue solid line is the PI curve of acetone from the literature40, the green solid line is the PI spectrum of propanal recorded by Cool,41 and the yellow solid line is the literature PI curve of glyoxal.42 The black solid line is the sum of these three literature spectra and when superimposed onto the experimental data match the overall shape. In this case, the m/z = 58 unknown can be identified as a mixture of acetone, propanal, and glyoxal.
Figure 3.6. Experimental photoionization spectrum at room temperature of m/z = 58 superimposed onto the integrated photoelectron spectra of some m/z = 58 isomers in the ETBE+Cl+O2 reaction. In 2-D time and m/z “slices,” which are the ion intensity plotted against reaction time and m/z, the consumption of reactants and formation of products can be observed. 1-D “slices,” which are the ion intensity vs. reaction time for a chosen m/z, show time traces of specific species providing indications on its “nature,” i.e., reactant (depletion), product (formation), primary product (fast formation comparable to the depletion time of the reactant), secondary product (slower formation when compared to the depletion time of the reactant), and radical (fast formation followed by a fast decay). Figure 3.7 is the time trace of m/z = 58, from which we can infer that this m/z consists of primary products. Coupling the time trace with the results obtained from the photoionization spectra shown in Figure 3.6, acetone, propanal and glyoxal are considered to be primary products of in the ETBE+Cl+O2 reaction at 293 K.
30
Figure 3.7. Experimental time-trace of m/z = 58 at room temperature showing product formation in the ETBE+Cl+O2 reaction. Reaction mechanisms could be inferred from the analysis of the reaction species and the branching fractions and/or ratios can be calculated (see Section 3.4). 3.9 Reference (1) Fessenden, R. W. J.phys.Chem 1964, 68, 1508. (2) Gross, J. H. Mass Spectrometry. A textbook from Springer. 2011. (3) Zador, J.; X.Fernandes, R.; Georgievskii, Y.; Meloni, G.; A.Taatjes, C.; A.Miller, J. Proceddings of the Combustion Institute 2009, 32, 271. (4) Meloni, G.; Selby, T. M.; Osborn, D. L.; Taatjes, C. A. J. Phys. Chem. A 2008, 112, 13444. (5) C.Neuman, R. Organic Chemistry, textbook, emeritus 2013. (6) L.Pavia, D.; M.Lampman, G.; S.Kriz, G.; R.Vyvyan, J. Introduction to Mass Spectroscopy, 4th edition, textbook of Brooks/Cole 2009. (7) http://www-als.lbl.gov/index.php/beamlines/beamlines-directory.html. (8) Kieffer, L. J. Natl.Bur.Stand.(U.S.), Spec.Publ. 1976, 426, 219. (9) J.W.Gallagher; C.E.Brion; J.A.R.Samson; P.W.Langhoff J.Phys.Chem.Ref.Data 1988, 17. (10) J.A.R.Samson Atomic Photoionization in Handbuch der Physik; Springer, Berlin, 1982; Vol. 31. (11) Samer Gozem, A. O. G., Takatoshi Ichino, David L. Osborn, John F. Stanton, and Anna I. Krylov, the journal of physical Chemistry letter 2015, 6, 4532. (12) Bobeldijk, M.; Zande, W. J. v. d.; Kistemaker, P. G. Chem. Phys. 1994, 179, 125. (13) Welz, O.; Zador, J.; Savee, J. D.; Ng, M. Y.; Meloni, G.; Fernandes, R. X.; Sheps, L.; Simmons, B. A.; Lee, T. S.; Osborn, D. L.; Taatjes, C. A. Physical Chemistry Chemical Physics 2012, 14, 3112. (14) Person, J. C.; Nicole, P. P. J. Chem. Phys. 1970, 53, 1767. (15) Atkins, P.; Paula, J. d. Physical Chemistry; W. H. Freeman and Company: New York, 2006.
31
(16) Atkins, P. W.; Friedman, R. S. Molecular Quantum Mechanics, Third Edition; Oxford University Press, 2000. (17) Coolidge, A. S.; James, H. M.; Present, R. D. journal of chemical physics 1936, 4, 193. (18) M. J. Frisch; G. W. Trucks; H. B. Schlegel; G. E. Scuseria; M. A. Robb; J. R. Cheeseman; G. Scalmani; V. Barone; B. Mennucci; G. A. Petersson; H. Nakatsuji; M. Caricato; X. Li; H. P. Hratchian; A. F. Izmaylov; J. Bloino; G. Zheng; J. L. Sonnenberg; M. Hada; M. Ehara; K. Toyota; R. Fukuda; J. Hasegawa; M. Ishida; T. Nakajima; Y. Honda; O. Kitao; H. Nakai; T. Vreven; J. A. Montgomery; Jr., J. E. P.; F. Ogliaro; M. Bearpark; J. J. Heyd; E. Brothers; K. N. Kudin; V. N. Staroverov; R. Kobayashi; J. Normand; K. Raghavachari; A. Rendell, J. C.; Burant, S. S. I.; J. Tomasi; M. Cossi; N. Rega; J. M. Millam; M. Klene; J. E. Knox; J. B. Cross; V. Bakken; C. Adamo; J. Jaramillo; R. Gomperts; R. E. Stratmann; O. Yazyev; A. J. Austin; R. Cammi; C. Pomelli; J. W. Ochterski; R. L. Martin; K. Morokuma; V. G. Zakrzewski; G. A. Voth; P. Salvador; J. J. Dannenberg; S. Dapprich; A. D. Daniels; Ö. Farkas; J. B. Foresman; J. V. Ortiz; J. Cioslowski; and D. J. Fox Gasussian, Inc. Wallingford Ct 2009. (19) Montgomery, J. A., Jr.; Frisch, M. J.; Ochterski, J. W.; Petersson, G. A. J. Chem. Phys. 1999, 110, 2822. (20) Hohenberg, P.; Kohn, W. Physical Review 1964, 136, 864. (21) Mei, L. Proceddings of the National Academy of Sciences 1979, 76, 6062. (22) L.J.Bartolotti Physical Review A 1981, 24, 1661. (23) Kohn, W.; Sham, L. J. Phys Rev. 1965, 140, 1133. (24) Parr, R. G.; Yang, W. Oxford Univeristy Press 1994, ISBN 978-0-19-509276-9. (25) Levine, I. N. Quantum Chemsitry 5th Ed; Prentice Hall:Upper Saddle River, New Jersey, 2000. (26) Jensen, F. Introduction to Computational Chemsitry; Wiley:Chichester, England, 1999. (27) Lewars, E. Computational Chemsitry: Introduction to the Theory and Applications of molecular and Quantum Mechanics; Kluwer Academic Publishers: Boston, 2003. (28) Becke, A. D. J. Chem. Phys. 1993, 98, 1372. (29) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. (30) Van Mourik, T.; Gdanitz, R. J. Journal of chemical physics 2002, 116, 9620. (31) Ochterski, J. W.; Petersson, G. A.; Montgomery, J. A., Jr. J. Chem. Phys. 1996, 104, 2598. (32) Nyden, M. R.; Petersson, G. A. J. Chem. Phys. 1981, 75, 1843. (33)Petersson, G. A.; Bennett, A.; Tensfeldt, T. G.; Al-Laham, M. A.; Shirley, W. A.; Mantzaris, J. J. Chem. Phys. 1988, 89, 2193. (34) Petersson, G. A.; Al-Laham, M. A. J. Chem. Phys. 1991, 94, 6081. (35) Petersson, G. A.; Yee, A. K.; Bennett, A. J. Chem. Phys. 1985, 83, 5105. (36) Montgomery, J. A., Jr.; Ochterski, J. W.; Petersson, G. A. J. Chem. Phys. 1994, 101, 5900. (37) Montgomery, J. A., Jr.; Frisch, M. J.; Ochterski, J. W.; Petersson, G. A. J. Chem. Phys. 2000, 112, 6532. (38) Solomons, T. W. G., Fryhle, Craig B. Organic Chemsitry (8th ed); John Wiley & Sons, Inc., 2004. (39) Fueno, T. Transition State: A Theoretical Approach; Gordon and Breach Science Publishers, 1999. (40) Wang, J.; Yang, B.; Cool, T. A.; Hansen, N.; Kasper, T. Int. J. Mass Spectrom. 2008, 269, 220.
32
(41) Cool, T. A.; Nakajima, K.; Mostefaoui, T. A.; Qi, F.; McIlroy, A.; Westmoreland, P. R.; Law, M. E.; Poisson, L.; Peterka, D. S.; Ahmed, M. J. Chem. Phys. 2003, 119, 8356. (42) Holmes, J. L.; Lossing, F. P. Org. Mass Spectrom. 1991, 26, 537.
33
Chapter 4 Synchrotron Photoionization Investigation of the Oxidation of Ethyl Tert-Butyl Ether
Abstract:
In this work the oxidation of ethyl tert-butyl ether (ETBE), a widely used fuel oxygenated additive, is investigated using Cl atoms as initiators in the presence of oxygen. The reaction is carried out at 293, 550, and 700 K. Reaction products are probed by a multiplexed chemical kinetics photoionization mass spectrometer coupled with the synchrotron radiation produced at the Advanced Light Source (ALS) of the Lawrence Berkeley National Laboratory. Products are identified based on mass-to-charge ratio, ionization energies, and shape of photoionization spectra. Reaction pathways are proposed together with detected primary products.
4.1 Introduction
A variety of pollutants exist in combustion exhaust, one category of which, particulate matter (PM),1 is often carcinogenic.2,3 PM frequently includes aromatic compounds such as benzene and polycyclic aromatic hydrocarbons. Oxygenates like alcohols, ethers, and esters have been shown to be able to reduce toxic exhaust.1 To promote the use of oxygenates as fuel additives, an amendment made in 1990 to the Clean Air Act mandates that fuels should be blended so that oxygen makes up 2.7% of the fuel by weight.4 Before 2004, methyl tert-butyl ether (MTBE) was the most widely used ether fuel additive.5 However, due to its high water solubility,6 it leaks into underground water, and then into drinking water easily.5,6,7,8 This poses some health concern because of MTBE’s potential toxicity, which has been demonstrated in animals.9,10 Starting from 2000, seventeen states, including California and New York, have either restricted or banned the use of MTBE in gasoline.11 Therefore, in order to make up for the required oxygenate level in gasoline, other ethers like ethyl tert-butyl ether (ETBE) have substituted MTBE. ETBE has already been in use as fuel additive since 1990s in Europe.12 ETBE has more favorable physical characteristics than MTBE, such as a lower Reid vapor pressure,13 which is a measure of its volatility, meaning that less ETBE could escape to the environment. ETBE’s solubility in water is 12 g/L14 versus MTBE’s solubility of 26 g/L at 20 oC.14 ETBE has low acute toxicity and its effect on human reproduction and development is under review.15 In addition, ETBE is “bio-obtainable” because it can be produced from renewable material. Traditionally, ETBE is synthesized from isobutene and ethanol,16 and bioethanol produced from biomass can be readily fed into its production. Also, it has been shown that isobutene can be generated from biomass as well via metabolism of genetic engineered microbial.17 Evaporation of ETBE when stored in large quantity is still a concern. As a consequence, researchers investigated the atmospheric lifetime of ETBE by studying the kinetics of its atmospheric reaction with OH.18,19,20,19 The recorded rate constant ranged from 5.6×10-12 to 9.7×10-12 cm3 molecules-1 s-1. In Wallington’s study21 of the atmospheric ETBE oxidation in the presence of NO, 80% of the reaction was ascribed to yield tert-butyl formate and formaldehyde and 20% 2-ethoxy-2-methylpropanal. Iuga and her coworkers22 studied the atmospheric oxidation of ETBE initiated by hydroxyl radicals; quantum chemistry methods were used to compute the overall reaction rate constants and potential energy surfaces. Xiongmin Liu and his coworkers23 investigated the oxidation characteristics and products of ETBE finding that tert-butyl alcohol, isobutylene, acetic acid, methyl formate, and acetone were the main products.
These present studies also provide valuable information to the oxidation of ETBE relevant for autoignition properties. Specifically, the oxidation of ETBE is investigated through the reaction of ETBE with Cl in the presence of O2 at 293 (room temperature), 550, and 700 K. This work aims to identify the primary oxidation products. In addition, electronic structure calculations of the potential energy surface of the initial H-abstraction radicals with O2 have been performed for the first time to provide insights on the product formation, while Iuga’s group computed the activation energy and transition states of the ETBE + OH system.22
34
4.2 Experimental Section
Experiments are carried out at the Chemical Dynamics beamline of the Advanced Light Source (ALS) located in the Lawrence Berkeley National Laboratories. Oxidation intermediates and products are detected by a multiplexed time- and energy-resolved mass spectrometer, coupled with the continuously tunable synchrotron radiation produced at the ALS as the photoionization source. The detailed description of the instrument has already been discussed elsewhere. 24, 25, 26 The vapor pressure of ETBE (≥ 99.0%, Sigma-Aldrich) is diluted to 0.714% with helium. Chlorine gas diluted to 1% with helium is used as chlorine radicals precursor. Using calibrated mass flow controllers, all gaseous reactants (ETBE, Cl2, and O2) flow into a slow-flow quartz reaction tube, which is 62-cm long and has an inner diameter of 1.05 cm. Wrapped by an 18 μm thick nichrome heating tape, the reaction tube provides opportunity to study reactions from room temperature to higher temperatures. Using a Roots pump that is connected to the reaction cell through a feedback controlled throttle valve, the pressure in the reaction cell is maintained at 4, 7.3, and 9.3 Torr at room temperature, 550, and 700 K, respectively. The gas mixture is photolyzed by a 4 Hz-pulsed unfocused 351 nm (XeF) excimer laser. The ranges of number densities for ETBE, Cl2, and O2 are 8.0×1012–1.1x1013, 3.6x1013–5.2x1013, and 1.0x1016–1.4x1016
molecules cm-3, respectively, in all ETBE + Cl + O2 reactions. Chlorine atoms are produced by photolysis of Cl2 according to the following reaction:
Cl2ℎ𝜈=351𝑛𝑚�⎯⎯⎯⎯⎯⎯⎯� 2Cl ∙ (R1)
IUPAC Subcommittee for Gas Kinetic Data Evaluation recommended 1.00 to be used as the quantum yield of the Cl2 photolysis.27 The absorption cross section of its photolysis at 351 nm is 1.82x10-19 cm2 according to Maric et al.28 Based on this value and the recommended quantum yield, the number density for Cl atoms after photodissociation is 2.2x1012–3.1x1012 molecules cm-3, which is about 1/3–1/4 the concentration of ETBE, the reactant. The authentic photoionization spectra of ETBE at the studied temperatures are also obtained for the identification of dissociative ionization fragments. The authentic photoionization spectra of methacrolein (Acros, 90%) and vinyl tert-butyl ether (VTBE) (Aldrich, 98%) are recorded as well to aid in the data analysis. The authentic spectra are recorded by flowing known amounts of the gaseous molecules of interest, helium, and a calibration gas mix that contains ethene, propene, and 1-butene through the reaction cell without photolysis. The reaction species effuse through a 650 μm-wide pinhole on the side of the reactor, which is then skimmed into the differentially vacuumed ionization region of the instrument, where the tunable synchrotron radiation photoionizes all the gas molecules according to their ionization energies. Gaseous ions are accelerated and detected by an orthogonal time-of-flight spectrometer pulsed at 50 kHz, which provides a mass resolution of approximately 1600 in the current experiments. Ion intensity (I), reaction time (t), and mass-to-charge ratio (m/z) are recorded simultaneously during the reaction. This process is repeated when the photon energy (E) is varied from 8 to 11.2 eV at increments of 0.025 eV. At each photon energy step, ion signal is background (pre-photolysis signal) subtracted and normalized by the ALS photocurrent measured by a calibrated photodiode. Integrating the ion intensity over the whole photon energy spectrum at a particular m/z yields kinetic plots. The reaction time starts when the photolysis laser is fired corresponding to t = 0 ms. On the other hand, when the ion signal of a specific m/z is integrated over a selected reaction time range, the photoionization (PI) curve or photoionization spectrum is obtained. Reaction species can be identified by comparing literature, measured, or calculated PI curves with the experimental data. Because photoionization spectra of different chemicals are different due to different geometries and vibrational modes, and therefore Franck-Condon (FC) factors, isomers can be in principle distinguished from each other. This makes PI curves a very useful tool in characterizing reaction intermediates. The adiabatic ionization energy (AIE) is obtained by a linear extrapolation of the signal onset. Based on the photon energy step used in data acquisition, complexity in analyzing multiple species within a PI spectrum, and possible presence of hot bands, an error of ±0.05 eV is assigned to the observed ionization energy values determined in this work.
35
4.3 Computational Methods
When the adiabatic ionization energy of a product is unknown, this value can be computed by performing electronic structure calculations of its neutral and cation ground states. Employing the Gaussian 09 program,29 all calculations are performed using the CBS-QB3 composite model30,31 to obtain optimized molecular structural parameters, such as bond lengths, bond angles, harmonic vibrational frequencies, and force constants, and energetics. The difference between the zero-point vibrational (ZPE) corrected total electronic energies (E) of the ground electronic states of the neutral and the cationic molecule yields the calculated adiabatic ionization energy. ZPE corrected total electronic energies are also used to calculate heat of reactions (ΔrHo) to show that proposed mechanisms are thermodynamically feasible. In addition, they are employed in connection with available literature enthalpies of formation (ΔfHo) of the elements in the gas phase to derive the enthalpy of formation of a reaction species at 0 or 298 K, according to the atomization reaction CxHyOz(g) → xC(g) + yH(g) + zO(g), where ∆𝑎𝑡𝐻𝑇𝑜 = 𝑥 𝐸𝑇𝑜(𝐶) + 𝑦 𝐸𝑇𝑜(𝐻) + 𝑧 𝐸𝑇𝑜(𝑂𝑂) −𝐸𝑇𝑜�𝐶𝑥𝐻𝑦𝑂𝑂𝑧� = 𝑥 ∆𝑓𝐻𝑇𝑜(𝐶,𝑔) + 𝑦 ∆𝑓𝐻𝑇𝑜(𝐻,𝑔) + 𝑧 ∆𝑓𝐻𝑇𝑜(𝑂𝑂,𝑔) − ∆𝑓𝐻𝑇𝑜(𝐶𝑥𝐻𝑦𝑂𝑂𝑧,𝑔) and T = 0 or 298 K.
If the photoionization spectrum of a reaction species has not been measured, the Franck-Condon32,33,34,35 (FC) and Franck-Condon-Herzberg-Teller (FCHT) methods34 are used through the Gaussian 09 program to simulate photoelectron (PE) spectra by approximating the Franck-Condon factors for the vibronic transitions from the neutral to the cationic state of the molecule. The FCHT method in Gaussian 09 expresses the vibrational normal modes using the Duschinsky rotation matrix.36 Then, a set of recursive formulae developed by Ruhoff,37 which are based on the Sharp-Rosenstock38 and Lermé method,39 is used to calculate FC overlap integrals. Simulated PI spectra are obtained by integrating the simulated photoelectron spectra.
4.4 Results
In combustion engines, the oxidation of a fuel molecule is typically initiated through hydrogen abstraction by any radical produced from an ignition source or from auto-ignition. In the current work, chlorine atoms are used as hydrogen abstractors to initiate the reaction. Cl addition is not possible in primary reactions in this investigation because ETBE does not have unsaturated C-C bonds.
36
Figure 1. Mass spectra for Cl-initiated oxidation of ETBE at room temperature, 550, and 700 K, obtained over the photon energy range 8.0–11.4 eV and over 0–70 ms. The negative peak at m/z = 87 is the main dissociative ionization fragment of ETBE (m/z = 102).
Fig.1 shows the mass spectra of the Cl atom-initiated ETBE oxidation reaction at room
temperature, 550, and 700 K. The reaction products are determined after subtracting the contributions of fragments from the dissociative photoionization of ETBE and also chlorinated high-mass products, as discussed in the following paragraph. The negative signal at m/z = 87 corresponds to the most prominent daughter ion generated from the dissociative photoionization of ETBE. The absence of ETBE parent signal (m/z = 102) is consistent with recent absolute photoionization cross section measurements reported by Qi and coworkers.40
Species detected at m/z = 50/52, 92/94, 107/109, 121/123, and 122/124 are assigned to as chlorinated products. The amplitudes of time profiles and photoionization spectra signal of the first set of species are three times that of the second set, in agreement with the 35Cl/37Cl isotopic ratio. These isotopologue pairs (50/52, 92/94, 107/109,121/123, as well as 122/124) are probably generated from the addition of Cl to radicals and unsaturated product species. The identity of m/z = 50/52 is likely to be chloromethane with an observed ionization energy of 11.2 eV, comparable to the literature value of 11.26 ± 0.03 eV.41,42 Nevertheless, the true identity of these chlorinated species is not investigated here because it is outside the scope of this study corresponding to secondary chemistry.
The mass-to-charge ratios, chemical formulae, and isomeric compositions of the most intense products are listed in Table 1, that are m/z = 30, 44, 56, 58, 86, and 100. Some of these signals contain contributions from dissociative ionization of higher-mass products. The correction for dissociative ionization contribution is done based on the reference photoionization spectra of authentic samples.
room temperature
37
Table 1. Mass-to-charge ratios of main products, chemical formulae, and identified products from the ETBE + Cl + O2 reaction at 293, 550, and 700 K.
m/z Chemical formula
293 K 550 K 700 K
30 CH2O Formaldehyde Formaldehyde Formaldehyde
44 C2H4O Acetaldehyde Acetaldehyde Acetaldehyde
56 C4H8 Isobutene Isobutene Isobutene
1-Butene - -
C3H4O Acrolein Acrolein Acrolein
58 C3H6O Acetone Acetone -
Propanal - -
86 C5H10O 2-Ethoxyprop-1-ene
-
100 C6H12O Vinyl tert-butyl ether
Vinyl tert-butyl ether
-
A m/z = 30 product is detected at room temperature, 550, and 700K. Fig. 2 shows the perfect match between the experimental m/z = 30 photoionization spectrum and the literature PI curve of formaldehyde.43 The formation of formaldehyde has a negative temperature dependence.
Figure 2. Literature photoionization spectrum of formaldehyde superimposed (black solid line43) onto the experimental m/z = 30 PI plot at room temperature.
Similarly, the literature PI spectrum of acetaldehyde recorded by Cool et al.44 is compared to the experimental m/z = 44 PI curve obtained in the present work (Fig. 3) with an experimental onset at
Ion Sing al (a.u.)
11.5 11.0 10.5 10.0 9.5 9.0 8.5 8.0
Photon Energy (eV)
m/z = 30 Authentic PIE of formaldehyde (Cooper et al.)
38
10.20 eV, which is in good agreement with the accepted value of 10.2 eV.45 The formation of acetaldehyde does not show obvious dependence on temperature.
Figure 3. Literature photoionization spectrum of acetaldehyde superimposed (solid black line44) onto the experimental m/z = 44 PI plot at 550 K.
As shown in Fig. 4 the PI curve of the m/z = 56 product shows a different overall shape depending on the reaction temperature value. At 293 K (Fig. 4a) the onset at 9.20 eV is attributed to isobutene based on the literature AIE measurement of 9.23 ± 0.02 eV.46 The CBS-QB3 AIE of isobutene is calculated to be 9.28 eV. In addition, a good agreement between the spectrum reported by Wang et al.47 and the first part of our plot confirms this assignment. The second feature at 9.6 eV is assigned to as 1-butene, for which the determined AIE from Traeger is 9.57 eV48 and 9.6 eV from Wang’s group.47 The third feature is assigned to acrolein at 9.96 eV, which is measured as 10.1 eV by Bock et al.49 The last part of the spectrum could be due to the dissociative fragment of a chlorinated species at m/z = 92, that is 2-chloro-butane, which has the C4H8
+ daughter ion with an appearance energy of 10.41 ± 0.08 eV.50 For the PI curve recorded at 550 (Fig. 4b) and 700 K, two spectral features are observed, the first one at 9.23 eV and the other one at 9.96 eV. The very good agreement of the experimental data with the superimposed literature spectra confirms the presence of isobutene and acrolein (second spectral feature).
Authentic PIE of acetaldehyde (Welz et al.) m/z = 44
39
a)
b)
Figure 4. a) PI plot of isobutene47 (blue line), acrolein (green line), and 1-butene47 (yellow line) superimposed onto the experimental photonionization spectrum of m/z = 56 product recorded at room temperature. b) PI plot of isobutene47 (blue line) and autherntic PIE of acrolein (green line) superimposed onto the experimental photonionization spectrum of m/z = 56 product recorded at 550 K.
40
a)
b)
Figure 5. a) PI plot of acetone51 (blue line) and propanal47 (green line) superimposed onto the experimental photonionization spectrum of m/z = 58 product recorded at room temperature. b) PI plot of acetone51 (black line) superimposed onto the experimental photonionization spectrum of m/z = 58 product recorded at 550 K.
In Fig. 5a, the summation of the literature photoionization spectra of acetone51 and propanal47 is in good agreement with the experimental m/z = 58 PI spectrum at 293 K. The first onset observed at 9.68 eV is assigned to as acetone, consistent with its literature AIE of 9.694 ± 0.006 eV.52 A second onset is found at 9.98 eV, in agreement with the accepted IE of propanal 9.96 ± 0.01 eV.53 Temperature effects are also observed for m/z = 58 products. Propanal is absent at 550 K showed in Fig. 5b, and the observed species at room temperature are absent at 700 K.
The m/z = 86 product photoionization spectrum taken at room temperature is compared with the 2-ethoxy-1-propene simulated FC plot (Fig. 6). There is not literature curve for this species and the pure chemical is not commercially available. The observed ionization of 2-ethoxy-1-propene is 8.3
41
eV and the calculated value is 8.20 eV. Also from the computed potential energy surfaces calculations, the 2-ethoxy-1-propene is one of the primary product from the ethoxy-tert-butyl (EOTB) reaction pathway, which will be presented later. Unfortunately, this assignment can be considered only tentative due to the low signal-to-noise ratio. The m/z = 86 product is absent at 550 and 700 K.
Figure 6. Franck-Condon simulated photoionization spectrum of 2-ethoxyprop-1-ene (black line) superimposed onto the experimental m/z = 86 PI plot at room temperature.
The m/z = 100 signal shows an onset at 8.32 eV at both room temperature and 550 K, which matches well with the calculated CBS-QB3 AIE value of 8.35 eV of vinyl tert-butyl ether and also agrees with the calculated results for the computed potential energy surfaces. Fig. 7 shows that the authentic photoionization spectrum of vinyl tert-butyl ether measured in the present investigation at room temperature is in good agreement with the experimental PI curve. Similarly to m/z = 86 product the m/z = 100 product disappears at 700 K.
Figure 7. Photoionization spectrum of VTBE measured in this investigation superimposed onto the experimental PI plot of m/z = 100 at room temperature.
The findings of this study in terms of product identification are different from the Wallington’s study20-21, 54 in that tert-butyl formate and 2-ethoxy-2-methylpropanal are not identified
42
among the current observed products. There is no information about the reaction time in Wallington’s study,20-21, 54 so the deviation in the two studies may be due to the difference in observation time. The two products mentioned by Wallington20-21, 54 may be formed at longer reaction time. Another possibility of the discrepancy may due to the presence of NO. Tert-butyl formate and 2-ethoxy-2-methylpropanal may form through a mechanism involving NO, which is absent in this study. Isobutene and acetone are the two products presented in both Xiongmin Liu’s23 and our work. Tert-butyl alcohol, acetic acid, methyl acetate, and methyl formate are identified in Xiongmin Liu’s work23, but they are not observed in this study. The different reaction conditions (liquid phase, higher temperature, and longer reaction time) could be the reason for the products difference.
4.5 Postulated mechanism
There are three possible sites for Cl to abstract H atoms from ETBE to form alkoxyalkyl radicals (ROR*) as shown in Scheme 1. When chlorine abstracts a hydrogen atom from the α-carbon on the ethoxy side, t-butoxy-1-ethyl radical (α-TBOE, or α-tert-butylethyl) (A) is produced along with HCl. The calculated heat for this process is -43 kJ mol-1. Two other radicals, t-butoxy-2-ethyl (β-TBOE, or β-tert-butylethyl) (B) and 2-ethoxy-2-methyl-1-propyl (EOTB or ethoxy-tert-butyl) (C) radicals, are generated in similar ways when chlorine atoms abstract a hydrogen from the β-carbon on the ethoxy side (B) or on the terminal methyl on the tert-butyl side (C), with CBS-QB3 calculated enthalpies of reaction of -9 and -5 kJ mol-1, respectively. The reactions are illustrated in scheme 1, including the O2 addition to form the various peroxy species. The CBS-QB3 calculated enthalpies of individual reactions are in the parenthesis. The formation of α-TBOE radical (A) is the most exothermic because the oxygen electronegativity stabilizes the radical electron by polling it closer, which makes the αC-O bond stronger. Indeed, the αC-O bond distance on the CBS-QB3 optimized structure of ETBE is 1.44 Å, whereas the same bond on the optimized structure of A is 1.37 Å. After forming ROR*, oxygen attacks the radical center, generating alkoxyalkylperoxy radicals (ROROO*), α-ROO (A), β-ROO (B), and t-ROO (C). These ROROO* species can undergo intramolecular hydrogen abstraction isomerizing to hydroperoxyalkoxyalkyl radicals (ROQOOH). Subsequent α-β cleavage produces either cyclic ethers with two oxygen atoms in the ring system and OH, or other smaller molecules. Cyclic ether channels are not observed in our experiment, while products are observed that correspond to the formation of OH via bond fission similar to the results of alcohol oxidation by Welz.55
Scheme 1. Three possible pathways to form alkoxyalkyl radicals (ROR*) and the corresponding peroxy species.
43
The 𝛂-TBOE reaction pathway
The α-TBOE radical reacts with oxygen to form the α-peroxy (A). The calculated enthalpy of this step is -153 kJ mol-1. This α-ROO radical (A) then undergoes intramolecular hydrogen abstraction depicted in Scheme 2. There are two possibilities for this rearrangement, abstracting hydrogen on the β-carbon or on the tert-butyl side.
Scheme 2 shows the decomposition pathway for (A). The red line indicates the energy level of -TBOE+ O2; any species with an energy level higher than the red line is deemed thermodynamically unfavorable.
When (A) abstract a hydrogen on the ethyl side, the β-carbon side, α-QOOH-β (A1) is produced, which is endothermic (79 kJ mol-1) using (A) as the zero energy level. The corresponding activation enthalpy of the reaction from (A) to (A1) is 121 kJ mol-1. This hydroperoxyalkoxyalkyl radical (A1) has two main routes to decomposition. The hydroperoxy group can leave the molecule, forming a double bond on the ethyl side yielding vinyl tert-butyl ether (VTBE) (m/z = 100, C6H12O), or 2-methyl-2-vinyloxypropane, which has already been discussed in the products identification section. The calculated enthalpy for this step is 25 kJ mol-1, while the enthalpy of activation is calculated to be 61 kJ mol-1. The other possibility is that the hydroxyl group leaves bonding to the oxygen attached to the original α-carbon on the ethyl side and the β-carbon on the ethyl side forming 2-tert-butyoxyoxirane, the formation of which is endothermic by 64 kJ mol-1 with the activation enthalpy of 42 kJ mol-1. It is not expected to be observed due to its unbound cation (CBS-QB3 calculation).
When (A) abstracts a hydrogen atom intramolecularly from the tert-butyl side, α-QOOH-t (A2) is formed. This particular step is endothermic (72 kJ mol-1) with the activation enthalpy of 108 kJ mol-1. A hydroxyl radical can be cleaved from (A2) to join the oxygen on the peroxy and the terminal methyl radical on the tert-butyl side, yielding a 5-membered ring compound, 2,4,4-trimethyl-1,3-dioxolane (m/z = 116, C6H12O2). The energy released during its formation is -172 kJ mol-1 with the activation energy of 47 kJ mol-1. It is not observed in our experiment because its cation is calculated to be unbound. Alternatively, if (A2) undergoes α-β cleavage on the C-O bond, which is on the tert-butyl side, isobutene (m/z = 56, C4H8) and 1-hydroperoxyethoxy radical can be formed, whereas 1-hydroperoxyethoxy is not detected. The CBS-QB3 enthalpy for this step is 75 kJ mol-1 and the activation enthalpy is 77 kJ mol-1. The OOH substituted ethoxy radical can also undergo α-β
-150
-100
-50
0
50
100
150
200
250
A
A1A2
A2I
TS1TS2
TS3
TS4TS5
TS6
TS7
O
O
OH
O
O
OH
O
OOH
+isobutene
O acetaldehyde+ OOH
O
O
2,4,4-trimethyl-1,3-dioxolane+ OH
O
O
2-tert-butoxyoxiane+ OH
O vinyl tert-butyl ether+ OOH
Rela
tive
Ene
rgy
(kJ/
mol
)
O
O
O
α
𝛼𝛼 − TBOE + 𝑂𝑂2
44
cleavage between the oxygen on the hydroperoxy group and the α-carbon of the oxy group. This cleavage results in a hydroperoxy radical and acetaldehyde with a calculated reaction enthalpy of -21 kJ mol-1and activation enthalpy of 57 kJ mol-1.
The β-TBOE reaction pathway
Scheme 3 shows the decomposition pathway for (B). The red line indicates the energy level of -TBOE+ O2 and any species with an energy level higher than the red line is deemed thermodynamically unfavorable.
The second possible site for Cl to abstract H from ETBE is on the ethyl side β-carbon. The resulting radical is β-TBOE and its reaction pathways are depicted in Scheme 3. Again, the energies of reaction in parentheses are calculated using β-ROO (B) as the reference point, which is about 140 kJ/mol marked as the red line in Scheme 3. After O2 addition to the H-abstraction radical, β-ROO (B) is formed with a calculated enthalpy of -140 kJ mol-1. Like α-TBOE, this β-radical has two possible rearrangement routes through intramolecular H-abstraction. If the hydrogen on the α-carbon is abstracted, β-QOOH-α (B1) is formed with an enthalpy of 30 kJ mol-1 and activation enthalpy of 111 kJ mol-1. There are two possible reaction pathways for (B1). The hydroperoxy group can leave, forming a double bond across the ethyl group yielding vinyl tert-butyl ether, whose production through the α-TBOE pathway is discussed above. The reaction heat for this particular step through β-TBOE is 74 kJ mol-1 with the activation enthalpy of 110 kJ mol-1. The second route for (B1) decomposition is through the removal of the hydroxyl group, exothermic by 58 kJ mol-1 with the activation enthalpy of 48, yielding 2-tert-butyoxyoxirane, which is not observed in the experiment due to its unbound cation using the CBS-QB3 composite model.
β-QOOH-t (B2) is formed through abstraction of the hydrogen on one of the methyl groups on the tert-butyl side by the peroxy group with a calculated CBS-QB3 enthalpy of 74 kJ mol-1 and activation enthalpy of 99 kJ mol-1. β-QOOH-t (B2) decomposes into 2,2-dimethyl-1,4-dioxane (m/z = 116, C6H12O2), which as 2-tert-butyoxyoxirane and other cyclic ethers could not be observed in this experiment since they dissociatively ionize. When the hydroperoxy oxygen attached to the β-carbon of the ethyl group binds to the radical center on the terminal carbon on tert-butyl side, the 6-membered ring 2,2-dimethyl-1,4-dioxane is produced together with OH with a calculated exothermicity of 178 kJ mol-1 and activation enthalpy of 50 kJ mol-1. While the above process is
-200
-100
0
100
200
300
400
O
O
2,2-dimethyl-1,4-dioxane+OH
BB1
B2B2I
O
O
HO
+
O
ethylene oxide+OOH
OO
vinyl tert-butyl ether+OOH
2-tert-butoxyoxiane+OH
O
O
O
OH
O
O
HO
Rela
tive
Ener
gy (k
J/m
ol)
TS8 TS9 TS10
TS11
TS12
TS13
TS14
β
𝛽𝛽 − TOBE + 𝑂𝑂2
45
achieved by breaking the O-O bond on the hydroperoxy group, another process can lead to the cleavage of the C-O bond between the tertiary carbon and the ethereal oxygen. This process yields isobutene and 2-hydroperoxyethoxy radical (B2I), which is endothermic by 84 kJ mol-1 with a calculated activation enthalpy of 87 kJ mol-1. 2-Hydroperoxyethoxy radical (B2I) can lose hydroperoxy radical to form ethylene oxide, with a calculated enthalpy of is 39 kJ mol-1 and activation enthalpy of 210 kJ mol-1. This activation energy is much higher than the initial reaction step -TBOE + O2 and, therefore, this reaction does not happen under our current experimental conditions.
The EOTB reaction pathway
Scheme 4 shows the decomposition pathway for (C). The red line indicates the energy level of EOTB + O2 and any species with an energy level higher than the red line is deemed thermodynamically unfavorable.
EOTB radical is obtained by abstracting a hydrogen of the methyl group of the tert-butyl side. Its reaction pathways are presented in scheme 4. When oxygen binds to this radical site, t-ROO radical (C) is formed exothermically (-147 kJ mol-1). In analogy with the previous peroxy species, there are three possible rearrangements through intramolecular H-abstraction.
The oxygen can abstract a hydrogen atom from one of the methyl groups on the tertiary carbon, resulting in a t-QOOH-t radical (C1) with a calculated enthalpy for this step of 69 kJ mol-1 and activation enthalpy of 105 kJ mol-1. This radical (C1) in turn has two feasible pathways. A possibility is to break the C-O bond between the hydroperoxy group and the methyl carbon to produce 2-ethoxy-1-propene (m/z = 86, C5H10O), which is observed at room temperature and discussed in the products identification section, formaldehyde and hydroxyl radical, with an overall CBS-QB3 exothermicity of 61 kJ mol-1 and activation enthalpy of 100 kJ mol-1. The second possible pathway for t-QOOH-t (C1) decomposition is breaking the O-O bond so that the oxygen binds to the methyl group on the tert-butyl side, yielding a 4-member-ring, 3-methoxy-3-methyloxetane (m/z = 116, C6H12O2) with an exothermicity of 82 kJ mol-1 and activation enthalpy of 72 kJ mol-1, which is not observed due to its poor Franck-Condon factors.
β
-200
-150
-100
-50
0
50
100
150
200
250
300
CC2
O
O
O+
acetaldehydeisobutene+ OOH
C3
O
O
O+ethylene oxideisobutene
+ OOH
C1
O
O
EOTB + O2
Rela
tive
Ener
gy (k
J/m
ol)
OOH2C+
2-ethoxy-1-propene + OH
TS15
TS16
TS17
TS19
TS20 TS21
TS22 TS18 TS23
TS24
3-methoxy-3-methyloxetane + OH
C2I
2, 2-dimethyl-1,4-dioxane + OH
2, 4, 4-trimethyl-1, 3-dioxolane + OH
46
The other possible rearrangement of t-ROO involves intramolecular H-abstraction on the α-carbon, which forms the hydroperoxyalkoxyalkyl t-QOOH-α (C2) species. The enthalpy of this step is 37 kJ mol-1, with the activation enthalpy of 68 kJ mol-1. Then, (C2) can be cleaved in two ways. The first route is to dissociate into acetaldehyde and hydroperoxy tert-butyl radical (C2I) by breaking the C-O bond between oxygen and the α-carbon on the tert-butyl side with a calculated enthalpy of 23 kJ mol-1 and enthalpy of activation of 69 kJ mol-1. The hydroperoxy tert-butyl radical (C2I) can degrade into isobutene and OOH radical, with the overall calculated enthalpy of 22 kJ mol-1 and activation enthalpy 61 kJ mol-1. The second possible (C2) decomposition is via the formation of the C-O bond by losing the OH radical to yield 2,4,4-trimethyl-1,3-dioxolane. The calculated reaction enthalpy of this reaction is -181 kJ mol-1 with the enthalpy of activation of 30 kJ mol-1.
The last possible rearrangement of t-ROO involves intramolecular H-abstraction on the β-carbon. A hydroperoxyalkoxyalkyl radical designated as t-QOOH-β (C3) is formed with an enthalpy of 73 kJ mol-1 and activation enthalpy of 102 kJ mol-1. (C3) then has two decomposition pathways. In the first one the peroxy oxygen and the β-carbon on the ethyl side join together by cleaving off the hydroxyl group, forming 2,2-dimethyl-1,4-dioxane. This reaction is exothermic by 179 kJ mol-1 with an activation enthalpy of 50 kJ mol-1. The second path is to cleave the C-O bond between the other oxygen and the α-carbon on the tert-butyl side to generate ethylene oxide and isobutene and OOH radical with a CBS-QB3 reaction enthalpy of 123 kJ mol-1 and activation enthalpy of 181 kJ mol-1.
Based on the above postulated reaction mechanism analysis, only the reaction pathway of EOTB involves the formation of acetaldehyde that is observed at all three temperatures, which indicates that this reaction pathway is open at all temperatures investigated in this study. Similarly, only EOTB reaction pathway generates the formaldehyde and 2-ethoxy-1-propene product, while 2-ethoxy-1-propene is only present at room temperature. As mentioned in the branching ratio part (see below), formaldehyde is the most abundant product at room temperature, and its relative concentration comparing with acetaldehyde at 550 K decreases, and disappears at 700 K. Thus, we can conclude that the EOTB reaction pathway involving generation of formaldehyde is only preferred at room temperature, while at higher temperatures decomposition occurs. The concentration of isobutene is enhanced as the temperature increases. All three reaction pathways can contribute to its formation. The calculation discussed above shows that the activation enthalpy of reaction to yield isobutene in the α-TBOE pathway is 77 kJ mol-1, in the β-TBOE reaction pathway is 87 kJ mol-1, and 61 kJ mol-1 in EOTB reaction pathway.
4.6 Branching Ratios
PI spectra, besides being important in species characterization, are useful in determining the concentration of products. These values relative to other products are also called branching ratios. The concentration (C) of a species is related to the photoionization cross-section (σE) by the following relationship:
𝑆𝐸 = 𝑘𝜎𝐸𝛿𝐶 (1)
where SE is the ion signal in the photoionization spectra at a specified photon energy and k represents the instrument constant. The ion signal is corrected by the mass-dependent response (δ), which in the current investigation is approximately equal to the mass (m) of the observed species to the power of 0.67.55 A plot of photon energy against photoionization cross-section is called an absolute photoionization spectrum or absolute PI curve, which, if unknown, can be obtained by comparing the species pure sample experimental PI spectrum with a known concentration to an absolute photoionization cross-section of a reference species, such as propene in this work. The absolute cross-section of propene used as a reference was recorded by Person and Nicole.56
In this investigation, from the result of the computed PES more acetaldehyde should be formed, because the activation barriers are lower for its formation than the barrier for formaldehyde. The origin of more formaldehyde observed could be from secondary chemistry. For this reason, branching ratios are presented. Here we have chosen acetaldehyde as the reference product (R),
47
relative to which the branching ratios of products (P) are obtained according to the following expression:
𝐶𝑃𝐶𝑅
=
𝑆𝑃𝜎𝑃𝛿𝑃𝑆𝑅𝜎𝑅𝛿𝑅
=𝑆𝑃𝜎𝑅𝛿𝑅𝑆𝑅𝜎𝑃𝛿𝑃
=𝑆𝑃𝜎𝑅𝑆𝑅𝜎𝑃
�𝑚𝑅
𝑚𝑃�0.67
=𝑆𝑃𝜎𝑅𝑆𝑅𝜎𝑃
𝑀𝐷𝐹 (2)
MDF symbolizes the mass discrimination factor. Since ion signal and photoionization cross-section are energy dependent, photoionization spectra are compared at the same photon energy at which possibly both PI signals of the product and reference exhibit a plateau to minimize signal fluctuations and, therefore, reading uncertainties.
The absolute photoionization cross-section of formaldehyde,39 acetaldehyde,44 isobutene,47 and acetone51 were taken from the literature. The absolute photoionization cross-section of 2-ethoxypropene (m/z = 86) was estimated using the semi-empirical model developed by Bobeldijk et al.57 The branching ratios relative to acetaldehyde are calculated and showed in Table 2. However, the branching ratio calculated for some products, such as 2-ethoxyprop-1-ene, do not reflect the actual values because of possible secondary reactions of these species with Cl radicals. Also, we want to point out that some possible products, such as cyclic ethers, cannot be quantified due their poor Franck-Condon factors, i.e., the geometry of the neutral is much different than the cationic structure.
Table 2 experimentally determined branching ratios at 298, 550, and 700 K relative to acetaldehyde.
Product Braching ratio relative to acetaldehyde
298 K 550 K 700 K
m/z = 44
acetaldehyde 1 1 1
m/z = 30
formaldehyde 2.07 0.086 -
m/z = 56
1-butene 1.19 - -
isobutene 0.146 0.644 0.730
acrolein - 0.304 0.334
m/z = 58
acetone 0.495 0.065 -
propanal 0.177 - -
m/z = 86
2-ethoxyprop-1-ene 0.092 - -
m/z = 100
VTBE 0.038 0.021 -
48
Since acetaldehyde is generated at all three temperatures, branching ratios are calculated with respect to its signal in the discussion. Six main products are observed at room temperature: acetaldehyde, formaldehyde, 1-butene, isobutene, acetone, and propanal with branching ratios of 1.00, 2.07, 1.19, 0.15, 0.50, and 0.18, respectively. One possible way to form acetone is through (CH3)3CO* breaking the bond between the Cγ and C yielding acetone and CH3*. Both (CH3)3CO* and CH3* are observed in our experiment and this result also agrees with the work of Bohm’s group.58 As temperature is increased there are changes in both main products and branching ratios. At 550 K three main products are observed: acetaldehyde, isobutene, and acrolein with branching ratios of 1, 0.64, and 0.30, respectively, in which the assignment of acrolein is only tentative since there is not a calculated mechanism to form it. Compared with the room temperature data, acetaldehyde rather than the formaldehyde becomes the most abundant product at 550 K. Also, formaldehyde, acetone, and propanal, which are observed at room temperature, almost disappear at 550 K. At 700 K the main products are identical to those at 550 K but with slightly different branching ratios of 1, 0.73, and 0.33. It can be noticed that the most abundant product of the Cl initiated ETBE oxidation reaction is 1-butene at room temperature and acetaldehyde at both 550 and 700 K.
The findings of this study in terms of product identification are different from the Wallington’s study20-21, 54 in that tert-butyl formate and 2-ethoxy-2-methylpropanal are not identified among the current observed products. There is no information about the reaction time in Wallington’s study,20-21, 54 so the deviation in the two studies may be due to the difference in observation time. The two products mentioned by Wallington20-21, 54 may be formed at longer reaction time. Another possibility of the discrepancy may due to the presence of NO. Tert-butyl formate and 2-ethoxy-2-methylpropanal may formed through a mechanism involving NO, which is absent in this study. Isobutene and acetone are two products observed in both Xiongmin Liu’s23 and our work. Tert-butyl alcohol, acetic acid, methyl acetate, and methyl formate are identified in Xiongmin Liu’s work23, but are not observed in this work. The different react conditions (liquid phase, higher temperature, and longer reaction time) could be the reason for the products difference.
4.7 Conclusions
Using the synchrotron radiation at the Lawrence Berkeley National Laboratory, the Cl initiated ETBE oxidation reaction has been studied at room temperature, 550, and 700 K. Both the products and their branching ratios are determined by using temporal resolved plots and absolute photoionization spectra. At room temperature acetaldehyde, formaldehyde, 1-butene, isobutene, acetone, and propanal are determined as the main products with branching ratios of 1.00, 2.07, 1.19, 0.15, 0.50, and 0.18, respectively. At 550 K the main products become acetaldehyde, isobutene, and acrolein with branching ratios of 1, 0.64, and 0.30, respectively. The same products as those at 550 K are observed at 700 K but with different branching ratios (1, 0.73, and 0.33). Theoretical calculations at the CBS-QB3 level are also performed for reaction mechanism analysis, in which three reaction pathways are postulated. Comparisons of the main experimental observations and the computations are also discussed.
4.8 References 1. Westphal, G. A.; Krahl, J.; Brüning, T.; Hallier, E.; Bünger, J., Ether oxygenate additives in gasoline reduce toxicity of exhausts. Toxicology 2010, 268 (3), 198-203. 2. Seagrave, J.; McDonald, J. D.; Gigliotti, A. P.; Nikula, K. J.; Seilkop, S. K.; Gurevich, M.; Mauderly, J. L., Mutagenicity and in Vivo Toxicity of Combined Particulate and Semivolatile Organic Fractions of Gasoline and Diesel Engine Emissions. Toxicol. Sci. 2002, 70 (2), 212-226. 3. Seagrave, J.; Mauderly, J. L.; Seilkop, S. K., IN VITRO RELATIVE TOXICITY SCREENING OF COMBINED PARTICULATE AND SEMIVOLATILE ORGANIC FRACTIONS OF GASOLINE AND DIESEL ENGINE EMISSIONS. J. Toxicol. Env. Heal. A 2003, 66 (12), 1113-1132. 4. Public Health and Welfare. US Code. 7401-7671, Title 42, 1977. 5. Sutherland, J.; Adams, C., Determination of Hydroxyl Radical Rate Constants for Fuel Oxygenates. Environ. Eng. Sci. 2007, 24 (8), 998-1005.
49
6. Squillance, P. J.; Pankow, J. F.; Korte, N. E.; Zogorski, J. S., Environmental behavior and fate of methyl tert-butyl ether (MTBE). Survey, U. D. o. I.-U. G., Ed. 1998. 7. Deeb, R. A.; Scow, K. M.; Alvarez-Cohen, L., Aerobic MTBE biodegradation: an examination of past studies, current challenges and future research directions. Biodegradation 2000, 11 (2), 171-185. 8. Stocking, A. J.; Deeb, R. A.; Flores, A. E.; Stringfellow, W.; Talley, J.; Brownell, R.; Kavanaugh, M. C., Bioremediation of MTBE: a review from a practical perspective. Biodegradation 2000, 11 (2), 187-201. 9. Sgambato, A.; Iavicoli, I.; De Paola, B.; Bianchino, G.; Boninsegna, A.; Bergamaschi, A.; Pietroiusti, A.; Cittadini, A., Differential toxic effects of methyl tertiary butyl ether and tert-butanol on rat fibroblasts in vitro. Toxicol. Ind. Health 2009, 25 (2), 141-151. 10. Li, D.; Liu, Q.; Gong, Y.; Huang, Y.; Han, X., Cytotoxicity and oxidative stress study in cultured rat Sertoli cells with Methyl tert-butyl ether (MTBE) exposure. Reprod. Toxicol. 2009, 27 (2), 170-176. 11. State Actions Banning MTBE; EPA420-B-07-013. United State Environmental Protection Agency: 2007. 12. Bartling, J.; Schloter, M.; Wilke, B.-M., Ethyl tert-butyl ether (ETBE) and tert-amyl methyl ether (TAME) can affect distribution pattern of inorganic N in soil. Biol. Fert. Soils 2009, 46 (3), 299-302. 13. Ancillotti, F.; Fattore, V., Oxygenate fuels: Market expansion and catalytic aspect of synthesis. Fuel. Process. Technol. 1998, 57, 163-194. 14. KHAROUNE, M.; PAUSS, A.; LEBEAULT, J. M., AEROBIC BIODEGRADATION OF AN OXYGENATES MIXTURE: ETBE, MTBE AND TAME IN AN UPFLOW FIXED-BED REACTOR. Wat.Res. 2001, 35 (7), 1665-1674. 15. (a) D, M., Ethyl tertiary-butyl ether: a toxicological review. Crit Rev Toxicol 2007, 37 (4), 287-312; (b) A, d. P., Ethyl t-butyl ether: review of reproductive and developmental toxicity. Birth Defects Res B Dev Reprod Toxicol. 2010, 89 (3), 239-263. 16. de Menezes, E. W.; Cataluña, R., Optimization of the ETBE (ethyl tert-butyl ether) production process. Fuel Process. Technol. 2008, 89 (11), 1148-1152. 17. Wackett, L. P., Biomass to fuels via microbial transformations. Curr. Opin. Chem. Biol. 2008, 12 (2), 187-193. 18. Bennett, P. J.; Kerr, J. A., Kinetics of the reactions of hydroxyl radicals with aliphatic ethers studied under simulated atmospheric conditions. J. Atmos. Chem. 1989, 8 (1), 87-94. 19. D. F. Smith, T. E. K., E. E. Hudgens, C. D. McIver, J. J. Bufalini, Kinetics and Mechanism of the Atmospheric Oxidation of Ethyl Tertiary Butyl Ether. Int J Chem Kinet 1991, 24 (2), 199-215. 20. Wallington, T. J.; Andino, J. M.; Skewes, L. M.; Siegl, W. O.; Japar, S. M., Kinetics of the reaction of OH radicals with a series of ethers under simulated atmospheric conditions at 295 K. Int. J. Chem. Kinet. 1989, 21 (11), 993-1001. 21. Wallington, T. J.; Japar, S. M., Atmospheric chemistry of diethyl ether and ethyl tert-butyl ether. Environ. Sci. Technol. 1991, 25 (3), 410-415. 22. Cristina Iuga , Liliana Osnaya-Soto , Elba Ortiz , Annik Vivier-Bunge Atmospheric oxidation of methyl and ethyl tert-butyl ethers initiated by hydroxyl radicals. A quantum chemistry study. fuel 2015, 159, 269-279. 23. Xiongmin Liu , Q. Z., Shunsuke Ito , Yuji Wada, Oxidation characteristics and products of five ethers at low temperature. fuel 2016, 165, 513-525. 24. Czekner, J.; Taatjes, C. A.; Osborn, D. L.; Meloni, G., Absolute photoionization cross-sections of selected furanic and lactonic potential biofuels. Int. J. Mass Spectrom. 2013, 348, 39-46. 25. Ng, M. Y.; Nelson, J.; Taatjes, C. A.; Osborn, D. L.; Meloni, G., Synchrotron Photoionization Study of Mesitylene Oxidation Initiated by Reaction with Cl((2)P) or O((3)P) Radicals. J Phys Chem A 2014, 118 (21), 3735-3748. 26. Ng, M. Y.; Bryan, B. M.; Nelson, J.; Meloni, G., Study of tert-amyl methyl ether low temperature oxidation using synchrotron photoionization mass spectrometry. Journal of Physical Chemistry A 2015, (32), 8667.
50
27. . IUPAC Subcommittee on Gas Kinetic Data Evaluation – Data Sheet PCl11. http://www.iupac-kinetic.ch.cam.ac.uk/datasheets/xhtml/PCl11_cl2+hv.xhtml_mathml.xml. 28. Maric, D.; Burrows, J. P.; Meller, R.; Moortgat, G. K., A study of the UV--visible absorption spectrum of molecular chlorine. J. Photoch. Photobio. A 1993, 70 (3), 205-214. 29. Frisch, M. J. e. a. Gaussian 09, Revision A.1; Wallingford, 2009. 30. Montgomery, J. A., Jr.; Frisch, M. J.; Ochterski, J. W.; Petersson, G. A., A complete basis set model chemistry. VI. Use of density functional geometries and frequencies. J. Chem. Phys. 1999, 110 (6), 2822-2827. 31. Montgomery, J. A., Jr.; Frisch, M. J.; Ochterski, J. W.; Petersson, G. A., A complete basis set model chemistry. VII. Use of the minimum population localization method. J. Chem. Phys. 2000, 112 (15), 6532-6542. 32. Santoro, F.; Lami, A.; Improta, R.; Barone, V., Effective method to compute vibrationally resolved optical spectra of large molecules at finite temperature in the gas phase and in solution. J. Chem. Phys. 2007, 126 (18), 184102/1-184102/11. 33. Oehlschlaeger, M. A.; Davidson, D. F.; Hanson, R. K., Experimental Investigation of Toluene + H H optical spectra of large molecu. J. Phys. Chem. A 2006, 110 (32), 9867-9873. 34. Baulch, D. L.; Bowman, C. T.; Cobos, C. J.; Cox, R. A.; Just, T.; Kerr, J. A.; Pilling, M. J.; Stocker, D.; Troe, J.; Tsang, W.; Walker, R. W.; Warnatz, J., Evaluated Kinetic Data for Combustion Modeling: Supplement II. J. Phys. Chem. Ref. Data 2005, 34 (3), 757-1397. 35. Cool, T. A.; McIlroy, A.; Qi, F.; Westmoreland, P. R.; Poisson, L.; Peterka, D. S.; Ahmed, M., Photoionization mass spectrometer for studies of flame chemistry with a synchrotron light source. Rev. Sci. Instrum. 2005, 76 (9), 094102. 36. Duschinsky, F., Physicochim. URSS 1937, 7, 551. 37. Ruhoff, P. T., Recursion relations for multi-dimensional Franck-Condon overlap integrals. Chem. Phys. 1994, 186 (2–3), 355-374. 38. Sharp, T. E.; Rosenstock, H. M., Franck—Condon Factors for Polyatomic Molecules. J. Chem. Phys. 1964, 41, 3453. 39. Lermé, J., Iterative methods to compute one- and two-dimensional Franck-Condon factors. Tests of accuracy and application to study indirect molecular transitions. Chem. Phys. 1990, 145 (1), 67-88. 40. Xie, M.; Zhou, Z.; Wang, Z.; Chen, D.; Qi, F., Determination of absolute photoionization cross-sections of oxygenated hydrocarbons. International Journal of Mass Spectrometry 2010, 293 (1-3), 28-33. 41. Nicholson, A. J. C., Photoionization efficiency curves. False, H. and genuine structure. J. Chem. Phys. 1965, 43 (4), 1171-7. 42. Werner, A. S.; Tsai, B. P.; Baer, T., Photoionization study of the ionization potentials and fragmentation paths of the chlorinated methanes and carbon tetrabromide. J. Chem. Phys. 1974, 60 (9), 3650-7. 43. Cooper, G.; Anderson, J. E.; Brion, C. E., Absolute photoabsorption and photoionization of formaldehyde in the VUV and soft X-ray regions (3–200 eV). Chem. Phys. 1996, 209 (1), 61-77. 44. Cool, T. A.; Nakajima, K.; Mostefaoui, T. A.; Qi, F.; McIlroy, A.; Westmoreland, P. R.; Law, M. E.; Poisson, L.; Peterka, D. S.; Ahmed, M., Selective detection of isomers with photoionization mass spectrometry for studies of hydrocarbon flame chemistry. J. Chem. Phys. 2003, 119 (16), 8356-8365. 45. Staley, R. H.; Wieting, R. D.; Beauchamp, J. L., Carbenium ion stabilities in the gas phase and solution. An ion cyclotron resonance study of bromide transfer reactions involving alkali ions, alkyl carbenium ions, acyl cations and cyclic halonium ions. J. Am. Chem. Soc. 1977, 99, 5964. 46. Watanabe, K.; Nakayama, T.; Mottl, J., Ionization potentials of some molecules. J. Quant. Spectry. Radiative Transfer 1962, 2, 369. 47. Wang, J.; Yang, B.; Cool, T. A.; Hansen, N.; Kasper, T., Near-threshold absolute photoionization cross-sections of some reaction intermediates in combustion. Int. J. Mass Spectrom. 2008, 269, 220. 48. Traeger, J. C., Heat of formation for the 1-methylallyl cation by photoionization mass spectrometry. J. Phys. Chem. 1986, 90, 4114.
51
49. Bock, H.; Mohmand, S.; Hirabayashi, T.; Semkow, A., Thioacrolein: Das stabilste C3H4S-Isomers und sein PE-spektroskopischer Nachweis in der gasphase. Chem. Ber. 1982, 115 (4), 1339-1348. 50. Maccoll, A.; Mathur, D., Hydrogen chloride elimination from the molecular ions of alkyl chlorides. Org. Mass Spectrom. 1980, 15, 483. 51. Cool, T. A.; Wang, J.; Nakajima, K.; Taatjes, C. A.; McLlroy, A., Photoionization cross sections for reaction intermediates in hydrocarbon combustion. Int. J. Mass. Spectrom. 2005, 247 (1-3), 18-27. 52. Trott, W. M.; Blais, N. C.; Walters, E. A., Molecular beam photoionization study of acetone and acetone-d6. J. Chem. Phys. 1978, 69, 3150. 53. Linstrom, P. J.; Mallard, W. G., NIST Chemistry WebBook, NIST Standard Reference Database Number 69. National Institute of Standards and Technology: Gaithersburg MD, 20899. http://webbook.nist.gov (accessed 2012). 54. Wallington, T. J.; Dagaut, P.; Liu, R.; Kurylo, M. J., Rate constants for the gas phase reactions of OH with C5 through C7 aliphatic alcohols and ethers: Predicted and experimental values. Int. J. Chem. Kinet. 1988, 20 (7), 541-547. 55. Welz, O.; Zador, J.; Savee, J. D.; Ng, M. Y.; Meloni, G.; Fernandes, R. X.; Sheps, L.; Simmons, B. A.; Lee, T. S.; Osborn, D. L.; Taatjes, C. A., Low-temperature combustion chemistry of biofuels: pathways in the initial low-temperature (550 K-750 K) oxidation chemistry of isopentanol. Physical Chemistry Chemical Physics 2012, 14 (9), 3112-3127. 56. Person, J. C.; Nicole, P. P., Isotope Effects in the Photoionization Yields and the Absorption Cross Sections for Acetylene, Propyne, and Propene. J. Chem. Phys. 1970, 53 (5), 1767-1774. 57. Bobeldijk, M.; Zande, W. J. v. d.; Kistemaker, P. G., Simple models for the calculation of photoionization and electron impact ionization cross sectionso f polyatomic molecules. Chem. Phys. 1994, 179, 125-130. 58. Bohm, H.; Baronnet, F.; El Kadi, B., Comparative modelling study on the inhibiting effect of TAME, ETBE and MTBE at low temperature. Phys. Chem. Chem. Phys. 2000, 2 (9), 1929-1933.
52
Chapter 5. Characterization of isooctane oxidation at low temperature via synchrotron photoionization mass spectrometry
Abstract: The photolytically Cl initiated oxidation reaction of isooctane was investigated by experimental and computational methods. Using the multiplex time-resolved mass spectrometer with tunable synchrotron photoionization data were collected at the low pressure (4-8 Torr) and temperature (298 – 550 K) regimes at the Advanced Light Source located in the Lawrence Berkeley National Laboratory. Data analysis was performed via characterization of the reaction species photoionization spectra and kinetic traces. A reaction schematic including study of relevant pathways on potential energy surfaces is presented and discussed. From the initiation reactions, isooctane + Cl to its subsequent reactions between alkyloxyalkyl radicals and O2, the actual reaction enthalpy of each individual step has been calculated using the CBS-QB3 composite model and a potential energy surface has been computed to explain the product formation.
5.1 Introduction
As a prototypical compound of branched alkanes, 2,2,4- trimethylpentane (isooctane) is one of several isomers of octane, which exists in gasoline, diesel, and jet fuel.1,2,3,4 Since its highly branched structure, isooctane as typical of end-gases in spark-ignited engines has negligible reactivity and resistance to ignition5 at low temperature and pressure compared to its straight alkanes (n-heptane) counterpart.6,7 In the octane rating scale used to describe the fuel’s performance on engine knocking, isooctane is the standard of 100 octane, whereas n-heptane is 0. All other compounds and blends of compounds then are graded against these two standards and assigned octane numbers. Therefore, the oxidation reaction of isooctane is very meaningful to investigate the combustion process, products, and potential pollutants.
Isooctane has been extensively investigated using experimental and computational methods by many groups. Isooctane is traditionally synthesized from distillation of petroleum or dimerization of isobutylene.8,9 From the ignition delay times from different engine machines to chemical kinetic models, from radicals to stable species, under a variety of experimental conditions and methods, measurements have been made on isooctane oxidation. Vermeer and Oppenheim10 revealed diverse auto-ignition features at different thermodynamic and mixture conditions relevant to practical combustion devices by employing optical techniques in a shock tube for air-dilute isooctane. Fieweger and his coworkers11 further studied air-diluted isooctane in a shock tube over a much broader range of initial thermodynamic conditions, observed both homogeneous and inhomogeneous ignition behaviors, and defined the strong ignition limit, which occurs over a high temperature threshold and generates a blast wave. Mansfield and Wooldridge12 explored low temperature ignition behavior of isooctane using the rapid compression facilities (RCF), mapped the auto-ignition behavior as a function of initial thermodynamic state and equivalence ratio combined with the result of Vermeer’s10 and Fieweger’s11 work. Experimental investigations of hydroxyl radical (OH) initialed isooctane oxidation have been studied at low temperature by He and his coworkers13 and at high temperatures by Davidson and Oehlschlaeger.14,15 The hydrocarbon (HC) intermediates of isooctane oxidation have been quantified by Minetti16 at the Rapid Compression Facility (RCF), by Dagaut17 and D’Anna18 using jet-stirred flow reactor, by Dryer 19 utilizing the plug flow reactor, and by Chen20 applying the high-pressure flow reactor. He and his coworkers21 amplified the detailed reaction kinetics especially focused on HC and oxygenated HC radicals obtained during RCF experiments. Radical species and intermediates on isooctane oxidation have been studied by Davidson15 and Zhang22 in shock tube, by Lockett23 using flames, and by He21 at the RCF. Measurements of stable species from isooctane oxidation have been made in a heated high pressure pulse shock tube by Malewicki and his group,2 and in jet stirred reactors by Ciajolo and his group6,18 and Dagaut,24 Dryer,19 and Chen20 studied stable species concentrations from flow reactors; whereas Bakali25 detected stable species during isooctane oxidation from flames. Kinetic models for isooctane oxidation have been established by Buda’s group,26 Bozzelli,27 and Curran.28
53
Our current multiplexed synchrotron photoionization measurements at 298 and 550 K coupled to ab initio computations provide fundamental insights into key reactions of the proposed mechanisms.26,27-28
5.2 Experimental Section
Multiplexed time- and energy-resolved mass spectrometry coupled with the continuously tunable synchrotron radiation is used to probe the products and intermediates of the reactions at the Chemical Dynamics Beamline of Advanced Light Source (ALS) of Lawrence Berkeley National Laboratory. More details of the instrument are presented elsewhere.29,30,31Here only a brief description is provided. The reactant isooctane (≥ 99.0%, Sigma-Aldrich), diluted to 1.14% by helium, is flown through a 62-cm long quartz slow-flow reactor with an inner diameter of 1.05 cm together with 1% chlorine gas and O2. The concentrations of the reactants are controlled by calibrated mass flow controllers. The number densities are 4×1012 for isooctane, 2×1016 for O2, and 5×1013 molecules cm-3 for Cl2. The reaction is photolytically initiated by a 4 Hz pulsed 351 nm (XeF) excimer laser. The reaction temperature can be adjusted by a 18 μm thick nichrome heating tape wrapped around the reactor. The pressure is maintained at 4 Torr at room temperature and 7.3 Torr at 550 K by a roots pump, which is connected to the reaction cell through a feedback controlled throttle valve. The explored photon energy range of tunable vacuum ultraviolet synchrotron radiation of the ALS used in this work is from 8.5 to 11.1 eV. The photon energy is incrementally changed with steps of 0.025 eV.
Chlorine atoms are produced by photolysis of Cl2 via the following reaction:
Cl2ℎ𝜈=351𝑛𝑚�⎯⎯⎯⎯⎯⎯⎯� 2Cl ∙ (R1)
IUPAC Subcommittee for Gas Kinetic Data Evaulation recommended 1.00 to be used as the quantum yield of the Cl2 photolysis.32 The absorption cross section at 351 nm is 1.82 × 10-19 cm2 according to Maric et al.33 Based on these two recommended quantities, the number density for Cl atoms after photodissociation is 4×1012 molecules cm-3, which is only about 1/10 the concentration of isooctane, making Cl atoms the limiting reagent. The authentic photoionization spectrum of ETBE at the studied temperatures is also obtained for identification of the dissociative ionization fragments. For calibration purposes a mixture of gases (ethene, propene, and 1-butene) with known concentration is used.
The reactant gas mixture effuses from a 650 μm-wide pinhole, which is positioned in front of a skimmer at 30 cm from the top, then it enters into a differentially pumped chamber where it is photoionized by the tunable synchrotron radiation. The resulting ions are then accelerated, focused, and mass selected by an orthogonal time-of-flight spectrometer pulsed at 50 kHz and detected by a time sensitive micro-channel plate (MCP) detector.
A three-dimensional data block consisting of time, mass-to-charge ratios, and photon energy is recorded. The ion signal is background subtracted and normalized by the ALS photocurrent measured at the ionization region by a calibrated photodiode. Time-resolved spectra and photoionization spectra are obtained by integration of the ion signal over a specified photon energy and time ranges, respectively. Furthermore, kinetic time traces are employed to differentiate between primary and secondary reaction products. Photoionization spectra are used to characterize potential products by comparing experimental spectra to literature, measured, or calculated photoionization curves.
5.3 Computational Methods
Electronic structure calculations are performed via the Gaussian 09 program34 at the CBS-QB3 composite model,35,36 which can provide accurate energy calculations at relatively low computational cost. Optimized structural parameters, including bond lengths, angles, force constants, energetics, and harmonic vibrational frequencies, are calculated. Adiabatic ionization energies are computed by taking
54
the difference between the calculated zero-point vibrational (ZPE) corrected total electronic energies of the cationic and neutral ground states. In the reaction:
AB A++ e- +B (1) the appearance energies (AE) at 0 K of dissociative fragment ions are determined from the 0 K energies of products (A+, dissociative ionization fragment or daughter ion, and B, neutral fragment) subtracted from the 0 K energy of the neutral parent (AB), assuming no barrier for dissociation is present (lower limit).
Enthalpies of reaction (∆rH00) are calculated using ZPE corrected total electronic energies to
verify that proposed mechanisms are thermodynamically feasible. Furthermore, computed ∆rH00 values
can be combined with the available literature enthalpies of formation (∆fH0 ) of the elements in the gas phase to derive the unknown enthalpy of the formation of a reaction species at 0 or 298 K, according to the atomization reaction CxHyOz(g) → xC(g) + yH(g) + zO(g), where
∆𝑎𝑡𝐻𝑇𝑜 = 𝑥 𝐸𝑇𝑜(𝐶) + 𝑦 𝐸𝑇𝑜(𝐻) + 𝑧 𝐸𝑇𝑜(𝑂) − 𝐸𝑇𝑜�𝐶𝑥𝐻𝑦𝑂𝑧� = 𝑥 ∆𝑓𝐻𝑇𝑜(𝐶,𝑔) + 𝑦 ∆𝑓𝐻𝑇𝑜(𝐻,𝑔) +𝑧 ∆𝑓𝐻𝑇𝑜(𝑂,𝑔) − ∆𝑓𝐻𝑇𝑜(𝐶𝑥𝐻𝑦𝑂𝑧,𝑔) and T = 0 or 298 K.
Relaxed potential energy surface (PES) scans are carried out at the B3LYP/CBSB7 level. Saddle points (first order transition states, energy maxima in PES scan with one imaginary vibrational frequency) and minima are then re-optimized using the CBS-QB3 composite model. PES minima and saddle points are verified by intrinsic reaction coordinate calculations (forward/reverse). The uncertainties of the computed values are estimated to be within 0.05 eV or 4-5 kJ mol-1 according to Montgomery et al.36
A Frank-Condon (FC) spectral simulation is obtained using the Gaussian program, when the photoionization spectrum is not available from the literature or cannot be measured experimentally. The simulation generates a photoelectron spectrum by approximating the Frank- Condon factors for the vibronic transitions from the neutral to the cationic state of the molecule. By a set of recursive formulae developed by Ruhoff,37 which are based on the Sharp-Rosenstock38 and Lerme39 methods, the FC overlap integrals are calculated. The calculated photoelectron spectrum then will be integrated into a simulated photoionization spectrum. 5.4 Results A time-dependent mass spectrum of Cl-initiated isooctane oxidation at 298 K and 4 Torr, Figure 1, is obtained by integration of the background-subtracted signal over photon energies from 8.9 to 11.0 eV. Several positive peaks appear after the photolysis laser pulse suggesting they are products, and the negative signal at m/z = 114, 56, and 99 represents isooctane and its primary dissociative fragments. The isooctane and its dissociative fragments photoionization spectra were acquired directly without photolysis laser and in absence of Cl2 and O2 (Figure 2). The spectra are in agreement with recently reported photoionization cross section measurements by Zhou’s group,40 which indicate that the isooctane cation is unstable.
55
Figure 1. Time-resolved background-subtracted photoionization mass spectrum of Cl-initiated isooctane oxidation at 298 K and 4 Torr over photon energies 8.9–11.0 eV. Negative signals represent the consumption of isooctane and its dissociative fragments. Positive masses produced by isooctane oxidation are shown.
Figure 2. Photoionization spectrum of isooctane and its dissociative fragments at room temperature and 4 Torr.
Difference mass spectra (Figure 3) are measured at 298 and 550 K, integrated over the photon energy range 8.9–11.0 eV and postphotolysis. The consumption of isooctane (m/z = 114) is reflected in the signal with a negative amplitude at m/z = 56, which arises from the dissociative ionization of isooctane. It should be noted that some of these signals contain contributions from dissociative ionization of higher mass products and also chlorinated products. The correction for dissociate ionization contribution is made based on the reference photoionization spectra of authentic samples.
Kine
tic ti
me
(ms)
56
Figure 3. Integrated difference mass spectra for Cl-initiated oxidation of isooctane at 298 and 550 K obtained by integration over photon energies from 8.9 to 11.0 eV and kinetic times of 20 ms after the photolysis laser fires. The peak with a negative amplitude at m/z = 56 comes from dissociative ionization of isooctane (m/z = 114) and is cut to emphasize product formation. The most intense product peaks appear at nominal m/z = 30, 56, 70, 72, 86, 110, 112, and 128 at both room temperature and 550 K. Similar products are observed at the two temperatures with different relative yields, and will be discussed later. The mass-to-charge ratios, chemical formulae, and isomeric compositions of the most intense reaction products are collected in Table 1.
298K
57
Table 1. chemical formula, species, and structures of main products identified in isooctane oxidation reaction. m/z Chemical formula Species structure 30 CH2O Formaldehyde 56 C4H6 Isobutene
70 C5H10 C4H6O
1-pentene 2-pentene 2-butenal
72 C4H8O Isobutanal 2-butanone 2,2-dimethyl oxirane
86 C5H10O 2,2-dimethyl-propanal
Figure 4 shows the perfect match of the experimental photoionization spectrum of m/z = 30 with the literature curve of formaldehyde.41
Figure 4. Experimental photoionization spectrum of m/z = 30 (red circles) at room temperature superimposed onto the absolute photoionization spectrum of formaldehyde (solid black line).41 The experimental photoionization spectrum of m/z = 70 (Figure 5) is matched with three compounds. The first onset at 9.1 eV is assigned to 2-pentene, which is consistent with the value of 9.04 eV from Cool et al.42 The second compound has an onset at around 9.5 eV and is assigned to 1-pentene, which is also matched at 9.49 eV from Wang et al.43 The last compound at 9.75 eV is attributed to 2- butenal, in perfect agreement with the value determined by Yang et al.44
O
O
O
O
O
O
58
Figure 5. Photoionization spectra of 2-pentene42 (green line), 1-pentene43(blue line), and 2-butenal44 (yellow line) superimposed onto the experimental spectrum of m/z = 70 product recorded at room temperature.
Figure 6 presents the experimental photoionization spectrum of m/z = 72. Four compounds are used to match the spectrum, i.e., 2-butanone at 9.47 eV, isobutanal at 9.7 eV, 2,2-dimethyl oxirane at 10.1 eV, and 2-propenoic acid at 10.6 eV. The onset at 9.47 eV agrees with the literature AIE of 9.54±0.04 eV.45 The assignment of isobutanol agrees with the AIE of 9.72±0.02 eV measured by Traeger et al.46 The assessed AIE of 10 eV for 2,2-dimethyl oxirane reported by McAlduff and Houk47 and the computed CBS-QB3 value of 10.05 eV are in a good agreement with the onset used to match the experimental spectrum. The last compound at 10.6 eV is consistent with the literature AIE of 10.6 eV for 2-propenoic acid by Katrib and Rabalais.48
Figure 6. Literature integrated photoelectron spectra of 2-butanone45 (light blue line), isobutanal46 (green line), 2,2-dimethyl oxirane47 (yellow line), and 2-propenoic acid 48 (blue line) superimposed onto the experimental spectrum of m/z = 72 at room temperature.
59
5.5 Postulated mechanism
There are eight carbon atoms in isooctane (all these carbons are numbered and shown in Scheme 1).
Scheme 1. Skeletal atom in isooctane.
There are four distinct molecular sites from which Cl can abstract a H atom from isooctane to form alkyl radicals (R*), which with the O2 generate alkylperoxy species (ROO*) showing in Scheme 2. When chlorine abstracts a hydrogen atom from the tertiary carbon C4, 2,2,4-trimethylamyl radical is produced along with HCl. The calculated enthalpy for this process is -44 kJ/mol, which is the lowest enthalpy and the most preferred way to form the R* radical compared to the other three routes. Then, O2 attaches 2,2,4-trimethylamyl radical yielding 2,2,4- trimethylamyl-4- peroxy (A). The enthalpy for this step is -261 kJ/mol. The second pathway is when Cl abstracts a hydrogen from the secondary carbon C3,with the CBS-QB3 calculated enthalpy of -28 kJ/mol. O2 addition to the C3 radical center yields 2,2,4-trimethylamyl-3- peroxy (B) with a calculated enthalpy of -262 kJ/mol. The other two R* radicals are formed by losing the hydrogen from the primary carbon C5 or C7. The CBS-QB3 calculated energies for these two reactions are -15 and -12 kJ/mol, respectively. After the R* radicals are formed, their reaction with O2 generate 2,2,4-trimethylamyl-5-peroxy (C) and 2,4- dimethylamyl-2-peroxy (D) with calculated enthalpies of -265 and -266 kJ/mol, respectively. These ROO* species will isomerize to hydroperoxyalkyl radicals QOOH by intramolecular hydrogen abstraction. The QOOH species could undergo the following possible reactions including: α-β cleavage producing smaller molecules, OH cleavage followed by one of the oxygen bonding to the radical center yielding cyclic ether, and OOH cleavage forming a unsaturated bond. These reactions will be discussed below.
CH3
CHCH3C
H2
CH3C
H3CCH31
23
4
5
6
7
8
60
CH2
CH2
HC
O
OO
OO
O
O
O
CH
C
O2
O2
O2
O2
Cl
-HCl
Cl
-HCl
Cl-HCl
Cl-HCl
(-28)
(-15)
(-12)
(-44)
(-262)
(-265)
(-266)
(-261)
A
B
C
Scheme 2. Isooctane four possible pathways yielding the alkyperoxy radical ROO* A, B, C, and D. The values in parentheses represent the calculated CBS-QB3 enthalpy changes for the considered step.
The ROO* radical (A), m/z = 154, can undergo three possible rearrangements through intramolecular H-abstraction depicted in Scheme 3, with the enthalpy changes and transition states energies calculated using the CBS-QB3 composite method and in agreement with Bozzelli et al.27 When a hydrogen is removed from the C3 site, 2,2,4-trimethyl-4-hydroperoxy-3-amyl (A1) is formed with the calculated enthalpy of 14 kJ/mol. Similarly, the hydrogen could also come from the C1 and C5 sites yielding 2,2,4-trimethyl-4-hydroperoxy-5-amyl (A2) and 2,2,4-trimethyl-4-hydroperoxy-1-amyl (A3), respectively. The enthalpy change for these two reactions is 17 and 16 kJ/mol, respectively, where the energy level of A is used as the zero energy (reference point). The corresponding activation energy of the reactions from A to A1, A2, and A3 are 30, 36, and 21 kJ/mol, all far below the initial abstraction radical 2,2,4-trimethyl-4-amyl + O2 channel (261 kJ/mol).
61
Scheme 3. Three possible reaction pathways of the 2,2,4-trimethylamyl-4-peroxy (A) radical decomposition calculated using the CBS-QB3 composite method.
A1 can undergo OH cleavage to generate 2,2-dimethyl-3-tert-butyl-ethylene oxide via breaking the O-O bond and forming the C3-O bond with the calculated enthalpy of -18 kJ/mol and activation enthalpy of 11 kJ/mol. The other route is the OOH cleavage, during which the C4-O bond is broken and the C3-C4 double bond is formed yielding 2,4,4-trimethyl-2-pentene with the calculated reaction enthalpy of 5 kJ/mol and activation enthalpy of 13 kJ/mol.
A2 also decomposes in two ways. From the OH cleavage the cyclic ether 2-methyl-2-tert-amyl-ethylene oxide is formed with the calculated enthalpy of -17 kJ/mol and activation barrier of 11 kJ/mol. From the OOH cleavage 2,4,4-trimethyl-1-pentene is formed with the calculated reaction enthalpy of 4 kJ/mol and activation barrier of 14 kJ/mol.
The last decomposition pathway for radical A is through the A3 peroxy, which can break into isobutene and 2-hydroperoxy-1-iso-propyl radical with the calculated enthalpy of 19 kJ/mol and activation barrier of 26 kJ/mol. A3 can also decompose yielding 2,2,4,4- tetramethyl-tetrahydrofuran and OH. This reaction is exothermic with the calculated enthalpy of -38 kJ/mol and activation energy of 14 kJ/mol.
For the A1 and A2 routes, 2,4,4-trimethyl-2-pentene and 2,4,4-trimethyl-1-pentene were postulated, but were not found in the products due to possible dissocative ionization and poor Franck-Condon factors. All cyclic oxirane and tetradrofuran have favorable transition states energies, but they were not observed because they have unbound cations.
-30
-20
-10
0
10
20
30
40
50
A
1
A1
5 4
2,4,4-trimethyl-2-pentene+OOH
2
3 A2 ,
A3
6 9
8
7
2,2,4,4-tetramethyl-tetrahydrofuran+OH
2,4,4-trimethyl-1-pentene+OOH
Relative Energy (kJ/mol)
HO O
O O
OH
O
HO O
2,2-dimethyl-3-tert-butyl- ethylene oxide + OH
2-methyl-2-tert-amyl-ethylene oxide + OH
62
Scheme 4. Three possible reaction pathways for 2,2,4-trimethylamyl-3-peroxy (B) radical decomposition calculated using the CBS-QB3 composite method.
When the oxygen binds to the C3 radical site, the 2,2,4-trimethylamyl-3-peroxy (B) species is formed. Three molecular sites are available for intramolecular hydrogen abstraction. 2,2,4-trimethy-3-hydroperoxy-1-amyl (B1), 2,2,4-trimethy-3-hydroperoxy-4-amyl (B2), and 2,2,4-trimethy-3-hydroperoxy-5-amyl (B3) are formed when B abstracts a hydrogen from C7, C4, and C5, respectively, as shown in Scheme 4. The calculated rearrangement enthalpies from B to B1, B2, and B3 are 76, 28, and 79 kJ/mol, respectively, where the energy level of B is set as reference zero energy. Their activation barriers are 92, 106, and 98 kJ/mol, respectively, all below the initial abstraction radical 2,2,4- trimethylamyl-3- peroxy (B) + O2 channel of -262 kJ/mol.
There are two pathways for B1 decomposition. The C2-C3 and O-O bonds can be broken, yielding isobutene (m/z = 56), 2-methylpropanal (m/z = 72), and OH. The enthalpy of this reaction is calculated as -97 kJ/mol, whereas its calculated activation barrier is 86 kJ/mol. The alternative B1 decomposition route yields 2-isopropyl-3,3-dimethyl oxetane and OH with the calculated reaction enthalpy of -110 kJ/mol and activation enthalpy of 41 kJ/mol.
Similarly, B2 has also two degradation pathways. The first is the formation of 2,4,4-trimethyl-2-pentene and OOH via the breaking of the C3-O bond and formation of the C3-C4 double bond. The calculated reaction enthalpy for this step is 55 kJ/mol and its activation barrier is 78 kJ/mol. B2 also can decompose to 2,3-epoxy-2,4,4-trimethyl pentane by losing OH. This reaction is exothermic (-66 kJ/mol) with a calculated activation enthalpy of 33 kJ/mol.
The final possible intramolecular hydrogen abstraction isomer of B is B3. Propene and 2,2-dimethylpropanal (m/z = 72) are the products, when B3 undergoes unimolecular decomposition. The calculated enthalpy for this process is -99 kJ/mol, whereas its activation enthalpy is 75 kJ/mol.
-50
0
50
100
150
200
B
B2
B3 10
11 12
13
14
18
15
16
17
2-Methylpropanal
2,4,4-Trimethyl-2-pentene+OOH
2-tert-butyl-3-methyl-oxetane+OH
2,3-Epoxy-2,4,4-trimethyl pentane+OH
+
B1
O
CH
OHO
OO
CH
HC
OOH
OHO
CH
OO
2,2-Dimethylpropanal+OH/Propene
O
Relative Energy (kJ/mol)
2-isopropyl-3,3-dimethyl oxetane + OH
63
Alternatively, B3 can break apart into 2-tert-butyl-3-methyl-oxetane and OH exothermically (-94 kJ/mol), with an activation barrier of 39 kJ/mol.
The identification of 2,2-dimethyl propanal in the products proved that the B3 routes occurred during the oxidation reaction, and this the only channel to form 2,2-dimethyl propanal. Likewise, the characterization of 2-methylpropanal suggests that the B1 routes are also one possible pathway of isooctane oxidation. While for the B2 channel, 2,4,4-trimethyl-2-pentene was formed but not shown in products, probably due to dissociative ionization. All oxiranes and oxetanes have favorable transition states energies, but they were not observed because of their unbound cation.
Scheme 5. Four possible decomposition pathways of 2,2,4- trimethylamyl-5- peroxy (C) calculated using the CBS-QB3 composite method.
Another alkylperoxy radical produced in the oxidation of isooctane is (C), which can isomerize to C1, C2, C3, and C4 as Scheme 5 shows. (C) is used as the reference point for the calculated enthalpy of reaction. If the hydrogen is abstracted from the tertiary carbon C4, C1 is formed. Similarly, if the hydrogen is removed from the secondary carbon C2, C7, or C5 of the tert-amyl side, C2, C3, and C4 are generated accordingly as shown in Scheme 5. The calculated enthalpy changes yielding C1, C2, C3, and C4 are 34, 51, 72, and 66 kJ/mol, respectively. The enthalpies of activation for the reactions from C to C1, C2, C3, and C4 are 125, 105, 89, and 93 kJ/mol, respectively.
C1 has two possible decomposition pathways presented in Scheme 5. When the bond between the tertiary carbon C4 and the ethereal oxygen is cleaved, 2,4,4-trimethyl-1-pentene and OOH are produced with a calculated enthalpy of 30 kJ/mol. The barrier of this reaction is calculated to be 66 kJ/mol. Alternatively, C1 can form 2-methyl-2-tert-amyl-ethelyne oxide through OH cleavage exothermically (-67 kJ/mol), with a barrier of 48 kJ/mol.
The second hydroperoxyalkyl isomer of C is C2, which can break apart into 4,4-dimethyl-2-pentene, formaldehyde, and hydroxyl radical. The calculated enthalpy of this step is -57 kJ/mol with a
-150
-100
-50
0
50
100
150
200
C
C3
C2 C1
C4
21
20 19
22
3,3,5-trimethyl-tetrahydropyran+OH
28
24 25
31
30
26
27
4,4-Dimethyl-2-pentene+ =O formaldhyde
2-tert-butyl-3-methyl-oxetane+ OH
isobutene
2,4,4-Trimethyl-1-pentene+ OOH
2-tert-amyl-oxetane+ OH
2,2-Dimethyl-4-pentene+formaldehyde
Relative Energy (kJ/mol)
O
O
O
OH
+
O
O
2-methyl-2-tert-amyl- ethylene oxide + OH
29
64
barrier of 107 kJ/mol. C2 can also yield a four-membered cyclic ether, 2-tert-butyl-3-methyl-oxetane, together with OH with a CBS-QB3 enthalpy of -77 kJ/mol and a barrier of 57 kJ/mol.
When the bond between quaternary C4 and secondary carbon C3 is broken, isobutene and 2-hydroperoxy-1-iso-propyl radical are formed. The reaction enthalpy for this process is 70 kJ/mol, with the activation enthalpy of 92 kJ/mol. The other pathway for C3 is to yield 3,3,5-trimethyl-tetrahydropyran by losing the hydroxyal radical exothermically (-164 kJ/mol), with a barrier of 47 kJ/mol.
The last hydroperoxyalkyl isomer of C is C4. 4,4-dimethyl-1-pentene, formaldehyde, and OH are formed via the C4-C5 and C5-O bond scissions with a CBS-QB3 enthalpy change of -63 kJ/mol and activation barrier of 110 kJ/mol. C4 can also decompose to 3-tert-amyl-oxetane and OH with a calculated reaction enthalpy of -89 kJ/mol and an activation enthalpy of 71 kJ/mol.
The presence of formaldehyde indicates that the C2 and C4 routes are possible. All the cyclic oxetanes and tetrahydropyran have preferred transition state energies, but they did not appear as products, probably because their unbound cation.
Scheme 6. Four possible decomposition pathways of 2,2,4- trimethylamyl-1- peroxy (D) radical calculated using the CBS-QB3 composite method.
The product of the reaction between oxygen and 2,2,4- trimethyl-1-amyl is (D), which will also undergo intramolecular hydrogen abstraction to form various QOOH species. There are four possible rearrangements of D. If the hydrogen on the secondary carbon C2 is abstracted, D1 is formed. If the hydrogen on the tertiary carbon C4 is abstracted, D2 is produced instead. Also, D3 can be generated when the hydrogen is removed from the C1 position of the tert-amyl side. The hydrogen on the C6 position also can be abstracted yielding D4. These four reactions are depicted in Scheme 6. The calculated enthalpies generating D1, D2, D3, and D4 are 46, 29, 64, and 67 kJ/mol, respectively, with D set as the zero energy
-150
-100
-50
0
50
100
150
200
D
D1
D4
D2
D3 32
35 33
44 36
37
38
39
40
41
42
43
3,3,5-trimethyl tetrahydropyran+ OH
2,2-dimethyl-3-isopropyl-oxetane+ OH
2,4-Dimethyl-2-pentene+ formaldehyde
2,2,4,4-Tetramethyltetrahydrofuran+ OH
2,4-Dimethyl-1-pentene+ formaldehyde
Relative Energy (kJ/mol)
O
O
O
OHO
Propene+
O
OHO
isobutene
3-methyl-3-iso-butyl-oxetane+ OH
65
level. The activation enthalpies for the formation of these hydroperoxyalkyl radicals are 76, 75, 97, and 93 kJ/mol, respectively.
Through the breaking of the C1-C2 and C1-O bonds in D1, 2,4-dimethyl-2-pentene, formaldehyde, and hydroxyal radical will be formed exothermically (-67 kJ/mol), with a calculated barrier of 116 kJ/mol. If the O-O bond in D1 is broken and C3-O bond is formed, 4-membered cyclic ether, 2,2-dimethyl-3-isopropyl-oxetane and OH will be the products. The enthalpy of this reaction is calculated as -100 kJ/mol, whereas the activation enthalpy is 57 kJ/mol.
The other decomposition pathway for D is via D2, which has also two possible reaction routes. The first one is the conversion to isobutene and 2- hydroperoxy-1-iso-propyl radical as shown in Scheme 6. The calculated CBS-QB3 energy for this step is 96 kJ/mol, with a barrier of 97 kJ/mol. The second pathway is the production of five-membered cyclic ether ring, 2,2,4,4-tetramethyl-tetrahydrofuran, and OH. The calculated enthalpy of this reaction is -160 kJ/mol, whereas the enthalpy of the activation is 38 kJ/mol.
Another possibility of D rearrangement is D3. 2,4-dimethyl-1-pentene, formaldehyde, and OH are produced, when D3 undergoes C1-C2 and O-O bonds scission. The calculated enthalpy for this process is -79 kJ/mol, and the activation enthalpy is 111 kJ/mol. The other pathway for D3 decomposition is via the O-O bond cleavage and C7-O bond formation, while the four-membered ring cyclic ether, 3-methyl-3-iso-butyl-oxetane, is formed exothermically (-86 kJ/mol), with the activation enthalpy of 70 kJ/mol.
The last possible isomerization for D is D4, which also has two decomposition pathways. The first is by breaking the C3-C4 bond to generate 2-hydroperoxy-1-iso-propyl radical and propene. The calculated heat of this step is 82 kJ/mol, with the activation enthalpy of 111 kJ/mol. The second pathway for D4 is through cleaving the two oxygen bond yielding 3,3,5- trimethyl-tetrahydropyran and OH with a calculated enthalpy of -162 kJ/mol and an activation barrier of 46 kJ/mol.
For all D routes, D2 is the most favorable pathway with the lowest transition state energy and reaction enthalpy. The identification of isobutene indicates that D2 route through barrier 39 is possible. The presence of formaldehyde, as one of products of D1 and D3 routes, suggests that D1 and D3 routes are both possible, but the calculated result hints that D3 is not as favorable as D1. The lack of propene implies that the D4 channel is not occurred, which is consistent with the calculated results. All the cyclic oxetanes and oxelanes have the favorable transition states energies, but they do not appear as products either because they have unbound cations.
5.6 Conclusions
The oxidation of isooctane is investigated via its reaction with Cl and O2 at room temperature and 550 K. All experiments were carried out at the Advanced Light Source (ALS) located in the Lawrence Berkeley National Laboratory. Oxidation products and intermediates are detected by a multiplex time- and energy- resolved mass spectrometer. The synchrotron radiation generated at the ALS was utilized as the ionization source of the mass spectrometer. Primary reaction products are identified on the basis of kinetic profiles, mass-to-charge ratios, ionization energies, and photoionization spectra. In addition, the potential energy surface for the reaction of the initial H abstraction radical of isooctane and O2 has been carried out using the CBS-QB3 composite method to substantiate the observed products. The primary products are detected at m/z = 30, 56, 70, 72, and 86, and are identified as formaldehyde, isobutene, 1-pentene, 2-pentene, 2-butenal, isobutanal, 2-butanone, 2,2-dimethyl oxirane, 2,2-dimethyl-propanal. Postulated mechanisms with the potential energy surface calculation are proposed to probe and confirm the formation of these observed products.
66
5.7 Reference
1. Wallington, T. J.; Kaiser, E. W.; Farrell, J. T., Automotive fuels and internal combustion engines: a chemical perspective. Chem. Soc. Rev. 2006, 35 (4), 335-347. 2. Malewicki, T.; Comandini, A.; Brezinsky, K., Experimental and modeling study on the pyrolysis and oxidation of iso-octane. Proc. Combust. Inst. 2013, 34 (1), 353-360. 3. Feng, H.; Zhang, C.; Wang, M.; Liu, D.; Yang, X.; Lee, C.-f., Availability analysis of n-heptane/iso-octane blends during low-temperature engine combustion using a single-zone combustion model. Energy Convers. Manage. 2014, 84, 613-622. 4. Mensch, A.; Santoro, R. J.; Litzinger, T. A.; Lee, S. Y., Sooting characteristics of surrogates for jet fuels. Combust. Flame 2010, 157 (6), 1097-1105. 5. Mayer, N., Manufacture of aviation gasoline. Oel u. Kohle ver. Petroleum 1940, 36, 274-80. 6. Ciajolo, A.; D'Anna, A., Controlling steps in the low-temperature oxidation of n-heptane and iso-octane. Combust. Flame 1998, 112 (4), 617-622. 7. D. Diarra∗, R. T. E. H., Characterizing the Ignition behavior of n-heptane and iso-octane by Eigenvalue Analysis of a Detailed Reaction Mechanism. In 8th International Seminar on Flame Structure, 2014. 8. Evans, T. I.; Karas, L. J.; Rameswaran, R. Selective isobutylene dimerization into isooctane using C3-4 alkane diluents in the presence of tert-butanol. WO2001051435A1, 2001. 9. Anon, Economic evaluation of production and blending of isooctene and isooctane produced by isobutylene dimerization. IP.com J. 2012, 12 (5B), 67. 10. Vermeer, D. J.; Meyer, J. W.; Oppenheim, A. K., Autoignition of hydrocarbons behind reflected shock waves. Combust. Flame 1972, 18 (3), 327-36. 11. Fieweger, K.; Blumenthal, R.; Adomeit, G., Shock-tube investigations on the self-ignition of hydrocarbon-air mixtures at high pressures. Symp. (Int.) Combust., [Proc.] 1994, 25th, 1579-85. 12. Mansfield, A. B.; Wooldridge, M. S.; Di, H.; He, X., Low-temperature ignition behavior of iso-octane. Fuel 2015, 139, 79-86. 13. He, X.; Zigler, B. T.; Walton, S. M.; Wooldridge, M. S.; Atreya, A., A rapid compression facility study of OH time histories during iso-octane ignition. Combust. Flame 2006, 145 (3), 552-570. 14. Oehlschlaeger, M. A.; Davidson, D. F.; Herbon, J. T.; Hanson, R. K., Shock tube measurements of branched alkane ignition times and OH concentration time histories. Int. J. Chem. Kinet. 2004, 36 (2), 67-78. 15. Davidson, D. F.; Oehlschlaeger, M. A.; Herbon, J. T.; Hanson, R. K., Shock tube measurements of iso-octane ignition times and OH concentration time histories. Proc. Combust. Inst. 2002, 29 (Pt. 1), 1295-1301. 16. Minetti, R.; Carlier, M.; Ribaucour, M.; Therssen, E.; Sochet, L. R., Comparison of oxidation and autoignition of the two primary reference fuels by rapid compression. Symp. (Int.) Combust., [Proc.] 1996, 26th (Vol. 1), 747-753. 17. Dagaut, P.; Reuillon, M.; Cathonnet, M., High pressure oxidation of liquid fuels from low to high temperature. 3. n-Decane. Combust. Sci. Technol. 1994, 103 (1-6), 349-59. 18. D'Anna, A.; Mercogliano, R.; Barbella, R.; Ciajolo, A., Low temperature oxidation chemistry of iso-octane under high pressure conditions. Combust. Sci. Technol. 1992, 83 (4-6), 217-32. 19. Dryer, F. L.; Brezinsky, K., A flow reactor study of the oxidation of n-octane and isooctane. Combust. Sci. Technol. 1986, 45 (3-4), 199-212. 20. Chen, J. S.; Litzinger, T. A.; Curran, H. J., The diluted stoichiometric oxidation of iso-octane in the intermediate temperature regime at elevated pressures. Combust. Sci. Technol. 2001, 172, 71-80. 21. He, X.; Walton, S. M.; Zigler, B. T.; Wooldridge, M. S.; Atreya, A., Experimental investigation of the intermediates of isooctane during ignition. Int. J. Chem. Kinet. 2007, 39 (9), 498-517. 22. Zhang, C.; Tang, H.; Zhang, C.; Zhao, Y.; Li, P.; Li, X., Diagnosis of iso-octane combustion in a shock tube by emission spectroscopy. Chem. Phys. Lett. 2013, 556, 13-17.
67
23. Lockett, R. D.; Woolley, R., Instabilities and soot formation in high-pressure, rich, iso-octane-air explosion flames. 1. Dynamical structure. Combust. Flame 2007, 151 (4), 601-622. 24. Dagaut, P.; Reuillon, M.; Cathonnet, M., High pressure oxidation of liquid fuels from low to high temperature. 2. Mixtures of n-heptane and isooctane. Combust. Sci. Technol. 1994, 103 (1-6), 315-36. 25. Bakali, A. E.; Delfau, J. L.; Vovelle, C., Experimental study of 1 atmosphere, rich, premixed n-heptane and iso-octane flames. Combust. Sci. Technol. 1998, 140 (1-6), 69-91. 26. Buda, F.; Bounaceur, R.; Warth, V.; Glaude, P. A.; Fournet, R.; Battin-Leclerc, F., Progress toward a unified detailed kinetic model for the autoignition of alkanes from C4 to C10 between 600 and 1200 K. Combust. Flame 2005, 142 (1/2), 170-186. 27. Snitsiriwat, S.; Bozzelli, J. W., Thermochemistry, Reaction Paths, and Kinetics on the tert-Isooctane Radical Reaction with O2. J. Phys. Chem. A 2014, 118 (26), 4631-4646. 28. Curran, H. J.; Gaffuri, P.; Pitz, W. J.; Westbrook, C. K., A comprehensive modeling study of iso-octane oxidation. Combust. Flame 2002, 129 (3), 253-280. 29. Meloni, G.; Zou, P.; Klippenstein, S. J.; Ahmed, M.; Leone, S. R.; Taatjes, C. A.; Osborn, D. L., Energy-Resolved Photoionization of Alkylperoxy Radicals and the Stability of Their Cations. J. Am. Chem. Soc. 2006, 128 (41), 13559-13567. 30. Ng, M. Y.; Nelson, J.; Taatjes, C. A.; Osborn, D. L.; Meloni, G., Synchrotron Photoionization Study of Mesitylene Oxidation Initiated by Reaction with Cl((2)P) or O((3)P) Radicals. J Phys Chem A 2014. 31. Ng, M. Y.; Bryan, B. M.; Nelson, J.; Meloni, G., Study of tert-amyl methyl ether low temperature oxidation using synchrotron photoionization mass spectrometry. Journal of Physical Chemistry A 2015, (32), 8667. 32. . IUPAC Subcommittee on Gas Kinetic Data Evaluation – Data Sheet PCl11. http://www.iupac-kinetic.ch.cam.ac.uk/datasheets/xhtml/PCl11_cl2+hv.xhtml_mathml.xml. 33. Maric, D.; Burrows, J. P.; Meller, R.; Moortgat, G. K., A study of the UV--visible absorption spectrum of molecular chlorine. J. Photoch. Photobio. A 1993, 70 (3), 205-214. 34. Frisch, M. J. e. a. Gaussian 09, Revision A.1; Wallingford, 2009. 35. Montgomery, J. A., Jr.; Frisch, M. J.; Ochterski, J. W.; Petersson, G. A., A complete basis set model chemistry. VI. Use of density functional geometries and frequencies. J. Chem. Phys. 1999, 110 (6), 2822-2827. 36. Montgomery, J. A., Jr.; Frisch, M. J.; Ochterski, J. W.; Petersson, G. A., A complete basis set model chemistry. VII. Use of the minimum population localization method. J. Chem. Phys. 2000, 112 (15), 6532-6542. 37. Ruhoff, P. T., Recursion relations for multi-dimensional Franck-Condon overlap integrals. Chem. Phys. 1994, 186 (2–3), 355-374. 38. Sharp, T. E.; Rosenstock, H. M., Franck—Condon Factors for Polyatomic Molecules. J. Chem. Phys. 1964, 41, 3453. 39. Lermé, J., Iterative methods to compute one- and two-dimensional Franck-Condon factors. Tests of accuracy and application to study indirect molecular transitions. Chem. Phys. 1990, 145 (1), 67-88. 40. Zhou, Z.; Xie, M.; Wang, Z.; Qi, F., Determination of absolute photoionization cross-sections of aromatics and aromatic derivatives. Rapid Commu. Mass Sp. 2009, 23 (24), 3994-4002. 41. Cooper, G.; Anderson, J. E.; Brion, C. E., Absolute photoabsorption and photoionization of formaldehyde in the VUV and soft X-ray regions (3–200 eV). Chem. Phys. 1996, 209 (1), 61-77. 42. Cool, T. A.; Wang, J.; Nakajima, K.; Taatjes, C. A.; McLlroy, A., Photoionization cross sections for reaction intermediates in hydrocarbon combustion. Int. J. Mass. Spectrom. 2005, 247 (1-3), 18-27. 43. Wang, J.; Yang, B.; Cool, T. A.; Hansen, N.; Kasper, T., Near-threshold absolute photoionization cross-sections of some reaction intermediates in combustion. Int. J. Mass Spectrom. 2008, 269, 220. 44. Yang, B.; Wang, J.; Cool, T. A.; Hansen, N.; Skeen, S.; Osborn, D. L., Absolute photoionization cross-sections of some combustion intermediates. Int. J. Mass Spectrom. 2012, 309 (0), 118-128. 45. Tam, W. C.; Yee, D.; Brion, C. E., Photoelectron spectra of some aldehydes and ketones. J. Electron Spectrosc. Relat. Phenom. 1974, 4, 77-80.
68
46. Traeger, J. C.; McAdoo, D. J., Decomposition thresholds and associated translational energy releases for eight C4H8O+ isomers. Int. J. Mass Spectrom. Ion Processes 1986, 68, 35. 47. McAlduff, E. J.; Houk, K. N., Photoelectron spectra of substituted oxiranes and thiiranes. Substituent effects on ionization potentials involving σ orbitals. Can. J. Chem. 1977, 55 (2), 318-332. 48. Rabalais, A. K. a. J. W., Electronic Interaction between the Vinyl Group and Its Substituents. The Journal of Physical Chemistry 1973, 77 (19).
69