The Interplay Between Hydrogen Bonding and Coulombic Forces in

Subscriber access provided by READING UNIV

Letter

The Interplay between Hydrogen Bonding and Coulombic Forces in Determining the Structure of Sulfuric Acid-Amine Clusters Sarah E. Waller, Yi Yang, Eleanor Castracane, Emily E. Racow, John J. Kreinbihl, Kathleen A. Nickson, and Christopher J Johnson J. Phys. Chem. Lett., Just Accepted Manuscript • DOI: 10.1021/acs.jpclett.8b00161 • Publication Date (Web): 21 Feb 2018 Downloaded from http://pubs.acs.org on February 21, 2018

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 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 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 Letters 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 22 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 Letters

The Interplay Between Hydrogen Bonding and Coulombic Forces in Determining the Structure of Sulfuric Acid-Amine Clusters Sarah E. Waller, Yi Yang, Eleanor Castracane, Emily E. Racow, John J. Kreinbihl, Kathleen A. Nickson, and Christopher J. Johnson* Department of Chemistry, Stony Brook University, 100 Nicolls Road, Stony Brook, New York 11794, United States Corresponding Author *E-mail: [email protected]

1 ACS Paragon Plus Environment

The Journal of Physical Chemistry Letters 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

ABSTRACT: Acid-base cluster chemistry drives atmospheric new particle formation (NPF), but the details of the growth mechanisms are difficult to experimentally probe. Clusters of ammonia, alkylamines, and sulfuric acid, species fundamental to NPF, are probed by infrared spectroscopy. These spectra show that substitution of amines for ammonia, which is linked to accelerated growth, induces profound structural rearrangement in clusters with initial compositions (NH4+)n+1(HSO4−)n (1 ≤ n ≤ 3). This rearrangement is driven by the loss of N–H hydrogen bond donors, yielding direct bisulfate-bisulfate hydrogen bonds, and its onset with respect to cluster composition indicates that more substituted amines induce rearrangement at smaller sizes. A simple model counting hydrogen bond donors and acceptors explains these observations. The presence of direct hydrogen bonds between formal anions shows that hydrogen bonding can compete with Coulombic forces in determining cluster structure. These results suggest that NPF mechanisms may be highly dependent on amine identity.

TOC GRAPHIC


Page 2 of 22

Page 3 of 22 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 largest uncertainty in climate models arises from the impact of aerosols on cloud albedo.1 Up to 50% of aerosols formed in the atmosphere are estimated to arise from new particle formation (NPF),2 the process by which trace gases condense to create clusters that can grow to climatically-relevant cloud condensation nuclei or ice nuclei.2 However, an incomplete understanding of the mechanisms underlying formation and growth impedes efforts to model this process.1, 3-6 Field campaigns, typically using mass spectrometric techniques developed to analyze the composition of new particles,7-10 indicate that key components in NPF are usually sulfuric acid and ammonia (NH3), with significant contributions from alkylamines and oxidized organic compounds.3, 4, 6, 11-15 The presence of amines is correlated with enhanced growth,6, 12 presumably due to increased stabilization by salt bridge formation with the more basic amines.12 Laboratory experiments have provided additional insight into the chemical composition, kinetics, and growth of new particles.16-18 Flow reactor studies indicate that amines stabilize sulfuric acid dimers19 and enhance nucleation, with an efficacy of ammonia < methylamine (MA) < trimethylamine (TMA) ≲ dimethylamine (DMA).20 Growth is suggested to occur first by formation of clusters of the more-abundant ammonia and sulfuric acid, followed by exchange of amines for ammonia.16, 21, 22 Another flow reactor study shows diamines form an equal or greater number of particles compared to the monoamines at the same concentration, implying they may be more effective nucleation species.23 Atmospheric simulation chambers, particularly CLOUD, have tracked cluster formation under controlled, atmospherically-relevant conditions.24-29 Based on these studies, a general mechanism has been proposed in which clusters grow by addition of an acid-base pair regardless of cluster charge, base-stabilization leads to decreased evaporation and increased nucleation rates, and ionic and neutral clusters both participate in cluster growth.30-32 Simulations based on these studies reinforce the roles of sulfuric acid, amines, and organic compounds in NPF but do not fully reproduce trends seen in ambient measurements.33, 34 Ion mobility spectrometry and mass spectrometry experiments probing energetics35-37 and growth/exchange mechanisms38-43 support these observations. Though the composition of NPF-relevant particles is becoming clearer, the structures, chemical mechanisms, and intermolecular interactions governing NPF are not directly probed in experiments mentioned thus far. This is primarily due to the difficulty in isolating and interrogating these aspects of the clusters in a dilute, compositionally heterogeneous mixture.

As a result, computational chemistry

techniques have been used extensively to predict structures21, 44, 45 and growth mechanisms,16, 38, 41, 46-49 finding that clusters contain primarily ionic moieties stabilized by hydrogen bonds (H-bonds). Predictions of the stability of these clusters indicate that DMA- and TMA-containing clusters exhibit greater stability and growth rates than MA or ammonia, and that these effects are most pronounced for the smallest clusters of only a few sulfuric acid molecules.50-52 It is typically assumed that the arrangement of ions is driven by 3 ACS Paragon Plus Environment


Coulombic interactions, and that H-bonds play a secondary role in the overall cluster structure.16, 31, 38 The plethora of potential structural isomers complicates these calculations, leading to some uncertainties over the most viable structures.37, 46-48, 53, 54 Spectroscopic techniques are valuable tools to investigate the structure and intermolecular interactions of molecular clusters.55-57 Infrared (IR) matrix isolation spectroscopy of acid-base,58, 59 sulfuric acid,60 and amine systems;61-63 IR multiple-photon dissociation spectroscopy of clusters containing sulfuric acid;64-67 and anion photoelectron spectroscopy68,

69

have provided structural insights and benchmark data. A

previous cryogenic ion vibrational predissociation (CIVP) study identified characteristic bisulfate- and ammonium-localized modes in the fingerprint region of ammonium/methylammonium bisulfate cationic clusters that are used to analyze the spectra presented here.54 The current work establishes the structural consequences of the exchange of ammonia for DMA or TMA to better characterize the roles of proton transfer and H-bonding in determining cluster structures. This study presents the CIVP spectra of doubly D2-tagged ammonium- and alkylaminium-bisulfate cationic clusters with compositions labeled (k,m,n): (alkylaminium+)k(NH4+)m(HSO4−)n. For clarity, the alkylamine species will be denoted for the alkylamine-containing clusters [ex. (3,0,2)TMA]. The previouslydetermined structure of the (0,3,2) cluster54 is shown at the top of Figure 1. Each ammonium donates one H-bond to a bisulfate S=O group, resulting in three bridging ammonium molecules tethering two bisulfates. In this high-symmetry arrangement, the characteristic bisulfate bands provide a sensitive probe of their local H-bonding environment. Figure 1 also presents the vibrational spectra of (3,0,2) clusters of ammonia and all alkylamines. We will focus on several characteristic bands throughout this paper. The narrow free OH (3621 cm−1) and S–OH stretching (855 cm−1) modes can be expected to report on H-bonding involving the bisulfate hydroxyl group, while the SO3 symmetric stretches (1055 cm−1) indicate the symmetry of the bridging H-bonds. These bands remain essentially unchanged when MA or DMA is substituted for NH3, since each methyl group can replace a free ammonium hydrogen. Changes observed in the ammonium bending and stretching regions of Figure 1 are consistent with expectations for the substitution of alkylamines for ammonia.70, 71 CH3 bends can be clearly seen in the MA spectrum at ~1470 cm−1 and are slightly blue shifted in the DMA spectrum (gray arrow). The band associated with NH2 bends blue shifts monotonically (blue arrows) with amine basicity. In the (0,3,2) spectrum, the free NH2 symmetric and asymmetric stretches near 3400 cm-1 are easily discerned. This reduces to a single remaining free N–H in the (3,0,2)MA spectrum, and no free NH stretches are observed in the DMA spectrum.


Page 4 of 22

Page 5 of 22 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


Figure 1: CIVP spectra of the (0,3,2)·2D2 and (3,0,2)·2D2 aminium bisulfate clusters, where the amine is ammonia (NH3), methylamine (MA), dimethylamine (DMA), or trimethylamine (TMA). The labels (k,m,n) denote (alkylaminium+)k(NH4+)m(HSO4−)n. Solid arrows indicate a band shift. The previously determined structure of the (0,3,2) cluster is also included.54 A clear spectral change is seen upon substitution by TMA, indicating a significant structural rearrangement.

Substitution by TMA results in a remarkably different spectrum. Most notably, all modes involving free OH groups are absent, and the SO3 symmetric and asymmetric stretching regions bear essentially no resemblance to the NH3, MA, and DMA spectra. The observation of this abrupt change is not consistent with the monotonic trends in basicity of the amines and suggests that a significant structural rearrangement has occurred. Analysis of the structure for the (0,3,2) cluster immediately reveals the cause of this rearrangement. The structure requires two H-bonds to bridge the bisulfates, but protonated TMA has only one H-bond donor, thus disrupting the bridging motif. This must necessitate a large-scale rearrangement to generate an optimal H-bonding network. Given the number of characteristic spectral markers resolved in these spectra, the putative structures can be deduced by inspection. Thus, we examine the evolution of the spectra upon sequential substitution of TMA for NH3. Figure 2 presents the spectra of (0,3,2) through (3,0,2)TMA clusters. The monotonic evolution of the broad, intense features in the H-bonded stretching region confirms that the spectrum of (3,0,2)TMA indeed originates from that of (0,3,2). The fingerprint region of the (1,2,2)TMA spectrum shows that the characteristic bisulfate stretches, while significantly perturbed, are still 5 ACS Paragon Plus Environment


recognizable when compared to (0,3,2). The new features observed in the (2,1,2)TMA and (3,0,2)TMA spectra are qualitatively different from those in the (0,3,2) spectrum, with no characteristic bisulfate bands, indicating that the bisulfates no longer reside in the high symmetry arrangement of the previous clusters.

Figure 2: CIVP spectra of the (0,3,2)·2D2 cluster with incremental substitution of TMA for NH3. The labels (k,m,n) indicate the chemical composition (TMAH+)k(NH4+)m(HSO4−)n of the cluster. Dashed arrows indicate a loss in band intensity. The first exchange of TMA results in a distorted but still qualitatively similar spectrum, while further exchanges yield very different spectra with no indication of the initial features. The H-bonded stretching region shows an essentially monotonic evolution with respect to substitution.

The free X-H stretching regions of these spectra provide insight into to the structures resulting from the three sequential TMA substitutions, with the proposed mechanism depicted in Scheme 1. NH2 symmetric and asymmetric stretching features are apparent until the third substitution, indicating that the remaining ammonium ions are still in bridging positions. The free OH stretch is apparent after the first TMA substitution but not after the second, suggesting that they are the likely binding sites of the first two TMA molecules. Proton transfer to TMA likely stabilizes this arrangement, but destabilizes a bridging ammonium, easing the loss of a neutral and resulting in a direct bisulfate-bisulfate H-bond. After the second substitution occurs, no other bisulfate H-bond donors are available, and the third TMA must substitute for an ammonia in a bridging position, necessarily breaking the bridge and yielding a free S=O group. In the final structure, no free OH or NH bonds remain, and indeed no corresponding stretches are seen in the spectrum. The H-bonded stretching region confirms that all three trimethylamines are protonated, with the lower-energy NH stretch of TMAH+ approaching the expected value of 2172 cm−1 for the TMAH+·HSO4– 6 ACS Paragon Plus Environment

Page 6 of 22

Page 7 of 22 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


dimer as identified in matrix isolation spectra.59 The direct binding of formally negatively charged bisulfates indicates that stabilization due to H-bonding is competitive with Coulombic repulsion, and thus the two compete in determining cluster structure. The (3,0,2)TMA cluster is produced in high yield in our instrument, suggesting that this arrangement is quite stable.

Scheme 1. The proposed structural consequences of sequential substitution of TMA for NH3 in the (0,3,2) cluster.

Given that the rearrangements in the (0,3,2) cluster are induced when a doubly H-bonded ammonium is replaced by a TMAH+ that can only donate one H-bond, we extrapolate that DMA should similarly disrupt clusters with triply H-bonded ammonium and MA for quadruply H-bonded ammonium. It has been previously shown that exchange rates of interior quadruply H-bonded ammonium for DMA should be much slower than for “surface” sites.16, 39, 53 We have developed a simple model to predict the onset of these effects, shown in Table 1 for several cationic and neutral clusters. This model makes no structural assumptions; it simply counts the number of H-bond donors on conjugate acids and acceptors on conjugate bases. Example calculations can be found in the Supporting Information. Shaded regions of the table indicate cluster compositions in which there are fewer donors than acceptors, nominally leaving at least one S=O site open. Ammonium-containing cationic clusters feature a surplus of H-bond donors for all cluster sizes, while more substituted amines induce this effect at smaller cluster sizes.



Table 1. Comparison of available H-bond acceptor and donor sites in cationic and neutral clusters containing 1–4 bisulfate residues.

Cluster Cationic (2,1) (3,2) (4,3) (5,4) Substituted (0,4,3) (1,3,3) (2,2,3) (3,1,3) (4,0,3) Neutral (1,1) (2,2) (3,3) (4,4) Substituted (0,2,2) (1,1,2) (2,0,2) (0,3,3) (1,2,3) (2,1,3) (3,0,3) (0,4,4) (1,3,4) (2,2,4) (3,1,4) (4,0,4) a b

Acceptor sitesa 3 6 9 12 9 9 9 9 9 3 6 9 12 6 6 6 9 9 9 9 12 12 12 12 12

NH4+ 8 12 16 20 NH4+ – – – – – NH4+ 4 8 12 16 NH4+ – – – – – – – – – – – –

H-bond donor sitesa MAH+ DMAH+ TMAH+ 6 4 2 9 6 3 12 8 4 15 10 5 + + MAH DMAH TMAH+ – – – 15 14 13 14 12 10 13 10 7 12 8 4 + + MAH DMAH TMAH+ 3 2 1 6 4 2 9 6 3 12 8 4 + + MAH DMAH TMAH+ – – – 7 6 5 6 4 2 – – – 11 10 9 10 8 6 9 6 3 – – – 15 14 13 14 12 10 13 10 7 12 8 4

See S3 for calculations of acceptor and donor sites. Clusters with fewer donor than acceptor sites are highlighted gray.

Once all aminium H-bond donors are bound, the next most favorable donor is a free bisulfate OH, suggesting the nature of the structural rearrangement is to form direct bisulfate-bisulfate H-bonds at the open S=O. These bisulfate OH donors are not counted with the amine NHs in our model so that this effect can be explicitly identified. The model predicts this rearrangement when TMA is substituted for NH3 in the (0,3,2) cluster as discussed above, and further predicts that DMA substitution should yield similar rearrangement in the next larger (0,4,3) cluster, specifically upon the fourth substitution.


Page 8 of 22

Page 9 of 22 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


Figure 3: Comparison of the fingerprint region of the CIVP spectra of clusters undergoing incremental substitution of MA (left) or DMA (right) for NH3 in the (0,4,3)·2D2 cluster. The labels (k,m,n) denote (DMA/MAH+)k(NH4+)m(HSO4−)n. Features unique to the (4,0,3)DMA spectrum are marked with an asterisk (*). For MA substitution, no appreciable change is seen in any cluster, while for DMA substitution, the fourth substitution gives rise to a structural change as evidenced by the new features in its spectrum.

To confirm this prediction, we examine the far-IR spectra for sequential substitution of DMA for NH3 as presented in Figure 3 (right column). As anticipated, the characteristic bisulfate bands remain intact through the first three substitutions, implying minimal structural change, while in the (4,0,3)DMA spectrum, new features arise in the fingerprint region as labeled by asterisks. Since our model indicates that MA substitution will not disrupt the (0,4,3) cluster structure, we examine the spectra of MA substituted (0,4,3) clusters in the left column of Figure 3. As anticipated, the characteristic bisulfate bands are unchanged upon substitution of any number of MA molecules, confirming that the cluster structure is not perturbed. A comparison of the (0,4,3), (4,0,3)MA, and (4,0,3)DMA spectra and a table containing the characteristic stretching frequencies can be found in the Supporting Information. Just as TMA substitution forces the bisulfate moieties to interact more strongly with each other due to the loss of H-bond donors and increased competition for the remaining protons, it is likely that DMA has the same effect on (0,4,3). This is supported by the fact that all of the new features in this spectrum arise in the region of bisulfate modes. The characteristic bisulfate features discussed above also remain in the (4,0,3)DMA spectrum, which is expected as the predicted structures for the (0,4,3) cluster feature both doubly- and triply-H-bonded ammonium 9 ACS Paragon Plus Environment


constituents,44, 53, 54 and DMA substitution at doubly-H-bonded sites is not disruptive. This model can be extended to neutral clusters, which are extremely difficult to probe experimentally. Neutrals with an (n,n) formula must undergo rearrangements if DMA and TMA are completely substituted for NH3. The point at which partially substituted neutral clusters rearrange can also be predicted and are included in Table 1 for clusters with 2–4 bisulfate moieties. The trend for neutrals is similar to that for the cationic clusters, suggesting that our structural conclusions generally hold for the neutral clusters as well. Examining a selection of previously proposed structures for neutral clusters shows that these bisulfate-bisulfate interactions are common in neutral clusters that are predicted to show fast growth rates.16, 22, 50-52, 72 In fact, one recently published computational study indicates that low energy structural motifs for neutral TMA-sulfuric acid clusters73 are qualitatively similar to our proposed structures in Scheme 1, supporting our extension of the model to neutral clusters. Our model may be extended to clusters containing diamines, which have been suggested as more effective nucleation agents than monoamines.23 While experimental studies of clusters containing diamines are beyond the scope of this paper, in our current model diamines would be exactly reproduced by two monoamines. For example, 1,2-dimethylethylenediamine would be equivalent to two DMA molecules. Treating a diamine as two monoamines clearly does not account for the geometric strain a diamine would have compared to two monoamines, but these effects are not considered in the model in any case. DMA and TMA were found to most significantly enhance sulfuric acid dimer stability and nucleation, with DMA having a slightly larger effect than TMA despite its lower gas phase basicity.19, 20, 22, 52

This work indicates that, despite this similarity, the specifics of their respective growth mechanisms may

differ. When amine substitution leads to fewer H-bonds, there is a loss of stability that could offset the gain due to increased proton transfer. This can explain the fact that the trend for NPF enhancement among the amines (NH3 < MA < DMA ≳ TMA) lies intermediate between the trend for gas phase basicity (NH3 < MA < DMA < TMA), in which hydrogen bonding is irrelevant, and that for pKa (NH3 < MA ≲ DMA > TMA).74 This explanation supports very recent computational work showing that the binding energy in small clusters is primarily determined by gas phase basicity, while slightly larger clusters follow more closely aqueous basicity.73 Amine exchange also mediates the balance of donor and acceptor sites on the exterior of the cluster, changing the availability of sites with H-bond donor-acceptor pairs for additional sulfuric acid or carboxylic acid moieties. The increased Coulomb repulsion from bisulfate-bisulfate interactions may also facilitate insertion of additional species into the interior of the cluster during growth. Of particular interest is the complex role of water, which could either insert between cluster constituents or bind to the cluster surface, and may present as neutral water or the hydronium ion (H3O+) according to recent computational results.49, 66, 72 This distinction will further clarify the competition between hydrogen 10 ACS Paragon Plus Environment

Page 10 of 22

Page 11 of 22 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


bonding and proton transfer in these clusters. Studies are currently underway in our group to probe the impact of these rearrangements on structure and uptake in these clusters. Experimental Details Ions are generated by positive-mode electrospray ionization (ESI) of solutions containing 1mM ammonium sulfate or 1 mM ammonium sulfate/2 mM MA or DMA in a 50/50 water/methanol solvent with 0.01%/vol. formic acid. The ESI emitter is housed in a gas-tight volume into which 1 atm of dry and CO2free air (0.0% RH,

The Interplay Between Hydrogen Bonding and Coulombic Forces in

Recommend Documents