UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Heterogeneous reactions between nitrate radicals + organic coatings and nitrate radicals + soot Mak, Jackson 2006

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
831-ubc_2006-0249.pdf [ 3.67MB ]
Metadata
JSON: 831-1.0059290.json
JSON-LD: 831-1.0059290-ld.json
RDF/XML (Pretty): 831-1.0059290-rdf.xml
RDF/JSON: 831-1.0059290-rdf.json
Turtle: 831-1.0059290-turtle.txt
N-Triples: 831-1.0059290-rdf-ntriples.txt
Original Record: 831-1.0059290-source.json
Full Text
831-1.0059290-fulltext.txt
Citation
831-1.0059290.ris

Full Text

HETEROGENEOUS REACTIONS BETWEEN NITRATE RADICALS + ORGANIC COATINGS AND NITRATE RADICALS + SOOT by Jackson Mak B.Sc , The University of British Columbia, 2003 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF M A S T E R OF SCIENCE in THE F A C U L T Y OF G R A D U A T E STUDIES (CHEMISTRY) THE UNIVERSITY OF BRITISH C O L U M B I A April 2006 © Jackson Mak, 2006 A B S T R A C T The reactive uptake of nitrate radicals (NO3) by self-assembled monolayers (SAM) and methane soot was investigated using a flow tube reactor coupled to a chemical ionization mass spectrometer (CIMS) as the detector. The detection of N O 3 was accomplished by using I" as the reagent ion, and the rate constant for the reaction of N 0 3 + I" was determined to be (6.5 ± 5.0) x 10"'° cm 3 • molecule"' • s"'. The performance of the experimental approach was evaluated by measuring the rate constant for the bimolecular reaction between N O 3 and NO. The rate constant for this reaction was determined to be (2.7 + 0.2) x 10"" cm3 - molecule"' -s"', which is in excellent agreement with the literature data. Self-assembled monolayers were used as proxies for organic aerosols. The reactive uptake coefficients were determined to be (8 ± 3) x 10"4 for octadecanethiol (ODT) on gold and (1.3 + 0.2) x 10"3 for octadecyltrichlorosilane (OTS) on glass. Although calculations show the loss of N O 3 to organic coatings is of minor importance as a sink of N O 3 , they do, however, suggest N O 3 is as efficient as OH in processing surfaces of saturated hydrocarbons. In addition, the surface of an OTS monolayer was probed with Fourier transform infrared (FTIR) spectroscopy prior to and following exposure to N O 3 . The spectra indicate there was no fragmentation of the monolayer. Heterogeneous reactions between N O 3 and methane soot were investigated. From the experimental results, a lower limit of 0.03 was determined for the reactive uptake coefficient for this reaction. Calculations show the depletion of N O 3 by soot may be an important N O 3 sink in the atmosphere in urban areas. Furthermore, infrared (IR) absorption spectra show significant changes to the soot surface after exposure to N O 3 . Functional groups such as epoxides, carbonyls, organic nitrites and nitrates were observed after treatment. 11 T A B L E O F C O N T E N T S ABSTRACT TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES ACKNOWLEDGEMENTS DEDICATION 1. INTRODUCTION 1.1 Introduction 1.2 Organic Aerosols 1.3 Aqueous Inorganic Aerosols Coated with Organic Films 1.4 Soot 1.5 Nitrate Radicals and Their Gas Phase Chemistry 1.6 Aerosols and Radicals 1.7 Thesis Overview 2. GENERAL EXPERIMENTAL 2.1 Introduction 2.2 N 0 3 Production 2.3 Experimental Apparatus for Measuring Reactive Uptake of N 0 3 2.4 Detection Method 2.5 Materials 2.6 Surface Preparation 2.7FTIR Spectroscopy : 3. CHARACTERIZATION OF THE FLOW SYSTEM 3.1 Introduction 3.2 Reagent Ion for Detecting N 0 3 3.3 Incomplete Dissociation of N 2 O 5 3.4 Recombination of N O 3 with N O 2 3.5 Determination of the Rate Constant for NO3 +1" 3.6 N O 3 Loss Processes 3.7 Validation of the Flow Tube - CIMS Apparatus and N 0 3 Source 3.8 Summary and Conclusions 4. HETEROGENEOUS REACTION BETWEEN NO3 + SAM 4.1 Introduction 4.2 Experimental 4.3 Data Analysis 4.4 Results and Discussion 4.5 Exposure Studies 4.6 Proposed Mechanism 4.7 Difference Between N 0 3 and OH Exposure Studies 4.8 Atmospheric Implications ; 4.9 Summary and Conclusions 5. H E T E R O G E N E O U S R E A C T I O N B E T W E E N N 0 3 + SOOT 58 5.1 Introduction 58 5.2 Experimental 59 5.3 Results and Discussion 60 5.3.1 N 2 0 5 + Methane soot 60 5.3.2 N 0 3 + Methane soot 63 5.4 Exposure Studies 65 5.5 Possible Reaction Mechanisms 69 5.6 Atmospheric Implications. 72 5.7 Summary and Conclusions 74 6. CONCLUSIONS A N D F U T U R E DIRECTIONS 75 6.1 Conclusions 75 6.2 Future Directions 76 7. R E F E R E N C E S 77 iv LIST OF TABLES Table 1.1 Typical soot concentrations found in various regions. 4 Table 3.1 Typical residence times in various regions of the flow system. 26 Table 3.2 N O 3 reactions and their rate constants at 298 K and 443 K (A pressure of 2 Torr is used where relevant). 35 Table 4.1 Summary of reactive uptake coefficients measured in this study. 44 Table 4.2 Comparison of N O 3 heterogeneous loss rate, Rneh by reaction with organic coatings to N O 3 homogeneous loss rate, RHOM, by reaction with a-pinene for different environments. 54 Table 4.3 Processing times for various atmospheric species to process a simulated organic coated aerosol particle with a surface concentration of 101 4 molecule-cm-2. 56 Table 5.1 Assignment of vibrational bands for methane soot exposed to N O 3 . 69 Table 5.2 Comparison of N O 3 heterogeneous loss rate, RHet, by reaction with soot to N O 3 homogeneous loss rate, Rnom, by reaction with a-pinene for different environments. 72 v LIST OF FIGURES Figure 2.1 Apparatus for the synthesis of N 2 O 5 . 11 Figure 2.2 Schematic of the experimental apparatus. 14 Figure 3.1 Observation of the 62 m/z signal (solid line) in real time using I" as the reagent ion in CIMS. The dashed line shows the external oven temperature history. 19 Figure 3.2 Observation of the 62 m/z signal (solid line) in real time using I" as the reagent ion in CIMS. The dashed line shows the external oven temperature history. The shaded regions indicate when excess NO was present. 21 Figure 3.3 Observation of the 46 m/z signal (solid line) in real time using SF6~ as the reagent ion in CIMS. The dashed line shows the external oven temperature history. The shaded regions indicate when excess NO was present. 23 Figure 3.4 Plots of percentage of N 2 O 5 dissociated versus time for various oven temperatures and at a fixed pressure of 5 Torr. 25 Figure 3.5 Plots of a) [N2O5] formed and b) percentage of N O 3 unreacted as a function of time for the combination reaction of N 0 3 + N 0 2 + M —> N 2 0 5 + M with initial N O 3 and N 0 2 concentrations equal to 1010 (dashed) and 1012 (solid) molecule-cm"3 at 298 K and 5 Torr. 27 Figure 3.6 Observation of the 127 m/z signal (solid gray line) and 62 m/z signal (solid black line) in real time using I" as the reagent ion in CIMS. The dash line shows the external oven temperature history. The shaded regions indicate when an excess amount of NO was present. 31 Figure 3.7 Concentrations of N 2 O 5 (solid black line), N O 2 (dashed gray line), and N 0 3 (dotted black line) in the flow system were simulated for the N 0 3 + I" rate determination study using a model with initial N 2 O 5 concentration equals to 2.0 x 10 molecule-cm and oven temperature equals 298 K. The double-headed arrows at the top indicate the time spent by the gas phase species in each of the indicated region in the flow system. 33 v i Figure 3.8 Concentrations of N 2 O 5 (solid black line), N O 2 (dashed gray line), and N O 3 (dotted black line) in the flow system were simulated for the N O 3 + I" rate determination study using a model with initial N 2 O 5 concentration equals to 2.0 x 10" molecule-cm"3 and oven temperature equals 443 K. The double-headed arrows at the top indicate the time spent by the gas phase species in each of the indicated region in the flow system. 34 Figure 3.9 Observation of the N O 3 signal (solid line) in real time at various injector positions (0 - 10 cm), which correspond to different reaction times between N O 3 and NO. The shaded region indicates when excess NO is present to completely react away all N O 3 to determine the background signal. 37 Figure 3.10 Examples of decay of N O 3 signal as a function of injector distance for various NO concentrations. 38 Figure 3.11 Plot of first-order loss rate coefficient versus NO concentration. The slope yields the second-order rate constant for the reaction of N O 3 + NO and has a value of (2.7 ± 0.2) x 10'" cm3-molecule"1-s"1. 40 Figure 4.1 Plots of natural log of N O 3 signal versus reaction time for the reactive uptake of N O 3 by bare gold (#), bare glass ( A ) , ODT monolayer on gold (*), and OTS monolayer on glass (•). The slopes yield the observed first-order rate coefficients, from which the reactive uptake coefficients, y, are calculated. 45 Figure 4.2 IR absorption spectra of OTS monolayer prior to (black) and following (gray) N O 3 exposure at 1.1 x 10"4 atm-s . 47 Figure 4.3 IR absorption spectra of OTS monolayer exposed to OH concentration of 10 molecule-cm for a) 0 min; b) 2.5 min; c) 5 min; d) 10 min. (Figure is adapted from Molina et al. [82]) 49 Figure 4.4 Reaction scheme for saturated hydrocarbons with N O 3 . 50 Figure 5.1 Plot of natural log of N2O5 signal versus reaction time for the reactive uptake of N2O5 by methane soot. The slope yields the observed first-order rate coefficient, from which the reactive uptake coefficient, y, is calculated. 61 Figure 5.2 The concentrations of H N O 3 (dotted line), N 2 0 5 (solid line), and N O 2 (dashed line) in the flow tube reactor were monitored in real time for the product vn studies of N 2 O 5 + methane soot. At times 2.8 - 7.5 min, the injector was pulled back exposing the methane soot to N 2 O 5 . Figure 5.3 Plot of natural log of N O 3 signal versus reaction time for the reactive uptake of N O 3 by methane soot. The slope yields the observed first-order rate coefficient, from which the reactive uptake coefficient, y, is calculated. Figure 5.4 The concentrations of H N 0 3 (dotted line) and N 0 2 (solid line) in the flow tube reactor were monitored in real time for the product studies of N O 3 + methane soot. At times 3 - 8 min, the injector was pulled back exposing the methane soot to N O 3 . Figure 5.5 IR absorption spectra of fresh methane soot (light gray), methane soot exposed to a gas mixture of N O 2 , H N O 3 , and 0 2 (gray), and methane soot exposed to N O 3 at exposure levels of 2.6 x 10"5 (dark gray) and 4.6 x 10"5 atm-s (black) in the presence of N O 2 , H N O 3 , and 0 2 . The spectra have been offset for clarity. Figure 5.6 Reaction scheme for the alkyl radical. Figure 5.7 Reaction scheme for the alkyl peroxy radical. A C K N O W L E D G E M E N T S The experience I gained over the past 2 + years in my graduate studies has been extremely invaluable and rewarding. I like to express my sincere gratitude and appreciation to Dr. Allan Bertram for the opportunity to work in his lab. I'm very grateful for having had him as a professor, a co-op supervisor, and a research supervisor for my graduate studies. His door was always open whenever I needed advice or guidance. I'm deeply thankful for that and for his support and encouragement throughout the years. I would also like to thank past and present members of the Bertram group for their assistance and valuable discussions over the years, but most importantly, for their friendship. Particularly, I would like to acknowledge: Matt Parsons for solving all my computer-related problems and for assisting in all the "side projects" that we worked on in the lab; Magda Dymarska for sharing her schematic drawing of the CIMS that is shown in this thesis; Lori Anthony, Simone Gross, and Daniel Knopf for their assistance in the N2O5 synthesis; and Lori again for introducing us to her energetic four-legged friend - Moka. Lastly, I would like to thank all the staff members from the Mech Shop, Electronic Shop, and Glass Shop for their assistance in various aspects of this project. IX DEDICATION To my parents for all their support. CHAPTER 1 INTRODUCTION 1.1 Introduction Aerosols are liquid or solid particles suspended in a gas and are abundant in the atmosphere. Atmospheric aerosols originate from a wide range of biogenic, natural, and anthropogenic sources [1]. Types of aerosol particles include: soot, mineral dust, inorganics, inorganic-organic mixtures, pure organics, and aqueous inorganics with organic coatings. Atmospheric aerosols range in size from 0.01 microns to several tens of microns in diameter [2]. Despite their small sizes, collectively they play a significant role in the atmosphere. Aerosols can change the climate by modifying cloud properties [3], rain formation [4], and the scattering and absorption of solar and terrestrial radiation [5, 6]. Moreover, these aerosol particles can alter the chemistry in the atmosphere through heterogeneous reactions (i.e. reactions involving more than one phase). Furthermore, aerosols can have a huge impact on the health and well-being of individuals and other living organisms [7]. Reactions involving aerosols and atmospheric gas phase species such as nitrate radicals, ozone, and hydroxyl radicals are particularly important because these reactions can affect the atmosphere's composition, the lifetime of the atmospheric aerosol particles, and physical properties of the aerosol particles. Therefore, in order to accurately predict all of these key elements, it is essential to understand the heterogeneous reactions between atmospheric species and different aerosol types. This thesis will improve the knowledge of the heterogeneous chemistry between nitrate radicals + organic surfaces. Specifically, these laboratory studies improve the fundamental understanding of the chemistry between nitrate radicals + organic particles and nitrate radicals + aqueous inorganic particles with organic coatings. In addition, a detailed study of the chemistry between nitrate radicals + soot is presented. These combined studies are necessary to better understand the oxidation of organic particles and 1 soot in the atmosphere. These studies should improve predictions of the role that aerosol particles play in the atmosphere. 1.2 Organic Aerosols Recent field measurements have shown that organic matter represents a significant component of aerosol particles [8, 9]. In certain regions of the troposphere, the amount of organics found in aerosols can exceed the amount of inorganic matter [10-12]. It has been found on average organic compounds comprise of 20 - 80 % of the particulate mass in the lower troposphere [12]. Studies have also shown organic aerosols can exist in the upper troposphere and lower stratosphere [10, 13]. The composition of organic aerosols varies depending on the source and aging of the aerosols. Some common organics found in aerosols include n-alkanes, ri-alkanoic acids, n-alkanals, aromatic polycarboxylic acids, polycyclic aromatic hydrocarbons (PAHs), dicarboxylic acids, ketones, and much more [1, 2, 14, 15]. Organic compounds in aerosol particles may exist as solid matter [2], viscous liquid [16], internally mixed with inorganic compounds [11, 17], or a thin film on an aqueous inorganic core [18-23]. A large fraction of gas phase organic matter found in the atmosphere originates from emissions such as biomass burning or fossil fuel combustion (i.e. primary organics) [12]. This organic material can condense onto existing aerosol particles as the particles pass through polluted areas or over biomass burning plumes [24, 25]. Low vapour pressure organics (formed by the oxidation of primary organics) have also been found to condense onto existing aerosol particles [26-28] or, alternatively, to homogeneously nucleate into a condensed form (secondary organic aerosols) [12]. The consequences of organics in or on aerosol particles have not been fully explored, but at the current level of understanding it is believed they may affect the aerosols' physical and chemical properties. For example, due to the hydrophobic nature of some organic molecules, they may affect the overall hygroscopic properties of the aerosols by hindering the absorption and desorption of water and trace species. This in turn can change the total mass and acidity of the particles [29]. In addition, the presence of organics may lower the particles' surface tension [19, 30, 31], which may lead to the 2 increase of cloud condensation nuclei (CCN) [32]. Lastly, the coagulation process that is involved in rain formation may be affected by organic matter since the presence of organic matter on water droplets may act as a barrier to prevent the droplets from coagulating [33, 34]. 1.3 Aqueous Inorganic Aerosols Coated with Organic Films There is direct evidence that supports the existence of organic layers on aerosols [11, 19, 35, 36]. Surface sensitive techniques such as time-of-flight secondary ion mass spectrometry (TOF-SIMS) and scanning electron microscopy (SEM) coupled with an energy dispersive X-ray (EDX) microanalyse have been used in analyzing marine aerosols to show the existence of organic-coated particles [35, 36]. The mass spectrometry investigations show a large amount of aliphatic hydrocarbons fragmented off from the surface of the marine aerosols. These hydrocarbons were later determined to have negligible solubility, implying they must exist on the surface rather than dissolved within the droplet. Subsequent studies with SEM on these aerosols also confirmed these particles were enclosed by a coating. The organics on these aerosol particles were expected to have originated from the decomposition of marine organisms such as phytoplankton. The breakdown of dead organisms and their cells releases a rich source of fatty acids such as stearic and palmitic acids that are commonly found in marine aerosols [36]. It has been estimated that 200 Tg of hydrophobic organic fragments are released onto the ocean surface each year [18]. Due to the hydrophobic nature of these organics, they tend to accumulate on the surface of the ocean. When marine aerosols are formed from air bubble bursting, they acquire a coating of organic surfactants with them upon ejection from the ocean surface to the atmosphere. One property of aerosols that indirectly affects the climate is their ability to absorb water. Although some studies have suggested a hydrophobic coating of organics on aqueous inorganic particles inhibits water uptake [37, 38], some have shown otherwise [39, 40]. Clearly more research is required in this area. There is, however, strong evidence suggesting changes in the hygroscopic properties of the aerosol particles 3 may occur following heterogeneous reactions [18, 37, 41, 42], leading to the assumption that surface functional groups may play an important role in water uptake. Another important aspect that needs to be considered is the release of gas phase byproducts following heterogeneous reactions. These byproducts may have a wide range of implications for aerosol formation, air pollution, and modification of the atmospheric composition. 1.4 Soot Soot is formed as a result of incomplete combustion and is ubiquitous in the atmosphere. The primary sources of anthropogenic soot are fossil fuel and biomass burnings. It has been estimated that as high as 24 Tg of soot is emitted globally per year [43]. The concentration of soot in the atmosphere varies from region to region. Table 1.1 presents a summary of typical soot concentrations found in various regions [44, 45]. Table 1.1 Typical soot concentrations found in various regions. Region Concentration of soot (ng m"3) In urban areas 1500 -- 20000 In rural areas 200- 2000 Over ocean areas 5-20 Over the poles 1 Upper troposphere and lower stratosphere 0.05- 4.2 The composition of soot is not well understood and is believed to vary depending on the source. It has been suggested that soot is composed of an elemental carbon inner core and an outer shell that may contain various functional groups and hydrocarbons such as polyaromatic hydrocarbons [46]. These surface species may play an important role in the reactivity, hygroscopic character, and potential catalytic activity of soot. Soot is believed to have a significant impact on the environment in several ways. Due to its large optical absorption coefficients, soot may influence the radiative budget 4 by absorbing incoming solar radiation and outgoing infrared radiation [47]. Soot may also affect the chemical composition of the atmosphere by reacting with atmospheric gas phase species. Additionally, soot may provide a medium to facilitate reactions that are too slow or would not otherwise proceed in the gas phase. Lastly, soot has been linked to human pulmonary diseases [48]. To date, soot remains one of the most poorly understood aerosol types. There is considerable uncertainty in terms of its role in the atmosphere. The lack of field and laboratory data and the complexity of the soot's structure and composition contribute a large part to this uncertainty. Knowledge in the soot's reactivity with various atmospherically relevant species will provide some understanding of the role soot plays in the atmosphere. Information on its kinetics, surface properties, and the products produced following reactions is also important in elucidating its importance in the atmosphere. 1.5 Nitrate Radicals and Their Gas Phase Chemistry Nitrate radicals like any other radicals have unpaired electrons. These unpaired electrons contribute to the radicals' high reactivity; hence, radicals participate in many chemical reactions. Following the practice throughout the literature, atmospheric radicals are written in reactions without showing the unpaired electrons in this thesis. For example, nitrate radical is written as N O 3 and the unpaired electron is implied. Other common radicals include nitrogen dioxide ( N O 2 ) , nitrogen monoxide (NO), hydroxyl radical (OH), alkyl peroxy radical ( R O 2 ) , hydroperoxyl radical ( H O 2 ) , alkoxy radical (RO), and alkyl radical (R). The existence of N O 3 in the atmosphere has been known for more than a century [49]. However, it is only until recently that extensive research has begun to explore the role N O 3 plays in the atmosphere [50-53]. Nitrate radicals are important nighttime tropospheric oxidants of gas phase species and, hence, participate in many reactions in the troposphere. Leighton [54] pointed out the potential reactivity of N O 3 with other oxides of nitrogen and gas phase organic species in the early nineteen-sixties. However, it was another twenty years after the first in situ measurements of N O 3 were made in the 5 stratosphere [55] and troposphere [56], that prompted a growing interest in N O 3 chemistry. Specifically, an interest in N O 3 reaction mechanisms and kinetics with other atmospheric species has emerged. The results of these important studies help researchers gain a better understanding of the role N O 3 plays in the atmosphere. To date, there have been numerous studies on gas phase reactions involving N O 3 but very few studies on heterogeneous reactions with N 0 3 . To get a more complete picture of the N 0 3 chemistry in the atmosphere, more studies on the heterogeneous reactions with N O 3 are needed. The concentration of N O 3 varies greatly in the troposphere since it is strongly dependent on the atmospheric mixing ratio of N0 2 , which can vary from 5 ppt to 100 ppb [49], as well as the mixing ratio of ozone (O3). NO3 concentrations can range from 2 -430 ppt [50]. At these concentrations, N O 3 may provide an important sink for organics such as isoprene or terpenes [57, 58]. The main tropospheric source of N O 3 is from the oxidation of N O 2 with 0 3 as given by Nitrogen pentoxide, N 2 0 5 , which can act as a temporary reservoir for N 0 3 , contributes as a source of N 0 3 when it thermally decomposes as follows: where M represents any inert molecules such as N 2 or O2. However, the formation of N2O5 depends on the presence of N O 3 as Reaction 1.1 is ultimately the main channel for the formation of N O 3 . N 2 0 5 is an important atmospheric species itself because it reacts with water in or on aerosols via: N 0 2 +0 3 -» N 0 3 + 0 2 (1.1) N 2 0 5 + M -> N 0 3 + N 0 2 + M (1.2) N 0 3 + N 0 2 + M - > N 2 0 5 + M (1.3) 6 N 2 0 5 + H 2 0 - > 2 H N 0 3 (1.4) to form nitric acid, an important sink of N O 3 and one of the major causes of acidification in rainwater. Other processes that are important in removing N O 3 from the troposphere, especially during daytime, are photolysis and reaction with NO. N O 3 photo-dissociates rapidly into NO and N O 2 via: NO 3 +hv(;i<630nm)->NO + O 2 (1.5) N 0 3 +hv(X < 6 3 0 n m ) - » N 0 2 +0 (1.6) Furthermore, the reaction between N O 3 and NO proceeds as N 0 3 + N O - > 2 N 0 2 (1.7) can dominate as the main removal process of N O 3 . NO is emitted directly into the atmosphere or formed by lightning. NO is also formed by the photolysis of N O 2 according to N0 2 +hv/(A<400nm)-^NO + O (1.8) As a result of all these loss processes, only a small amount of N O 3 is present during the day. During nighttime, however, when these processes are negligible, N O 3 becomes relatively abundant (2 - 430 ppt) and it can react with a range of gas phase organic species through two routes: hydrogen abstraction (reaction 1.9) and addition to unsaturated bonds (reaction 1.10). N 0 3 + RH -> FTN03 + R (1.9) 7 The products from these reactions may lead to the formation of stable nitrate compounds, which could be toxic and have adverse health effects associated with them [1] . 1.6 Aerosols and Radicals Aerosols are believed to be transformed physically and chemically as a result of their interactions with radicals in the atmosphere. Therefore, in order to accurately predict the state of the particles and their role in the atmosphere, a better understanding of these processes is necessary. Numerous kinetic studies have focused on the reactive uptake of different oxidants such as O 3 , OH, and N O 3 by various organic species with assorted functional groups in liquids and on surfaces [52, 59-63]. Furthermore, kinetic studies have also been investigated for the reactive uptake of gas phase species by soot [64-66]. Although these measurements are crucial, very few studies have concentrated on the changes to the surfaces and the products released following reactive uptake. Currently, there are no studies involving the reactive uptake of N 0 3 by organic coatings and only one study to our knowledge on the reactive uptake of N 0 3 by soot. This thesis will provide the preliminary study on reactions of N O 3 + organic coatings, which will improve the knowledge of the heterogeneous chemistry on organic-coated aerosols and on the interface of pure organic aerosols. Additionally, this thesis will advance the current level of understanding of the chemistry between N O 3 + soot. 8 1.7 Thesis Overview This thesis consists of six chapters. This chapter covers general descriptions of atmospheric aerosols and N O 3 . Chapter 2 outlines the general procedures, which include descriptions of instruments and methods used to carry out the experiments. In Chapter 3, the characterization and validation of the experimental approach are presented. Chapters 4 and 5 contain results for the reactive uptake of N O 3 by self-assembled monolayers (SAM) and by methane soot, respectively. Also included in these chapters are results of quantitative exposure studies and surface analysis studies using Fourier transform infrared (FTIR) spectroscopy. Finally, a brief summary of the results is presented in Chapter 6. 9 CHAPTER 2 GENERAL EXPERIMENTAL 2.1 Introduction Cylindrical flow tube reactors have been used extensively by many researchers for conducting kinetic studies [52, 63, 67-69]. The flow dynamics in these systems are well characterized and understood. Therefore, kinetic information can be obtained. The flow tube technique was used to study the reactive uptake of N O 3 by various organic surfaces. N O 3 exposure studies, which involved probing the organic surfaces before and after treatment with N O 3 , were also investigated. This chapter covers the methods for generating and detecting N O 3 and the preparation of the organic surfaces. As well, techniques used for the reactive uptake and exposure studies will be discussed in this section. 2.2 N0 3 Production The production of N O 3 involved the synthesis of N 2 O 5 followed by thermal dissociation to yield N O 3 and N O 2 . The setup used for the synthesis of N 2 O 5 is shown in Figure 2.1. Prior to the production, the system was pumped down for at least an hour to remove trace amounts of water that may have adsorbed onto the walls of reaction vessels and tubing. The lines were also conditioned with the gases used for a minimum of one hour. Excess amount of O3 was generated by passing dry O2 at approximately 500 STP cm 3 • min"1 (STP = 1 atm, 273 K) through an ultraviolet (UV) lamp (Jelight, model #600), which converted a fraction of the O2 to O3. To remove water vapour from the O2 carrier gas, a Drierite trap was placed immediately before the UV lamp. A small flow (not measured) of N O 2 that had previously passed through a P2O5 trap (Aldrich, 97 %) to remove trace amounts of water was introduced through a second line to the glass reactor. The 10 cm i.d. by 30 cm long glass reactor allowed enough time for N O 2 to react with O3, 10 UVlarnp Glass reactor Dnente trap Ah N O A i== To exhaust P 2 0 5 t r a p ^ A ^ N 2 O 5 crystals To pump Ethanol bath at -80°C Valve P 2 O 2 trap Figure 2.1 Apparatus for the synthesis of N 2 0 5 . which was in excess over N O 2 . The amount of N O 2 entering the glass reactor was adjusted until brown N O 2 gas was not observed at the exit of the vessel, indicating a complete consumption of N O 2 . Reactions that occurred in the reactor were as follows: NO z +0 3 - > N 0 3 +O z (2.1) N 0 3 + N 0 2 + M - > N 2 0 5 + M (2.2) O3 oxidized N O 2 to yield N O 3 and O2. Subsequently, N O 3 combined with N O 2 to form the stable N 2 0 5 , which has a sublimation point of 240.8 K at 1 atm [70]. The N 2 0 5 flow was passed through a P2O5 trap to reduce the concentration of nitric acid, then to a glass trap immersed in an ethanol bath cooled to 193 K with an immersion cooler (Neslab, model CC-100 II), and N 2 O 5 condensed as white crystals inside the glass trap. During an experiment which the reactive uptake of N O 3 was studied, the trap containing N 2 O 5 was transferred to an ethanol/dry ice bath. The stable bath temperature at about 195 K helped maintain the vapour pressure of N 2 O 5 constant inside the trap. A slow flow of dry helium (He) was introduced through an inlet into the trap and the gas was saturated with N 2 O 5 as it passed over the crystals. The He/N2C>5 gas mixture was then directed to a glass oven heated to above 433 K to thermally dissociate N2O5 according to N 2 0 5 + M - > N 0 3 + N 0 2 +M (2.3) A more detailed description on the setup for producing N O 3 and measuring the reactive uptake of N O 3 is presented in the next section. When the trap was not in use, it was kept at 193 K and pressurized to slightly above 1 atm with He to prevent water vapour from entering and reacting with N 2 O 5 to form nitric acid: 12 N 2 0 5 + H 2 0 - > 2 H N 0 3 (2.4) 2.3 Experimental Apparatus for Measuring Reactive Uptake of N O 3 Uptake experiments were performed with a flow tube reactor coupled to a chemical ionization mass spectrometer (CIMS). A schematic of the experimental apparatus is shown in Figure 2.2. As mentioned in the previous section, He was introduced through an inlet into the trap containing the N 2 O 5 . It is important to note that the effluents from the N 2 O 5 trap contained roughly equal amount of H N O 3 and N 2 O 5 as determined with the CIMS. Immediately at the exit of the trap, additional helium was added through a second inlet. The purpose of this was to dilute the He/N 2 0 5 mixture and to minimize the recombination of N O 3 with N O 2 later downstream. The glass oven used for dissociating N 2 O 5 had a dimension of 2.2 cm i.d. by 22 cm long and was heated by wrapping a heat tape around the outside. The oven temperature was measured by a thermocouple placed in between the exterior surface of the glass and the heat tape. The oven and the flow tube were connected by a movable injector, which had a dimension of 0.48 cm i.d. by 60 cm long. The flow tube, constructed of borosilicate glass, had a dimension of 2.4 cm i.d. by 40 cm long. The temperature of the flow tube was regulated by circulating a 50:50 mixture of water and ethylene glycol from a constant temperature bath through the jacket on the exterior of the flow reactor. Helium, which served as the carrier gas, was added through a port located at the upstream end of the flow tube. All flow rates were determined with electronic mass flow meters (MKS Instruments, Inc.), and the flow reactor pressure was monitored by a 10 Torr pressure gauge (MKS Baratron). 2.4 Detection Method N O 2 , N O 3 , H N O 3 , and N 2 O 5 were chemically ionized by selected reagent ions and detected with the mass spectrometer. The reagent ions were generated by introducing trace amounts of an ion precursor (SF6 or C H 3 I ) in 1000-2000 STP cm 3 - min"' of N2 210 through a radioactive polonium ( Po) source (NRD, model Po-2031). The reagent ion 13 .Glass, oven wrapped in heat tape Movable reactor injector •SE 6/ 'N aor C H 3 I / N 2 i 31<>Po J, b l h e - (Ion Source) Front Vacuum Entrance Chamber Lens C M S Flow* N2Q5 crystals -ethanolbath - 195 K He ts . --1.1V Surface of Ion-molecule interest Region •20:V -3V r^i 18 V I ! Cage ,6V! 35V 2 V . Exit Lens OV 18 V Quadruple Filter 18 V Lenses j-Pre Filter Post Filter Pin Rotary Turbomolecular h o l e :Bump& Pump'-' Mechanical Pump Channeltron Mulitplier Turbomolecular Pump . Rear Vacuum Chamber Figure 2.2 Schematic of the experimental apparatus. flow was added to the main flow through a side port positioned at the exit of the flow reactor. This combined flow, which contained both ionized and neutral molecules, entered the ion-molecule region where reactions took place. Typical reaction times in the ion-molecule region were several milliseconds. The majority of the gases exited from the ion-molecule region, which was biased at -15 V to improve sensitivity, were pumped away with a rotary pump backed by a mechanical pump (see Figure 2.2). Only a small fraction was sampled through a 100 um size pinhole held at -3 V before entering the vacuum chamber, which was differentially pumped in two stages. The front chamber, which was pumped by a turbomolecular pump, operated at 10"5 Torr. Inside the front chamber, it consisted of a cage that operated at a potential of 18 V and a series of ion optics to focus the ions into the quadrupole mass filter, which was housed in the rear chamber. The rear vacuum chamber was maintained at 10"6 Torr during operation by a second turbomolecular pump. The quadrupole separated the ions by their mass-to-charge ratios and the sorted ions were accelerated to a channeltron multiplier detector for detection. Two reagent ions were used for the majority of the studies in this thesis: SF6" and I". The reagent ions had typical count rates of approximately 300000 count -s~'. SF6~ and r gave signals at 146 and 127 amu, respectively. N O 2 was detected as NO 2 by undergoing charge transfer with SF6~ according to N 0 2 +SF6" -> N 0 2 +SF6 (2.5) The rate constant for this reaction was reported to be 1.4 x 10" 1 0 cm 3 - molecule'1 -s~' [71]. HNO3 and N2O5 were detected as H N 0 3 F ~ and NO3, respectively, after reacting with SF6~. Their respected rate constants are 2 x 10"9 and 7.5 x 10"'° cm 3 - molecule"1 -s_ 1 [71]. N2O5 was also detected as N 0 3 with I" as the reagent ion [72] according to 15 N 2 0 5 + r - > N 0 ~ + I N 0 2 (2.6) The rate constant for this reaction has a reported value of 1.3 x 10"9 cm 3 • molecule-1 • s~' [71]. It has been shown that H N O 3 does not react with I" [72, 73], which agrees with current experimental observations. N O 3 was detected as N O 3 using I" as the reagent ion and had a detection limit of 108 molecule-cm-3. The rate constant for this reaction was measured to be 6.5 x 10 :'° cm3 - molecule-1 s - ' . The determination of this rate constant is presented in Chapter 3. 2.5 Materials A 5 % (v/v) mixture of CH3I (Acros Organics, 99 %) in N 2 was used for the reagent ion flow. Al l other gases used were commercially available: He (99.999%, Praxair), N 2 (99.999%, Praxair), NO (3.72 %, certified NO-N 2 gas mixture, Praxair), N 0 2 (99.5%, Matheson), 0 2 (99.993%, Praxair), and SF 6 (99.995%, Praxair). Octadecyltrichlorosilane, CH 3(CH 2)i 7SiCl3 (Aldrich, > 90 %), and octadecanethiol, CH3(CH2)i7S (Aldrich, 98%), were used for the monolayer preparation. 2.6 Surface Preparation Depending on the system under investigation, two different methods were used for coating a cylindrical tube insert for studying heterogeneous chemistry. For soot studies, methane soot was deposited on the inner wall of a Pyrex glass tube (1.75 cm i.d. by 15 cm long) by exposing it to a methane flame produced with a standard torch. Air entrained in the flame was the only source of oxidant, resulting in an oxygen-lean flame. For organic monolayer studies, self-assembly technique was employed. This technique can be applied to a glass or gold substrate and has been reported in detail elsewhere [74, 75]. Briefly, prior to deposition, the Pyrex glass tube (1.75 cm i.d. by 15 cm long) and the gold tube (Epner Technology, Inc., 1.91 cm i.d. by 17.5 cm long) were cleaned with piranha solution (3:1 H2S04:H202) by immersing the tabes in the solution for 15 min. 16 The tubes were then rinsed with Millipore water and finally rinsed with the solvent used for preparing the monolayer. The clean Pyrex tube was immersed in a 1 mM solution of octadecyltrichlorosilane (OTS) in toluene for 1 hour. Excess OTS molecules that may have adsorbed on the surface were removed by sonicating the Pyrex glass tube in fresh toluene for 5 minutes. The last step was repeated twice. The deposition of a monolayer on the gold tube was prepared by immersing the clean gold tube in a 1 mM solution of octadecanethiol (ODT) in pure ethanol for 6 - 2 4 hours. Excess ODT molecules were removed by sonicating the gold tube in fresh pure ethanol for 5 minutes, three times. After having coated the inner wall of the Pyrex glass tube or gold tube with the organic material of interest, the tube was inserted into the flow reactor to study heterogeneous reactions between the coating and N O 3 . 2.7 FTIR Spectroscopy FTIR spectra were collected prior to and following exposure of the OTS monolayer and methane soot to N O 3 . IR spectroscopy allowed the identification of functional groups that formed as a result of oxidation reactions with N O 3 for the methane soot study. For the monolayers, IR was used to determine if there were any changes in the chain length of the hydrocarbons after reactions with N O 3 . Methane soot was deposited on a barium fluoride (BaFa) window by exposing it to a methane flame, and a clean BaF2 window was used for the background spectrum. The same procedure used for coating the Pyrex glass tube with OTS was also applied to a small piece of Pyrex glass substrate (1 cm by 1.5 cm) for exposure studies involving the monolayer. A clean Pyrex glass substrate was used for the background spectrum. All IR spectra were recorded by averaging 32 scans at 4 cm"1 resolution using a Bruker Equinox 55 FTIR spectrometer equipped with a deuterated triglycine sulphate (DTGS) detector. 17 CHAPTER 3 CHARACTERIZATION OF THE FLOW SYSTEM 3.1 Introduction Using a newly built system requires the knowledge and understanding of its feasibilities and limitations. Although flow tube reactors coupled to CIMS have been used extensively, their usage with N O 3 has not. Several challenges emerged when the two were used in conjunction. One challenge was finding a reagent ion for measuring N O 3 . The second challenge was avoiding the possibility of incomplete dissociation of N 2 O 5 . Lastly, the recombination of N O 3 with N O 2 after dissociation could have also posed a similar problem with the measurements. These key issues will be addressed in this chapter and the work on validating the experimental approach and the N O 3 source will be presented. 3.2 Reagent Ion for Detecting N O 3 For the majority of the studies in this thesis, I" was chosen as the CIMS reagent ion. The use of I" as the reagent ion to measure N O 3 with CIMS has not been previously reported. To ensure the reaction between I" and N O 3 did proceed, the 62 mass-to-charge ratio (m/z) signal was monitored as a function of oven temperature while having a continuous and constant flow of IS^Os/He mixture flowing into the CIMS. As mentioned in the previous chapter, N 2 O 5 is also detected asNOj, which has a 62 m/z, using I" as the reagent ion. By increasing the oven temperature, N 2 O 5 will dissociate to N O 3 and N0 2 . If N O 3 also reacts with I" to give 62 m/z signal, then the signal should still remain after N 2 O 5 is fully dissociated. Shown in Figure 3.1 is the 62 m/z signal monitored in real time, designated by a solid line. The dashed line shows the external temperature history of the oven. During the first 10 minutes when the oven temperature is between 298 - 333 K, the signal remains relatively unchanged. As soon as the oven temperature 18 100000 r 4 0 0 75000 50000 25000 H T i m e ( m i n ) Figure 3.1 Observation of the 62 m/z signal (solid line) in real time using I as the reagent ion in CIMS. The dashed line shows the external oven temperature history. 19 reaches 353 K,a significant drop in the signal is observed, signifying a possible depletion of N 2 O 5 . Above 373 K, the signal remains constant, which may indicate all the N2O5 has been dissociated. It is worth mentioning that the signal never reaches zero. This may indicate three things: 1) N O 3 reacts with I" to give N O 3 as mentioned above, 2) N O 3 recombines with N O 2 to give N 2 O 5 back, or 3) N 2 O 5 does not completely dissociate under these conditions. The approach used to verify whether N O 3 converts to NO3 by I" in the ion-molecule region was to react N O 3 with NO while monitoring for the changes in the 62 m/z signal. NO was chosen for this test since NO selectively reacts with N O 3 according to the following reaction: N 0 3 + N O - > 2 N 0 2 (3.1) Prior to the test, a mass spectrum of NO was obtained using I" as the reagent ion and showed no observable peaks in the regions of interest (62 amu). The results of this study are shown in Figure 3.2. The solid and dashed lines are the 62 m/z signal and the oven temperature history, respectively, measured in real time. The shaded regions indicate when an excess amount of NO was introduced to the main flow at the exit of the flow tube reactor. Initially, the oven temperature is set to approximately 413 K and a signal of 100000 is observed, presumably from N O 3 . At times 2.5 - 5 min, during which excess NO is present, the signal drops close to zero, which confirms the signal prior to the addition of NO indeed originated from N O 3 . As the oven temperature cools down to below 383 K, the signal begins to rise, suggesting the presence of N 2 O 5 . The signal rises since CIMS is more sensitive to N 2 O 5 than N 0 3 when using I" as the reagent ion. At time 21 min, when the oven temperature drops to about 348 K, NO is again added to the main flow. The signal decreases but does not reach zero, which indicates the coexistence of N2O5 and N O 3 under these conditions. When NO is introduced to the flow system at lower oven temperatures, the decrease in the 62 m/z signal is less apparent, as expected, since the dissociation of N 2 O 5 to generate N O 3 is temperature dependent. 20 500000 400000 4 300000 4 200000 H 1000004 10 15 20 25 30 35 Time (min) Figure 3.2 Observation of the 62 m/z signal (solid line) in real time using l~ as the reagent ion in CIMS. The dashed line shows the external oven temperature history. The shaded regions indicate when excess NO was present. 21 To further confirm the presence of N O 3 in the flow tube reactor, a similar test was implemented using SF6" as the reagent ion. The presence of N O 3 was monitored indirectly by observing the presence of N O 2 since each time an N 2 O 5 molecule dissociates, one N O 2 molecule is generated. As shown in Figure 3.3, the 46 m/z signal, which is associated with N O 2 , is close to zero when the oven temperature is between 298 - 333 K. As the oven temperature ramps up to about 353 K, a stable signal of 60000 is observed. At this point (time =12 min), excess NO is added to the main flow. The initial spike of the signal may be due to a sudden burst of N 0 2 that may have formed and accumulated in the NO line. The signal goes down after a brief period once the excess N O 2 is removed and equilibrium is established. As expected, the signal obtained in the presence of NO is approximately 180000, about 3 times the signal before NO is added. This increase in the N 0 2 signal from the addition of NO as a result of reaction 3.1 verifies the presence of N O 3 . A similar trend is observed at a higher temperature as shown in the same figure. 3.3 Incomplete Dissociation of N 2 O 5 The degree of dissociation of N 2 O 5 in the flow system was controlled by the oven temperature. In order to estimate the amount of dissociation, both the residence time, r , for the N 2 O 5 spent in the oven and the rate at which it dissociates need to be established. The former can be accomplished by applying the following equation: where F(T,P) is the total flow rate (measured in volume per unit time), which is a function of temperature, T, and total pressure, P, flowing through the oven with volume, V. Typical N2O5 residence times in the oven for the majority of the experiments ranged from 55 - 100 ms. The next step is to determine the rate for the dissociation reaction: 22 Time (min) Figure 3.3 Observation of the 46 m/z signal (solid line) in real time using SF6 as the reagent ion in CIMS. The dashed line shows the external oven temperature history. The shaded regions indicate when excess NO was present. 23 N 2 0 5 + M -> N 0 3 + N 0 2 + M (3.3) Reaction 3.3 is not a simple dissociation reaction. Its rate not only depends on temperature, but also on the concentration of the third body, [M]. Based on the work of Atkinson el al. [76], who fitted the rate constant for this reaction as a function of temperature and pressure (i.e. [M]), the rate for reaction 3.3 can be described by where [N2Os] is the concentration of N 2 O 5 and kj(T,P) is the dissociation rate constant, which is a function of both temperature, T, and total pressure, P. By integrating equation 3.4, the degree of dissociation can be calculated as a function of time. Figure 3.4 shows a plot of the percentage N 2 O 5 dissociated as a function of time for various oven temperatures and at a fixed pressure of 5 Torr, an approximate pressure in the oven. As illustrated, when the oven temperature is at 393 K, close to 100 % of the N 2 O 5 molecules are dissociated under the typical residence time of 100 ms in the oven. To be conservative, all experiments were conducted at an oven temperature of 433 K or above to ensure complete dissociation of N 2 O 5 . The plot also shows that an appreciable amount of dissociation begins at about 353 K, which agrees with experimental observations (see Figures 3.1 and 3.3). 3.4 Recombination of N03with NO2 After the dissociation of N 2 O 5 in the oven, the possibility of N O 3 recombining with N O 2 can potentially interfere with the reactive uptake measurements. In order to estimate the degree of recombination, residence times for N O 3 spent in the injector and flow tube reactor need to be evaluated. In addition, the rate at which the recombination occurs to form N 2 O 5 needs to be determined. Using equation 3.2, the residence times can be calculated for the two regions. The results are tabulated in Table 3.1. d[N2Os] kd(T,P)-[N205] (3.4) dt 24 T i m e (ms) Figure 3.4 Plots of percentage of N 2 O 5 dissociated versus time for various oven temperatures and at a fixed pressure of 5 Torr. 25 Table 3.1 Typical residence times in various regions of the flow system. Region T (ms) Oven 55-100 Injector 10-25 Flow tube reactor 25-120 The formation of N 2 O 5 according to the following reaction is also temperature and pressure dependent. N 0 3 + N 0 2 +M -> N 2 O s + M (3.5) The rate constant for this reaction is also estimated based on the work by Atkinson et al. as above [76]. The rate for the combination reaction is given by = k,(T,P)- ([N03 ] 0 -[N205])- ([N02]0 -[N205]) (3.6) where kf(T,P) is the rate constant for the formation of N 2 O 5 and is a function of temperature, T, and total pressure, P; [NO% ] 0 and [N02 ] 0 are the initial concentrations of N 0 3 and N O 2 , respectively; and [N205] is the concentration of N 2 O 5 . By integrating equation 3.6, the concentration of N2O5 formed can be calculated as a function of time. The integrated form [77] by assuming [NOy ] 0 = [N02 ] 0 is k(TP) = -- [ N M l (37) f ' [^ 3 ] 0 - ( [7VC7 3 ] 0 - [ iV 2 6) 5 ] , ) where [A^205], is the concentration of N2O5 at reaction time t. The estimated concentration of N O 3 or N O 2 formed after dissociation in the oven ranged from 1010 - 1012 molecule-cm"3 for a typical experiment. Figure 3.5A shows 26 ~ 3.0x10 ,u-] O l . u - | 1 1 1 1 , 1 1 1 1 1 , - , 1 1 — 0 20 40 60 80 100 120 140 B) Time (ms) Figure 3.5 Plots of a) [N2O5] formed and b) percentage of N O 3 unreacted as a function of time for the combination reaction of N 0 3 + N 0 2 + M -» N 2 0 5 + M with initial N O 3 and N O 2 concentrations equal to 1010 (dashed) and 1012 (solid) molecule-cm"3 at 298 K and 5 Torr. 27 plots of concentration of N 2 O 5 formed as a function of time for [M? 3 ] 0 = 1010 and 1012 molecule-cm"3 at 298 K and at a pressure of 5 Torr. Figure 3.5B shows plots of percentage of unreacted NO3 as a function of time for the same parameters used for the previous plots. As the plots suggest, even under the worst-case scenario with a combined residence time of 145 ms in the injector and flow tube reactor, less than 3 % of the NO3 are consumed to re-form N 2 O 5 . This is consistent with experimental observations, which showed when NO3 was reacted with NO, the signal dropped close to zero, suggesting minimal amount of N 2 O 5 formation (see Figure 3.2). 3.5 Determination of the Rate Constant for NO3 + I In order to determine the concentration of NO3 in the flow reactor, the rate constant for the reaction of NO3 + I" occurring in the ion-molecule region needs to be measured. This is particularly important for exposure studies (see Chapters 4 and 5). The rate constant for N O 3 +1" was measured relative to the rate constant for N 2 O 5 +1", which is known. This was accomplished by measuring the signals for N O 3 and N2O5 under the same flow conditions while varying the oven temperature from 295 to 443 K. In other words, the experiment consisted of having the oven at room temperature to measure the N 2 O 5 signal, then increasing the oven temperature to 443 K to measure the NO3 signal. The following derivation shows how the ratio of the rate constants is related to the ratio of their respected signals. The rate law for the reaction N 2 0 5 +1" —> NO3 + IN0 2 (reaction 2.6) can be expressed as ^ = -kN20i-[n-[N205] (3.8) at where kN205 is the rate constant for the reaction o f N 2 0 5 +1" and and [N205] denote the concentrations of I" and N 2 O 5 , respectively. Since the concentration of N2O5 28 is much greater than I in the ion-molecule region, a pseudo first-order rate is assumed and the expression above can be integrated to give In = -kN2Q5-[N205]-t (3.9) where [I ] 0 and [I ], denote the concentrations of I at reaction times equal to 0 and t, respectively. Based on the conservation of charge, an expression for [/"], can be formulated as U'l=[r]0-[NO-], (3.10) where [NO^]t is the concentration of NO 3 at reaction time / formed as a result of reaction 2.6. Substituting expression 3.10 into 3.9, the following expression is obtained. In v [no j kN205-{N205yt (3.11) It can be shown that ln ( l -x )=-x for x<0.1. Using this approximation, expression 3.11 can be rewritten as [ M U ~kN205-[N205]-t (3.12) [ ' " J o by assuming the ratio on the left hand side of expression 3.12 is less than 0.1. Following the same approach, a similar expression can be obtained as [NO:]\ l—22i_^kNm.[NOi].t (3.13) L-* Jo 29 for the reaction of N 0 3 + I according to N 0 3 +T ->NC>3~ +1 (3.14) where [N03]\ is the concentration of NO3 at reaction time t formed by reaction 3.14; kN03 is the rate constant for the reaction; [/"]* is the concentration of I" at reaction time t* = 0; and [N03 ] is the concentration of NO3. The relative rate constant is obtained by the ratio of expression 3.12 to expression 3.13 to get U']0 [No~t kN03-[No3]-r ( 3 1 5 ) Expression 3.15 can be further simplified if the signals due to NO3 and N2O5 are measured under the same flow conditions. In other words, by fixing all the flows constant, their respected residence times in the ion-molecule region and initial I~ concentrations will be identical. In addition, by assuming all N 2 O 5 molecules are converted to N O 3 at 443 K (oven temperature) and that there is not a major N O 3 loss process that exists, the N2O5 concentration with the oven at 295 K and NO3 concentration with the oven at 443 K should be equal. The validity of this assumption will be examined more closely in the next section. For now, expression 3.15 can be reduced to [NO;], =kN205 [No-y.- kN03 (3-16) The data for determining the rate constant for reaction 3.14 is shown in Figure 3.6. The plot shows signals of I~ (127 m/z; solid gray) and NO3 (62 m/z; solid black) measured in real time. As illustrated, the approximations that [N03], /[I']0 <0.1 and [N03]\ /[I~]*Q < 0.1 used to derive expressions 3.12 and 3.13, respectively, are valid 30 18000000-, CO 0 5 10 15 20 T i m e (min) Figure 3.6 Observation of the 127 m/z signal (solid gray line) and 62 m/z signal (solid black line) in real time using l~ as the reagent ion in CIMS. The dash line shows the external oven temperature history. The shaded regions indicate when an excess amount of NO was present. 31 under these conditions. To verify all the N2O5 molecules have converted to NO3, an excess amount of NO is added, as indicated by the shaded region in the figure, to completely react away the N O 3 when the oven temperature is set to around 443 K. Once it has been confirmed the signal is strictly due to N O 3 , the signal prior to the addition of NO is substituted for [NO^]*. in expression 3.16. Independent studies have shown the signal is directly proportional to concentration. Similarly, the value used for [NO^],, which is strictly from the product of N 2 0 5 +1", was the 62 m/z signal when the oven temperature was at 295 K. Under these conditions, it was estimated that less than 1 % of the N2O5 dissociated. Using the rate constant of (1.3 ± 0.5) x 10"9 cm 3 - molecule"1 -s"' from the literature [71] for the reaction of N 2 O s + 1" , the rate constant for N 0 3 +1" is estimated to be (5.2 ± 2.0) x 10"'° cm 3 • molecule"1 • s"1. 3.6 N O 3 Loss Processes To verify whether other processes may have contributed to the loss of N O 3 for the rate constant determination study, box model simulations of the reaction conditions were carried out using Acuchem (a program for solving a system of differential equations describing the behavior of multi-component chemical reaction systems). Al l relevant and significant reactions involving N O 3 were examined and entered into the model along with their rate coefficients and initial concentrations of the species involved. Based on the input data, it calculates the concentrations of all the species (reactants and products) as a function of time. The results from the model were used to determine if N O 3 may have been depleted by other gas phase reactions. Table 3.2 presents a summary of the reactions and their respective rate coefficients at 298 and 443 K used in the box model. Where relevant, a pressure of 2 Torr was used for determining the rate coefficients. The results generated by the model are shown in Figures 3.7 and 3.8 for oven temperatures of 298 and 443 K, respectively. The plots show the concentrations of N 2 O 5 (solid black line), N O 2 (dashed gray line), and N O 3 (dotted black line) as a function of time as they enter different regions of the flow 32 Oven Injector Reactor 2.0x10 . o | 1 .5x10" H o o g i . o x i o " H CO i_ "c CD g 5.0x10 0.0 4 -120 Time (ms) Figure 3.7 Concentrations of N 2 O 5 (solid black line), N O 2 (dashed gray line), and N O 3 (dotted black line) in the flow system were simulated for the N O 3 + I~ rate determination study using a model with initial N 2 O 5 concentration equals to 2.0 x 1011 molecule-cm-3 and oven temperature equals 298 K. The double-headed arrows at the top indicate the time spent by the gas phase species in each of the indicated region in the flow system. 33 O v e n Injector Reac to r 1* ;—x x * 1 ' 1 • 1 1 1 1 1 1 1 1 1 0 20 40 60 80 100 120 T i m e (ms) Figure 3.8 Concentrations of N 2 O 5 (solid black line), N O 2 (dashed gray line), and N O 3 (dotted black line) in the flow system were simulated for the N O 3 +1" rate determination study using a model with initial N 2 O 5 concentration equals to 2.0 x 1011 molecule-cm"3 and oven temperature equals 443 K. The double-headed arrows at the top indicate the time spent by the gas phase species in each of the indicated region in the flow system. 34 system as indicated by the double-headed arrows at the top of each figure. A concentration of 2.0 x 1011 molecule-cm"3 for N 2 O 5 was used as the initial condition. Rate coefficients calculated for 443 K were used for simulating the conditions in the oven region in Figure 3.8. For all other regions, rate coefficients evaluated for 298 K were used for the simulations. The sharp drop in the concentrations for all species at 85 ms in both figures is due to dilution by helium carrier gas entering the flow tube reactor. Table 3.2 NO3 reactions and their rate constants at 298 K and 443 K (A pressure of 2 Torr is used where relevant). Reaction k(298) k(443) N 0 3 + N 0 2 + M - > N 2 0 5 + M *1.0x 10"'3 [76] *2.2 x 10"14 [76] N 2 0 5 + M - > N 0 3 + N 0 2 + M +3.6 x 10"3 [76] +1.6x 102 [76] N 0 3 ->NO + 0 2 +3.2 x 10"3 [49] +2.6 [49] N 0 3 +NO->2N0 2 *2.6x 10"" [76] *2.3 x 10"" [76] * 1 — 1 units: cm -molecule s units: s Figure 3.7 suggests the dissociation of N 2 O 5 is minimal at 298 K. In contrast, N 2 O 5 completely dissociates for the duration of residence in the oven when set to 443 K as illustrated in Figure 3.8. The same figure also shows there is some loss of NO3 in the oven resulting from thermal dissociation. Under these conditions, the model estimated a loss of about 20 % in NO3 concentration, which translates to an underestimation of about 25 % for the previously determined rate constant for the reaction of N 0 3 +1", modifying it from (5.2 ± 2.0) x 10"10 to (6.5 ± 5.0) x 10"'° cm3 • molecule"1 • s"1. 3.7 Validation of the Flow Tube - CIMS Apparatus and N0 3 Source To evaluate the performance of the flow tube technique and to validate the NO3 source, the rate constant for the gas phase reaction of NO3 + NO according to reaction 3.1 35 was measured and compared to the literature value. The experiments were carried out using the flow tube technique and the detection of NO3 was monitored with CIMS using I~ as the reagent ion. The reaction was studied between 296 - 298 K and at a pressure of 5 Torr. NO3 was introduced via the movable injector and NO along with He (helium was the carrier gas) was added through a port located at the upstream end of the flow tube (see Figure 2.2). A typical experiment had a total flow of approximately 2500 STP cm3 • min - 1 in the flow reactor. Under these conditions, the flows were laminar. Measurements of the second-order rate constant were performed under pseudo first-order conditions with NO being in excess. For a typical experiment, as shown in Figure 3.9, the decay of the NO3 signal was monitored at various injector positions, which correspond to different reaction times; At the end of each experiment, the injector was pushed back to the original position to ensure there was no drift in the signal. Furthermore, NO3 was completely reacted away with an excess amount of NO to acquire the background signal, which was subtracted from the NO3 signal when determining the kinetics. The decay of N 0 3 as a result of reaction with NO is described by - kobs • z ln[7V03 ] z = — ^ + ln[A^03 (3.17) v where [NOz ] z and [N03 ] z = 0 are the NO3 concentrations at injector positions z and z = 0, respectively; v is the linear gas flow velocity; and k"^ is the observed pseudo first-order rate coefficient. By plotting the natural log of [N03 ] versus relative injector position, z, k°bs was determined from the slope of the plot, which is equal to(- k°bs/v), and from the linear gas flow velocity. Figure 3.10 shows plots of the natural log NO3 signal versus reaction distance for various NO concentrations. To determine the true first-order loss rate coefficient, kUt, from k"bs, the approach described by Brown was used [78]. It corrects for wall loss and diffusion of NO3. First-order rate coefficients were measured for several NO concentrations, which were calculated based on the pressure in the flow 36 400000 - i 350000 -\ 300000• 250000• C O o 200000 H 1 5 0 0 0 0 H 100000^ 50000H 0 cm 0 cm N 0 3 signal ; Addit ion of excess N O for background determination 5 10 15 20 Time (min) Figure 3.9 Observation of the NO3 signal (solid line) in real time at various injector positions (0 - 10 cm), which correspond to different reaction times between NO3 and NO. The shaded region indicates when excess NO is present to completely react away all NO3 to determine the background signal. 37 e 8 J TN01 molecule/cm 3 • 0 A 2.57 x 1 0 1 2 • 7.04 x 1 0 1 2 1.37 x 1 0 1 3 4 2 6 8 10 - T -12 14 Dis tance (cm) Figure 3.10 Examples of decay of N 0 3 signal as a function of injector distance for various NO concentrations. 38 reactor and the flow of NO. Since the reaction of N O 3 + NO is a bimolecular reaction and the experiments were carried out with NO being in excess, kUl is given by K,=k2nd.[NO] (3.18) where k2nd is the second order rate coefficient for reaction 3.1 and [NO] is the concentration of NO. The results of the corrected first-order rate coefficients were plotted against the concentration of NO as shown in Figure 3.11. According to equation 3.18, the slope of this plot yields k2nd, which was estimated to be (2.7 ± 0.2) x 10"11 cm 3 • molecule"1 • s"'. The uncertainty comes from the scatter in the data and represents a 95 % confidence level. The errors associated with flow rate, pressure, and temperature are negligible. The determined bimolecular rate constant is in excellent agreement with the literature data of (2.65 ± 1.1) x 10"" cm3 - molecule"' -s"' [79]. 3.8 Summary and Conclusions The use of CH3I as an ion precursor in CIMS was proven to be an ideal candidate for measuring N O 3 . The rate constant for the reaction of N 0 3 + I" was measured relative to the reaction of N 2 O 5 + I". The rate constant for the former reaction was determined to be (6.5 ± 5.0) x 10"'° cm 3 • molecule"'• s"1. Calculations show under the current conditions used in the flow tube, N 2 O 5 undergoes complete dissociation in the oven during heating to produce N O 3 and N O 2 . Moreover, the recombination reaction following thermal dissociation is estimated to be negligible. These calculations agree well with experimental observations. Prior to experimentation, the flow Uibe technique and the N O 3 source were validated by measuring the rate constant for the bimolecular gas phase reaction between N O 3 and NO. The measured value is in excellent agreement with the literature value. The successful application of the flow tube technique was expanded to measuring heterogeneous reactions between N 0 3 + organic coatings and N O 3 + soot as presented in the next two chapters. 39 Figure 3.11 Plot of first-order loss rate coefficient versus NO concentration. The slope yields the second-order rate constant for the reaction of N 0 3 + NO and has a value of (2.7 ± 0.2) x 10"11 cm 3 • molecule"1 • s"1. 40 CHAPTER 4 HETEROGENEOUS REACTION BETWEEN N0 3 + SAM 4.1 Introduction It has long been speculated that organic compounds can act as surfactants and form organic layers on the surface of aerosol particles [21]. However, it is only recently that researchers have started to focus on organic-coated aerosols and their roles in the atmosphere. For example, Rudich et al: [39] investigated the interfacial interactions between water and organic surfaces, which provided insight to the hygroscopic properties of coated aerosols. Other researchers have measured the reactive uptake of tropospheric species such as O3 [62, 80], OH [63], CI radicals, and Br radicals [81] by organic surfaces. The oxidation of organic surfaces is of interest because it may alter the surface properties, which can affect the hygroscopic character and cloud condensation nuclei activity of the organic aerosols. NO3 has been recognized as an important nighttime oxidant. Numerous laboratory studies have focused on NO3 reactions with gas phase organics [60] and more recently with condensed phase organics [52]. However, little is known about its interactions with organic surfaces. The organic surfaces chosen as proxies for coated aerosols and for the interface of organic aerosols were octadecyltrichlorosilane (OTS) deposited on glass substrates and octadecanethiol (ODT) supported on gold substrates. Self-assembled monolayers (SAM) have been used by other researchers in the past for similar studies [63, 82, 83] and it has been suggested some atmospheric aerosols may possess a similar layer on the aerosol surface [18]. In the following, results are presented for two sets of experiments. In the first set, uptake experiments were conducted to determine the reaction probability of N 0 3 with SAM. From the reaction probability, the heterogeneous loss rate can be calculated and compared to the homogeneous loss rate to determine the relative importance of 41 heterogeneous loss as a N O 3 sink. Moreover, the time needed for N O 3 to process a simulated organic-coated aerosol particle can be calculated from the reaction probability and compared to the process times for other atmospheric radicals to determine their relative efficiency in processing. In the second set of experiments, exposure studies were investigated. In these studies, OTS monolayers were exposed to N 0 3 and the surfaces were characterized with FTIR before and after exposure. Finally, atmospheric implications of all the findings are discussed. 4.2 Experimental The uptake experiments were carried out using the flow tube technique and the detection of NO3 was monitored with CIMS using I" as the reagent ion. The experiments were done at 298 K and at a pressure of 2 Torr. The coated glass or gold tube was inserted into the flow tube reactor for heterogeneous reaction with N O 3 . N O 3 was introduced via the movable injector (see Figure 2.2). A typical experiment had a total flow of approximately 100 STP cm3 - min"' in the flow tube reactor. The flows were laminar under these conditions. Exposure studies were performed on a 1 cm by 1.5 cm Pyrex glass substrate coated with OTS. This involved inserting the coated glass substrate into the flow tube reactor while passing a constant flow of N O 3 over the substrate. The detection of N O 3 was monitored with CIMS using I" as the reagent ion, and the concentration of NO3 in the flow tube was determined based on the rate constant measured in Chapter 3 for the reaction of NO3 + I". Infrared spectroscopy was used to characterize the surfaces prior to and following reaction with NO3. 4.3 Data Analysis Heterogeneous reactions refer to reactions involving more than one phase. An example of such is the loss of gas phase species as a result of physical transfer or chemical reaction in or on a condensed phase. The rate of a heterogeneous reaction is normally expressed in terms of a dimensionless parameter, the reactive uptake 42 coefficient. It is defined as the fraction of gas phase molecules, which collide with the surface, that result in a reactive uptake (i.e. the reaction probability). The advantage of describing the kinetics in terms of reactive uptake coefficients instead of rate coefficients is that the results can be compared to other experiments that may have been conducted with a significantly different surface-to-volume ratio. The reactive uptake coefficient of NO3 was calculated from the first-order rate coefficient, kw, for the removal of NO3 by the organic surface. The heterogeneous loss of N O 3 to the surface is described by ln[MU=-C-' + ln[M?3],=o C4-1) where [N03 ], and [N03 ],= 0 are the NO3 concentrations at reaction times / and / = 0, respectively; and k°^s is the observed first-order loss rate. By plotting the natural log of [N03 ] versus relative reaction time, t, k°bs was determined from the slope of the plot. Figure 4.1 shows plots of the ln(N03 signal) versus reaction time for the heterogeneous loss of NO3 to the OTS and ODT monolayers and to the bare surfaces. The true first-order loss rate coefficient, kw, was calculated from k°J" by applying the approach developed by Brown [78], which corrects for the radial concentration gradient as a result of N O 3 uptake by the surfaces. Finally, the reactive uptake coefficients, y, were calculated from kw using the following equation [63]: 2-r-kw r = — (4.2) cavg+r-k„ where r is the radius of the gold or glass tube insert (cm) and c is the average thermal velocity (cm-s"') ofN03. 43 4.4 Results and Discussion Typical data for the uptake measurements are shown in Figure 4.1. For the two monolayers, the ln(NC>3 signal) decreases with reaction time. In contrast, the ln(N03 signal) for the bare surfaces remains almost unchanged with reaction time. The reactive uptake coefficients were calculated from the decay of NO3 signal as described in the previous section. Shown in Table 4.1 are the values for the reactive uptake coefficients of N O 3 for the two monolayers and the bare surfaces. These values were obtained from averaging several experiments. The uncertainty comes from the scatter in the data and represents a 95 % confidence level. The errors associated with flow rate, pressure, and temperature are negligible compared to the uncertainty from the scatter of data in Figure 4.1. Table 4.1 Summary of reactive uptake coefficients measured in this study. Surface Y Bare gold (8 ± 5) x lO - 5 ODT on gold (8 + 3) x 10"4 Bare glass (1.010.6) x lO - 4 OTS on glass (1.3 ± 0.2) x 10"3 The exposure time for each uptake measurement was slightly under 10 minutes and the typical NO3 concentration was between 5 x 10 n and 8 x 10" molecule-cm"3. Under these conditions, it would take approximately 3 minutes to oxidize about 95 % of the available reactive sites, assuming a single reactive site per molecule (i.e. per carbon chain). When uptake measurements were performed on previously exposed organic surfaces (i.e. surfaces that had been performed uptake measurement once), they only showed on average a 20 % decrease from their previously determined y values obtained from "fresh" surfaces. These findings suggest that there was more than one reactive site available per molecule and/or that multiple N O 3 radicals may be consumed by one reactive site in order to explain the minor decrease in the y values (see section 4.7 for a reaction mechanism involving uptake of two NO3 radicals per reactive site). Note that 44 0.1 -, 0.0 -0.1 --0.2-h -0.3-C / ) n O -0.4--0.5--0.6--0.7-• Bare Gold A Bare Glass • ODT on Gold • OTS on Glass 0.000 —I— 0.005 T 0.010 0.015 0.020 Reaction Time (s) — i — 0.025 0.030 Figure 4.1 Plots of natural log of N O 3 signal versus reaction time for the reactive uptake of N O 3 by bare gold (•) , bare glass ( A ) , ODT monolayer on gold (*), and OTS monolayer on glass (•). The slopes yield the observed first-order rate coefficients, from which the reactive uptake coefficients, y, are calculated. 45 additional experiments were done with shorter times and at lower NO3 concentrations and the results were the same within uncertainty. Nonetheless, the reactive uptake coefficients listed in Table 4.1 represent lower limits for the reactive uptake of NO3 by fresh organic surfaces since a fraction of the surface was oxidized during the experiment. The small reactive uptake coefficients (« 10"4) for the bare surfaces show NO3 does not react efficiently with gold or glass. The ODT and OTS monolayers have reactive uptake coefficients of 8 x 10"4 and 1.3 x 10"3, respectively. It is interesting to note that even though both the OTS and ODT consist of 18-saturated-hydrocarbon chains, their reactive uptake coefficients are statistically different. This discrepancy could be explained by the fact that OTS has a slightly higher packing density than ODT. It has been reported the inter-chain distance for OTS is approximately 4.4 A, which corresponds to a surface density of 6.0 x 1014 molecule' cm"2 [84]. The spacing between two neighboring sulfur molecules for alkanethiols on gold is reported to be 4.97 A, corresponding to a surface density of 4.7 x 1014 molecule• cm"2 [84]. The slightly higher surface density in OTS, which translates to more reactive sites per area available for reaction, resulted in a higher loss of NO3, which was reflected by a larger reactive uptake coefficient. 4.5 Exposure Studies Shown in Figure 4.2 are IR spectra obtained prior to (black) and following (gray) exposure of the OTS monolayer to NO3. The OTS monolayer was subjected to an exposure level of 1.1 x 10"4 atm-s, which corresponds to exposing the surface to a typical NO3 concentration found in highly polluted areas of 150 ppt (24-hour average) for 8.5 days. The IR spectra show - C H 2 symmetric (vs) and asymmetric (vas) stretch modes at 2851 and 2918 cm"1, respectively. The spectra also exhibit small absorbance peaks due to - C H 3 asymmetric stretch at 2957 cm"1. Unfortunately, due to absorption of infrared in the lower wavenumber region by the glass, it was not possible to determine the functional groups that may have formed on the surface following oxidation. However, since IR intensity is proportional to concentration, the fact that the intensities for the - C H 2 bands 46 0.150 n J n i i 1 1 1 2750 2800 2850 2900 2950 3000 W a v e n u m b e r ( c m 1 ) Figure 4.2 IR absorption spectra of OTS monolayer prior to (black) and following (gray) N O 3 exposure at 1.1 x 10"4 atm-s . 47 remained almost unchanged following N O 3 exposure suggests there was little or no fragmentation of the monolayer. This is contrary to a similar study done by Molina et al. [82] who observed decreasing intensity in the CH X bands with increasing exposure time when the OTS monolayer was exposed to OH as illustrated in Figure 4.3. The difference between these findings will be discussed in section 4.7. 4.6 Proposed Mechanism It has been suggested that reactions with organic surfaces undergo similar mechanisms to analogous gas phase reactions [80, 85]. Under this assumption, N O 3 most likely undergoes hydrogen abstraction with the organic surface to form an alkyl radical and H N O 3 as described by reaction 1.9. The fate of the alkyl radical, R, is summarized in Figure 4.4. In the presence of oxygen, the alkyl radical forms the alkyl peroxy radical, R0 2 , by R + 0 2 - > R 0 2 (4.3) As illustrated in Figure 4.4, R O 2 reacts with N O 2 , N O 3 and itself. The reaction between R O 2 and N O 2 leads to the formation of alkyl peroxynitrate (R-O -ONO2), which is an important reservoir for N O 2 since it can thermally decompose back to the reactants. R O 2 could react with N 0 3 to form an alkoxy radical (RO). At high concentrations of R O 2 , self-reaction becomes an important loss process that can also lead to the production of RO, which can result in the formation of an alcohol, a carbonyl, and an organic nitrate by isomerization, reaction with O 2 , and reaction with N O 2 , respectively. Alternatively, RO may undergo decomposition to form an aldehyde and alkyl radical. The dashed arrows in Figure 4.4 indicate the products could subsequently react by cycling through similar reaction channels. Current experimental results suggest, however, the decomposition channel is not important under the present conditions used for the experiments. 48 2800 2850 2900 2950 3000 Wavenumbers, cm"1 Figure 4.3 IR absorption spectra of OTS monolayer exposed to OH concentration of 108 molecule • cm"3 for a) 0 min; b) 2.5 min; c) 5 min; d) 10 min. (Figure is adapted from Molina et al. [82]) 49 I' / / / ' / / t / / / / / / / ' / ' / RON0 2 ^ Nitrate O RH r N 0 3 or OH v HN0 3 o r H 2 0 0 2 RO, R' R" Carbonyl NO, X R-0-0N0 2 \ RO, RO + 0 2 / (isomerization) *~ R ' O H (decomposition) O + R" R' H Aldehyde 1 * Figure 4.4 Reaction scheme for saturated hydrocarbons with N 0 3 . 50 4.7 Difference Between N O 3 and OH Exposure Studies As mentioned previously, Molina et al. [82] observed that the alkane surface decomposes when exposed to OH. The mechanism for the degradation of the alkane surface as proposed by Molina et al. was first initiated by hydrogen abstraction by OH to form an alkyl radical, R. The alkyl radical promptly transforms to R O 2 in the presence of O2. R O 2 then undergoes self-reaction to form RO followed by decomposition, which lead to the breaking of a C-C bond to form an aldehyde and alkyl radical. The similarity in the chemistry of N O 3 and OH is evident as the mechanism for the degradation of the alkane surface by OH is one of the reaction pathways (the one leading to decomposition) shown in Figure 4.4. Although the chemistry of OH and N O 3 is very similar - both undergo hydrogen abstraction with saturated hydrocarbons to form alkyl radicals - the discrepancy in the findings perhaps can be explained by how the experiments were conducted. For both the OH and N O 3 exposure studies, a much higher concentration of the radicals was used in the lab compared to their respected typical concentrations in the troposphere. The exposure period used was reduced to a fraction of the typical residence time of an aerosol particle to achieve a similar exposure level in the atmosphere. Clearly, the benefit of conducting the exposure studies this way is the time-saving aspect. The drawback of this is that at a higher exposure concentration, it may introduce chemistry that may not otherwise occur at a lower exposure concentration. Needless to say, the closer the conditions used in the lab are to the conditions in the atmosphere, the more atmospherically relevant it becomes. The fact that the fragmentation of the monolayer was observed for the OH exposure studies possibly could be explained by the high concentration of OH used for the experiments. A high concentration of OH may result in the formation of high concentration of R 0 2 on the surface. Consequently, this would increase the extent of R O 2 to self-react to form RO, which is an important intermediate for the decomposition of the monolayer. At a lower OH concentration, however, the R O 2 surface concentration may be lower. The extent of R O 2 to self-react to form RO is also expected to decrease, thus, limiting the pathway to fragmentation. Likewise, the same argument can also be applied to reactions initiated by N O 3 , inferring that the R O 2 surface 51 concentration is dependent on the N O 3 concentration used during exposure. The fact that the decomposition of the monolayer was not observed for exposure studies involving N O 3 possibly could be due to the conditions used for these experiments, which may have limited the R O 2 surface concentration. In other words, in these experiments, N O 3 reactive uptake may have been relatively low which would have limited the R O 2 surface concentration and the extent for R O 2 to self-react to form RO follow by decomposition. The reaction mechanism for the N03-initiated oxidation of the alkane monolayers may follow the reaction scheme shown in Figure 4.4. The only channel that can be excluded based on FT1R results is the one leading to decomposition. Based on the discussion by Molina et. al. [82], RO predominately undergoes decomposition at the surface rather than isomerization or reaction with O2 or N O 2 . Since decomposition was not observed for the reaction of N O 3 + OTS, the formation of RO was expected to be minimal. This implies the reaction between R O 2 and N O 3 to form RO may also be slow or insignificant. Therefore, the formation of R - O - O N O 2 could be one of the main channels for the reaction. Further measurements will be helpful in determining its importance. A second reaction mechanism, which is consistent with the results of the uptake measurements as discussed in section 4.4, is proposed based, in part, on analogous gas phase chemistry as follows: R 1 Ri C H 2 + N O 3 +- -pn + H N O 3 ( 4 - 4 ) / ' / Ri \ \ • C H + O, H C O (4.5) R? R? O 52 R i \ H C- -o / V + NO, \ / (4.6) , C = 0 + N 0 2 + 0 2 + H 0 2 The reaction is initiated by N O 3 through hydrogen abstraction of the alkane monolayer to yield the alkyl radical. The alkyl radical promptly transforms to R O 2 in the presence of 0 2 . R O 2 then combines with N O 3 to form the intermediate shown by the bracketed term in reaction 4.6. In the presence of 0 2 , the intermediate dissociates to form a carbonyl, N O 2 , O2, and H O 2 . The only difference between this mechanism and the channel leading to the formation of a carbonyl depicted in Figure 4.4 is reactions 4.4 - 4.6 involve the formation of a long-lived complex that goes directly to produce a carbonyl rather than forming the RO intermediate follow by reaction with O2. At this point, these proposed mechanisms are only based on speculations. Further work is needed to confirm its relevance and to completely understand the difference between oxidation of organic monolayers by N O 3 and by OH. However, based on these findings, it is unlikely the organics on a coated aerosol particle or the organics at the interface of an organic aerosol particle would undergo decomposition by radicals since the concentrations of radicals are much lower in the atmosphere (limiting RO from forming and, hence, decomposing). Furthermore, the removal of these organic aerosols by decomposition as suggested by Molina et al. [82] may not be an important removal pathway under tropospheric conditions. 53 4.8 Atmospheric Implications Although the results show there is heterogeneous loss of NO3 to the organic surfaces, the extent of this loss is not significant enough to alter the concentration of NO3 in the troposphere. To demonstrate this, the rate for the heterogeneous loss of NO3 to coated aerosols was compared to the rate for the homogeneous loss of NO3 to oc-pinene (in the gas phase), which is one of the major loss processes of NO3 in the troposphere [58]. The results are listed in Table 4.2 for various environments. Table 4.2 Comparison of NO3 heterogeneous loss rate, R^,, by reaction with organic coatings to N 0 3 homogeneous loss rate, RHom, by reaction with a-pinene for different environments. Environment Aerosol surface area density (cm2 cm"3) RHe/RHom Urban 1.1 x 10"5 0.001 Rural 1.4 xlO" 6 0.0002 Remote 1.2 xlO" 7 0.00001 The heterogeneous loss rate, Rnel, was calculated using the following equation [63]. RH«=^--A-[N03] (4.7) where c is the average thermal velocity of N 0 3 (cm-s - 1), A is the aerosol surface area densities (cm2-cm"3), y is the reactive uptake coefficient, and [N03 ] is the concentration of NO3 (molecule-cm-3). This equation assumes gas phase diffusion is negligible. The assumption is of minor importance for aerosol particles under 0.5 um in diameter, which account for more than 80 % of all aerosol particles in the troposphere [2]. For larger particles, the equation overestimates the heterogeneous loss rate. Values for typical aerosol surface areas found in various environments were adapted from 54 Bertram et al. [63] and as reported, these values represent upper limits to the surface areas of organic particles. The reactive uptake coefficient used for calculating Rnet was the average reactive uptake coefficient value determined from the OTS and ODT monolayer studies. The homogeneous loss rate, RHOM, was determined based on nighttime oc-pinene concentration of 0.5 ppb [51] and a rate constant of 5.82 x 10"12 cm3 - molecule"1 -s"1 [50] for the reaction of NO3 + ot-pinene. As shown in Table 4.2, the heterogeneous loss rate of NO3 to organic coated aerosols is considerably less than the homogeneous loss rate of NO3 to oc-pinene for all environments, implying NO3 loss through heterogeneous reactions is of minor importance under tropospheric conditions. The importance of heterogeneous reactions, therefore, shifts to the potential for surface modifications. As discussed previously, the reaction between NO3 and SAM may lead to the formation of carbonyls and organic nitrates. Consequently, the presence of these functional groups may cause the surface to become more hydrophilic, which may affect the hygroscopic properties of the aerosols. The extent to which the organic surface is oxidized, assuming the uptake is proportional to the fraction of reactive sites, is given by [63] F = exp N ^ total (4.8) where F is the fraction of surface unoxidized; y0 is the reactive uptake coefficient of NO3 for a fresh organic surface; Z is the collision frequency of NO3 with the surface (molecule - cm - 2 • s"1); / is the reaction time (s); and Ntotal is the total number of surface reactive sites (moleculecm-2). Equation 4.8 assumes the hydrogen abstraction reaction initiated by NO3 is the rate-determining step. Additionally, the gas phase diffusion is assumed to be negligible, which, as mentioned before, is valid for aerosols under 0.5 (am. The aerosol processing time, r , which is defined as the time it takes to have the surface sites reach 1/e of their initial concentration, was calculated using expression 4.8 for NO3 and other commonly found tropospheric species for oxidizing a simulated organic aerosol 55 particle with a surface concentration of 1014 molecule-cm"2. The results are listed in Table 4.3. The tropospheric radical concentrations, which were used for determining the processing times, were adapted from Moise et al. [81]. Al l the reactive uptake coefficients used for the calculations were determined based on reactions with OTS on glass. The reactive uptake coefficient for NO3 was determined based on this study while all others were as referenced. The results indicate processing by N O 3 is as efficient as by OH, which is believed to be a dominant route for processing organics. These results suggest N O 3 should not be overlooked as an important nighttime atmospheric oxidant and its potential for other chemistry. Table 4.3 Processing times for various atmospheric species to process a simulated organic coated aerosol particle with a surface concentration of 101 4 molecule- cm"2. Species [Species] in the troposphere (molecule cm"3) Y T N O 3 l x l O 9 1.3 x 10"3 2.7 hr OH 1 x 106 0.5 [63] 3.6 hr Br 1 x 106 4x 10"2 [81] 4.1 day Cl 1 x 104 0.5 [81] 22.0 day 4.9 Summary and Conclusions The reactive uptake of N 0 3 by self-assembled monolayers was measured for the first time. The results from the monolayer studies indicate the packing of the monolayer affects the reactive uptake of N O 3 . The reactive uptake coefficients were determined to be (8 ± 3) x 10"4 for ODT and (1.3 ± 0.2) x 10"3 for OTS, which has a higher packing density. Surface analysis studies show there is no significant fragmentation of the monolayer following exposure to N0 3 , which is contrary to studies done with OH [82]. This suggests that decomposition is not an important removal channel for organic aerosols containing mostly saturated hydrocarbons. Calculations show the small uptake 56 of N 0 3 by these organic surfaces does not significantly alter the N 0 3 concentration in the troposphere. However, from the calculations of processing times, they indicate N 0 3 is as efficient as OH in processing organic surfaces of saturated hydrocarbons. 57 CHAPTER 5 HETEROGENEOUS REACTION BETWEEN N0 3 + SOOT 5.1 Introduction Soot particles are important atmospheric aerosols. They play an intricate and diverse role in the troposphere responsible for heterogeneous reactions and cloud formation, and play a role in the global radiation balance. For example, it has been reported that soot particles provide reactive sites for gases such as O3 [64], N O 2 [65], and H N O 3 [86]. Their interactions with these gas phase species may lead to a change in the soot particles' surface properties, which may allow them to act as cloud condensation nuclei, a prerequisite for cloud formation [87]. Furthermore, the ability of soot particles to absorb and scatter solar and terrestrial radiations with high efficiency has led to the inclusion of soot in models for determining the earth's radiation budget [88]. In order to gain a better understanding of the role soot plays in the atmosphere, it is important to examine all the potential chemical processes that can occur on the soot particles' surface. As an example, Lelievre et al. [64] have studied heterogeneous reactions of ozone with soot and showed that such reactions do not significantly deplete the ozone concentration in the troposphere or stratosphere. However, it has been pointed out that oxidation reactions may cause soot to become more hygroscopic, which is a prerequisite to the removal of soot from the atmosphere [89]. Reactions on soot particles with other important atmospheric gases have also been studied and provided important knowledge in terms of their kinetics and product formation [44, 63, 68, 90]. However, there is still a lack of laboratory data - specifically for the reaction between soot and N O 3 . This is despite the fact that N O 3 is an important nighttime oxidant and is abundant in the troposphere. N 2 O 5 is also an important species in the atmosphere. Removal of N 2 O 5 is thought to be a major sink of N O and N O 2 , which are closely related to O3 and O H concentrations 58 in the atmosphere. As a result, it is imperative to quantify the sinks of N 2 O 5 . N 2 O 5 loss on soot has only been investigated in a few previous studies. In the following, results of the reactive uptakes of N O 3 and N 2 O 5 by methane soot and the products released from these reactions are presented. In addition, spectra from surface analysis by F U R prior to and following exposure of methane soot to N O 3 are shown. These results will add to only a handful of studies on N 2 O 5 + soot and will be the first for examining exclusively the reaction between N O 3 + soot. Finally, atmospheric implications of all the results are discussed. 5.2 Experimental The uptake experiments were conducted using a flow tube reactor coupled to a CIMS. N O 3 was monitored with CIMS using I~ as the reagent ion. The experiments were done at 298 K over a pressure ranging from 3 to 4 Torr. A glass tube coated with methane soot was inserted into the flow tube reactor for heterogeneous reaction with N O 3 , which was introduced via the movable injector (see Figure 2.2). A typical experiment had a total flow of approximately 1000 STP cm 3 - min"1 in the flow tube reactor with He as the carrier gas. The flows were laminar under these conditions. Gas phase products were investigated with the flow tube and CIMS. A fresh sample of methane soot coated on the inner wall of a Pyrex tube was inserted into the reactor and exposed to N O 3 . Volatile products were monitored with CIMS using SF6" as the reagent ion. Surface products were studied with a barium fluoride ( B a F 2 ) window coated with methane soot on one side. The coated window was inserted into the flow tube reactor while passing a constant flow of N O 3 over the surface. He and O2 were used as carrier gases. The detection of N O 3 was monitored with CIMS using I" as the reagent ion, and the concentration of N O 3 in the flow tube was determined based on the rate constant measured in Chapter 3 for the reaction of N O 3 + I~. Infrared spectroscopy was used to examine surface products following reaction with N O 3 . 59 5.3 Results and Discussion 5.3.1 N 2 0 5 + Methane soot Prior to the study of N O 3 + methane soot, uptake measurements and product studies were conducted for the reaction of N 2 O 5 + methane soot and the results were compared to the literature data. Results of the reactive uptake of N 2 O 5 by methane soot are shown in Figure 5.1. Typical N2O5 concentrations used in the flow tube were 5 x 101 0 - 8 x 101 0 molecule-cm"3. The reactive uptake coefficient was calculated from the decay of the N 2 O 5 signal as described in section 4.3 and was determined to be 0.026 ± 0.009 by assuming a geometric surface area for the soot. This assumption underestimates the available surface area of soot for reactive uptake due to porosity in the soot. Several uptake measurements were made with the same methane soot sample and showed no substantial decrease in the reactive uptake coefficient with increased exposure time. This suggests that over a short exposure time and under these conditions, only a small fraction of the reactive sites on the methane soot surface underwent reaction. Current results for the N 2 O 5 reactive uptake coefficient are in good agreement with previous studies by Longfellow et al. [68], who reported a reactive uptake coefficient of 0.016. However, an accurate measurement of the available surface area for reaction on the methane soot is necessary for comparing y values from two different experiments since y is dependent on the surface area accessible to N 2 O 5 in the case of soot samples. Shown in Figure 5.2 are the results for the product studies of N 2 O 5 + methane soot. Initially, the injector was pushed in so that the methane soot particles were not exposed to N2O5. At times 2.8 - 7.5 min, the injector was pulled back exposing the methane soot to N 2 O 5 . During this period, production of N O 2 and the consumption of N2O5 were observed. Note that the production of N O 2 exceeds the consumption of N 2 O 5 . This may be due to N O 3 , which may have formed as a result of the conversion of N 2 O 5 to N O 2 , reacting on the surface to produce more N O 2 (see next section that shows N O 2 is in fact produced when methane soot is exposed to NO3). It is also interesting to note that during exposure, there was a fast initial uptake of H N O 3 followed by a rapid surface saturation, which led to the H N O 3 concentration to relax back to its initial value. Upon returning the injector to its initial position, desorption of H N O 3 was evident. The 60 1 1 1 i 1 — 0 . 0 0 0 0.001 0 . 0 0 2 Reaction Time (s) Figure 5.1 Plot of natural log of N 2 O 5 signal versus reaction time for the reactive uptake of N 2 O 5 by methane soot. The slope yields the observed first-order rate coefficient, from which the reactive uptake coefficient, y, is calculated. 61 Figure 5.2 The concentrations of H N 0 3 (dotted line), N 2 0 5 (solid line), and N 0 2 (dashed line) in the flow tube reactor were monitored in real time for the product studies of N 2 0 5 + methane soot. At times 2.8 - 7.5 min, the injector was pulled back exposing the methane soot to N 2 O 5 . 62 integrated area of the desorbed H N O 3 accounts for 70 ± 20 % of the absorbed H N O 3 integrated area. It appears that the majority of H N O 3 simply underwent adsorption-desorption at the methane soot surface, while a small percentage of H N O 3 may have permanently adsorbed onto the methane soot surface or have undergone reactive uptake. Further research is needed to verify this. Nevertheless, the amount of H N O 3 taken up, which was estimated to be 1.0 x 1016 molecules, would cover roughly 1 % of the methane soot assuming the surface area to be approximately 30 times the geometric surface area to account for the porosity in the methane soot [68]. These observations are consistent with previous studies [68]. 5.3.2 N O 3 + Methane soot Results of the reactive uptake of N O 3 by methane soot are shown in Figure 5.3. Typical N O 3 concentrations used in the flow tube were 4 x 1010 - 7 x 1010 molecule - cm"3. The reactive uptake coefficient was determined from the first-order loss rate as described in section 4.3 and had a measured value of 0.1 by assuming a geometric surface area for the methane soot. This assumption may not be an accurate representation of the methane soot, which is known to be porous. The true surface area available for reaction is, in fact, greater than the geometric surface area. However, it has been shown that the correction factor for y is small (estimated to be between 1/3 and 1) for porous films with y > 0.1 [91, 92]. Furthermore, the measured reactive uptake coefficient is close to the diffusion-limited loss rate; therefore, the reported value corresponds to a lower limit of the actual reactive uptake coefficient. Based on these assumptions, a conservative value of > 0.03 for the reactive uptake coefficient for N O 3 + methane soot was obtained. The possible interference from N O 2 was investigated by measuring the reactive uptake of N O 2 on methane soot. A reactive uptake coefficient of < 10~3 was obtained, which is consistent with the literature data [90]. The small reactive uptake coefficient suggests N O 2 does not react appreciably with methane soot. In addition, interference from H N O 3 was expected to be minimal (see below). 63 Figure 5.3 Plot of natural log of N 0 3 signal versus reaction time for the reactive uptake of N 0 3 by methane soot. The slope yields the observed first-order rate coefficient, from which the reactive uptake coefficient, y, is calculated. 64 The results for the product studies of N O 3 + methane soot are shown in Figure 5.4. During the times (at times 3 - 8 min) when the methane soot was exposed to N O 3 , there was an increase in N O 2 concentration. It was estimated that more than 20 % of the N O 3 converted to N O 2 , which is consistent with the previous discussion on the mechanism for N2O5 + methane soot. This amount was calculated by assuming the concentrations of N O 3 and N 0 2 before exposure to the methane soot were equal. As seen previously in the N2O5 product studies, the majority of H N O 3 in this study also underwent adsorption-desorption at the surface while the rest may have permanently adsorbed onto the methane soot surface or may have undergone reactive uptake. As pointed out before, this would not interfere with the measurements as only a small percentage (< 1 % ) of the surface was covered by HN0 3 . An attempt was made to investigate the possible formation of NO due to thermal decomposition of N O 3 on the methane soot surface via the following reaction: N 0 3 - > N O + 0 2 (5.1) Although the decomposition reaction is known to be slow in the gas phase, it may be enhanced or catalyzed by the surface. Using 0 2 as the reagent ion in the CIMS, NO can be detected as NO + . During exposure of methane soot to N O 3 , production of NO was not detected. Calculations show that even if only 1 % of all the N O 3 that collided with the surface get converted to NO, this amount will be well above the detection limit. However, it is possible that the NO generated may have promptly reacted with other species on the surface or with surrounding N O 3 , which may help explain the increase in N O 2 . A more detailed discussion on the reaction mechanism is presented in section 5.5. 5.4 Exposure Studies Shown in Figure 5.5 are FTIR spectra obtained prior to (light gray) and following exposure of methane soot to N O 3 in the presence of O2 and trace amount of N O 2 and H N O 3 . The methane soot was subjected to NO3 exposure levels of 2.6 x 10"5 atm-s 65 J. J i I i I i I i L 1.60E+011 4 = 1 . 4 0 E + 0 1 H 8.00E+010 -j 1 1 1 1 1 1 . 1 1 1 1 r 0 2 4 6 8 10 12 Time (min) Figure 5.4 The concentrations of H N 0 3 (dotted line) and N O 2 (solid line) in the flow tube reactor were monitored in real time for the product studies of NO3 + methane soot. At times 3 - 8 min, the injector was pulled back exposing the methane soot to NO3. 6 6 Figure 5.5 IR absorption spectra of fresh methane soot (light gray), methane soot exposed to a gas mixture of N 0 2 , H N O 3 , and 0 2 (gray), and methane soot exposed to N O 3 at exposure levels of 2.6 x 10"5 (dark gray) and 4.6 x 10"5 atm-s (black) in the presence of N 0 2 , H N O 3 , and 0 2 . The spectra have been offset for clarity. 67 (dark gray) and 4.6 x 10"5 atm • s (black), which correspond to exposing the methane soot to N O 3 concentrations observe in typical urban environment (50 ppt, 24-hour average) and polluted urban (90 ppt, 24-hour average) environments, respectively, for about 6 days. The broad absorption band at 1600 cm"1 that is present in all spectra is attributed to the C = C stretching modes of the polyaromatic system, which constitutes the overall carbon skeleton in the methane soot. Absorption bands of an organic nitrite (R-ONO) and organic nitrate ( R - O N O 2 ) at 1645 and 1281 cm"1 were identified on the reacted methane soot. Additionally, the absorption band at 1281 cm"1 could also be assigned to the symmetric ring stretch of an epoxide. As expected, these bands are more intense in methane soot that was exposed to a higher N O 3 level. The formation of a carbonyl group was evident by the absorption band at 1726 cm"1. At this point, the band at 1352 cm"1 has not been identified, as further research is required to assign this band. Also included in this figure is the FTIR spectrum of methane soot that was exposed to a gas mixture of O2, N 0 2 , and H N O 3 at a similar exposure level as for the N O 3 studies (gray). The spectrum clearly shows that under these conditions, N O 2 and H N O 3 contribute only a small amount to the formation of organic nitrite and nitrate, indicated by weak absorption bands at 1645 and 1281 cm"1. In addition, the peak attributed to the carbonyl group at 1726 cm"1 is much weaker in comparison to studies with N O 3 present. This suggests N O 3 is more reactive and oxidizes the surface more efficiently than N O 2 and H N O 3 . A summary of all the absorption bands and their assignments is given in Table 5.1. 68 Table 5.1 Assignment of vibrational bands for methane soot exposed to N 0 3 . Frequency (cm"1) Species Assignment 1281 Nitrite R-ONO Nitrate R-ON0 2 Epoxide C-0 R-0 stretch N O 2 symmetric stretch Symmetric ring stretch 1352 ? ? 1600 Aromatic C=C stretch 1645 Nitrite R-ONO Nitrate R-ON0 2 N=0 stretch in alkyl nitrites Asymmetric stretch in alkylnitrate 1726 Carbonyl C=0 C=0 stretch 5.5 Possible Reaction Mechanisms The following reaction mechanisms are proposed to explain the IR results. Assuming the reaction mechanisms for the heterogeneous reactions are analogous to the homogenous gas phase reactions, N O 3 most likely undergoes addition reactions with the unsaturated carbons in methane soot as shown by reaction 1.10. The fast reactive uptake of N O 3 by methane soot compared to the monolayer is expected since addition reactions are more favourable than hydrogen abstraction reactions based on gas phase chemistry. The alkyl radical generated as a result of reaction 1.10 may undergo reactions with various species as illustrated in Figure 5.6. The alkyl radical can react with N 0 2 to form a nitrite. The alkyl radical can also react with another N 0 3 to form a dinitrate species, which may dissociate by cleaving the C-C and N-0 bonds to yield two carbonyls and N O 2 . The formation of an epoxide and N O 2 results from the cleavage of the N-0 bond in the alkyl radical. In the presence of molecular oxygen, the alkyl radical forms an alkyl peroxy radical. As shown in Figure 5.7, the alkyl peroxy radical may undergo self-69 o °~\ °\ V / C C R 3 / \ R 2 R 4 See Figure 5.7 Epoxide O, 0 & / O \/°./' R 2 R 4 N O , N O , O O © ©/ \ © © O N N O A / V / y C C — R 3 R 2 R4 Dinitrate .O © &/ O N \ R 1 P R 3 \ / / JZ C ~ — O N O R 2 R 4 Nitrite / Nitrate 0 R r R 2 + 0 R 4 + 2 N 0 2 Carbonyl Figure 5.6 Reaction scheme for the alkyl radical. 70 .o 0 W . V / A l k o x y Alkyl Peroxynitrate Alkoxy Carbonyl Figure 5.7 Reaction scheme for the alkyl peroxy radical. 71 reaction and reactions with N O 3 and N 0 2 . Self-reaction and reaction with N O 3 may lead to the formation of an alkoxy radical, which may transform to a carbonyl in the presence of 0 2 . The reaction between an alkyl peroxy radical and N 0 2 leads to the formation of alkyl peroxynitrate, which may dissociate back to the reactants. The products discussed here are consistent with the FTIR measurements. The formation of organic nitrites and nitrates, carbonyls, and epoxides can be observed in FTIR spectra (Figure 5.5). Based on observations from the product studies, it is believed reactions involving the release of N 0 2 are important. However, quantitative measurements are necessary to determine their relative contributions. 5.6 Atmospheric Implications The contribution of methane soot to NO3 loss processes was determined by comparing the heterogeneous loss rate to the homogeneous loss rate using the same approach as in section 4.5. The surface area density, A , of the atmospheric soot is given by A = [soot] x surface -to- mass ratio (5.2) where [soot] is the concentration of soot (g-m~3) and surface-to-mass ratio of 140 m 2 -g"' was used for the calculations [64]. By applying equation 4.3 to determine RHel and using the same parameters as before for determining RHOM, the ratios RHet/Rnom were calculated for different environmental conditions using reactive uptake coefficients equal to 0.1 and 1 as listed in Table 5.2. Table 5.2 Comparison of N 0 3 heterogeneous loss rate, Rnei, by reaction with soot to N O 3 homogeneous loss rate, RHOM, by reaction with oc-pinene for different environments. Environment [soot] (g VOL3) RHe/RHom (Y = 0.1) RHe/RHom (Y = 1) Urban 1 x 10"5 [64] 0.15 1.5 Rural 8 x 10"7 [45] 0.012 0.12 72 Contrary to the results of organic monolayers, the contribution of soot to N O 3 loss processes may be of significance. For example, if all collisions between N O 3 and soot are reactive (i.e. y = 1), calculations show this loss rate is 1.5 times faster than the N O 3 homogeneous loss rate to a-pinene in urban areas. Therefore, heterogeneous loss may be the dominant N 0 3 sink. In contrast, in rural areas, the heterogeneous loss rate is reduced due to less soot particles; thus, soot may play a lesser role (although may still be important) in N O 3 loss processes. It has been reported that aging (i.e. constant exposure to trace amounts of reactive atmospheric gas phase species) may lead to the deactivation of soot [64]. Therefore, in order to predict the amount of N O 3 degradation accurately, laboratory data on the time dependency for the process of soot aging is necessary. For this reason, the results above correspond to an upper limit since calculations were based on unreacted methane soot. Nonetheless, based on these results it is unlikely the global N O 3 level would be affected significantly but N O 3 reaction with soot may play an important role regionally. Recently, Geyer et al. [58] conducted a field study to determine the relative contribution of all major NO3 sink mechanisms. In their study, they determined that the NO3 loss to volatile organic compounds (VOCs), which consisted predominately a-pinene, contributed about 50 % of all NO3 degradation. However, they were unable to account for about 40 % of the NO3 loss. They assumed the missing sink was from the hydrolysis of N 2 O 5 . The results in Table 5-2 indicate that the heterogeneous loss may be equally important and, thus, a likely candidate to partially explain the "missing sink". The modification of the soot surface may also be of importance in terms of changing its physical properties. FTIR spectra show the presence of oxygen and nitrogen containing functional groups following exposure to NO3. The coverage of the methane soot surface by these species may alter the light absorption and scattering properties along with its hygroscopicity and toxicity. The aging of soot by exposure to N O 3 may convert the hydrophobic soot into potential cloud condensation nuclei, leading to the formation of cloud droplets at lower supersaturations of water vapour in the atmosphere. This has been observed for soot that was exposed to ozone [89]. Consequently, 73 atmospheric lifetime of soot may be reduced as a result of the enhancement of water absorption by these surface species that facilitate in its removal by wet deposition. 5.7 Summary and Conclusions The reaction of N2O5 + methane soot was investigated and the results were compared to the literature data. The reactive uptake coefficient was measured and has an upper limit of 0.026 ± 0.009, which is in good agreement with that reported by Longfellow et al. [68]. Results of the product studies are also consistent with previous studies [68], which show N O 2 as the only significant product. The reactive uptake of N O 3 by methane soot is at least one order of magnitude faster than that of N2O5. A lower limit of 0.03 was obtained for the reactive uptake coefficient for N O 3 + methane soot. Calculations show soot may be a significant N O 3 loss process in regions where there are high concentrations of soot. Product studies reveal N O 2 is the major product formed for the reaction of N O 3 + methane soot. FTIR absorption spectra show the presence of organic nitrite and nitrate groups, carbonyls, and epoxides on the methane soot surface after exposure to N O 3 . The existence of these surface species may affect the physical properties of soot through light absorption and scattering. In addition, hygroscopicity may be affected, which may lead to an increase in cloud formation and reduction in the atmospheric lifetime of soot. 74 CHAPTER 6 CONCLUSIONS AND FUTURE DIRECTIONS 6.1 Conclusions It has been demonstrated that the flow tube technique coupled to a CIMS is extremely useful in determining the reactive uptake of NO3 by various organic species. In Chapter 4, the reactive uptake of NO3 by organic monolayers was measured. In Chapter 5, the reactive uptake of NO3 by methane soot was determined. Using IR spectroscopy, the bands associated with CH X were monitored for the monolayer studies prior to and follow exposure to NO3. With the same technique, functional groups formed as a result of oxidation reaction with NO3 were observed on the methane soot. Based on the experimental results, it was shown that the concentration of NO3 in the troposphere would not be significantly depleted by organic monolayers. However, N 0 3 loss to soot may be of significance in areas where there are high concentrations of soot. The efficiency of NO3 oxidizing a saturated organic surface was compared to that of OH and other atmospheric species. Calculations show NO3 is on par with OH at oxidizing saturated hydrocarbon surfaces in the troposphere, suggesting NO3 may be an important oxidant during the nighttime. Results of the IR studies reveal no significant carbon loss for the monolayers following exposure to N 0 3 , implying the hydrocarbon chains remain in tact. Although these results are different from a similar study done with OH, the discrepancy may be explained by how the experiments were conducted as discussed in section 4.7. Based on these current findings, the removal of organic aerosols containing saturated hydrocarbons by decomposition may not be a significant channel under tropospheric conditions. For the methane soot, functional groups such as organic nitrates, organic nitrites, carbonyls, and epoxides were observed after NO3 exposure. The presence of these functional groups on the methane soot may have a wide range of implications for cloud formation, the light absorption and scattering properties, and the lifetime of soot. 75 6.2 Future Directions This thesis has provided the preliminary studies for the heterogeneous reactions involving organic coatings and soot. These studies have contributed to our understanding of heterogeneous chemistry, but additional studies are necessary to form a more complete picture. Several important studies can be conducted using the current setup to further our knowledge. For example, by performing uptake measurements of NO3 on monolayers containing different functional groups (e.g. alkene, carboxylic acid, and alcohol) would provide insights to the interactions between NO3 and the diversity of organics that are found in atmospheric aerosols. Furthermore, it is possible that a mixture of organics may contain on a single aerosol particle; hence, the uptake of NO3 by mixed-monolayers would be more appropriate proxies for these atmospheric aerosols. One of the drawbacks with the current setup is that we were unable to observe the surface products formed on the monolayer following exposure to NO3. Using techniques such as polarization modulation infrared reflection absorption spectroscopy (PMIRRAS) and X-ray photoelectron spectroscopy (XPS), this problem can be resolved. Another improvement that can be made is on the uptake coefficient for the reaction of NO3 + methane soot. Currently, only a lower limit of the actual value has been determined due to diffusion limitations. This can be improved by performing the measurements in a flow tube using submicron-sized soot particles suspended in a buffer gas since diffusion is small under these conditions. Different types of soot should also be used to determine whether they would influence the results since it has been reported that the composition of soot can vary depending on the source. In addition, the hygroscopic properties of reacted soot should be investigated. The formation of surface functional groups can enhance the absorption of water, thus changing the hydrophobic soot to potential cloud condensation nuclei. These are only a handful of suggestions for future work. The information gained from this thesis and future developments could be extremely useful in elucidating the role heterogeneous chemistry plays in the atmosphere. 76 REFERENCES 1. Finlayson-Pitts, B.J. and J.N.J. Pitts, Chemistry of the Upper and Lower Atmosphere. 2000, San Diego: Academic Press. 2. Seinfeld, J.H. and S.N. Pandis, Atmospheric Chemistry and Physics from Air Pollution to Climate Change. 1998, New York: John Wiley & Sons, Inc. 3. Facchini, M.C., M . Mircea, S. Fuzzi, and R.J. Charlson, Cloud albedo enhancement by surface-active organic solutes in growing droplets. Nature, 1999. 401(6750): p. 257-259. 4. Levin, Z., E. Ganor, and V. Gladstein, The effects of desert particles coated with sulfate on rain formation in the eastern Mediterranean. Journal of Applied Meteorology, 1996. 35(9): p. 1511-1523. 5. Murphy, D.M., J.R. Anderson, P.K. Quinn, L . M . Mclnnes, F.J. Brechtel, S.M. Kreidenweis, A . M . Middlebrook, M . Posfai, D.S. Thomson, and P.R. Buseck, Influence of sea-salt on aerosol radiative properties in the Southern Ocean marine boundary layer. Nature, 1998. 392(6671): p. 62-65. 6. Sokolik, I.N., O.B. Toon, and R.W. Bergstrom, Modeling the radiative characteristics of airborne mineral aerosols at infrared wavelengths. Journal of Geophysical Research-Atmospheres, 1998.103(D8): p. 8813-8826. 7. Kappos, A.D., P. Bruckmann, T. Eikmann, N. Englert, U. Heinrich, P. Hoppe, E. Koch, G.H.M. Krause, W.G. Kreyling, K. Rauchfuss, P. Rombout, V. Schulz-Klemp, W.R. Thiel, and H.E. Wichmann, Health effects of particles in ambient air. International Journal of Hygiene and Environmental Health, 2004. 207(4): p. 399-407. 8. Murphy, D.M. and D.S. Thomson, Chemical composition of single aerosol particles at Idaho Hill: Negative ion measurements. Journal of Geophysical Research-Atmospheres, 1997. 102(D5): p. 6353-6368. 9. Murphy, D.M. and D.S. Thomson, Chemical composition of single aerosol particles at Idaho Hill: Positive ion measurements. Journal of Geophysical Research-Atmospheres, 1997. 102(D5): p. 6341-6352. 10. Murphy, D.M., D.S. Thomson, and T.M.J. Mahoney, In situ measurements of organics, meteoritic material, mercury, and other elements in aerosols at 5 to 19 kilometers. Science, 1998. 282(5394): p. 1664-1669. 11. Middlebrook, A . M . , D.M. Murphy, and D.S. Thomson, Observations of organic material in individual marine particles at Cape Grim during the First Aerosol Characterization Experiment (ACE 1). Journal of Geophysical Research-Atmospheres, 1998. 103(D13): p. 16475-16483. 12. Jacobson, M.C., H.C. Hansson, K.J. Noone, and R.J. Charlson, Organic atmospheric aerosols: Review and state of the science. Reviews of Geophysics, 2000. 38(2): p. 267-294. 13. Novakov, T., D.A. Hegg, and P.V. Hobbs, Airborne measurements of carbonaceous aerosols on the East Coast of the United States. Journal of Geophysical Research-Atmospheres, 1997. 102(D25): p. 30023-30030. 77 14. Rogge, W.F., M.A. Mazurek, L . M . Hildemann, G.R. Cass, and B.R.T. Simoneit, Quantification of urban organic aerosols at a molecular-level -identification, abundance and seasonal-variation. Atmospheric Environment Part a-General Topics, 1993. 27(8): p. 1309-1330. 15. Didyk, B.M., B.R.T. Simoneit, L.A. Pezoa, M.L. Riveros, and A.A. Flores, Urban aerosol particles of Santiago, Chile: organic content and molecular characterization. Atmospheric Environment, 2000. 34(8): p. 1167-1179. 16. Jang, M. , R.M. Kamens, K.B. Leach, and M.R. Strommen, A thermodynamic approach using group contribution methods to model the partitioning of semivolatile organic compounds on atmospheric particulate matter. Environmental Science & Technology, 1997. 31(10): p. 2805-2811. 17. Tseng, R.S., J.T. Viechnicki, R.A. Skop, and J.W. Brown, Sea-to-air transfer of surface-active organic-compounds by bursting bubbles. Journal of Geophysical Research-Oceans, 1992. 97(C4): p. 5201-5206. 18. Ellison, G.B., A.F. Tuck, and V. Vaida, Atmospheric processing of organic aerosols. Journal of Geophysical Research-Atmospheres, 1999. 104(D9): p. 11633-11641. 19. Gill, P.S., T.E. Graedel, and C.J. Weschler, Organic films on atmospheric aerosol-particles, fog droplets, cloud droplets, raindrops, and snowflakes. Reviews of Geophysics, 1983. 21(4): p. 903-920. 20. Posfai, M . , H.F. Xu, J.R. Anderson, and P.R. Buseck, Wet and dry sizes of atmospheric aerosol particles: An AFM-TEM study. Geophysical Research Letters, 1998. 25(11): p. 1907-1910. 21. Giddings, W.P. and M.B. Baker, Sources and effects of monolayers on atmospheric water droplets. Journal of the Atmospheric Sciences, 1977. 34(12): p. 1957-1964. 22. Chesselet, R., M . Fontugne, P. Buatmenard, U . Ezat, and C E . Lambert, The origin of particulate organic-carbon in the marine atmosphere as indicated by its stable carbon isotopic composition. Geophysical Research Letters, 1981. 8(4): p. '345-348. 23. Hoffman, E.J. and R.A. Duce, Factors influencing organic-carbon content of marine aerosols - laboratory study. Journal of Geophysical Research-Oceans and Atmospheres, 1976. 81(21): p. 3667-3670. 24. Jaffe, D., I. McKendry, T. Anderson, and H. Price, Six 'new' episodes of trans-Pacific transport of air pollutants. Atmospheric Environment, 2003. 37(3): p. 391-404. 25. Formenti, P., M.O. Andreae, T.W. Andreae, C. Ichoku, G. Schebeske, J. Kettle, W. Maenhaut, J. Cafmeyer, J. Ptasinsky, A. Karnieli, and J. Lelieveld, Physical and chemical characteristics of aerosols over the Negev Desert (Israel) during summer 1996. Journal of Geophysical Research-Atmospheres, 2001. 106(D5): p. 4871-4890. 26. Hoffmann, T., C D . O'Dowd, and J.H. Seinfeld, Iodine oxide homogeneous nucleation: An explanation for coastal new particle production. Geophysical Research Letters, 2001. 28(10): p. 1949-1952. 78 27. O'Dowd, C D . , P. Aalto, K. Hameri, M . Kulmala, and T. Hoffmann, Aerosol formation - Atmospheric particles from organic vapours. Nature, 2002. 416(6880): p. 497-498. 28. O'Dowd, C D . , J.L. Jimenez, R. Bahreini, R .C Flagan, J.H. Seinfeld, K. Hameri, L. Pirjola, M . Kulmala, S.G. Jennings, and T. Hoffmann, Marine aerosol formation from biogenic iodine emissions. Nature, 2002. 417(6889): p. 632-636. 29. Daumer, B., R. Niessner, and D. Klockow, Laboratory studies of the influence of thin organic films on the neutralization reaction of H2S04 aerosol with ammonia. Journal of Aerosol Science, 1992. 23(4): p. 315-325. 30. Facchini, M . C , S. Fuzzi, S. Zappoli, A. Andracchio, A. Gelencser, G. Kiss, Z. Krivacsy, E. Meszaros, H .C Hansson, T. Alsberg, and Y. Zebuhr, Partitioning of the organic aerosol component between fog droplets and interstitial air. Journal of Geophysical Research-Atmospheres, 1999. 104(D21): p. 26821-26832. 31. Feingold, G. and P. Y. Chuang, Analysis of the influence offilm-forming compounds on droplet growth: Implications for cloud microphysical processes and climate. Journal of the Atmospheric Sciences, 2002. 59(12): p. 2006-2018. 32. Gorbunov, B., R. Hamilton, N. Clegg, and R. Toumi, Water nucleation on aerosol particles containing both organic and soluble inorganic substances. Atmospheric Research, 1998. 48: p. 271-283. 33. Seidl, W., Model for a surface film offatty acids on rain water and aerosol particles: Atmospheric Environment, 2000. 34(28): p. 4917-4932. 34. Garrett, W.D., Impact of organic material on cloud and fog processes. Pure and Applied Geophysics, 1978. 116(2-3): p. 316-334. 35. Tervahattu, H , K. Hartonen, V .M. Kerminen, K. Kupiainen, P. Aarnio, T. Koskentalo, A.F. Tuck, and V. Vaida, New evidence of an organic layer on marine aerosols. Journal of Geophysical Research-Atmospheres, 2002. 107(D7-8), 4053, doi:10.1029/2000JD000282. 36. Tervahattu, H., J. Juhanoja, and K. Kupiainen, Identification of an organic coating on marine aerosol particles by TOF-SIMS. Journal of Geophysical Research-Atmospheres, 2002.107(D16), 4319, doi:10.1029/2001JD001403. 37. Saxena, P., L . M . Hildemann, P.H. Mcmurry, and J.H. Seinfeld, Organics alter hygroscopic behavior of atmospheric particles. Journal of Geophysical Research-Atmospheres, 1995. 100(D9): p. 18755-18770. 38. Vartiainen, M . , S.R. McDow, and R.M. Kamens, Water uptake by sunlight and ozone exposed diesel exhaust particles. Chemosphere, 1996. 32(7): p. 1319-1325. 39. Rudich, Y., I. Benjamin, R. Naaman, E. Thomas, S. Trakhtenberg, and R. Ussyshkin, Wetting of hydrophobic organic surfaces and its implications to organic aerosols in the atmosphere. Journal of Physical Chemistry A, 2000. 104(22): p. 5238-5245. 40. Thomas, E., Y. Rudich, S. Trakhtenberg, and R. Ussyshkin, Water adsorption by hydrophobic organic surfaces: Experimental evidence and implications to the atmospheric properties of organic aerosols. Journal of Geophysical Research-Atmospheres, 1999. 104(D13): p. 16053-16059. 79 41. Niessner, R., B. Daeumer, and D. Klockow, Investigation of surface-properties of ultrafine particles by application of a multistep condensation nucleus counter. Aerosol Science and Technology, 1990.12(4): p. 953-963. 42. Kotzick, R. and R. Niessner, The effects of aging processes on critical supersaturation ratios of ultrafine carbon aerosols. Atmospheric Environment, 1999. 33(17): p. 2669-2677. 43. Penner, J.E., H. Eddleman, and T. Novakov, Towards the development of a global inventory for black carbon emissions. Atmospheric Environment Part a-General Topics, 1993. 27(8): p. 1277-1295. 44. Stadler, D. and M.J. Rossi, The reactivity of N02 and HONO on flame soot at ambient temperature: The influence of combustion conditions. Physical Chemistry Chemical Physics, 2000. 2(23): p. 5420-5429. 45. Alcala-Jornod, C , H. van den Bergh, and M.J. Rossi, Reactivity of N02 andH20 on soot generated in the laboratory: a diffusion tube study at ambient temperature. Physical Chemistry Chemical Physics, 2000. 2(24): p. 5584-5593. 46. Akhter, M.S., A.R. Chughtai, and D.M. Smith, The structure ofhexane soot-I -spectroscopic studies. Applied Spectroscopy, 1985. 39(1): p. 143-153. 47. Horvath, H., Atmospheric Light-Absorption - a Review. Atmospheric Environment Part a-General Topics, 1993. 27(3): p. 293-317. 48. Mauderly, J.L., Toxicological and epidemiologic evidence for health risks from inhaled engine emissions. Environmental Health Perspectives, 1994. 102: p. 165-171. 49. Wayne, R.P., I. Barnes, P. Biggs, J.P. Burrows, C E . Canosamas, J. Hjorth, G. Lebras, G.K. Moortgat, D. Perner, G. Poulet, G. Restelli, and H. Sidebottom, The nitrate radical - physics, chemistry, and the atmosphere. Atmospheric Environment Part a-General Topics, 1991. 25(1): p. 1-203. 50. Atkinson, R., Kinetics and mechanisms of the gas-phase reactions of the N03 radical with organic-compounds. Journal of Physical and Chemical Reference Data, 1991. 20(3): p. 459-507. 51. Golz, C , J. Senzig, and U. Piatt, N03-initiated oxidation of biogenic hydrocarbons. Chemosphere-Global Change Science, 2001. 3: p. 339-352. 52. Moise, T., R.K. Talukdar, G.J. Frost, R.W. Fox, and Y. Rudich, Reactive uptake ofN03 by liquid and frozen organics. Journal of Geophysical Research-Atmospheres, 2002. 107(D1-D2), 4014, doi:10.1029/2001JD000334. 53. Suh, I., W.F. Lei, and R.Y. Zhang, Experimental and theoretical studies of isoprene reaction with N03. Journal of Physical Chemistry A, 2001. 105(26): p. 6471-6478. 54. Leighton, P.A., Photochemistry of air pollution. 1961, New York: Academic Press. 55. Noxon, J.F., R.B. Norton, and W.R. Henderson, Observation of atmospheric N03. Geophysical Research Letters, 1978. 5(8): p. 675-678. 56. Noxon, J.F., R.B. Norton, and E. Marovich, N03 in the troposphere. Geophysical Research Letters, 1980. 7(2): p. 125-128. 80 57. Starn, T.K., P.B. Shepson, S.B. Bertman, D.D. Riemer, R.G. Zika, and K. Olszyna, Nighttime isoprene chemistry at an urban-impacted forest site. Journal of Geophysical Research-Atmospheres, 1998. 103(D17): p. 22437-22447. 58. Geyer, A., B. Alicke, S. Konrad, T. Schmitz, J. Stutz, and U. Piatt, Chemistry and oxidation capacity of the nitrate radical in the continental boundary layer near Berlin. Journal of Geophysical Research-Atmospheres, 2001. 106(D8): p. 8013-8025. 59. Moise, T. and Y. Rudich, Reactive uptake of ozone by aerosol-associated unsaturatedfatty acids: Kinetics, mechanism, and products. Journal of Physical Chemistry A, 2002. 106(27): p. 6469-6476. 60. Rudich, Y. , R.K. Talukdar, R.W. Fox, and A.R. Ravishankara, Rate coefficients for reactions ofN03 with a few olefins and oxygenated olefins. Journal of Physical Chemistry, 1996. 100(13): p. 5374-5381. 61. Baker, J., J. Arey, and R. Atkinson, Kinetics of the gas-phase reactions of OH radicals, NO3 radicals and 03 with three C-7-carbonyls formedfrom the atmospheric reactions of myrcene, ocimene and terpinolene. Journal of Atmospheric Chemistry, 2004; 48(3): p. 241-260. 62. Dubowski, Y., J. Vieceli, D.J. Tobias, A. Gomez, A. Lin, S.A. Nizkorodov, T.M. Mclntire, and B.J. Finlayson-Pitts, Interaction of gas-phase ozone at 296 K with unsaturated self assembled monolayers: A new look at an old system. Journal of Physical Chemistry A, 2004. 108(47): p. 10473-10485. 63. Bertram, A.K., A.V. Ivanov, M . Hunter, L.T. Molina, and M.J. Molina, The reaction probability of OH on organic surfaces of tropospheric interest. Journal of Physical Chemistry A, 2001. 105(41): p. 9415-9421. 64. Lelievre, S., Y. Bedjanian, N. Pouvesle, J.L. Delfau, C. Vovelle, and G. Le Bras, Heterogeneous reaction of ozone with hydrocarbon flame soot. Physical Chemistry Chemical Physics, 2004. 6(6): p. 1181-1191. 65. Lelievre, S., Y. Bedjanian, G. Laverdet, and G. Le Bras, Heterogeneous reaction ofN02 with hydrocarbon flame soot. Journal of Physical Chemistry A, 2004. 108(49): p. 10807-10817. 66. Bedjanian, Y., S. Lelievre, and G. Le Bras, Experimental study of the interaction ofH02 radicals with soot surface. Physical Chemistry Chemical Physics, 2005. 7(2): p. 334-341. 67. Knopf, D.A., L . M . Anthony, and A.K. Bertram, Reactive uptake of 03 by multicomponent and multiphase mixtures containing oleic acid. Journal of Physical Chemistry A, 2005. 109(25): p. 5579-5589. 68. Longfellow, C.A., A.R. Ravishankara, and D.R. Hanson, Reactive and nonreactive uptake on hydrocarbon soot: HN03, 03, andN205. Journal of Geophysical Research-Atmospheres, 2000. 105(D19): p. 24345-24350. 69. Thornton, J.A., C.F. Braban, and J.P.D. Abbatt, N205 hydrolysis on sub-micron organic aerosols: the effect of relative humidity, particle phase, and particle size. Physical Chemistry Chemical Physics, 2003. 5(20): p. 4593-4603. 70. Weast, R.C., Handbook of Physics and Chemistry. 70th ed. 1990, Boca Raton: CRC Press. 81 71. Huey, L.G., D.R. Hanson, and C.J. Howard, Reactions of SF6- and I- with atmospheric trace gases. Journal of Physical Chemistry, 1995. 99(14): p. 5001-5008. 72. Hu, J.H. and J.P.D. Abbatt, Reaction probabilities for N205 hydrolysis on sulfuric acid and ammonium sulfate aerosols at room temperature. Journal of Physical Chemistry A, 1997. 101(5): p. 871-878. 73. Hanson, D.R. and A.R. Ravishankara, The reaction probabilities of CION02 and N205 on polar stratospheric cloud materials. Journal of Geophysical Research-Atmospheres, 1991. 96(D3): p. 5081-5090. 74. Mcgovern, M.E., K.M.R. Kallury, and M . Thompson, Role of solvent on the silanization of glass with octadecyltrichlorosilane. Langmuir, 1994. 10(10): p. 3607-3614. 75. Ishida, T., S. Tsuneda, N . Nishida, M . Hara, H. Sasabe, and W. Knoll, Surface-conditioning effect of gold substrates on octadecanethiol self-assembled monolayer growth. Langmuir, 1997. 13(17): p. 4638-4643. 76. Atkinson, R., D.L. Baulch, R.A. Cox, J.N. Crowley, R.F. Hampson, R.G. Hynes, M.E. Jenkin, M.J. Rossi, and J. Troe, Evaluated kinetic and photochemical data for atmospheric chemistry: Volume I - gas phase reactions of Ox, HOx, NOx and SOx species. Atmospheric Chemistry and Physics, 2004. 4: p. 1461-1738. 77. Laidler, K.J. and J.H. Meiser, Physical Chemistry. 3 rd ed. 1999, Boston, MA: Houghton Mifflin Company. 78. Brown, R.L., Tubular flow reactors with lst-order kinetics. Journal of Research of the National Bureau of Standards, 1978. 83(1): p. 1-8. 79. Sander, S.P., Friedl, W.B. Demore, D.M. Golden, M.J. Kurylo, R.F. Hampson, R.E. Huie, G.K. Moortgat, A.R. Ravishankara, C E . Kolb, and M.J. Molina, Chemical kinetics and photochemical data for use in stratospheric modeling evaluation number 13. 2000, Pasadena: JPL Publication 00-3. 80. Thomas, E.R., G.J. Frost, and Y. Rudich, Reactive uptake of ozone by proxies for organic aerosols: Surface-bound and gas-phase products. Journal of Geophysical Research-Atmospheres, 2001. 106(D3): p. 3045-3056. 81. Moise, T. and Y. Rudich, Uptake of Cl and Br by organic surfaces - a perspective on organic aerosols processing by tropospheric oxidants. Geophysical Research Letters, 2001. 28(21): p. 4083-4086. 82. Molina, M.J., A.V. Ivanov, S. Trakhtenberg, and L.T. Molina, Atmospheric evolution of organic aerosol. Geophysical Research Letters, 2004. 31(22), L22104, doi:10.1029/2004GL020910. 83. Fiegland, L.R., M . M . Saint Fleur, and J.R. Morris, Reactions of C=C terminated self-assembled monolayers with gas-phase ozone. Langmuir, 2005. 21(7): p. 2660-2661. 84. Ulman, A., Formation and structure of self-assembled monolayers. Chemical Reviews, 1996. 96(4): p. 1533-1554. 85. Atkinson, R., Gas-phase degradation of organic compounds in the troposphere. Pure and Applied Chemistry, 1998. 70(7): p. 1327-1334. 82 86. Munoz, M.S.S. and M.J. Rossi, Heterogeneous reactions ofHN03 with flame soot generated under different combustion conditions. Reaction mechanism and kinetics. Physical Chemistry Chemical Physics, 2002. 4(20): p. 5110-5118. 87. Zuberi, B., K.S. Johnson, G.K. Aleks, L.T. Molina, and A. Laskin, Hydrophilic properties of aged soot. Geophysical Research Letters, 2005. 32(1), L01807, doi: 10.1029/2004GL021496. 88. Hohenegger, C. and P.L. Vidale, Sensitivity of the European climate to aerosol forcing as simulated with a regional climate model. Journal of Geophysical Research-Atmospheres, 2005. 110(D6), D06201, doi:10.102-9/2004JD005335. 89. Kotzick, R., U. Panne, and R. Niessner, Changes in condensation properties of ultrafine carbon particles subjected to oxidation by ozone. Journal of Aerosol Science, 1997. 28(5): p. 725-735. 90. Longfellow, C.A., A.R. Ravishankara, and D.R. Hanson, Reactive uptake on hydrocarbon soot: Focus on N02. Journal of Geophysical Research-Atmospheres, 1999. 104(D11): p. 13833-13840. 91. Keyser, L.F., S.B. Moore, and M.T. Leu, Surface-reaction and pore diffusion in flow-tube reactors. Journal of Physical Chemistry, 1991. 95(14): p. 5496-5502. 92. Leu, M.T., S.B. Moore, and L.F. Keyser, Heterogeneous reactions of chlorine nitrate and hydrogen-chloride on type-I polar stratospheric clouds. Journal of Physical Chemistry, 1991. 95(20): p. 7763-7771. 83 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0059290/manifest

Comment

Related Items