Subscriber access provided by CORNELL UNIVERSITY LIBRARY
Article
Competitive Adsorption of Acetic Acid and Water on Kaolinite Valerie J Alstadt, James David Kubicki, and Miriam Arak Freedman J. Phys. Chem. A, Just Accepted Manuscript • DOI: 10.1021/acs.jpca.6b06968 • Publication Date (Web): 05 Oct 2016 Downloaded from http://pubs.acs.org on October 13, 2016
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
The Journal of Physical Chemistry A is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 34
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The Journal of Physical Chemistry
Competitive Adsorption of Acetic Acid and Water on Kaolinite Valerie J. Alstadt1, James D. Kubicki2, Miriam Arak Freedman1,* 1) Department of Chemistry, The Pennsylvania State University, University Park, PA 16802 2) Department of Geological Sciences, University of Texas at El Paso, El Paso, TX 79968 Revised for the Journal of Physical Chemistry A August 8, 2016 * To whom all correspondence should be addressed:
[email protected] and (814) 867-4267 Abstract Mineral dust is prevalent in the atmosphere due to emissions from natural and anthropogenic sources. As mineral dust particles undergo long distance transport, they are exposed to trace gases and water vapor. We have characterized the interactions of acetic acid on kaolinite using Diffuse Reflectance Infrared Fourier Transform Spectroscopy and molecular modeling to determine the chemisorbed species present. After the addition of acetic acid, gas phase water was introduced to explore how water vapor competes with acetic acid for surface sites. We have found that four chemisorbed acetate species are found on kaolinite after exposure to acetic acid in which acetate bonds through a monodentate, bidenatate, or bidentate bridging linkage with an aluminum atom. These species exhibit varying levels of stability after the introduction of water, indicating that water vapor affects the adsorption of organic acids. These results indicate that the type of chemisorbed species determines its stability toward competitive adsorption, which has potential implications for atmospheric composition and ice nucleation.
1
ACS Paragon Plus Environment
The Journal of Physical Chemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 2 of 34
Introduction Wind action over arid soils results in the emission into the atmosphere of 1000-3000 Tg of mineral dust every year.1, 2 Up to 50% of mineral dust may be anthropogenic in origin,3 yet more information is needed regarding the natural sources and emission amounts of these components.3, 4
Mineral dust less than 1.8 µm in diameter can be subject to long distant transport over thousands
of kilometers and can be exposed to many atmospheric constituents.5,
6
The heterogeneous
chemistry of mineral dust is of particular interest because it can be a source or sink for atmospheric gas phase species. For example, mineral dust has been found to be a sink for reactive sulfur and nitrogen species.7 Mineral dust may also be a source for new atmospheric constituents as reactive metal oxides in mineral dust can produce OH radicals, which in the presence of SO2, can lead to the production of H2SO4 in the atmosphere.8 Additionally, mineral dust aerosol is a sink for carboxylic acids such as oxalic and malonic acid.1,
9, 10
These interactions may result in the
production and removal of compounds from the environment, affecting atmospheric composition and radiative balance. In general, the reaction mechanisms for these heterogeneous processes are not well understood. Organic compounds are often associated with mineral dust. For instance, sampling during the Atlanta SuperSite Project found particles categorized as mineral dust due to the presence of alumina or aluminosilicate and 95% of these collected mineral dust particles contained watersoluble organics, likely organic acids.11 Carboxylic acids comprise 25% of nonmethane hydrocarbon loading in the atmosphere, of which formic and acetic acids are the first and second most abundant, respectively.12 Acetic acid is present in the atmosphere due to emissions from biomass burning, automobile exhaust, and biogenic emissions from soil and vegetation,13-16 and is
2
ACS Paragon Plus Environment
Page 3 of 34
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The Journal of Physical Chemistry
present in urban, rural and marine environments.12,
17, 18
Estimates of gas phase acetic acid
concentrations in the troposphere range from 0.05–16 ppbv.17 Mineral dust is comprised of aluminosilicate clay minerals in addition to oxides, carbonates, quartz, and feldspars. Approximately 85% by weight of Asian mineral dust and 60% by weight of Saharan mineral dust is comprised of aluminosilicate clay minerals.7, 19 Kaolinite is one of the most common types of aluminosilicate clay minerals present in mineral dust aerosol.20 Kaolinite is composed of alternating layers of tetrahedrally coordinated silicon and octahedrally coordinated aluminum, where the aluminosilicate layers are held together with hydrogen bonds.21 Kaolinite has a basal plane as well as edge sites which together contain many different types of surface sites, although the edge sites contain more reactive hydroxyl groups compared to the hydroxylated and non-hydroxylated basal planes. A variety of different types of hydroxyl groups are found on the surface of kaolinite, including AlOH, AlOH2+ and hydroxyl groups that bridge two aluminum atoms as well as SiOH groups.21 Due to the complex nature of aluminosilicate clay minerals, prior studies have focused on the interaction of mineral dust proxy materials such as Al2O3 and silica with small organic acids and water.22-26 It has been found that acetic acid adsorbs to the surface of Al2O3 as acetate.25 After modeling the interactions of Al2O3 using Al2(OH)4 with acetate, Tong et al. found that the frequencies are best explained by acetate forming a chemisorbed structure bridging the two aluminum atoms.25 Although mineral dust analogs have been probed, acetic acid interactions on kaolinite have not been well characterized outside of aqueous reactions.27 Kaolinite is a more complex surface with reactive sites on the edge and the basal plane, while Al2O3 is frequently modeled as Al2(OH)4. Modeling the interactions of acetic acid with kaolinite can provide us with
3
ACS Paragon Plus Environment
The Journal of Physical Chemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
a more realistic model of the interactions of small organic acids with aluminosilicate clay minerals in the atmosphere. Mineral dust has been shown to interact with carboxylic acids in aqueous solution as well as adsorb onto dry particles.12, 25 Surface adsorbed water influences how much acid adsorbs to the mineral surface, although it has different effects, depending on the exact system used. For example, when the relative humidity is increased, some mineral dust species exhibit increased adsorption of organic acids,22-24, 28 while others report a decrease in adsorption after the addition of water.25 The specific surface interaction of a mineral with water determines how the reactivity of the surface will change toward the adsorption of trace gas species. Typically, when the relative humidity is increased, multilayers of water form on the surface.29 As a result, some of the interactions between water, trace gases, and minerals are due to liquid phase water, and others are presumably due to gas phase water. By introducing gas phase water in a purged cell heated above 100 °C, we can separate the effects of gas phase water from surface adsorbed water to develop a greater understanding of heterogeneous chemistry in these systems. In this study, we use Diffuse Reflectance Infrared Fourier Transform Spectroscopy (DRIFTS) coupled with molecular modeling to probe how acetic acid interacts with the kaolinite surface. We subsequently probe how gas phase water competes for surface sites occupied by chemisorbed acetic acid. The introduction of acetic acid followed by water allows us to understand how acetic acid adsorbs to the surface without the presence of liquid water and how gas phase water can compete for surface sites. In comparison to literature studies, we can hypothesize the respective roles of gas and liquid phase water on the adsorption of this small organic molecule on kaolinite.
4
ACS Paragon Plus Environment
Page 4 of 34
Page 5 of 34
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The Journal of Physical Chemistry
Experimental DRIFTS Low defect kaolinite (KGa-1b, Washington County, GA) with a known anatase impurity (Chiper and Bush. 2001) was obtained from the Source Clays Repository of the Clay Mineral Society (West Lafayette, IN). Kaolinite samples were size selected to less than 75 µm by sieving to remove quartz impurities. IR spectra of the sample were obtained using Diffuse Reflectance Infrared Fourier Transform Spectroscopy (DRIFTS; Bruker Optics Vertex 70, Harrick Praying Mantis DRIFTS accessory) in a 1:25 mass ratio with KBr (International Crystal Laboratories, FTIR Grade) to reduce signal saturation. The sample holder on the DRIFTS spectrometer is enclosed by a hood with CaF2 windows. To minimize physisorption of water on the kaolinite surface, the sample holder was heated with a variable temperature control and maintained under an argon purge flow during sample preparation and experiments. Samples were loaded at 50 °C and heated to a set point temperature that corresponds to a sample surface temperature of 185°C initially and 175 °C for experiments. This heating does not change the structure of or remove hydroxyl groups from kaolinite as measurements from thermogravimetric analysis – mass spectrometry (TGA-MS) have determined that hydroxyl bonds in the mineral are not broken until 545 °C.30 Six µL of glacial acetic acid (Sigma Aldrich, >99% ReagentGrade®) were introduced into the sample cell via an injection port heated to 175 °C, in order that acetic acid entered the cell in the vapor phase. After dosing, spectra were obtained at 2-3 minute intervals for 15 minutes while the temperature of the sample was held constant. The spectra reported were recorded at the end of these 15 minutes. Spectra confirm that acetic acid that does not adsorb to the sample surface remains in the gas phase and is carried away by the purge flow. These samples are referred to as kaolinite after acetic acid exposure. Excess acetic acid was removed by the argon flow. After the 15 minute interval was
5
ACS Paragon Plus Environment
The Journal of Physical Chemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
complete, 2 µL of water (Millipore grade) was introduced into the injection port. Additional water amounts were injected at 5 minute intervals until the total exposure was 3, 4, 5, 6, 8, 10, and 14 µL. 5 min was sufficient time for the spectra to stabilize. These samples are referred to as kaolinite after acetic acid and water exposure. Note that the relative humidity within the chamber is not held constant during the course of an experiment. However, the maximum relative humidity that the sample experiences can be calculated assuming that all of the injected water vapor reaches the sample chamber at the same time. Using the volume of the DRIFTS chamber and the saturation vapor pressure of water vapor at 175 °C,31 we find that the samples are exposed to 1.1 to 4.4 %RH during the course of the experiment. According to Schuttlefield et al., this corresponds to submonolayer coverages to approximately 1.5 monolayers of water.29 Such low concentrations of water vapor may be relevant to conditions in the upper troposphere, where the relative humidity can vary widely between nearly 0 % due to entrainment of stratospheric air to 100 % in storm systems.32 A sample of pure KBr was also exposed to acetic acid. Spectra are reported as the – log(Obtained Spectrum/Reference) where the reference is the sample immediately prior to acetic acid exposure unless otherwise noted. Referencing the sample in this way provides the spectrum of the surface and removes the peaks from the bulk kaolinite structure. Reported spectra have a resolution of 6 cm-1 and are comprised of 400 scans. The spectra were baseline corrected at a nonadsorbing wavenumber and CO2 was subtracted from the spectrum.
Computational Studies Molecular modeling was used to interpret the observed changes in the DRIFTS spectra. The kaolinite base structure used was courtesy of C. Ignacio Sainz-Diaz and consisted of a cluster of six Si atoms and two Al atoms with the appropriate number of O and H atoms.33 Nine structures
6
ACS Paragon Plus Environment
Page 6 of 34
Page 7 of 34
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The Journal of Physical Chemistry
were constructed in which acetic acid or acetate molecules were bonded to Al or Si atoms in the kaolinite structure (Fig. S1). These structures represent all the possible configurations of acetic acid and acetate on kaolinite. Six of these structures were duplicated and hydrated. One or two H2O molecules were added so that hydrogen bonds could form with two H2O molecules (Fig. S2). The positions of the atoms in the structures were optimized using Gaussian with the B3-LYP functional using a 6-31G(d,p) basis set as well as the M05-2x functional using a 6-31G(d,p) basis set.34-36 After optimization, the vibrational frequencies of the structures were calculated using the same basis set. Frequencies that originated entirely from kaolinite were eliminated. As stated above, the bulk spectra were ratioed to kaolinite prior to dosing to remove bulk kaolinite frequencies. Only frequencies that involved acetic acid or acetate were considered. As described by Merrick et al., a scalar factor of 0.9648 was applied to the frequencies calculated using B3-LYP and a scalar factor of 0.9419 was applied to the frequencies calculated using M05-2x for direct comparison with the observed results, as these factors are known to better correlate theoretical and observed frequencies.37 These factors have been shown to be appropriate for acetic acid.38 For the structures that produced the observed frequencies, we confirmed that all the frequencies for these structures were present in the measured spectra that were ratioed to remove bulk kaolinite frequencies. In addition, any structure that produced a vibration that was not observed was discarded. To predict the stability of the bound acetate species on the kaolinite surface in the presence of water, the difference between the free energies of the free protonated acetic acid and the bound acetate was determined by using the calculated free energy from the Gaussian frequency calculation. The bound acetate was modeled in the presence of water to represent the water introduced to the sample. The kaolinite structure with the free acetic acid was represented where the H2O molecules have reacted with the Al atoms on the kaolinite surface to form OH and OH2
7
ACS Paragon Plus Environment
The Journal of Physical Chemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
groups. Results and Discussion Acetic Acid Adsorption on Kaolinite Figure 1 compares the spectrum of KGa-1b diluted with KBr before and after the introduction of acetic acid, where the spectra are referenced to KBr. All kaolinite samples discussed in this manuscript were diluted with KBr to minimize saturation in the FTIR spectra. Kaolinite exhibits peaks characteristic of the hydroxyl region from 3000-4000 cm-1 and peaks characteristic of Si-O and Al-O stretching from 1000-2000 cm-1. These peaks are consistent with the spectrum observed for KGa-1b by Madejova et al. except for the addition of an anatase (TiO2) impurity at 2600 cm-1.39 The anatase impurity has also been identified in X-ray diffractograms of this sample.40 Literature samples are more highly diluted with KBr and as a result, the anatase impurity is not observed.39 Note that if quartz impurities have not been completely removed by the sieving, then the vibrational modes of quartz will overlap with KGa-1b in the Si-O stretching region.41 The peaks shown in Figure 1 are indicative of the bulk structure of kaolinite. As a result, no differences are present between the spectra of kaolinite before and after the addition of acetic acid when KBr is used as a reference. To determine the effect of the presence of KBr on the kaolinite spectra with acetic acid, the adsorption of acetic acid on KBr was investigated. Acetic acid was introduced to a pure KBr surface and the resultant spectra were referenced to KBr prior to acetic acid exposure (Figure 2). Two peaks are present at 1589 and 1406 cm-1, indicating that some acetic acid adsorbs on KBr in the diluted kaolinite, but this interference is confined to two distinct peaks. When the spectrum of kaolinite with acetic acid is referenced to kaolinite immediately prior to dosing, the surface functional groups of kaolinite that interact with acetic acid are visible
8
ACS Paragon Plus Environment
Page 8 of 34
Page 9 of 34
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The Journal of Physical Chemistry
because the features that arise due to bulk functional groups are the same for both samples and are removed by using kaolinite as the reference (Figure 3). In Figure 3, the interactions of acetic acid with the hydroxyl groups of kaolinite are featured from 3000-3500 cm-1 while the interactions of acetic acid with the aluminum and silicon atoms of kaolinite could arise in the 1000-1800 cm-1 region. There is some interference due to the presence of gas phase water. To determine the interactions of acetic acid with the surface that lead to the peaks in the DRIFTS spectra, we have used molecular modeling.
Molecular Modeling of Acetic Acid on Kaolinite We have performed computational studies on acetic acid adsorption to kaolinite to assign the peaks in the 1000-1800 cm-1 region of Fig. 3 (Table 1 and Table 2). While the hydroxyl region of kaolinite exhibits changes after the addition of acetic acid, we are focusing on the 1000-1800 cm-1 region, as the changes in the hydroxyl region indicate that hydroxyl groups are lost, but provide little insight into the exact structures that have formed at these sites and whether or not these are due to aluminum or silicon hydroxyl groups. The region that we have focused on does not overlap with the vibrational modes of the anatase impurity. Molecular models were considered that included acetate or acetic acid interactions with the kaolinite cluster model as well as corresponding hydrated structures. Four models account for all the peaks observed in the 10001800 cm-1 region: Al-Acetate1, Al-Acetate2, Al2-bidentate, and hydrated Al2-bidentate (Figure 4). In addition, every vibrational frequency calculated for these structures is consistent with the experimental spectrum (Table 1). Each of the remaining constructed structures shown in Fig. S1 were eliminated because they have vibrational frequencies that were not observed experimentally (Figure S1).42 Note that all of the interactions of acetic acid with the surface involve an Al-O
9
ACS Paragon Plus Environment
The Journal of Physical Chemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
linkage rather than Si-O, indicating that any low concentration quartz impurity is not the primary species involved in the observed vibrational mode. Theoretical peak positions can be determined and plotted vs. the observed frequencies (Fig. 5). Tables 1 and 2 show the slope and R2 values of the correlation between the theoretical and observed frequencies using these four models using different functionals. Some frequencies have contributions from multiple model structures. The four structures shown in Figure 4 are responsible for the observed peaks. The agreement between the different functionals indicate that the assignment of the model conformations to the observed frequencies is correct. Interestingly, they include acetate that bonds through monodentate, bidentate, and bridging structures with Al-O. Figure 6 is a compilation of the experimental and computational results. The peaks shown arise from acetate interactions with Al-O, and include peaks at frequencies of 1658 (A), 1589 (B), 1481 (C), 1432 (D), 1331 (F), and 1197 (G) cm-1. The peaks at 1589 (B) and 1406 (E) have contributions from acetic acid adsorbed to KBr and water. The peaks seen after the introduction of acetic acid on kaolinite are similar to those seen by Kubicki et al. in shape and position even though their sample preparation was in the aqueous phase.27 Tong et al. reacted Al2O3 with acetic acid and found that bridging structures formed between acetate and kaolinite on the surface in similar positions as Figure 6 for Al2-bidentate.25 Acetate formation was also observed on Al2O3 upon the addition of acetic acid by Ma et al. and Tang et al.22, 43 Rubasinghege et al. reacted formic acid with kaolinite and Al2O3 and found that formic acid reacts with surface hydroxyl groups to form chemisorbed molecules on Al2O3.24 Our observed models are consistent with the structures modeled in Rubasinghege et al. for the adsorption of formic acid on aluminum sites.24 Our results are also consistent with the chemisorption and loss of a hydroxyl group seen by Prince et al. for
10
ACS Paragon Plus Environment
Page 10 of 34
Page 11 of 34
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The Journal of Physical Chemistry
the interaction of calcite and acetic acid where two acetic acid molecules bind to CaCO3 to form water and carbon dioxide.28
Addition of Water Vapor Also shown in Figures 3 and 6 are spectra of acetic acid adsorbed on kaolinite with subsequent exposure to water vapor. Note that if a given amount of water is dosed prior to acetic acid exposure, then an identical spectrum results as for acetic acid with subsequent exposure to that given amount of water. Figure S3 depicts a peak fitted spectrum as well as the six individual peaks fitted to form the observed spectrum. Peak E is not included, because it does not have a determinable peak intensity to match in the peak fitting of the original spectrum, as the KBr-acetic acid interaction is weak compared to the Al-acetate interaction. We have peak fit the spectrum of kaolinite after exposure to acetic acid to determine which peaks are decreasing and increasing in intensity with increasing water exposure (Table S1). As the kaolinite surface is exposed to increasing amounts of water vapor, the peaks at 1432, 1481, and 1658 cm-1 (D, C, and A) decrease in intensity. The peak at 1198 cm-1 (G) becomes more distinct after the addition of water vapor and then is constant. The peak at 1331 cm-1 (F) decreases at the initial addition of water vapor and then remains at the same intensity. However, some of this decrease is due to the overlap of peak F with peak E, which becomes more distinct and narrower (Fig. S3). The peak at 1406 cm-1 (E), due to acetic acid interactions with KBr, is clearly visible after the addition of water vapor due to the decrease in intensity of peak D. The presence of negative changes in peak intensity indicate that some interactions between kaolinite and acetic acid are unstable in the presence of water vapor, while the presence of other peaks that remain constant may indicate that some interactions are stable.
11
ACS Paragon Plus Environment
The Journal of Physical Chemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The continued presence of some peaks while other peaks decrease in intensity and area is in contrast to the literature that tends to observe a total change or no change in the adsorption of carboxylic acids on metal oxides as relative humidity increases.24, 25 Rubasinghege et al. observed an increase in the uptake of formic acid on Al2O3 after the relative humidity was increased from
