Citation
Heterogenous reaction kinetics of succinic acid and methyl amine

Material Information

Title:
Heterogenous reaction kinetics of succinic acid and methyl amine
Creator:
Vizcaino, Christopher
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English

Thesis/Dissertation Information

Degree:
Master's ( Master of science)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
Department of Chemistry, CU Denver
Degree Disciplines:
Chemistry
Committee Chair:
Liu, Yong
Committee Members:
Maron, Marta
Wang, Xiotai

Notes

Abstract:
The atmosphere is becoming an increasingly complex gaseous system. As the human population continues its exponential trajectory, the emissions as a result of human interactions are creating an intricate chemical mixture in the biosphere. Many of these pollutants are poorly understood but may have large implications to cloud condensation that can lead to climate change, and the formation of large amounts of particulate matter in the troposphere ultimately effecting human health. It was previously thought that many aerosols in the atmosphere were either solid or liquid phase particles. However in many cases it was found that there is a dual phase, solid particles with a thin film layer of organic material. In the global amounts of organics in the atmosphere, dicarboxylic acids have been found to be in relatively plentiful concentrations.1 Another major source of the atmospheric aerosols are nitrogen compounds, in particular amines. These are emitted into the atmosphere through a variety of sources including industrial processes, biomass burning, livestock, and marine organisms.2 The combination of these two plentiful aerosols is inevitable. Understanding how these organics interact with one another can provide valuable information on the potential formation of much of the aerosols in our biosphere. It is advantageous to have a quick throughput method to characterize the kinetics of the aerosol formation and the corresponding products of the reaction between different aerosols.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
Copyright Christopher Vizcaino. Permission granted to University of Colorado Denver to digitize and display this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.

Downloads

This item has the following downloads:


Full Text
HETEROGENOUS REACTION KINETICS OF
SUCCINIC ACID AND METHYL AMINE by
CHRISTOPHER VIZCAINO M.S, University of Colorado, Denver 2016
A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Science Chemistry Program
2016


This thesis for the Master of Science degree by Christopher Vizcaino has been approved for the Chemistry Program
by
Yong Liu Marta Maron Xiaotai Wang
December 17, 2016


Vizcaino, Christopher (M.S., Chemistry Program)
Heterogeneous Reaction Kinetics of Methyl Amine and Succinic Acid Thesis directed by Associate Professor Yong Liu
ABSTRACT
The atmosphere is becoming an increasingly complex gaseous system.
As the human population continues its exponential trajectory, the emissions as a result of human interactions are creating an intricate chemical mixture in the biosphere. Many of these pollutants are poorly understood but may have large implications to cloud condensation that can lead to climate change, and the formation of large amounts of particulate matter in the troposphere ultimately effecting human health.
It was previously thought that many aerosols in the atmosphere were either solid or liquid phase particles. However in many cases it was found that there is a dual phase, solid particles with a thin film layer of organic material. In the global amounts of organics in the atmosphere, dicarboxylic acids have been found to be in relatively plentiful concentrations.1 Another major source of the atmospheric aerosols are nitrogen compounds, in particular amines. These are emitted into the atmosphere through a variety of sources including industrial processes, biomass burning, livestock, and marine organisms.2 The combination of these two plentiful aerosols is inevitable. Understanding how these organics interact with one another can provide valuable information on the potential formation of much of the aerosols in our biosphere. It is advantageous to have a quick throughput method to characterize the kinetics of the aerosol formation and


the corresponding products of the reaction between different aerosols.
The form and content of this abstract are approved. I recommend its publication.
Approved: Yong Liu
IV


I would like to thank Kaldi and the spirited goats for their wonderful discovery.
v


TABLE OF CONTENTS
CHAPTER
1. INTRODUCTION..................................................1
Particulate matter and aerosols............................2
Cloud condensation nucleation..............................6
Anthropogenic sources and smog..................8
Climate change............................................10
Human health effects......................................11
2. EXPERIMENTAL SECTION.........................................14
Experimental set up.......................................14
Procedure.................................................17
Conditions................................................19
3. METHODS......................................................20
Theory....................................................20
Attenuated total reflectance infrared spectrometry........21
4. RESULTS & DISCUSSION.........................................24
Predicted aerosol formation...............................24
Infrared spectra..........................................25
Reaction kinetics.........................................35
Uptake coefficients.......................................39
5. CONCLUSION...................................................43
6. REFERENCES...................................................45
VI


LIST OF TABLES
TABLE
1. Major sources of aerosol species (AR5 of IPCC)..................3
2. Limits of National Ambient Air Quality Standards (NAAQS) for ambient
aerosols (CPCB, 2009)...........................................5
3. IR feature assignments of SA-MA at room temperature.............30
4. Summary of pseudo first order rate constants k-i, derived from the
temporal integrated absorbance taken at different locations, and the corresponding calculated uptake coefficients..............40
vii


LIST OF FIGURES
FIGURE
1. Schematic of the overall gas movement..............................15
2. Schematic of the flow through reactor coupled with the ATR-IR......17
3. Michelson Interferometer...........................................22
4. Conventional reaction of methyl amine and succinic acid in an aqueous
medium..............................................................24
5. MA (red) and SA (black) each analyzed alone using an ATR-IR........25
6. ATR-IR spectra of SA (red) and the MAS product (black).............27
7. ATR-IR spectrum of SA with a 50 seem flow through of MA............28
8. FTIR spectra of SA thin film at different exposure times to a 50-sccm flow
rate of MA focusing on 3445-3432 cm'1...............................34
9. Temporal changes in the infrared spectrum focusing on the NH2
vibrational stretch, reaction was carried out using 50 seem of MA..35
10. Temporal changes in the infrared spectrum focusing on the NH2
vibrational stretch, reaction was carried out using 25 seem of MA..36
11. Temporal changes in the infrared spectrum focusing on the NH2
vibrational stretch, reaction was carried out using 10 seem of MA..37
12. FTIR spectra of thin film SA at different exposure times to 50 seem of MA
focusing on 1537-1528 cm-1..........................................38
13. Temporal changes in the infrared spectrum focusing on the NH2 scissor
vibrational stretch, reaction was carried out using 50 seem of MA..38
14. Reaction rates of each peak analysis as a function of MA concentration. 41
viii


LIST OF EQUATIONS
EQUATION
1. Basic bimolecular reaction.......................................32
2. Pseudo first order rate..........................................32
3. Exponential function used to model reaction......................33
4. Uptake coefficient...............................................39
IX


CHAPTER I
INTRODUCTION
Atmospheric aerosols are major contributors to air quality and can have vast impacts on the biosphere. There are five principle layers to the atmosphere, and each layer has a specific function in maintaining life on planet Earth. The lowest layer of the atmosphere is the troposphere, where all life exists. The troposphere extends from the earth’s surface to 14.5 km high and is where the predominant forms of weather exist, as well as the gas composition needed to facilitate life.3 The second layer is the stratosphere, extending 50 km high and contains the majority of earth’s ozone. This layer supports the shield needed to protect life from the sun’s violent radiation, all while maintaining a confortable temperature in the biosphere4. The third layer of the atmosphere is the mesosphere, extending to 85 km. This is the layer where most meteors entering the earth’s atmosphere meet their fate with oxygen and burn up. The fourth layer is the thermosphere, where fast moving electrons from space collide and excite the gases in the earth’s atmosphere stretching 600 km above the earth. When the gases return back to the ground state from their excited state, they emit small bursts of energy in the form of photons, which are in the visible range of the electromagnetic spectrum.5 The outer most layer of the atmosphere is called the exosphere, and it is the first line of defense when it comes to any harmful radiation trying to enter the lower atmospheric levels from outer space. It extends far beyond any other layer with a height upwards of 10,000 km above the earth’s surface. The atmosphere is held into place by earth’s gravitational field. Without
1


gravity, the air we breathe would dissipate into the expanding universe and life would cease to exist. It is the atmosphere that protects all life from the cosmic rays, and is the perfect ether to allow all life to thrive.
The gas composition in the troposphere is predominantly composed of nitrogen (78.0%), oxygen (20.9%), and argon (0.9%). All other gases are so low that they are often defined in the ppm range. Although not considered a gas, water vapor is one of the most important constituents of the atmosphere. Water vapor in large concentrations, or colloquially known as clouds, act as earth’s atmospheric reaction mediums. The formation of these reaction epicenters depends highly on particulates and aerosols. The colloidal particulates and aerosols function as the center point for water molecules in the atmosphere to condense.6 Besides cloud condensation nucleation, aerosols can dictate smog levels, ozone levels, influence climate change and overall has a major implication to human health.
Aerosols can be classified as suspensions in a gas. They can exist discretely as small particles in the liquid or solid phase and they are dispersed and suspended throughout the atmosphere. Aerosols are best characterized coming from three major sources: marine, crustal and anthropogenic.7 In addition to the major sources, there are many different species of aerosols as shown in Table 1. Aerosols can be defined as a complex mixture of particles that can be composed of liquid droplets, organic materials such as metals, dust, organic chemicals, and organic acids.8 The solid phase of aerosols are primarily referred
2


to as particulate matter, which is also known as PM. The sizes of these particles are usually in the micron range, spanning from 0.001 microns to greater than 100 microns. An example of a visible aerosol would be smoke resulting from an incomplete combustion process. These particles of smoke for example, are typically heterogeneous particles and many of them are anthropogenic emissions resulting from human combustion processes.
Table 1 Major sources of aerosol species (AR5 of IPCC)
Aerosol Species Main Sources
Sulfate Primary sulfate: Marine and volcanic emissions. Secondary sulfate: formed by the oxidation of sulfur oxides
Nitrate Oxidation of NOx
Black Carbon Emitted from fossil fuel combustion, biomass and biofuel burning
Organic aerosols Fossil fuel combustion, biomass and biofuel burning, Continental and marine ecosystems. Noncombustion sources including biogenic origin
Brown carbon Biofuel and biomass burning. Humic-like materials of the biosphere origin
Terrestrial primary biological aerosol particles Terrestrial ecosystems
Mineral dust Wind erosion and suspension of soil. Selected agriculture activities and industrial units
Sea Spray Bubble rupturing and wind erosion
Marine primary organic aerosol Aerosols injected with sea spray in the biologically active regions of the oceans
Because aerosols contribute to such critical functions in the earth’s atmosphere, there are many efforts put forth to understand them. The investigation begins with the formation of aerosols and PM, and ends with their
3


transformation and reactivity in the atmosphere. The human population is so sizeable, that it is now contributing to the global volumes of aerosol. Because of the complexity in pollution, it can be fairly difficult to track these anthropogenic emissions. In particular, the EPA has been focusing much of its environmental effort on PM2.5, which is the particulate matter in the sub 2.5-micron range. The reasoning behind this range of particles is that there are extreme health effects associated with its size, which if inhaled in large quantities can lead to morbidity and mortality.9 In the last ten years the advancements of measuring aerosols have increased dramatically. From a macro point of view on aerosols, the National Aeronautics and Space Administration has provided data from two sensors aboard the Terra satellite. The Moderate Resolution Imaging Spectroradiometer and the Multiangle Imaging Spectroradiometer instruments have provided global observations of aerosols by optical depth, a process of measuring light extinction by aerosol in the atmospheric column above the earth’s surface.10 These detectors from the Moderate Resolution Imaging Spectroradiometer and the Multiangle Imaging Spectroradiometer instruments are measuring the gross and macro scale amounts of aerosols in the human habitable zone. When taking a micro scale point of view on aerosols in or around the community, there are instruments that can measure the amounts of PM emissions from industrial processes. For example, a major process that requires PM controls is the coal fired power industry. These combustion processes are heavily regulated by the EPA to emit less than 12 micrograms per cubic meter.11
4


The instruments being employed in the electric utility industry, which apply to the example of coal fired power, are usually located at the emission stack and are measured by a light scattering technique. According to the National Ambient Air Quality Standards, there are already ambient levels we should not exceed with the various PM emissions per the location. These are used as guidelines for any industrial or residential setting, as well as power facilities located in or near ecological sensitive areas. A table of these limits is shown below using 24 hour time measurement basis.
Table 2 Limits of National Ambient Air Quality Standards (NAAQS) for ambient aerosols (CPCB, 2009)
Types of aerosols Time period Concentration in ambient air (pg/mA3)
Industrial, Residential, Rural and other area Ecological sensitive area
PM10 Annual/24 hours 60/100 60/100
PM2.5 Annual/24 hours 40/60 40/60
Observing even more diminutively, before their inception, information on the reactivity of the components that form aerosols and PM can explain the chemical and physical attributes of the soon-to-be particles. The reactions that form aerosols can be measured on much more complicated instrumentation, and new methods and approaches have advanced aerosol chemistry to a higher sophistication. Some of the more expensive and costly methods to study aerosol chemistry may involve the use of a mass spectrometer, in which a smog chamber can be coupled to an aerosol mass spectrometer or a chemical
5


ionization mass spectrometer.12 An alternative to the mass spectrometer, which is a more available method to measure the kinetics and uptake coefficients of certain aerosol formations, can come from samples deposited on inert substrates where the reactants can interact while being measured by IR. In some recent studies, a flow reactor was coupled to an attenuated total reflectance infrared spectrometer (ATR-IR). The flow through the reactor was used to study the reaction kinetics of ozone and linoleic acid.13 This is a relatively cost effective set up where temperature dependence and relative humidity can be observed in conjunction with the reaction kinetics.13 The ease of these alternative techniques make them advantageous from other costly instruments, and the relatively fast scan in these approaches provide a lot of high-resolution information. Characterizing these reactions with widely available techniques can lead to an in-depth understanding of their formation by a larger scientific community. Aerosol characterization is imperative to distinguishing the implications of this complex chemistry that will affect the biosphere.
Aerosol formation can play a critical role in the nucleation of clouds. The physical properties of aerosols can influence cloud density and their reflective properties. Averaged around the globe, anthropogenic sources of aerosols are responsible for around 10% of the total aerosols in our atmosphere.14 The typical aerosols responsible for cloud condensation nucleation come from natural sources such as dust, sea spray, and volcanoes. The chemical species mostly associated with natural sources are sulfates, sea salts, and ammonium salts.15
6


The aerosols can provide a surface for which water can condense and depending on the hydrophilic nature of certain organic compounds there is a correlation to cloud nucleation. Aerosols have a clear implication to affect convective clouds and precipitation by altering cloud development and latent heat profiles.6 In order to facilitate the formation of rain the cloud droplets need a threshold radius of 14 pm. Many of the aerosols are found in the troposphere where they can be washed out with rainfall and brought back to the ground, however in other cases such as volcano eruptions the aerosols can reach as far as the stratosphere, which is the atmospheric layer above the troposphere. There are some organic fatty acids that can be found in many of the marine aerosols as well as large urban areas. The physical properties of the organic acids are unique as they contain both a hydrophobic and a hydrophilic moiety. Previously aerosols were mainly though to be mostly inorganic however evidence now shows many aerosols to have a thin organic film. Oleic acid is one of the most common fatty acids found in many plant materials, and is also very prevalent in cooking oils.16 Of the many organic acids in the atmosphere oleic acid is a long chain single proton carboxylic acid. However throughout the years in aerosol research it was determined that dicarboxylic acids are some of the most commonly found.17 Many of these organic acids range in carbon length C2-C10, and exist primarily as the result of the oxidation of larger volatile organic matter.18 The focus of this study explores the reactivity of a C4 carbon length dicarboxylic acid commonly known as succinic acid. Because of the small size the smaller dicarboxylic acids
7


have a low vapor pressure allowing them to travel and condense onto preexisting aerosols. These thin film layers can be responsible for drastically changing the behavior of the aerosols ability to uptake water and can affect the optical properties of clouds making them more reflective. Many of these organic aerosols can increase the threshold of the radius needed to form rain and increase the lifetime of the cloud and suppress rainfall.7 The hydrophilic properties of the organic aerosols can take yet another form as many of these aerosols can undergo oxidation reactions in the atmosphere. Ozone reactions are one of the most common oxidation reactions to occur. Depending on the aerosol substrate the oxidation reaction can drastically change the physiochemical and radiative properties of the aerosol.13,16 In some cases the ozone oxidation of aerosols can decrease the likelihood of cloud condensation nucleation. Having the ability to characterize these heterogeneous reactions would be advantageous in understanding the behavior of aerosols and their ability to change key properties of clouds that can directly affect the climate.
In urban areas the reactions that can take place on aerosols can get increasingly complex as the plethora of aerosols being emitted from industry. In the early 1900’s the term smog resulting from the contraction of the words smoke and fog was introduced and used mainly in reference to coal combustion products. Smog is mostly anthropogenic and in the more developed countries is now predominantly created by automobiles in addition to industrial emissions. Aerosols resulting from combustion processes have higher concentrations of
8


organic matter and are of high interest due to their impacts on global warming. The aerosols resulting from the combustion processes most always contribute to aerosol formation. In such cases pertaining to the transportation industry the soot emissions from diesel vehicles have very high carbon content. The high organic content of these aerosols are now thought to rival the global warming effects of C02, where before the black carbon associated with these aerosols were thought to have much greater scattering impacts.19 There are two types of scattering, Rayleigh and Mie. Rayleigh scattering occurs when the wavelength of the incoming radiation is much larger than the size to the particles it comes into contact with. Rayleigh scattering is responsible for the appearance of the blue color in the sky and the color of the sun to appear yellow. This type of scattering is when the incident electromagnetic radiation hits gases that are smaller than the wavelength of the incoming radiation.20 Mie scattering is when the particle is much larger than the wavelengths of the incoming electromagnetic radiation. This type of scattering is responsible for poor visibility and can cause the radiation to scatter in all directions including backwards known as backscatter.21,22 Large concentrations of aerosol particles in the range of 0.1-1.5 microns can scatter visible light and reduce visibility. In some locations the amount of these particles can range from 20-200 pg/m3 The annual standard for PM2.5 is met whenever the 3 year average of the annual mean PM2.5 concentrations for designated monitoring sites in an area is less than or equal to 15.0 pg/m3, however this depends on the designation area as characterized by the National Ambient Air
9


Quality Standards (NAAQS).8,23
There are significant implications to climate change by the introduction of aerosols into the atmosphere. Aerosols exist in two primary locations in the atmosphere, the stratosphere and the troposphere. Because aerosols play a major role as cloud condensation nuclei, they can greatly influence the formation, optical properties, and saturation of clouds.6 Unlike greenhouse gases aerosols can cause major forces in the climate change in either a cooling or a heating manner. Through solar forcing the irradiance of aerosols at the tropopause can cause major environmental effects either indirectly or directly by scattering and absorption of solar energy and its thermal radiation. In the case of major volcanic eruptions the fine particulate matter and aerosols can actually block and reflect the sun causing a cooling effect. Prime examples of major eruptions large enough to affect the earth’s temperature are eruptions such as that of Tambora (1815), Krakatoa (1883) and Agung in (1963), which was associated with global cooling of around a few tenths of a Celsius. This cooling can last for up to several years after some very large volcanic eruptions, caused by the reflective behavior of the suspended aerosols.24 On the other end of the temperature spectrum the aerosols can cause heating by preventing a reflective behavior when they are deposited or concentrated in areas such as large bodies of water, snow and ice.7 These depositions can affect the albedo of the different surfaces and may increase the adsorption of solar energy. Some of the latest climate changing research uses complex earth system climate models. It is well known that the
10


ocean temperatures are linked to the climate in the near by landmasses.25 Such correlations have linked the Atlantic sea surface temperatures to the droughts in Africa and the rainfall in Brazil.26,27 The power of these models can help predict the macro level of aerosols impact on climate change. However the modeling still uses inputs for aerosol formation such as oxidants, biomass, and carboniferous properties of aerosols.25 This leads back to the characterization of different aerosols by analytical means. The better we can characterize the physical properties of aerosols such as optical properties, ability to absorb moisture, and combine with any large anthropogenic emissions, the better we can fit these macro models to climate change.
There has been much research conducted around PM2.5 lung inhalation, most of this research stems from cigarette use and the health effects of cigarette smoking and the implications of second hand smoke. The research clearly shows how PM25 cardiovascular disease (CVD) and lung cancer are linked. The first time the American Heart Association published a scientific statement for air pollution and CVD was in 2004.28 The study details the duration of PM2.5 exposure and the triggers of CVD. The most frightening aspect of the research is the shear numbers of deaths caused by CVD. The reports and research reveal that one in every three deaths in America is caused by CVD.28,29 Although much of the research focuses on cigarette smoke our lungs cannot differentiate the PM2.5 emissions, and the main emissions leading to a deadly source of PM2.5 can come from fossil fuels containing arsenic, selenium, mercury, and sulfates from
11


sulfur dioxide. These emissions are created by the combustion of coal, oil, diesel and gasoline, and are concentrated in major cities.30 The coal and natural gas industries together supply more than 50% of the energy used in America.31 Larger cities have an increased likelihood that combustion byproducts can elevate to dangerous concentrations of PM2.5.15 In many circumstances the inefficiencies of power transmission caused the placement of coal-fired power plants to be near large cities, and even if the plants were built further from the heavily populated cities their emissions can travel hundreds and in some cases thousands of miles from their source. With population growth the demand for power is increasing and the development of homes near large industrial settings are becoming more prevalent.
The atmospheric air quality is becoming increasingly complex in the evergrowing industrialization of developing societies. These multifaceted reactions can occur naturally however many of them are influenced by human emissions and can be concentrated in urban areas, industrial settings and large agricultural areas. Human activities and heavily populated areas are likely to have amplified the influences to air quality. The EPA is designed to regulate harmful environmental factors to human health but with the complexity of the reactions it is becoming increasingly difficult to identify what factors are contributing to the largest impacts in the biosphere. Overall air quality has been improving in developed countries since the 1990s because of organizations helping to create public awareness. This can be attributed to growing evidence that there are
12


strong correlations between air quality and human health. Although the improvements are progressing in the right direction identifying the constituents of air quality are becoming more demanding. The method proposed in this research is a relatively quick and effective way of determining the reaction kinetics of a two-component system along with uptake coefficients. With some additional components the proposed tests using the same equipment can investigate the relative humidity effects. These physical properties can be inputs into well-established models to obtain the potential for cloud condensation nucleation.
13


CHAPTER II
EXPERIMENTAL SECTION
The experiment was designed to react liter volumes of clean dry air containing low concentrations of vapor over a thin film of organic crystalline solid. This custom set up can be used to simulate the reactions contributing to aerosol formation. However the set up its not limited to just exploring aerosol formation. Once the aerosol is formed or other constituents already considered aerosols could be further studied with water vapor or reacted with ozone. The aerosol can then be treated with different relative humidity’s to understand the uptake coefficients. The aerosols can also be reacted with ozone to determine the atmospheric ozonolysis products. The flow through chamber can also be combined with equipment to simulate moderate temperatures and pressures.
One of the key features of this design is the ability to combine analytical equipment and monitor each set of reactions on a time-based scale. The experimental set up can be best described in three parts, the instrumental analysis, the vapor pathway, the flow through reactor, and an overall detailed procedure of the experiment.
Using a very well known method to analyze organic molecules, an ATR-IR (Nicolet 6700 manufactured by ThermoScientific) was set up to monitor the uptake of MA onto the thin film SA. During this type of experimentation, the ATR-IR data can be taken over very brief time intervals to better resolve any kinetic information associated with these organic reactions. The advantage of this set-up is the real time capabilities to measure changes within an organic structure.
14


Bonding information can reveal exactly what is happening during the course of the reaction. The temporal spectra generated can be reduced to identify any kinetic information, and from that data the determination of uptake properties can be calculated. Covering a wide range of organic molecules IR is a relatively useful instrument that can analyze gases, liquids and solids. It is a nondestructive test with a high resolution and can complete a scan within a couple of seconds.
The experiment was constructed using a custom set up of mass flow controllers (MFC) manufactured by Alicat Scientific and a liquid bubbler produced by SKC INC. Eighty Four, PA used to introduce the MA (Alfa Aesar 40%) as a vapor form into the closed cell flow through reaction cell containing the ZnSe crystal. There are additional components that can be included into this set up to explore the relative humidity uptake and the reaction products from an ozone generator. A schematic of the experimental set up is shown in Figure 1.
990 seem
Figure 1 Schematic of the overall gas movement
The plumbing system fully contained inside of a fume hood and begins
with a clean dry air source. Air is conveyed through MFC 1 at a volume of five
15


standard cubic centimeters per minute (seem). To create a MA vapor the clean dry air is then sent through a bubbler containing MA the vapor pressure of MA is fairly high and successful generation of MA vapor is attained easily. The vapor carried out of the bubbler by MFC 1 is now combined with 995 seem of clean dry air from MFC 2. The combined flow on both of these MFC’s equals a total flow of one standard liter per minute. Those two flow controllers are then connected to a vent and another MFC. The amount of gas flow that doesn’t go through the vent traverses through MFC 3 at a flow rate of 10, 25, and 50 seem. Depending on that flow, for example if it is 10 seem it will be combine with 990 seem clean dry air to make up a total of one standard liter per minute. The purpose of having this configuration of flow controllers is to dilute the MA vapor down to the parts per million (ppm) range, thus simulating the concentrations of the aerosols that can be present in atmospheric reactions.
To determine the reaction kinetics, the MA vapor and the SA is monitored in real time using an ATR-IR spectrometer. The flow through reactor coupled with the ATR-IR is shown in Figure 2 below. The ATR-IR is equipped with a mercury cadmium telluride (MCT) detector, which must be cooled with liquid nitrogen.
The multireflection sampler uses a zinc selenide (ZnSe) crystal with a refractive index of 2.4 and dimensions of 5 cm in length, 1 cm in width and 0.5 cm in depth. The ZnSe crystal was selected for its large penetration depth, which is estimated to be 1.1 microns assuming a 45° degree angle of incidence.
16


Vent
CH3NH2 in
Figure 2 Schematic of the flow through reactor coupled with the ATR-IR
The SA deposition was prepared by first making a reagent of SA in a
volumetric flask using 1.476 g of SA (99% purity, Aldrich) and 500mL of acetone (99.99% ACS reagent grade purity from Fisher Scientific) creating a 0.025 g/mol solution. The ZnSe crystal was cleaned extensively with a lens wipe and a micropipette was used to transfer 150 pL of the SA solution onto the ZnSe crystal. The acetone evaporates at room temperature and 442 pg of the SA is deposited onto the ZnSe crystal. The thin film of SA is dispersed sufficient to completely cover the path of the incident beam. The thickness of the film is estimated to be less than 0.1 pm, which is much less than the penetration depth of the evanescent wave. The importance of the SA thickness ensures proper probing of the thin film by the IR.
Before the experiment is started, nitrogen gas is sent to the ATR-IR, this nitrogen gas purges the path of the laser beam used to make the measurements. Liquid nitrogen is also transferred to the MCT detector cell, the working
17


conditions of the MCT need to be operated at very low temperatures to reduce noise. The diagnostics signals are monitored using the Experiment Setup feature. The Bench and Diagnostics tab reveal the MCT stabilization prior to initializing the analysis. After approximately fifteen minutes the live interferogram signal from the MCT detector is stabilized and an experiment can begin. The ATR-IR is set to collect data every three minutes with 64 scans per data collection. A macro is created to loop 70 data collections over the course of a 4-hour period equaling one data collection every three minutes. A background of the clean ZnSe crystal is taken prior to depositing any SA onto the ZnSe crystal. This background will be used to blank any artifacts from the ZnSe crystal. A thin film of SA is deposited onto the ZnSe crystal with a micropipette and the complete evaporation of the acetone occurs in less than 5 minutes. The clean dry air is turned on and sent over to the MFC’s and the bubbler containing MA. The MFC’s are powered on and the MA vapors are sent through the bypass line. The lines are conditioned for 0.5 hours prior to starting an experiment and the first couple of data collections are taken without the introduction of MA. This establishes the baseline SA spectrum without any interference from MA. After three data collections the bypass valve is closed and the valve to send MA vapor over to the flow through cell connected to the ATR-IR is opened. The MA vapor traverses through the closed flow through cell containing the ZnSe crystal with a thin film of SA deposited on it. The reaction between the MA vapor and the thin film of SA has a residence time of less than a tenth of a second. The length of the ZnSe crystal is
18


5 cm and the custom stainless steel chamber has a volume of approximately 1 ml_. To explore the reaction kinetics the amount of MA is introduced in varying concentrations. The mass transfer of MA onto the SA takes place over the course of 4 hours and is usually completely saturated by the end of the experiment. Depending on the flow rates coming from the bubbler containing MA the saturation times can vary. The reaction kinetics and uptake coefficients can be slightly different depending on the concentration of MA flowing through the reaction cell. However the test was designed to simulate a pseudo first order reaction so the concentration of MA is very dilute compared to the amount of SA. The conditions at which each component of this custom set up are standard temperatures and pressures, 1 atm and 293 K.
The entire system was leak checked prior to beginning any experimentation and a rotometer was installed post reaction cell on the vent line to verify the one standard liter per minute flow rate. The data was taken using OMNIC™ Spectra Software by Thermo Fisher Scientific. The spectrum was configured to deliver the results in an absorption manner. The program used to write the macro was also OMNIC™ and was set up with a loop sequence to collect sample, delay, and restart the loop. The data was then manipulated in either the OMNIC™ Spectra Software or Essential FTIR®. Within the OMNIC™ Spectra Software an experimental parameter was set up to average one data point from 64 scans with a resolution of 4 cm-1. The range of IR signals was taken over 650-4000 cm-1.
19


CHAPTER III
METHODS
Heterogeneous reactions of succinic acid and methyl amine have been examined by attenuated total reflectance infrared spectroscopy (ATR-IR). A preparation of succinic acid (SA) was deposited as a thin film onto a zinc selenide crystal (ZnSe). A dilute vapor of methyl amine (MA) was conveyed into a closed cell flow reactor containing a thin film deposit of SA. The MA vapor was introduced into the flow reactor via a clean dry air passing through a bubbler containing MA. Throughout the experiment, the reaction was subject to standard temperature and pressures. Over the course of a 4 to 8 hour period, the ATR-IR was set to take absorbance readings every 1 to 3 minutes. As the reaction progressed a temporal spectrum was established. An analysis of the temporal spectrum was used to calculate the reaction rates k based on the modeling of a pseudo first order reaction. In addition to kinetic information, uptake coefficients y were also acquired based on the kinetic information and the changes in IR signals near the 3445, 1537, 1060 cm'1 range. These IR values correspond to the —NH2 stretching from the primary amine group, the -NH2 bending from the primary amine, and the C-N stretching from the carbon bonding with the nitrogen within the MA molecule. As more MA absorbed onto the thin film of SA the ATR-IR spectra began to display some of the characteristics associated with the MA organic functionalities. Furthermore, the results revealed some loss in the carbonyl bond characterization in the 1780-1710 cm-1 range caused by an elongation in the carbonyl bonding, and an increase of the absorbance signals in
20


the 3500-3300 cm-1 range due to the -NH2 stretching adding to the peaks along with the electron density near the -OH vibrational stretching.
Modern infrared spectroscopy commonly uses the wavenumber as the x-axis value and either an absorption or transmittance on the y-axis. The wavenumber is the number of waves in a length of one centimeter and is given by the following relationship, v=1/A. This unit has a unique linear relationship to with energy. The energy associated with infrared radiation can be used to create signature absorptions within an atom or molecule. For a molecule to exhibit these signature absorptions it must be able to possess electric dipole features. These absorptions are actually resonant features within the bonding structures that can be modeled by a harmonic oscillator.32 There are many vibrational modes that can be in the form of symmetric or antisymmetric stretching, scissoring, rocking, wagging, and twisting.
Traditional infrared spectrometers have been commercially available since the 1940’s. However they relied on prisms to act as dispersive elements, this was eventually replaced with diffraction gradients and then again with interferometers.32 Michelson interferometers are most commonly associated with FTIR and consist of two perpendicular plane mirrors. Where one mirror is stationary and the other one can move in a single plane. Figure 3 below shows the Michelson interferometer. There is incident light source directed over to a semi reflective mirror. The mirror effectively splits the beam to where half of the light beam travels through the mirror and other half is reflected. Usually one path
21


travels over to a reference and the other half travels through a sample cell with the molecules of interest. Both beams are then recombined and are directed to a detector. The movable mirror can move to create an optical path difference between the two arms of the interferometer.
Stationary Mirror
Detector
Figure 3 Michelson Interferometer
The purpose of having the movement is to create path differences where
the frequencies are interconvertable by the mathematical Fourier transform method, hence the name FT-IR. The two light beams are passed through a beam splitter that quickly alternates the signals to the detector. The comparison of the
22


beams can later be resolved with computational software. Because one beam is always directed to a reference and the other beam through a sample it will give relatively accurate data even if the light source drifts over time. Another important feature of FTIR is the versatility of the sample types that can be analyzed. Examples of samples range from gases to liquids, and solids. There are many applications where IR is used, and many of the applications are still being employed to more information about the galaxy.33,34
23


CHAPTER IV
RESULTS AND DISSCUSSION
The most likely result of the reaction between MA and SA under the standard temperatures and pressures is the formation of a carboxylate salt. The differences in pH properties are large enough to be the driving force for salt formation. This happens when the amine deprotonates of the carboxylic acid, and by doing so takes on a proton forming a methyl ammonium ion. Because SA has two protons the completion of the reaction will require two moles of MA to every mole of SA. The pKa of SA is 4.2 and 5.6 respectively and both protons will react readily with MA, having a pKa of 10.66. The result is a stable succinate salt as depicted in Figure 4.
Figure 4 Conventional reaction of methyl amine and succinic acid in an aqueous medium
However as the reaction continues the direct evidence of the methyl ammonium succinate (MAS) product is somewhat masked by the overwhelming amounts of SA. To confirm the formation of the MAS each of the above reactants and products were analyzed separately from one another. The MA was analyzed as a thin film liquid dispersed directly from the reagent bottle onto the ZnSe crystal and the SA was analyzed alone as a thin film solid deposit as described in the
24


above experimental section. The two ATR-IR spectra are displayed in Figure 5
below.
ATR-IR Spectra
o
3650 3150 2650 2150 1650 1150 650
Wavenumber [cm-1]
Figure 5 MA (red) and SA (black) each analyzed alone using an ATR-IR
The main IR assignments for MA stem from the amine functionality, the -NH2
stretching near 3280 cm'1, -NH2 scissor vibration near 1670 cm'1, and the C-N stretching near 1334 cm'1. The main assignments from SA revolve around the carboxylic acid functionality. The very broad -OH stretch between 3000 - 2500 cm'1, C=0 stretching near 1675 cm'1, and the C-0 stretch near 1196 cm'1. The IR assignments are listed in Table 3 below with a designation of functional group, peak location, and peak intensity.
25
Absorbance


Table 3 IR assignments for MA and SA, br = broad, s = strong, m = medium, w = weak
Functional Group Peak Position Peak Intensity
Assignment Wavenumber [cm'1]
NH2 stretching 3438.4 br
NH2 bending 1534.7 s
C-N stretching 1334.5 m
-OH stretching 3000-2500 w
C=0 carboxylic stretching 1720.9 s
C-0 1410.6 m
To carry out the MAS product evaluation, a liquid - liquid reaction of MA and SA was accompanied by an ATR-IR analysis. In the liquid - liquid reaction, 0.3103 g of SA was reacted with 0.1645 g of MA in an aqueous medium. A heat evolution and a white crystalline precipitate followed the resulting reaction. Considering the two spectra in Figure 5 the prediction of where the formation of MAS will fall in the ATR-IR analysis is unsurprising. In Figure 6 below the MAS is displayed with an overlay of SA to forecast of where the peaks are expected to arise from the reaction of MA vapor with the overpowering amount of SA.
26


0
3650 3150 2650 2150 1650 1150 650
Wavenumber [cm-1]
Figure 6 ATR-IR spectra of SA (red) and the MAS product (black)
In an attempt to force the reaction to completion the flow rate of MA was
increased from 10, to 25, and to 50 seem. Increasing the flow rate effectively increases the overall concentration of MA and the exposure of MA to SA. With varying concentrations of MA flowing into the closed cell reactor the most dramatic spectra was displayed with the highest concentration of MA. To verify reproducibility and variability each concentration was directed a minimum of 3 times. The lower the concentration the longer the reaction takes to complete and alternatively the higher the concentration the shorter it takes to complete the reaction. Taking this into account, to make sure the data being captured was sufficient for the proper kinetic modeling and uptake coefficient calculations the OMNIC macro program was set to take 64 scans every 3 minutes all the way down to 64 scans every minute, and the length of the measurements varied from 4 to 8 hrs. After the experiment was carried out three times at each
27
Absorbance


concentration, the reaction kinetics were analyzed, uptake coefficients were calculated and furthermore the reaction rates were examined as a function of MA concentration.
The experiment began with the highest concentration of MA introduced into the system, as this was the securest condition for the reaction to form the desired products. The ATR-IR spectra in Figure 7 show SA with no exposure to MA and the final spectra of the SA with the longest exposure to MA. The spectrum in Figure 6 has provided a forecast of areas in the spectrum below where the MAS formation can be identified with lower interference from the overpowering SA peaks.
ATR-IR Spectra
----SA
0
3650 3150 2650 2150 1650 1150 650
Wavenumber [cm-1]
Figure 7 ATR-IR spectrum of SA with a 50 seem flow through of MA
Besides the easily distinguished carboxylic C=0 stretching vibration at 1710cm'1
other features that can be attributed to the SA IR spectrum are the -CH2
28
Absorbance


antisymmetric stretching observed at 2919 cm'1 and the -CH2 symmetric stretching at 2850 cm'1. In addition to those stretching vibrations the -CH2 moiety also exhibits a bending at 1410 cm'1 and 1306 cm'1 while its wagging vibration is shown at 1197 cm'1. Some other unique characteristics of the SA spectrum comes from its carboxylic acid functionality, the broad -OH stretch around 2500-3300 cm'1 and the -COH out-of-plane bending at 910 cm'1 as shown in the black spectrum labeled SA in Figure 7. Not only showing SA by itself Figure 7 shows the end result of the reaction where MA has deposited and reacted with the thin film of SA to a relatively complete saturation. Nevertheless, there is an inconvenience of identifying the contributions from SA and MA in the IR spectrum due to the fairly puzzling overlap of peaks. To help identify peaks in the above reaction each compound was not only analyzed alone but also accompanied by reference IR spectrums from the NIST Webbook. With a forecast of where the MAS peaks are located the clearly identifiable peaks coming from the MA vibrations are the -NH2 stretching vibrations at 3438 cm'1 the -NH2 bending vibration at 1537 cm'1 and the C-N vibrational stretch at 1068 cm'1. The documentation of the infrared vibration assignments of SA saturated with MA are identified in Table 4 below and correlate to the red IR spectra labeled SA-MA in Figure 7.
29


Table 4 IR feature assignments of SA-MA at room temperature
Functional Group Assignment Peak at Wavenumber [cm'1] Absorbance Peak Height
NH2 stretching 3438.4 0.4
CH2 antisymmetric stretching 2919.6 0.6
CH2 symmetric stretching 2850.2 0.6
C=0 carboxylic stretching 1720.9 0.9
NH2 bending 1534.7 0.6
CH2 bending 1410.6 1.1
CH2 bending 1306.7 1.1
CH2 wagging 1197.0 1.3
C-N stretching 1068.5 0.5
-COH carboxylic out-of-plane bending 910.3 1.1
When the lower concentrations of MA were reacted with SA the same resulting spectra formed but the temporal absorbance profiles changed and will be discussed in detail below. As the MA is introduced into the flow through reactor containing the thin film of SA, the spectrum began to take on some characteristics of MA. This is most prominently detected as new peaks arise from the MA depositing and reacting with the SA surface. However the separation of the spectral peaks becomes increasingly difficult as the SA and the MA have absorbance bands that overlap. To overcome this commonality in absorption bands between the SA and MA, isolated functional groups specific to each compound were identified and the change in absorbance was used to calculate the reaction rates. The most notable changes are found in the regions where the
30


SA spectrum shows very low absorbance peaks and where the MA shows strong absorbance as identified in the spectra labeled SA-MA in Figure 7. Primary amines exhibit vibrational stretching with two bands in the 3500-3300 cm'1 range, these peaks are often very weak. Fortuitously the IR spectrum of SA has a large broad -OH stretch with a peak cut off of 3300 cm'1. As the MA further absorbs onto the SA the development of one of the weak -NH2 vibrational stretches occurs at 3438 cm'1. This falls outside of the broad -OH stretching vibration flaunted by SA and is clearly evident as the reaction is complete. Another area where the primary amine’s character is revealed can be found in the development of an -NH2 scissor vibration at 1534 cm'1.
The method for deriving the uptake coefficient employed in the present work is similar to the methods used in previous studies of ozone reacting with oleic acid, linoleic acid, and squalene.13,35,36 The details of this approach was developed in research elsewhere and follows an absorption model.37 The derivation of the uptake coefficient y stems from the calculation of each reaction rate. The analysis began with the modeling of a second order reaction. However, establishing the rate constant of a bimolecular reaction can be challenging. Considering Equation 1 below the rate constant can become pragmatic and in most cases needs to be determined experimentally. Typical methods to determine the reaction rate constant will employ an experimental method where both reactants can be introduced into the reaction in different concentrations. The
31


difficulty in this practice is that both reactants need to be measured simultaneously.
A + B -> c
Equation 1 Basic bimolecular reaction
Because of the challenge in measuring the two reactants in Equation 1 simultaneously a common solution is applied. The pseudo first order approximation can resolve the complications in measurement by supplying one of the reactants in great excess. If the experimental reaction is designed to be a pseudo first order reaction and the concentration of one reactant is in great excess the bimolecular reaction can take the form of Equation 2 where the [SA] is in great excess over the reactant [MA], This form is a result of the [SA] concentration remaining mostly unchanged during the course of the reaction pathway. The main measurement is the character development of [MA] taking shape over time with k-i denoting the pseudo first order rate constant.
d[SA]
^ = k-iiSA], where k^SA] = k[MA]
Equation 2 Pseudo first order rate
The uptake coefficient can be estimated with the changes in the IR absorbance profiles over time.38 The most discernible absorbance profiles will be taken from the distinct wavenumbers of the amine’s -NH2 vibrational stretch at 3438 cm'1, the -NH2 bending at 1534 cm'1 and the C-N stretching at 1068 cm'1. Moreover the peak development at each of these wavenumbers will be associated with the absorbance data to estimate the uptake coefficient y. The temporal response of
32


the absorbance peaks will be plotted using Equation 3 below. Where B is the absorbance difference at time zero and k is the pseudo first order rate coefficient. To avoid any errors in baseline shifting the absorbance value B will be integrated absorbance values at time t as calculated from the baseline.37,39 As the data is plotted an exponential fit will be applied and the reaction rate k will be determined in Equation 3. With the changes in MA concentration the most discernable peaks are still developed however the amount each temporal integrated absorbance increase varies. The higher the concentration of MA used in the reaction the larger the intensification between absorbance measurements per the same amount of time.
AA(t) = -Be~kt
Equation 3 Exponential function used to model reaction
The peak development at each wavenumber was manipulated using the
Essential FTIR ® software. An example of the peak evolution during the exposure of MA onto a SA thin film is shown in Figure 8. The data begins with a relatively unexciting line and then begins to take shape as the -NH2 vibrational stretching grows into a well-defined peak. As mentioned above the same peak arises from the different MA concentrations however with a different temporal response that can be best explained using Equation 3 to model the data.
33


Figure 8 FTIR spectra of SA thin film at different exposure times to a 50-sccm flow rate of MA focusing on 3445-3432 cm'1
The chronological changes in absorbance that created uninhibited and well-defined peaks were shown in Figure 9 where the x-axis is the time in minutes and the y-axis is the integrated absorbance at time zero. It is from these plots that the reaction rate constant can be determined. When changing the MA concentrations as a function of flow the reaction rate /calso changes. Several experiments were carried out using variable MA concentrations and the resulting reaction rates are directly proportional to the concentration of the MA. Each plot discloses the function coefficients to fit the data with a greater than ±95% confidence interval. The integrated absorbance values were fitted into an exponential function using Igor Pro v7.01. The relatively good fit of the integrated temporal absorbance data to the function in Equation 3 confirms the application of a pseudo first order reaction rate model to be appropriate for the estimation of the uptake coefficients.
34


Figure 9 Temporal changes in the infrared spectrum focusing on the NH2 vibrational stretch, reaction was carried out using 50 seem of MA
There was a fairly good agreement of calculated reaction rates k-i using the integrated absorbance peaks in the above plot from 3445-3432 cm'1. Analyzing relatively the same location in another reaction using half of the MA concentration Figure 10 shows a decreased rate constant k2.
35


Figure 10 Temporal changes in the infrared spectrum focusing on the NH2 vibrational stretch, reaction was carried out using 25 seem of MA
As shown in Figure 11 the rate constant k2 was decreased as the concentration of the MA in the reaction was lowered. Compared to the reaction rate k-i using 50 seem the reaction rate using 25 seem k2 was nearly cut in half. The reaction was carried out again at 10 seem and the data was analyzed using the same method above however the reaction at 10 seem was much slower than the previous concentrations of MA. The data in Figure 11 shows the difficulty of measuring the lowest concentration of MA. The data doesn’t fit as well as the previous two concentrations and k3 may have a higher variability due to several reasons. One potential reason is the MFC working at such a low flow rate increasing the noise and inconsistency of the MA delivery.
36


Figure 11 Temporal changes in the infrared spectrum focusing on the NH2 vibrational stretch, reaction was carried out using 10 seem of MA
To verify the reaction rate k, FTIR spectra was examined in all of the absorbance bands where peak changes were uninhibited by the excess amounts of SA. In an additional observation and to confirm the reaction kinetics calculated from the 3455-3432 cm'1 region another peak area was integrated from 1537-1528 cm'1, Figure 12.
37


Figure 12 FTIR spectra of thin film SA at different exposure times to 50 seem of MA focusing on 1537-1528 cm-1
The IR spectrum at this particular location displays a shift in several locations but the most notable shift occurs at the peak between 1537- 1528 cm'1. The bar that highlights the peak is the area where each spectra was integrated.
Figure 13 Temporal changes in the infrared spectrum focusing on the NH2 scissor vibrational stretch, reaction was carried out using 50 seem of MA
38


This reaction rate analysis was carried out with all of the MA concentrations 10, 25, and 50 seem in both peak areas. However, the reaction rate of the high concentration MA exposure k-i from each of the above peak locations has a relatively good correlation and confirms the pseudo first order rate environment.
For heterogeneous reactions involving a gas-solid or a gas-liquid reaction the likelihood of the reaction occurring will depend on the surface reaction probability also known as the uptake coefficient.38 The mean speed of the MA molecules in the gas phase coming from the contact with the thin film of SA will be measured by ATR-IR and the surface area to volume ratio of the SA thin film will be used to estimate the geometric surface of the reaction. Using Equation 2 where the rate k-i is calculated from the temporal absorbance changes and the concentration of [SA] is equal to the initial concentration. The uptake coefficient can be solved for using Equation 4.
d [SA] _ ^ /Pmac\$a dt ~ ~Y\4Rt) V
Equation 4 Uptake coefficient
The rate constant ki multiplied by the initial concentration of [SA] is equal to the uptake coefficient multiplied by the MA pressure Pma, cthe mean speed of the MA particles in the gas phase [cm s'1]; R is the gas constant, Tis the temperature; and S^the surface area to Vvolume ratio of the SA thin film [cm'1], where the surface area is estimated to be about 4.5 cm2 based on the available working surface area of the ZnSe crystal used in the flow through reactor. In the work presented here the uptake coefficient was calculated using the rate
39


constants derived from each of the MA concentrations 10, 25, and 50 seem and
exploiting two of the three absorbance bands that were not obscured by the SA absorbance peaks, the -NH2 vibrational stretch at 3445 cm'1 and the -NH2 bending at 1537 cm'1 as shown in Table 5. The standard deviation was calculated as a percentage from the average. As mentioned above the lower the concentration of MA the higher the variability.
Table 5 Summary of pseudo first order rate constants k, derived from the temporal integrated absorbance taken at different locations, and the corresponding calculated uptake coefficients
Physical State MA flow rate [seem] Reaction rate k1 x 10'J [s-1] taken at different peak locations Uptake coefficient y x10'3 per the peak location
3455 - 3432 cm-1 1537-1528 cm-1 3455 - 3432 cm-1 1537-1528 cm-1
Solid/Vapor 10 0.187 0.327 0.182 0.317
Solid/Vapor 10 0.219 0.219 0.212 0.212
Solid/Vapor 10 0.130 0.063 0.126 0.061
AVG 10 seem 0.179 0.203 0.173 0.197
% STDEV 10sccm 20.49% 53.37% 20.49% 53.37%
Solid/Vapor 25 0.423 0.546 0.411 0.530
Solid/Vapor 25 0.650 0.568 0.630 0.551
Solid/Vapor 25 0.640 0.623 0.621 0.604
AVG 25 seem 0.571 0.579 0.554 0.562
% STDEV 25 seem 10.45% 3.22% 10.13% 3.12%
Solid/Vapor 50 1.225 1.085 1.188 1.052
Solid/Vapor 50 1.003 1.580 0.972 1.532
Solid/Vapor 50 1.338 1.238 1.297 1.200
AVG 50 seem 1.189 1.301 1.153 1.262
% STDEV 50 seem 13.92% 20.69% 13.50% 20.06%
Overall the two areas of peak development were in good agreement, however the second NH2 vibrational stretch near 1537 cm'1 may have a higher level of interference due to the location of the signal in the region of the IR field. When
40


comparing the data to literature values of MA uptake onto ammonium salts by Liu et al,2 the reaction rates are in same order of magnitude however the application is not quite the same.2 In addition to the comparison of reaction rates at different peak integration, the reaction rates were also plotted to determine their dependency on the MA concentration, Figure 14.
Reaction Rates
MA Flow Rate [seem]
Figure 14 Reaction rates of each peak analysis as a function of MA concentration
The reaction rates as a function of MA concentration shows a relatively good fit with both sets of data from each respective peak analyzed having a linearity R2 correlation greater than 0.9. Although the less inhibited peak near 3445 cm'1 demonstrates a better correlation, which can be due to much less interference from overlapping peaks in the finger print region. Likewise, the lower
41


concentration of MA contains variability from the MFC limitations, and with the lower the MA concentration eventually all reaction rates will converge to zero.
42


CHAPTER V
CONCLUSION
The above study applies the use of an ATR-IR to characterize heterogeneous reaction rates and uptake coefficients of a thin film solid phase SA and a vapor phase form of MA. The kinetics of the reaction was found to model a pseudo first order rate with a correlation coefficient R2 greater than 0.95. The overall uptake coefficients were verified by conducted calculations on two separate peak locations near 3445 and 1537 cm'1. Using the highest concentration of MA the two peaks analyzed were observed to be within less than 10% of each other. The reaction rates were presented to have a relatively high level of precision as they also correlated well to the MA concentration.
The advantages of this experimental set up prove the capability to explore the dependency of reaction rates on MA concentrations. Reproducibility and the repeatability between individual analysis are less than 25% of the average value, excluding the lower MA concentration runs. Although the predicted reaction between SA and MA to form the dimethyl ammonium salt was difficult to decipher there was strong evidence of MAS formation as shown in the Figure 7 overlay.
The main learning from the research is the demonstration of a high throughput method to characterize the formation of potential aerosols as a result of organic acids reacting with amines. The importance of this technique stems from the need to understand the ever-changing atmospheric environment. With climate change arguable being influenced by the growing human population it is advantageous to couple a physical analysis of aerosol formation with other macro
43


spectroscopy approaches. Understanding the physical characteristics on a micro level will shed light into multiple areas, including but not limited to aerosol formation, cloud condensation nucleation, understanding anthropogenic emission sources influencing climate change, and the changes in the atmosphere ultimately effecting the health of plants and animals.
44


REFERENCES
(1) Kawamura, K.; Bikkina, S. Atmos. Res. 2016, 170, 140-160.
(2) Liu, Y.; Han, C.; Liu, C.; Ma, J.; Ma, Q.; He, H. Atmos. Chem. Phys. 2012, 12, 4855-4865.
(3) Fallis, A. . J. Chem. Inf. Model. 2013, 53, 1689-1699.
(4) Lewis, A.; Carslaw, N.; Marriott, P.; Kinghorn, R.; Morrison, P.; Lee, A.; Bartle, K.; Pilling, M. Nature 2000, 405, 778-781.
(5) Way, M.; Eather, R. 2012.
(6) Li, Z.; Niu, F.; Fan, J.; Liu, Y.; Rosenfeld, D.; Ding, Y. 2011, 4.
(7) Manzoor, S.; Kulshrestha, U. 2015, 10, 738-746.
(8) U.S. Epa. U.S. Environ. Prot. Agency, Off. AirQual. Plan. Stand. Res. Triangle Park. North Carolina 2011, EPA 452/R-.
(9) Arden Pope, C.; Burnett, R. T.; Turner, M. C.; Cohen, A.; Krewski, D.; Jerrett, M.; Gapstur, S. M.; Thun, M. J. Environ. Health Perspect. 2011, 119, 1616-1621.
(10) van Donkelaar, A.; Martin, R. V.; Brauer, M.; Kahn, R.; Levy, R.; Verduzco, C.; Villeneuve, P. J. Environ. Health Perspect. 2010, 118, 847-855.
(11) Weinhold, B. Environ. Health Perspect. 2013, 121, 2013.
(12) Rudich, Y.; Donahue, N. M.; Mentel, T. F. Annu. Rev. Phys. Chem. 2007, 58, 321-352.
(13) Zeng, G.; Holladay, S.; Langlois, D.; Zhang, Y.; Liu, Y. J. Phys. Chem. A 2013, 117, 1963-1974.
(14) Kahn, R. Earth Obs. 2010, November 1.
(15) Valavanidis, a.; Fiotakis, K.; Vlachogianni, T. Urban Airborne Particulate Matter: Origin, Chemistry, Fate and Health Impacts; 2011.
(16) Schwier, A. N.; Sareen, N.; Lathem, T. L.; Nenes, A.; Mcneill, V. F. 2011, 116, 1-12.
45


(17) Rogge, W. H. L. G. C. 1997, 32, 13-22.
(18) Lightstone, J. M.; Onasch, T. B.; Imre, D.; Oatis, S. J. Phys. Chem. A 2000, 104, 9337-9346.
(19) Saleh, R.; Robinson, E. S.; Tkacik, D. S.; Ahern, A. T.; Liu, S.; Aiken, A. C.; Sullivan, R. C.; Presto, A. a.; Dubey, M. K.; Yokelson, R. J.; Donahue, N. M.; Robinson, A. L. Nat. Geosci. 2014, 7, 1-4.
(20) Chakraborti, S. 2008, 2-4.
(21) Cox, a. J.; DeWeerd, A. J.; Linden, J. Am. J. Phys. 2002, 70, 620.
(22) Kostylev, V. I. Bistatic Radar Prime. Pract. 2007, 193-223.
(23) Rule, F. 78 Fed. Regist. 3086 2013, 78, 1-27.
(24) Hansen, J.; Lacis, A. Nature 1990, 346, 183-187.
(25) Booth, B. B. B.; Dunstone, N. J.; Halloran, P. R.; Andrews, T.; Bellouin, N. Nature 2012, 484, 228-232.
(26) Ackerley, D.; Booth, B. B. B.; Knight, S. H. E.; Highwood, E. J.; Frame, D. J.; Allen, M. R.; Rowell, D. P. J. dim. 2011, 24, 4999-5014.
(27) Knight, J. R.; Folland, C. K.; Scaife, A. A. Geophys. Res. Lett. 2006, 33, 1-4.
(28) Brook, R. D.; Rajagopalan, S.; Pope, C. A.; Brook, J. R.; Bhatnagar, A.; Diez-Roux, A. V.; Holguin, F.; Hong, Y.; Luepker, R. V.; Mittleman, M. A.; Peters, A.; Siscovick, D.; Smith, S. C.; Whitsel, L.; Kaufman, J. D. Circulation 2010, 121, 2331-2378.
(29) American Heart Association; American Stroke Association. 2012.
(30) Pope, C. A.; Turner, M. C.; Burnett, R. T.; Jerrett, M.; Gapstur, S. M.; Diver, W. R.; Krewski, D.; Brook, R. D. Circ. Res. 2015, 116, 108-115.
(31) EIA. U.S. Energy Inf. Adm. 2015.
(32) Stuart, B. H. Infrared Spectroscopy: Fundamentals and Applications', 2004; Vol. 8.
(33) Leisawitz, D.; Rinehart, S. A. .
46


(34) Steele, J. D. 1885, 9-16.
(35) Stewart, D. J.; Griffiths, P. T.; Cox, R. A. Atmos. Chem. Phys. 2004, 4, 1381-1388.
(36) Fu, D.; Leng, C.; Kelley, J.; Zeng, G.; Zhang, Y.; Liu, Y. Environ. Sci. Technol. 2013, 47, 10611-10618.
(37) Segal-Rosenheimer, M.; Dubowski, Y. J. Phys. Chem. C 2007, 111, 11682-11691.
(38) Akimoto, H. Atmospheric Reaction Chemistry, 2015.
(39) Hung, H. M.; Katrib, Y.; Martin, S. T. J. Phys. Chem. A 2005, 109, 4517-4530.
47


Full Text

PAGE 1

HETEROGENOUS REACTION KINETICS OF SUCCINI C ACID AND METHYL AMINE by CHRISTOPHER VIZCAINO M.S, University of Colorado, Denver 2016 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Science Chemistry Program 2016

PAGE 2

! ii This thesis for the Master of Science degree by Christopher Vizcaino has been approved for the Chemistry Program by Yong Liu Marta Maron Xiaotai Wang December 17, 2016

PAGE 3

! iii Vizcaino, Christopher (M.S ., Chemistry Program ) Heterogeneous Reaction Kin etics of Methyl Amine an d Succin ic Acid Thesis directed by Associate Professor Yong Liu ABSTRACT The atmosphere is becoming an increasingly complex gaseous system. As the human population continues its exponential trajectory, the emissions as a result of human interactions are creating an intricate chemical mixture in the biosphere. Many of these pollutants are poorly understood but may have large implications to cloud condensation that can lead to climate change, and the formation of large amounts of particulate matter in the troposphere ultimately effecting human health . It was previously thought that many aerosols i n the atmosphere were either solid or liquid phase particles. However in many cases it was found that there is a dual phase, solid particles with a thin film layer of organic material. In the global amounts of organics in the atmosphere, dicarboxylic acids have been found to be in relatively plentiful concentrations . 1 Another major source of the atmospheric aerosols are nitrogen compounds, in particular amines. Thes e are emitted into the atmosphere through a variety of sources including industrial processes, biomass burning, livestock, and marine organisms . 2 The combination of these two plentiful aerosols is inevitab le . Understanding how these organics interact with one another can provide valuable information on the potential formation of much of the aerosols in our biosphere. It is advantageous to have a quick throughput method to characterize the kinetics of the ae rosol formation and

PAGE 4

! iv the corresponding products of the reaction between different aerosols. The form and content of this abstract are approved. I recommend its publication. Approved: Yong Liu

PAGE 5

! v I would like to thank Kaldi and the spirited goats for their wonderful discovery.

PAGE 6

! vi TABLE OF CONTENTS CHAPTER 1. INTRODUCTIONÉÉÉÉ ÉÉÉÉÉÉÉÉÉÉÉÉÉ É.. ÉÉÉÉÉ.... 1 Particulate matter and aerosols ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ2 Cloud condensation nucleation É... ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ.6 Anthropogenic sources and smog ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ8 Climate change É... ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ..10 Human health effects ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ...1 1 2. EXPERIMENTAL SECTION ÉÉÉ... ÉÉÉÉÉÉÉÉÉÉ É.. ÉÉÉÉ 14 Experimental set up ÉÉÉÉÉÉÉÉÉÉÉÉÉÉ... ÉÉÉÉÉ . É. 14 Procedure ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ. .17 Conditions ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ..19 3. METH ODSÉÉÉÉÉÉ ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ É.. ÉÉ.20 T heory É ÉÉÉÉÉ ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ... É.20 Attenuated total reflectance infrared spectrometry ÉÉÉÉÉÉÉÉ.21 4. RESULTS & DISCUSSION ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ É.. .24 Predicted aerosol formationÉÉÉÉÉÉÉÉÉÉÉÉ ÉÉÉÉÉÉ24 Infrared spectraÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ. 25 Reaction k ineticsÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ. . . 35 U ptake coefficients ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ...39 5. CONCLUSION ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ ÉÉ 43 6. REFERENCES ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ... ÉÉÉ ÉÉ. 45

PAGE 7

! vii LIST OF TABLES TABLE 1. Major sources of aerosol species (AR5 of IPCC) ................................ ....... 3 2. Limits of National Ambient Air Quality Standards (NAAQS) for ambient aerosols (CPCB, 2009) ................................ ................................ .............. 5 3. IR feature assignments of SA MA at room temperature ........................... 30 4. Summary of pseudo first order rate constants k 1 , derived from the temporal integrated absorbance taken at different locations, and the corresponding calculated uptake coefficients ................................ ........... 40

PAGE 8

! viii LIST OF FIGURES F IGURE "# Schematic of the overall gas movement ! ################################ ################################ ##### ! "$ ! ! %# Schematic of the flow through reactor coupled with the ATR IR ! ################### ! "& ! ! '# Michelson Interferometer ! ################################ ################################ ################################ #### ! %% ! ! (# Conventional reaction of methyl amine and succinic acid in an aqueou s medium ! ################################ ################################ ################################ ################################ ######### ! %( ! ! $# MA (red) and SA (black) each analyzed alone using an ATR IR ! ################## ! %$ ! ! )# ATR IR spectra of SA (red) and the MAS product (black) ! ################################ # ! %& ! ! &# ATR IR spectrum of SA with a 50 sccm flow through of MA ! ############################ ! %* ! ! *# FTIR spectra of SA thin film at different exposure times to a 50 sccm flow rate of MA focusing on 3445 3432 cm 1 ! ################################ ################################ ###### ! '( ! ! +# Temporal changes in the infrared spectrum focusing on the NH 2 vibrational stretch, reaction was carried out using 50 sccm of MA ! ############### ! '$ ! ! ",# Temporal c hanges in the infrared spectrum focusing on the NH 2 vibrational stretch, reaction was carried out using 25 sccm of MA ! ############### ! ') ! ! ""# Temporal changes in the infrared spectrum focusing on the NH 2 vibrational stretch, reaction was carried out using 10 sccm of MA ! ############### ! '& ! ! "%# FTIR spectra of thin film SA at different exposure times to 50 s ccm of MA focusing on 1537 1528 cm 1 ! ################################ ################################ ############################ ! '* ! ! "'# Temporal changes in the infrared spectrum focusing on the NH 2 scissor vibrational stretch, reaction was carried out using 50 sccm of MA ! ############### ! '* ! ! "(# Reaction rates of each peak analysis as a function of MA concentration ! # ! (" !

PAGE 9

! ix LIST OF EQUATIONS EQUATION 1. Basic bimolecular reaction ................................ ................................ ....... 32 2. Pseudo first order rate ................................ ................................ .............. 32 3. Exponential function used to model reaction ................................ ............ 33 4. Uptake coefficient ................................ ................................ ..................... 39

PAGE 10

! 1 CHAPTER I INTRODUCTION Atmospheric aerosols are major contributors to air quality and can have vast impacts on the biosphere . There are five principle layers to the atmosphere, and each layer has a specific function in maintaining life on planet E arth. The lowest layer of the atmosphere is the troposphere, wh ere all life exists. The troposphere extends from the earth ' s surface to 14.5 km high and is where the predominant forms of weather exist , as well as the gas composition needed to facilitate life . ' The second layer is the stratosphere , extending 50 km high and contains the majority of earth's ozone . This layer supports the shield needed to protect life from the sun's violent radiation , all while maintaining a confortable temperature in the biosphere 4 . The third layer of the atmosphere is the mesosphere , extending to 85 km . This is the layer where most meteors entering the earth's atmosphere meet their fate with oxygen and burn up. The fourth layer is the thermosphere , where fast moving electrons from space collide and excite th e gases in the earth's atmosphere stretching 600 km above the earth . When the gases return back to the ground state from their excited state , they emit small bursts of energy in the form of photo ns , which are in the visible range of the electromagnetic spectrum . 5 The outer most layer of the atmosphere is called the exosphere , and it is the first line of defense when it comes to any harmful radiation trying to enter the lower atmospheric levels from outer space . It extends far beyond any other layer with a height upwards of 10,000 km above the earth's surface. The atmosphere is held into place by earth's gravitational field . W ithout

PAGE 11

! 2 gravity , the air we breathe would dissipate into the expanding universe and life would cease to exist . It is the atmosphere that protect s all li fe from the cosmic rays , and is the perfect ether to allow all life to thrive . The gas composition in the troposphere is predominantly composed of nitrogen ( 78.0% ) , oxygen ( 20.9% ), and argon ( 0.9% ) . All other gases are so low that they are often defined in the ppm range. Although n ot considered a gas, water vapor is one of the most i mportant constituents of the atmosphere. Water vapor in large concentrations , or colloquiall y known as clouds , act as earth' s atmosph eric reaction medi ums. T he formation of these reaction epicenters depends highly on particulate s and aerosol s . The colloidal particulates and aerosols function as the center point for water molecules in the atmosphere to condense . 6 Besides cloud condensation nucleation , aerosols can dictate smog levels, ozone levels, influence climate change and overall has a major implication to human health. Aerosols can be classified as suspensions in a gas . T hey can exist discretely as small particles in the liquid or solid phase and they are dispersed and suspended throughout the atmosphere . Aerosols are best characterized com ing from three major sources : marine, crustal and anthropogenic . 7 In addition to the major sources, there are many different species of aerosols as shown in Table 1. Aerosols can be defined as a complex mixture of particles that can be composed of liquid droplets, organic materials such as metals, dust, organic chemicals, and organic acids . 8 The solid phase of aerosols are primarily referred

PAGE 12

! 3 to as particulate matter, which is also known as PM. The sizes of these particles are usually in the micron range , spanning from 0.001 microns to greater than 100 microns . An example of a visible aerosol would be smoke resulting from an incomplete combustion process. These particles of smoke for example, are typically heterogeneous particles and many of them are anthropogenic emissions resulting from human combustion processes . Table 1 Major sources of aerosol species (AR5 of IPCC) Aerosol Species Main Sources Sulfate Primary sulfate: Marine and volcanic emissions. Secondary sulfate: formed by the oxidation of sulfur oxides Nitrate Oxidation of NO x Black Carbon Emitted from fossil fuel combustion, biomass and biofuel burning Organic aerosols Fossil fuel combustion, biomass and biofuel burning, Continental and marine ecosystems. Non combustion sources including biogenic origin Brown carbon Biofuel and biomass burning. Humic like materials of the biosphere origin Terrestrial primary biological aerosol particles Terrestrial ecosystems Mineral dust Wind erosion and suspension of soil. Selected agriculture activities and industrial units Sea Spray Bubble rupturing and wind erosion Marine primary organic aerosol Aerosols injected with sea spray in the biologically active regions of the oceans Because aerosols contribute to such critical function s in the earth 's atmosphere , there are m any efforts put forth to understan d the m. The investigation begins with the formation of aerosols and PM , and ends with their

PAGE 13

! 4 transformation and reactivity in the atmosphere. The human population is so sizeable , that it is now contr ibuting to the global volume s of aerosol . B ecause of the complexity in pollution , it can be fairly difficult to track these anthropogenic emissions . I n particular , the EPA has been focusing much of its environmental effort on PM 2.5, which is the particulate matter in the sub 2.5 micron range. The reasoning behind this range of particles is that the re are extreme health effects associated with its size , which if inhaled in large quantities can lead to morbidity and mortality . 9 I n the last ten years the advancements of measuring aerosols have increased dramatically. From a macro point of view on aerosols, the National Aeronautics and Space Administration h as provided data from two sensors aboard the Terra satellite. The Moderate Resolution Imaging Spectroradiometer and the Multiangle Imaging Spectroradiometer instruments have provided global observations of ae rosols by optical depth , a process of measur ing light extinction by aerosol in the atmospheric column above the earth's surface . 10 These detectors from the Moderate Resolution Imaging Spectroradiometer and the Multiangle Imaging Spectroradiometer instruments are measuring the gross and macro scale amounts of aerosols in the human habitable zone. When tak ing a micro scale point of view on aerosols in o r around the community , there are instruments that can measure the amounts of PM emissions from industrial processes. For example, a major process that requires PM control s is the coal fired power industry. These combustion processes are heavily regulated by the EPA to emit less than 12 micrograms per cubic meter . 11

PAGE 14

! 5 The instruments being employed in the electric utility industry , which apply to the example of coal fired power, are usually located at the emission stack and are measured by a light scattering technique . According to the National Ambient Air Quality Standards , there are already ambient levels we should not exceed with the various PM emissions per the location. These are used as guidelines for any industria l or residential setting , as well as power facilities located in or near ecological sensitive areas. A t able of these limits is shown below using 24 hour time measurement basis. Table 2 Limits of National Ambient Air Q uality Standards (NAAQS) for ambient a erosols (CPCB, 2009) Types of aerosols Time period Concentration in ambient air (!g/m^3) Industrial, Residential, Rural and other area Ecological sensitive area PM10 Annual/24 hours 60/100 60/100 PM2.5 Annual/24 hours 40/60 40/60 Observing even more diminutively , before their inception , information on the reactivity of the components that form aerosols and PM can explain the chemical and physical attributes of the soon to be particles. The reactions that form aerosols can be measured on much more complicated instrumentation , and new methods and approaches have advanced aerosol chemistry to a higher sophistication. Some of the more expensive and costly methods to study aerosol chemistry may involve the use of a mass spectrometer, in which a smog cha mber can be coupled to an aerosol mass spectrometer or a chemical

PAGE 15

! 6 ionization mass spectrometer . 12 An alternative to the mass spectrometer , which is a more availab le method to measure the kinetics and uptake coefficients of certain aerosol formations, can come from samples deposited on inert substrates where the reactants can interact while being measured by IR. In some recent studies , a flow reactor was coupled to an attenuated total reflectance infrared spectrometer (ATR IR). The flow through the reactor was used to study the reaction kinetics of ozone and linoleic acid . 13 This is a relatively cost effective set up where temperature dependence and relative humidity can be observed in conjunction with the reaction kinetics . 13 The ease of these alternative techniques make them advantageous from other costly instruments , and the relatively fast scan in these approaches provide a lot of high r esolution information. Characterizing these reactions with widely available techniques can lead to an in depth understand ing of their formation by a larger scientific community . Aerosol characterization is imperative to distinguish ing the implications of this complex chemistry tha t will affect the biosphere . Aerosol formation can play a critical role in the nucleation of clouds . The physical properties of aerosol s can influence cloud density and the ir reflective properties . Averaged around the globe, anthropogenic sources of aerosols are responsible for around 10% of the total aerosols in our atmosphere . 14 The typical aerosols responsible for c loud condensation nucleation come from natural sources such as dust , sea spray , and volcanoes . T he chemical species mostly associated with natural sources are sulfates, sea salts , and ammonium salts . 15

PAGE 16

! 7 The aerosols can provide a surface for which water can condense and depending on the hydrophilic nature of certain organic compounds there is a corre lation to cloud nucleation. Aerosols have a clear implication to affect convective clouds and precipitation by altering cloud development and latent heat profiles . 6 In order to facilitate the formation of rain the cloud droplets need a threshold radius of 14 ! m . Many of the aerosols are found in the tropos phere where they can be washed out with rainfall and brought back to the ground, however in other cases such as volcano eruptions the aerosols can reach as far as the stratosphere , which is the atmospheric layer above the troposphere. There are some organic fatty acid s that can be found in many of the marine aerosols as well as large urban areas . The physical propert ies of the organic acids are unique as they contain both a hydrophobic and a hydrophilic moiety. Previously aerosols were mainly tho ugh to be mostly inorganic however evidence now shows many aerosols to have a thin organic film. Oleic acid is one of the most common fatty acids found in many plant materials, and is also very prevalent in cooking oils . 16 Of the many organic acids in the atmosphere o leic acid is a long chain single proton carbox y lic acid . However throughout the years in aerosol research it was determined that dicarboxylic acids are some of the most commonly found . 17 Many of these organic acids range in carbon le ngth C2 C10 , and exist primarily as the result of the oxidation of larger volatile organic matter . 18 The focus of this study explores the reactivity of a C4 carbon length dicarboxylic acid commonly known as succinic acid. Because of the small size th e smaller dicarboxylic acids

PAGE 17

! 8 have a low vapor pressure allowing them to travel and condense onto preexisting aerosols. T hese thin film layers can be responsible for drastically changing the behavior of the aerosols abil ity to uptake water and can affect th e optical properties of cloud s making them more reflective. Many of these organic aerosols can increase the threshold of the radius needed to form rain and increase the lifetime of the cloud and suppress rainfall . 7 The hydrophilic propert ies of the organic aerosols can take yet another form as m any of these aero sols can undergo oxidation reactions in the atmosphere . O zone reactions are one of the most common oxidation reactions to occur. Depending on the aerosol substrate the oxidation reaction can drastically change the physiochemical and radiative properties of the aerosol . 13 , 16 In some cases the ozone oxidation of aerosols can decrease the likelihood of cloud condensation nucleation. Having the ability to characterize these heterogeneous reactions would be advantageous in understanding the behavior of aerosols and the ir ability to change key properties of clouds that can directly affect the climate. In urban areas the reactions that can take place o n aerosols can get increasingly complex as the plethora of aerosols being emitted from industry . In the early 1900's the term smog resulting from the contraction of the words smok e and fog was introduced and used mainly in refer ence to coa l combustion products. Smog is mostly anthropogenic and in the more developed countries is now predominantly created by automobile s in addition to industrial emissions . Aerosols resulting from combustion processes have higher concentrations of

PAGE 18

! 9 organic matter and are of high i nterest due to their impacts on global warming. The aerosols resulting from the combustion processes most always contribute to aerosol formation. In such cases pertaining to the transportation industry the soot emissions fro m diesel vehicles have very high carbon content . The high organic content of these aerosols are now though t to rival the global warming effects of CO 2 , where before the black carbon associated with these aerosols were thought to have much greater scatterin g impacts . 19 There are two types of scattering, Rayleigh and Mie. Rayleigh scattering occurs when the wavelength of the incoming radiation is much larger than the size to the particles it comes into contact with. Rayleigh scattering is responsible for the appearance of the blue color in the sky and the color of the sun to appear yellow. T his type of scattering is when the incident electromagnetic radiation hits gases that are smaller than the wavelength of the incoming radiation . 20 Mie scattering is when the particle is much larger than the wavelengths of the incoming electromagnetic ra diation. This type of scattering is responsible for poor visibility and can cause the radiation to scatter in all directions including backwards known as backscatter . 21 , 22 Large concentrations of aerosol particles in the ran ge of 0.1 1.5 microns can scatter visible light and redu c e visibility . I n some locations the amount of these particles can range from 20 200 ! g/m 3 . The annual standard for PM 2.5 is met whenever the 3 year average of the annual mean PM 2.5 concentrations for designated monitoring sites in an area is less than or equal to 15. 0 ! g/m 3 , however this depends on the designation area as characterized by the National Ambient Air

PAGE 19

! 10 Quality Standards (NAAQS) . 8 , 23 There are significant implication s to cli mate change by the introduction of aerosols into the atmosphere. Aerosols exist in two primary locations in the atmosphere, the stratosphere and the troposphere. Because aerosols play a major role as cloud condensation nuclei, the y can greatly influence the formation, optic al properties , and saturation of clouds . 6 Unl ike greenhouse gases aerosols can cause major forces in the climate change in either a cooling or a heating manner. Through solar forcing the irradiance of aerosols at the tropopause can cause major environmental effects either indirectly or directly by scattering and absorption of solar energy and its thermal radiation. In the case of major volcanic eruptions the fine particulate matter and aerosols can actually block and reflect the sun causing a cooling effect. Prime examples of major eruptions large enough to affect the earth's temperature are eruptions such as that of Tambora (1815), Krakatoa (1883) and Agung in (1963), which was associated with global cooling of around a few tent hs of a Celsius. This cooling can last for up to several years after some very large volcanic eruptions , caused by the reflective behavior of the suspended aerosols . 24 On the other end of the temperature spectrum the aerosols can cause heating by preventing a reflective behavior when they are deposited or concentrated in areas suc h as large bodies of water, snow and ice . 7 These depositi ons can affect the albedo of the different surfaces and may increase the adsorption of solar energy. Some of the latest climate changing research uses complex earth system climate models. It is well known that the

PAGE 20

! 11 ocean temperatures are linked to the climate in the near by landmasses . 25 Such correlations h ave linked the Atlantic sea surface temperatures to the droughts in Africa and the rainfall in Brazil . 26 , 27 The power of these models can help predict the macro level of aerosols impact on climate change. However the modeling still uses inputs for aerosol formation such as oxidants, biomass, and carboniferous properties of aerosols . 25 This leads back to the characterization of different aerosols by analytical means. The better we can characterize the physical propertie s of aerosols such as optical properties, ability to absorb moisture, and combine with any large anthropogenic emissions, the better we can fit these macro models to climate change. There has been much research conducted around PM 2.5 lung inhalation, m ost of this research stems from cigarette use and the health effects of cigarette smoking and the implications of second hand smoke . The research clearly shows how PM 2.5 cardio vascular disease (CVD) and lung cancer are linked. The first time the American Heart Association published a scientific statement for air pollution and CVD was in 2004 . 28 The study details the duration of PM 2.5 exposure and the triggers of CVD. The most frightening aspect of the research is the shear numbers of deaths caused by CVD. The reports and research reveal that one in every three deaths in America is caused by CVD . 28 , 29 Although much of the research focuses on cigarette smoke our lungs cannot differentiate the PM 2.5 emissions, and the main emissions leading to a deadly source of PM 2.5 can come from fossil fuels containing arsenic, selenium, mercury, and sulfates from

PAGE 21

! 12 sulfur dio xide. These emissions are creat ed by the combustion of coal, oil, diesel and gasoline, and are concentrated in major cities . 30 The coal and natural gas industries together supply more than 50% of the energy used in America . 31 Larger cities have an increased likelihood that combustion byproducts can elevate to dangerous concentrations of PM 2.5 . 15 In many circumstances the ineffici encies of power transmission cause d the placement of coal fired power plants to be n ear large cities, and e ven if the plants were built further from the heavily populated cities their emissions can travel hundreds and in some cases thousands of miles from their source. With population growth the demand for power is increasing and the deve lopment of homes near large industrial settings are becoming more prevalent. The atmospheric air quality is becoming increasingly complex in the ever growing industrialization of developing societies. These multifaceted reactions can occur naturally howev er many of them are influenced by human emissions and can be concentrated in urban areas, industrial settings and large agricultural areas. Human activities and heavily populated areas are likely to have amplified the influences to air quality. The EPA is designed to regulate harmful environmental factors to human health but with the complexity of the reactions it is becoming increasingly difficult to identify what factors are contributing to the largest impacts in the biosphere. Overall air quality has bee n improving in developed countries since the 1990s because of organizations helping to create public awareness. This can be attributed to growing evidence that there are

PAGE 22

! 13 strong correlations between air quality and human health. Although the improvements ar e progressing in the right direction identifying the constituents of air quality are becoming more demanding. The method proposed in this research is a relatively quick and effective way of determining the reaction kinetics of a two component system along with uptake coefficients . With some a dditional components the proposed tests using the same equipment can investigate the relative humidity effects . These physical properties can be inputs into well established models to obtain the potential for cloud condensation nucleation.

PAGE 23

! 14 CHAPTER II EXPERIMENTAL SECTION The experiment was designed to react liter volumes of clean dry air containing low c oncentrations of vapor over a thin film of organic crystalline solid . This custom set up can be used to simulate the rea ctions contributing to aerosol formation. However the set up its not limited to just exploring aerosol formation. Once the aerosol is fo rmed or other constituents already considered aerosols could be further studied with water vapor or reacted with ozone . The aerosol can then be treated with different relative humidity's to understand the uptake coefficients . The aerosols can also be react ed with ozone to determine the atmospheric ozonolysis products . The flow through chamber can also be combined with equipment to simulate moderate temperatures and pressures. One of the key features of this design is the ability to combine analytical equipment and monitor each set of reaction s on a time based scale . The experimental set up can be best described in three parts, the instrumental analysis , the vapor pathway, the flow through reactor , and an overall detailed procedure of the experiment. U sing a very well known method to analyz e organic molecules, an ATR IR (Nicolet 6700 manufactured by ThermoScientific ) was set up to monitor the uptake of MA onto the thin film SA . During thi s type of experimentation, the ATR IR data can be taken over very brief time interval s to better resolve any kinetic information associated with these organic reactions. The advantage of this set up is the real time capabilities to measure changes within an organic structure .

PAGE 24

! 15 B onding information can reveal exactly what i s happening during the course of the reaction. The temporal spectra generated can be reduced to identify any kinetic information, and from that data the determination of uptake properties can be calculated. Covering a wide range of organic molecules IR is a relatively useful instrument that can analyze gases, liquids and solids. It is a nondestructive test with a high resolution and can complete a scan within a couple of seconds. The experiment was constructed using a custom set up of mass flow controllers (MFC) manufactured by Alicat Scienti fi c and a liquid bubbler produced by SKC INC. Eighty Four, PA used to intr oduce the MA (Alfa Aesar 40 %) as a vapor form into the closed cell flow through reaction cell containing the ZnSe crystal . There are additional components that can be included into this set up to explore the relative humidity uptake and the reaction products from an ozone generator. A schematic of the experiment al set up is shown in Figure 1. ! Figure 1 Schematic of the overall gas movement The plumbing system fully contained inside of a fume hood and begins with a clean dry air source. Air is conveyed through MFC 1 at a volume of five

PAGE 25

! 16 standard cu bic centimeters per minute (sccm ) . To create a MA vapor th e clean dry air is then sent throug h a bubbler containing MA the vapor pressure of MA is fairly high and successful generation of MA vapor is attained easily . The vapor carried out of the bubbler by MFC 1 is now combined with 995 sccm of clean dry air from MFC 2 . The combined flow on both of these MFC's equals a total flow of one standard liter per minute. Those two flow controllers are then connected to a vent and another MFC. The amount of gas flow that doesn't go through the vent tr averses through M FC 3 at a flow rate of 10, 25, and 50 sccm. Depending on that flo w, for example if it is 10 sccm it will be combine with 99 0 sccm clean dry air to make up a total of one standard lite r per minute. The purpose of having this configuration o f flow controllers is to dilute the MA vapor down to the parts per million ( ppm ) range , thus simulating the concentrations of the aerosols that can be present in atmospheric reactions. To determine the reaction kinetics, the MA vapor and the SA is monitor ed in real time using an ATR IR spectrometer. The flow through reactor coupled with the ATR IR is shown in Figure 2 below. The ATR IR is equipped with a mercury cadmium telluride (MCT) detector, which must be cooled with liquid nitrogen. The multireflection sampler uses a zinc selenide (ZnSe) crystal with a refractive index of 2.4 and dimensions of 5 cm in length, 1 cm in width and 0.5 cm in depth. The ZnSe crystal was selected for its large penetration depth, which is estimated to be 1.1 mic rons assuming a 45 ¡ degree angle of incidence .

PAGE 26

! 17 ! Figure 2 Schematic of the flow through reactor coupled w i th the ATR IR The SA deposition was prepared by first making a reagent of SA in a volumetric flask using 1.476 g of SA (99% purity, Aldrich ) and 500 mL of acetone (99 .99 % ACS reage nt grade purity from Fisher Scientific ) creating a 0.025 g/mol solution. The ZnSe crystal was cleaned extensively with a lens wipe and a micropipette was used to transfer 150 ! L of the SA sol ution onto the ZnSe crystal. The acetone evaporates at room temperature and 442 ! g of the SA is deposited on to the ZnSe crystal. The thin film of SA is dispersed sufficient to completely cover the path of the incident beam. The thickness of the film is estimated to be less than 0.1 ! m , which is much less than the penetration de pth of the evanescent wave. The importance of the SA thickness ensures proper probing of the thin film by the IR. Before the experiment is started , nit rogen gas is sent to the ATR IR, this nitrogen gas purges the path of the laser beam used to make the measurements. L iquid nitrogen is also transferred to the MCT detector cell, the working

PAGE 27

! 18 conditions of the MCT need to be operated at very low temperatures to reduce noise. T he d iagnostics signals are monitored using the Experiment Setup featur e. T he Bench and Diagnostics tab reveal the MCT stabilization prior to initializing the analysis . After approximately fifteen minutes the live interferogram signal from the MCT detector is s tabilized and an experiment can begin . The ATR IR is set to collect data every three minutes with 64 scans per data collection. A macro is created to loop 70 data col lections over the course of a 4 hour period equaling one data collection every three minut es. A background of the clean ZnSe crystal is taken prior to depositing any SA onto the ZnSe crystal . This background will be used to blank any artifacts from the ZnSe crystal. A thin film of SA is deposited onto the ZnSe crystal with a micropipette and th e complete evaporation of the acetone occurs in less than 5 minutes. The clean dry air is turned on and sent over to the MFC's and the bubbler containing MA . The MFC's are powered on and the MA vapors are sent through the bypass line. The lines are conditioned for 0.5 hours prior to starting an experiment and the first couple of data collections are taken without the introduction of MA. This establishes the baseline SA spec trum without any interf erence from MA. After three data collections the bypass valve is closed and the valve to send MA vapor over to the flow through cell connected to the ATR IR is opened. The MA vapor traverses through the closed flow through cell containing the ZnSe crystal with a thin film of SA deposited on it. The reaction between the MA vapor and the thin film of SA has a residence time of less than a tenth of a second. T he length of the ZnSe crystal is

PAGE 28

! 19 5 cm and the custom stainless steel chamber has a volume of approximately 1 mL. T o explore the reaction kinetics the amount of MA is introduced in varying concentrations . The mass trans fer of MA onto the SA takes place over the course of 4 hours and is usually completely saturated by the end of the experiment. Depe nding on the flow rates coming from the bubbler containing MA the saturation times can vary. The reaction kinetics and uptake coefficients c an b e slightly different depending on the concentration of MA flowing through the reaction cell . However the test wa s designed to simulate a pseudo first order reaction so the concentration of MA is very dilute compared to the amount of SA . The conditions at which each component of this custom set up are standard temperatures and pressures , 1 atm and 293 K . The entire system was leak checked prior to beginning an y experimentation and a rotometer was installed post reaction cell o n the vent line to verify the one standard liter per minute flow rate . The data was taken using OMNIC TM Spectra Software by Thermo Fisher Scientific. The spectrum was configured to deliver the results in an absorption manner. The program used to write the macro was also OMNIC TM and was set up with a loop sequence to collect sample, delay, and restart the lo op . The data was then manipulated in either the OMNIC TM Spectra Software or E ssential FTIR ¨. Within the OMNIC TM Spectra Software an experimental parameter was set up to average one data point from 64 scans with a resolution of 4 cm 1. The range of IR signals was taken over 650 4000 cm 1.

PAGE 29

! 20 CHAPTER III METHODS Heterogeneous reactions of succinic acid and methyl amine have been examined by attenuated total reflectance infrared spectroscopy (ATR IR). A preparation of succinic acid (SA) was deposited as a thin film onto a zinc selenide crystal (ZnSe). A dilute vapor of methyl amine (MA) was conveyed into a closed cell fl ow reactor containing a thin film deposit of SA. The MA vapor was introduced into the flow reactor via a clean dry air passing through a bubbler containing MA. Throughout the experiment, the reaction was subject to standard temperature and pressures. Over the course of a 4 to 8 hour period, the ATR IR was set to take absorbance readings every 1 to 3 minutes. As the reaction progressed a temporal spectrum was established. An analysis of the temporal spectrum was used to calculate the reaction rates k based o n the modeling of a pseudo first order reaction. In addition to kinetic information, uptake coefficients ! were also acquired based on the kinetic information and the changes in IR signals near the 3445, 1537, 1060 cm 1 range. These IR values correspond to the Ð NH 2 stretching from the primary amine group, the Ð NH 2 bending from the primary amine, and the C N stretching from the carbon bonding with the nitrogen within the MA molecule. As more MA absorbed onto the thin film of SA the ATR IR s pectra began to di splay some of the characteristics associated with the MA organic functionalities. Furthermore, the results revealed some loss in the carbonyl bond characterization in the 1780 1710 cm 1 range caused by an elongation in the carbonyl bonding , an d an increase of the absorbance signals in

PAGE 30

! 21 the 3500 3300 cm 1 range due to the Ð NH 2 stretching adding to the peaks along with the electron density near the Ð OH vibrational stretching. Modern infrared spectroscopy commonly uses the wavenumber as the x axis val ue and either an absorption or transmittance on the y axis. The wavenumber is the number of waves in a length of one centimeter and is given by the following relationship, "=1/# . This unit has a unique linear relationship to with energy. The energy associa ted with infrared radiation can be used to create signature absorptions within an atom or molecule. For a molecule to exhibit these signature absorptions it must be able to possess elec tric dipole features. These absorptions are actually resonant features within the bonding structures that can be modeled by a harmonic oscillator . 32 There are many vibrational modes that can be in the form of symmetric or antisymmetric stretching, scissoring, rocking, wagging, and twisting. Traditional infrared spectrometers have been commercially available since the 1940's. However they relied on prisms to act as dispersive elements, this was eventually replaced with diffraction gradients and then a gain with interferometers . 32 Michelson interferometers are most commonly associated with FTIR and consist of two perpendicular plane mirrors. Where one mirror is stationary and the other one can move in a single plane. Figure 3 below shows the Michelson interferometer. There is inc ident light source directed over to a semi reflective mirror. The mirror effectively splits the beam to where h alf of the light beam travels through the mirror and other half is reflected. Usually o ne path

PAGE 31

! 22 travels over to a reference and the other half travels through a sample cell with the molecules of interest. Both beams are then recombined and are directed to a detector. The movable mirror can move to create an optical path difference between the two arms of the inte rferometer. Figure 3 Michelson Interferometer The purpose of having the movement is to create path differences where the frequencies are interconvertable by the mathematical Fourier transform method, hence the name FT IR. The tw o light beams are passed through a beam splitter that quickly alternates the signals to the detector . T he comparison of the

PAGE 32

! 23 beams can later be resolved with computational software. Because one beam is always directed to a reference and the other beam through a sample it will give relatively accurate data even if the light source drifts over time. Another important feature of FTIR is the versatility of the sample types that can be analyzed. Examples of samples range from gases to liquids, and solids . Th er e are many applications where IR is used, and many of the applications are still being employed to more information about the galaxy . 33 , 34 ! !

PAGE 33

! 24 CHAPTER IV RESULTS AND DISSCUSSION The most likely result of the reaction between MA and SA under the standard temperatures and pressures is the formation of a carboxylate salt. The differences in pH properties are large enough to be the driving force for salt formation . This happens when the amine deprotonates of the carboxylic acid , and by doing so takes on a proton forming a methyl ammonium ion . Because SA has two protons the completion of the reaction will require two moles of MA to every mole of SA. The pK a of SA is 4.2 and 5.6 respectively and both protons will react readily with MA, having a pK a of 10.66 . The result is a stable succinate salt as depicted in Figure 4 . ! Fig ure 4 Conventional reaction of methyl amine and succinic acid in an aqueous medium However as the reaction continues the direct evidence of the methyl ammonium succinate (MAS) product is somewhat masked by the overwhelming amounts of SA . To confirm the formation of the M AS each of the above reactants and products were analyzed separately from one another. The MA was analyzed as a thin film liquid dispersed directly from the reagent bot tle onto the ZnSe cryst al and t he SA was analyzed alone as a thin film solid deposit as described in the

PAGE 34

! 25 above experimental section . T he two ATR IR spectra are displayed in Figure 5 below. ! Figure 5 MA (red) and SA (black) each analyzed alone using an ATR IR The main IR assignments for MA stem from the amine functionality, the NH 2 stretching near 3280 cm 1 , NH 2 scissor vibration near 1670 cm 1 , and the C N stretching near 1 334 cm 1 . The main assignments from SA revolve around the carboxylic acid functionality. The very broad OH stretch between 3000 2500 cm 1 , C=O stretching near 1675 cm 1, and the C O stretch near 1196 cm 1 . The IR assignments are listed in Table 3 below with a designation of functional group, peak location, and peak intensity. ,! ,#%! ,#(! ,#)! ,#*! "! "#%! "#(! "#)! "#*! !"#$ %%"#$ %!"#$ &%"#$ &!"#$ '%"#$ '!"#$ -./01.2345! 6275389.51!:49;";?>!@A54B12! C-! @-!

PAGE 35

! 26 Table 3 IR assignments for MA and SA, br = broad, s = strong, m = medium, w = weak Functional Group Assignment Peak Position Wavenumber [cm 1 ] Peak Intensity NH2 stretching 3438.4 br NH2 bending 1534.7 s C N stretching 1334 .5 m OH stretching 3000 2500 w C=O carboxylic stretching 1720.9 s C O 1410.6 m To carry out the MAS product evaluation , a liquid Ð liquid reaction of MA and SA was accompanied by an ATR IR analysis . In the liquid Ð liquid reaction , 0.3103 g of SA was reacted with 0.1645 g of MA in an aqueous medium. A heat evolution and a white crystalline precipitate followed the resulting reaction . Considering the two spectra in Figure 5 the prediction of where the formati on of MAS will fall in the ATR IR analysis is unsurprising. In Figure 6 below the MAS is displayed with an overlay of SA to forecast of where the peaks are expected to arise from the reaction of MA vapor with the overpowering amount of SA.

PAGE 36

! 27 ! Figure 6 ATR IR spectra of SA (red) and the MAS product (black) In an attempt to force the reaction to completion the flow rate of MA was increased from 10, to 25, and to 50 sccm . Increasing the flow rate effectively increases the overall concentration of MA and the exposure of MA to SA . With varying concentrations of MA flowing into the closed cell reactor the most dramat ic spectra was displayed with the highest concentration of MA . To verify reproducibility and variability e ach concentra tion was directed a minimum of 3 times. The lower the concentration the longer the reaction takes to complete and alternatively the higher the concentration the shorter it takes to complete the reaction. Taking this into account, t o make sure the data bein g captured was sufficient for the proper kinetic modeling and uptake coefficient calculations the OMNIC macro program was set to take 64 scans every 3 minutes all the way down to 64 scans every minute , and t he length of the measurements varied from 4 to 8 hrs . After the experiment wa s carried out three times at each ,! ,#%! ,#(! ,#)! ,#*! "! "#%! "#(! "#)! !"#$ %%"#$ %!"#$ &%"#$ &!"#$ '%"#$ '!"#$ -./01.2345! 6275389.51!:49;";?>!@A54B12! C-@! @-!

PAGE 37

! 28 concentration, the reaction kinetics were analyzed, uptake coefficients were calculated and furthermore the reaction rates were examined as a function of MA concentration. The experiment began w ith the high est concentration of MA introduced into the system, as this was the securest condition for the reaction to form the desired products. T he ATR IR spectra in Figure 7 show SA with no exposure to MA and the final spectra of the SA w ith the longest expos ure to MA. The spectrum in Figure 6 has provided a forecast of are as in the spectrum below where the MAS formation can be identified with lower interference from the overpowering SA peaks . ! Figure 7 ATR IR spectrum of SA with a 50 sccm flow through of MA Besides the easily distinguished carboxylic C =O stretching vibration at 1710 cm 1 o ther featur es tha t can be attributed to the SA IR spectrum are the CH 2 ,! ,#%! ,#(! ,#)! ,#*! "! "#%! "#(! "#)! "#*! !"#$ %%"#$ %!"#$ &%"#$ &!"#$ '%"#$ '!"#$ -./01.2345! 6275389.51!:49;";?>!@A54B12! @-! @-;C-!

PAGE 38

! 29 antisymmetric stretching observed at 2919 cm 1 and the CH 2 symmetric stretching at 2850 cm 1 . In addition to those stretching vibrations the CH 2 moiety also exhibits a bending at 1410 cm 1 and 1306 cm 1 while its wagging vibration is shown at 1197 cm 1 . Some other unique characteristics of the SA spectrum comes from its carboxylic acid functionality, the broad Ð OH stretch around 2500 3300 cm 1 and the Ð COH out of plane bending at 910 cm 1 as shown in the black spectrum labeled SA in Figure 7. Not only showing SA by itself Figure 7 shows the end res ult of the reaction where MA has deposited and reacted with the thin film of SA to a relatively complete saturation . Nevertheless, t he re is an inconvenience of identifying the contributions from SA and MA in the IR spectrum due to the fairly puzzling overlap of peaks. To help identify peaks in the above reaction each compound was not only analyzed alone but also accompanied by reference IR spectrum s from the NIST W ebbook . With a forecast of w h ere the MAS peaks are located t he clearly identifiable peaks coming from the MA vibrations are the Ð NH2 stretching vibrations at 3438 cm 1 th e Ð NH2 bending vibration at 1537 cm 1 and the C N vibrational stretch at 1068 cm 1 . The d ocumentation of the infrared vibration assignments of SA saturated with MA are i dentified in Table 4 below and correlate to the red IR spectra labeled SA MA in Figure 7 .

PAGE 39

! 30 Table 4 IR feature assignments of SA MA at room temperature Functional Group Assignment Peak at Wavenumber [cm 1 ] Absorbance Peak Height NH2 stretching 3438.4 0.4 CH2 antisymmetric stretching 2919.6 0.6 CH2 symmetric stretching 2850.2 0.6 C=O carboxylic stretching 1720.9 0.9 NH2 bending 1534.7 0.6 CH2 bending 1410.6 1.1 CH2 bending 1306.7 1.1 CH2 wagging 1197.0 1.3 C N stretching 1068 .5 0.5 COH carboxylic out of plane bending 910.3 1.1 When the lower concentrations of MA were reacted with SA the same resulting spectra formed but the temporal absorbance profiles changed and will be discussed in detail below. As the MA i s introduced into the flow through reactor containing the thin film of SA, the spectrum began to take on some characteristics of MA. This is most prominently detected as new peaks arise from the MA depositing and reacting with the SA surface. However the separation of the spectral peaks becomes increasingly difficult as the SA and the MA have absorbance bands that overlap. To overcome this commonality in absorption bands between the SA and MA , isolated functional groups specific to each compound were identified and the c hange in abs orbance wa s used to calculate the reaction rate s . The most notable changes are found in the regio ns where the

PAGE 40

! 31 SA spectrum shows very low absorbance peaks and where the MA shows strong absorbance as identified in the spectra labeled SA MA in Figure 7 . Prima ry amines exhibit vibrational stretchi ng with two bands in the 3500 33 00 cm 1 range, these peaks are often very weak. Fortuitously the IR spectrum of SA has a large broad OH stretch with a peak cut off of 3300 cm 1 . As the MA further absorbs onto the SA the development of one of the weak Ð NH 2 vibr ational stretches occurs at 3438 cm 1 . T his falls outside of the broad Ð OH stretching vibration flaunted by SA and is clearly evident as the reaction is complete . Another area where the primary amine's character is revealed can be found in the development of a n Ð NH 2 scissor vibration at 1534 cm 1 . The method for deriving the uptake coefficient employed in the present work is similar to the methods used in previous studies of ozone reacting with oleic acid, linoleic acid, and squalene . 13 , 35 , 36 The details of this approach was developed in research elsewhere and follows an absorption model . 37 The derivation of the upt ake coefficient ! stems from the calculation of each reaction rate. T he analysis began with the modeli ng of a second order reaction. However, establishing the rate constant of a bimolecular reaction can be challenging. Considering Equation 1 below the rate constant can become pragmati c and in most cases needs to be determined experimentally. Typical methods to determine the reaction rate constant will employ an experimental method where both reactants can be introduced into the reaction in different c oncentrations . The

PAGE 41

! 32 difficulty in th is practice is that both reactants need to be measured simultaneously . ! ! ! ! ! Equation 1 Basic bimolecular reaction Because of the challenge in measuring the two reactants in Equation 1 simultaneously a common solution is applied. The pseudo first order approximation can resolve the complications in measurement by supplying one of the reactants in great excess. If the experimental reaction is designed to be a pseudo first order reaction and the concentration of one reactant is in great excess the bimolecular reaction can take the form of E quation 2 where the [SA] is in great excess over the reactant [MA]. This form is a result of the [SA] conce ntration remaining mostly unchanged during the course of the reaction pathway. The m ain measurement is the character development of [MA] taking shape over time with k 1 denoting the pseudo first order rate constant. ! ! ! ! ! !" ! ! ! !" ! ! ! !"! ! ! ! !" ! ! ! !" ! ! ! Equation 2 Pseudo first order rate T he uptake coefficient can be estimated with the changes in the IR absorbance profiles over time . 38 The most discernible ab sorbance prof iles will be taken from the distinct wavenumbers of the a mine 's NH 2 vibrational stretch at 3438 cm 1 , the Ð NH 2 bending at 1534 cm 1 and the C N stretching at 1068 cm 1 . Moreover the peak development at each of these wavenumbers will be associated with the absorbance data to estimate the uptake coefficient ! . The temporal response of

PAGE 42

! 33 the absorban ce peaks will be plotted using E quation 3 below. Where B is the absorbance difference at time zero and k is the pseudo first order rate coefficient . To avoid any errors in baseline shifting the absorbance value B will be integrated absorbance values at time t as calculated from the baseline . 37 , 39 As the data is plotted an exponential fit wi ll be applied and the reaction rate k will be determined in Equation 3 . With the changes in MA concentration the most discernable peaks are still developed however the amo unt each temporal integrated absorban ce increase varies. The higher the concentration of MA used in the reaction the larger the intensification between absorbance measurements per the same amount of time. !" ! ! ! ! ! ! !" ! Equation 3 Exponential function used to model reaction The peak development at each wavenumber was manipulated using the Essential FTIR ¨ software. An example of the peak evolution during the exposure of MA onto a SA thin film is shown in Figure 8 . The data begins with a relatively unexciting line and then begins to take shape as the Ð NH 2 vibrational stretching grows into a well defined peak. As mentioned above the same peak arises from the different MA concentrations however with a different temporal response th at can be best explained using E quation 3 to m odel the data.

PAGE 43

! 34 ! Figure 8 FTIR spectra of SA thin film at different exposure times to a 50 sccm flow rate of MA focusing on 3445 3432 cm 1 The chronological changes in absorbance that created uninhibited and well defined peaks were shown in Figure 9 where the x axis is the time in minutes and the y axis is the integrated absorbance at time zero . It is from these plots that the reaction rate c onstant can be determined. When changing the MA concentrations as a function of flow the reaction rate k also changes. Several experiments were carried out using variable MA concentration s and the resulting reaction rates are directly proportional to the concentration of the MA . Each plot discloses the function coefficients to fit the data with a greater than ± 95% confidence interval . The integrated absorbance values were fitted into an exponential function using Igor Pro v7.01. The relatively good fit of the integrated temporal absorbance data to the function in Equation 3 confirm s the application of a pseudo first order reaction rate model to be appropriate for the estimation of the uptake coefficients.

PAGE 44

! 35 ! Figure 9 Temporal changes in the infra red spectrum focusing on the NH 2 vibrational stretch , reaction was carried out using 50 sccm of MA There was a fairly good agreement of calculated reaction rates k 1 using the integrated absorbance peaks in the above plot from 3445 3432 cm 1 . Analyzin g relatively the same location in another reaction using half of the MA concentration Figure 10 shows a decreased rate constant k 2 .

PAGE 45

! 36 ! Figure 10 Temporal changes in the infrared spectrum focusing on the NH 2 vibrational stretch, reaction was carried out using 25 sccm of MA As shown in F igure 11 the rate constant k 2 was decreased as the concentration of the MA in the reaction was lowered. Compared to the reaction rate k 1 using 50 sccm the reaction rate using 25 sccm k 2 was nearly cut in half. The reaction was carried out again at 10 sccm and the data was analyzed using the same method above however the reaction at 10 sccm was much slower than the previous concentrations of MA. The data in Figure 11 shows the difficulty of measuring the lowest concentration of MA. The data doesn't fit as well as the previous two concentrations and k 3 may have a higher variability due to several reasons. One potential reason is the MFC working at such a l ow flow rate increasing the noise and inconsistency of the MA delivery .

PAGE 46

! 37 Figure 11 Temporal changes in the infrared spectrum focusing on the NH 2 vibrational stretch, r eaction was carried out using 10 sccm of MA To verify the reaction rate k, FTIR spectra was examined in all of the absorbance bands where peak changes were uninhibited by the excess amounts of SA. In an additional observation and t o confirm the reaction kinetics calculated from the 3455 3432 cm 1 re gion another pea k area was integrated from 1537 1528 cm 1 , Figure 12 .

PAGE 47

! 38 ! Figure 12 FTIR spectra of thin film SA at different exposure times to 50 sccm of MA focusing on 1537 1528 cm 1 The IR spectrum at this particular location displays a shift in several locations but the most notable shift occurs at the peak between 1537 1528 cm 1 . The bar that highlights the peak is the area where each spectra was integrated. Figure 13 Temporal changes in the infrared spectrum focusing on the NH 2 scissor vibrational stretch , reaction was carried out using 50 sccm of MA

PAGE 48

! 39 This reaction rate analysis was carried out with all of the MA concentrations 10, 25, and 50 sccm in both peak areas. H owever, t he reaction rate of the high concentration MA exposure k 1 from each of the above peak locations has a relatively good correlation and confirms the pseudo first order rate environment. For heterogene ous reactions involving a gas Ð solid or a gas Ð liquid reaction the likelihood of the reaction occurring will depend on the surface reaction probability also kn own as the uptake coefficient . 38 The mean speed of the MA molecules in the gas phase coming from the contact with the thin film of SA will be measured by ATR IR and the surface area to volume ratio of the SA thin film will be used to estimate the geometric surface of the reaction. Using Equation 2 where the rate k 1 is calculated from the t emporal absorbance changes and the concentration of [S A] is equal to the initial concentration. T he uptake coefficie nt can be solved for using Equation 4. ! ! ! !" ! !" ! ! ! ! ! !" ! ! !" ! ! ! ! Equation 4 Uptake coefficient The rate c onstant k 1 multiplied by the initial concentration of [ SA ] is equa l to the uptake coefficient multiplie d by the MA pressure P MA ; c the mean speed of the MA particles in the gas phase [cm s 1 ]; R is the gas constant, T is the temperature; and S A the surface area to V volume ratio of the SA thin film [cm 1 ], where the surface area is estimated to be about 4.5 cm 2 based on the available working surface area of the ZnSe crystal used in the flow through reactor. In the work presented here the uptake c oefficient was calculated using the rate

PAGE 49

! 40 constants derived from each of the MA concentrations 10, 25, and 50 sccm and exploiting two of the three absorbance bands that were not obscured by the SA absorbance peaks, the NH 2 vibrational stretch at 3445 cm 1 and the Ð NH 2 bending at 1537 cm 1 as shown in Table 5 . ! The standard deviation was calculated as a percentage from the average. As mentioned above the lower the concentration of MA the higher the variability. Table 5 Summary of pseudo first order r ate con stants k, derived from the temporal integrated absorbance taken at different locations, and the corresponding calculated uptake coefficients Physical State MA flow rate [sccm] Reaction rate k1 x 10 3 [s 1] taken at different peak locations Uptake coefficient " x10 3 per the peak location 3455 3432 cm 1 1537 1528 cm 1 3455 3432 cm 1 1537 1528 cm 1 Solid/Vapor 10 0.187 0.327 0.182 0.317 Solid/Vapor 10 0.219 0.219 0.212 0.212 Solid/Vapor 10 0.130 0.063 0.126 0.061 AVG 10 sccm 0.179 0.203 0.173 0.197 % STDEV 10sccm 20.49% 53.37% 20.49% 53.37% Solid/Vapor 25 0.423 0.546 0.411 0.530 Solid/Vapor 25 0.650 0.568 0.630 0.551 Solid/Vapor 25 0.640 0.623 0.621 0.604 AVG 25 sccm 0.571 0.579 0.554 0.562 % STDEV 25 sccm 10.45% 3.22% 10.13% 3.12% Solid/Vapor 50 1.225 1.085 1.188 1.052 Solid/Vapor 50 1.003 1.580 0.972 1.532 Solid/Vapor 50 1.338 1.238 1.297 1.200 AVG 50 sccm 1.189 1.301 1.153 1.262 % STDEV 50 sccm 13.92% 20.69% 13.50% 20.06% Overall the two areas of peak devel opment were in good agreement, however the second N H 2 vibrational stretch near 1 537 cm 1 may have a higher level of interference due to the location of the signal in the region of the IR field. When

PAGE 50

! 41 comparing the data to literature values of MA uptake onto ammonium salts by Liu et al , 2 the reaction rates are in same order of magnitude however the application is not quite the same . 2 In addition to the com parison of reaction rates at different peak integration, the reaction rates were also plotted to determine their dependency on the MA concentration, Figure 14. ! Figure 14 Reaction rates of each peak analysis as a function of MA concentration The reaction rates as a function of MA concentration shows a relatively good fit w ith both sets of data from each respective peak analyzed having a linearity R 2 correlation greater than 0.9 . Although t he less inhibited peak near 3445 cm 1 demonstrate s a better correlation, which can be due to much less interference from overlapping peaks in the finger print region. Likewise, the lower D!E!,#,%$%F!;!,#,)&&! >G!E!,#+(%%*! D!E!,#,%&)F!;!,#,*&$! >G!E!,#+")*$! ,#,,,! ,#%,,! ,#(,,! ,#),,! ,#*,,! "#,,,! "#%,,! "#(,,! "#),,! "#*,,! ,! $! ",! "$! %,! %$! ',! '$! (,! ($! $,! $$! ),! !"#$%&'()!#%")*)+,-./) 01)23'4)!#%")+,$$5/) !"#$%&'()!#%",) '($$!;!'('%!49;"! "$'&!;!"$%*!49;"!

PAGE 51

! 42 concentration of MA contains variability from the MFC limitations, and with the lower the MA concentration eventually all reaction rates will converge to zero.

PAGE 52

! 43 CHAPTER V CONCLUSION The above study applies the use of an ATR IR to characterize heterogeneous reaction rates and uptake coefficients of a thin film solid phase SA and a vapor phase form of MA. T he kinetics of the reaction was found to model a pseudo first order rate with a correlation coefficient R 2 greater than 0.95. T he overall uptake coefficients were verified by conducted calculations on two separate peak locations near 3445 and 1537 cm 1 . Using the highest concentration of MA the two peaks analyzed were observed to b e w ithin less than 10 % of each other. The reaction rates were presented to have a relatively high level of precision as they also correlated well to the MA concentration . The advantages of this experimental set up prove the capability to explore the dependency of reaction rates on MA concentrations. Reproducibility and the repeatability between individual analysis are less than 25% of the average value, excluding the lower MA concentration runs. A lthough the predicted re action between SA and MA to form the di methyl ammonium salt was difficult to decipher there was strong evidence of MA S formation as shown in the Figure 7 overlay. T he main learning from the research is the demonstr ation of a high throughput method to char acterize the formation of potential aerosols as a result of organic acids reacting with amines . The importance of this technique stems from the need to understand the ever changing atmospheric environment. With climate change arguable being influenced by t he growing human population it is advantageous to couple a physical analysis of aerosol formation with other macro

PAGE 53

! 44 spectroscopy approaches. Understanding the physical characteristics on a micro level will shed light into multiple areas, including but not l imited to aerosol formation , cloud condensation nucleation, understanding anthropogenic emission sources influencing climate change, and the changes in the atmosphere ultimately effecting the health of plants and animals.

PAGE 54

! 45 REFERENCES (1) Kawamura, K.; Bikkina, S. Atmos. Res. 2016 , 170 , 140 Ð 160. (2) Liu, Y.; Han, C.; Liu, C.; Ma, J.; Ma, Q.; He, H. Atmos. Chem. Phys. 2012 , 12 , 4855 Ð 4865. (3) Fallis, A. . J. Chem. Inf. Model. 2013 , 53 , 1689 Ð 1699. (4) Lewis, A.; Carslaw, N.; Marriott, P.; Kinghorn, R.; Morrison, P.; Lee, A.; Bartle, K.; Pilling, M. Nature 2000 , 405 , 778 Ð 781. (5) Way, M.; Eather, R. 2012 . (6) Li, Z.; Niu, F.; Fan, J.; Liu, Y.; Rosenfeld, D.; Ding, Y. 2011 , 4 . (7) M anzoor, S.; Kulshrestha, U. 2015 , 10 , 738 Ð 746. (8) U.S. Epa. U.S. Environ. Prot. Agency, Off. Air Qual. Plan. Stand. Res. Triangle Park. North Carolina 2011 , EPA 452/R . (9) Arden Pope, C.; Burnett, R. T.; Turner, M. C.; Cohen, A.; Krewski, D.; Jerrett , M.; Gapstur, S. M.; Thun, M. J. Environ. Health Perspect. 2011 , 119 , 1616 Ð 1621. (10) van Donkelaar, A.; Martin, R. V.; Brauer, M.; Kahn, R.; Levy, R.; Verduzco, C.; Villeneuve, P. J. Environ. Health Perspect. 2010 , 118 , 847 Ð 855. (11) Weinhold, B. Env iron. Health Perspect. 2013 , 121 , 2013. (12) Rudich, Y.; Donahue, N. M.; Mentel, T. F. Annu. Rev. Phys. Chem. 2007 , 58 , 321 Ð 352. (13) Zeng, G.; Holladay, S.; Langlois, D.; Zhang, Y.; Liu, Y. J. Phys. Chem. A 2013 , 117 , 1963 Ð 1974. (14) Kahn, R. Earth Obs. 2010 , November 1 . (15) Valavanidis, a.; Fiotakis, K.; Vlachogianni, T. Urban Airborne Particulate Matter: Origin, Chemistry, Fate and Health Impacts ; 2011. (16) Schwier, A. N.; Sareen, N.; Lathem, T. L.; Nenes, A.; Mcneill, V. F. 2011 , 116 , 1 Ð 12.

PAGE 55

! 46 (17) Rogge, W. H. L. G. C. 1997 , 32 , 13 Ð 22. (18) Lightstone, J. M.; Onasch, T. B.; Imre, D.; Oatis, S. J. Phys. Chem. A 2000 , 104 , 9337 Ð 9346. (19) Saleh, R.; Robinson, E. S.; Tkacik, D. S.; Ahern, A. T.; Liu, S.; Aiken, A. C.; Sullivan, R. C.; Prest o, A. a.; Dubey, M. K.; Yokelson, R. J.; Donahue, N. M.; Robinson, A. L. Nat. Geosci. 2014 , 7 , 1 Ð 4. (20) Chakraborti, S. 2008 , 2 Ð 4. (21) Cox, a. J.; DeWeerd, A. J.; Linden, J. Am. J. Phys. 2002 , 70 , 620. (22) Kostylev, V. I. Bistatic Radar Princ. Pract. 2007 , 193 Ð 223. (23) Rule, F. 78 Fed. Regist. 3086 2013 , 78 , 1 Ð 27. (24) Hansen, J.; Lacis, A. Nature 1990 , 346 , 183 Ð 187. (25) Booth, B. B. B.; Dunstone, N. J.; Halloran, P. R.; Andrews, T.; Bellouin, N. Nature 2012 , 484 , 228 Ð 232. (26) Ackerley, D.; Booth, B. B. B.; Knight, S. H. E.; Highwood, E. J.; Frame, D. J.; Allen, M. R.; Rowell, D. P. J. Clim. 2011 , 24 , 4999 Ð 5014. (27) Knight, J. R.; Folland, C. K.; Scaife, A. A. Geophys. Res. Lett. 2006 , 33 , 1 Ð 4. (28) Brook, R. D.; Rajagopala n, S.; Pope, C. A.; Brook, J. R.; Bhatnagar, A.; Diez Roux, A. V.; Holguin, F.; Hong, Y.; Luepker, R. V.; Mittleman, M. A.; Peters, A.; Siscovick, D.; Smith, S. C.; Whitsel, L.; Kaufman, J. D. Circulation 2010 , 121 , 2331 Ð 2378. (29) American Heart Association; American Stroke Association. 2012 . (30) Pope, C. A.; Turner, M. C.; Burnett, R. T.; Jerrett, M.; Gapstur, S. M.; Diver, W. R.; Krewski, D.; Brook, R. D. Circ. Res. 2015 , 116 , 108 Ð 115. (31) EIA. U.S. Energy Inf. Adm. 2015 . ( 32) Stuart, B. H. Infrared Spectroscopy: Fundamentals and Applications ; 2004; Vol. 8. (33) Leisawitz, D.; Rinehart, S. A. .

PAGE 56

! 47 (34) Steele, J. D. 1885 , 9 Ð 16. (35) Stewart, D. J.; Griffiths, P. T.; Cox, R. A. Atmos. Chem. Phys. 2004 , 4 , 1381 Ð 1388. (36) Fu, D.; Leng, C.; Kelley, J.; Zeng, G.; Zhang, Y.; Liu, Y. Environ. Sci. Technol. 2013 , 47 , 10611 Ð 10618. (37) Segal Rosenheimer, M.; Dubowski, Y. J. Phys. Chem. C 2007 , 111 , 11682 Ð 11691. (38) Akimoto, H. Atmospheric Reaction Chemistry ; 2015. (39) Hung, H. M.; Katrib, Y.; Martin, S. T. J. Phys. Chem. A 2005 , 109 , 4517 Ð 4530.