Article pubs.acs.org/JPCA
Interaction of Glycine with Common Atmospheric Nucleation Precursors Jonas Elm,*,† Mehrnoush Fard,† Merete Bilde,‡ and Kurt V. Mikkelsen† †
Department of Chemistry, H. C. Ørsted Institute, University of Copenhagen, Universitetsparken 5, DK-2100 Copenhagen, Denmark Department of Chemistry, Aarhus University, Langelandsgade 140, DK-8000 Århus C, Denmark
‡
S Supporting Information *
ABSTRACT: The interaction between the simplest amino acid glycine in three different protonation states and common atmospheric nucleation precursors (H2O, NH3, and H2SO4) has been investigated using computational methods. Each nucleation step has been thoroughly sampled, and statistical Gibbs free energies of formation have been calculated using M06-2X/6-311++G(3df,3pd). From the stepwise ΔG values, the stabilities of the molecular clusters have been evaluated. Glycine in all three protonation states is found to have a favorable interaction with sulfuric acid with a higher cluster stabilizing effect than ammonia. The deprotonated glycine molecule is found to yield the highest stabilizing effect on the sulfuric acid clusters through the interaction of both the amino and carboxylic moieties, while the protonated glycine molecule is found to have a high stabilizing effect on the addition of water and ammonia. Furthermore, we find that a single sulfuric acid molecule is capable of stabilizing the glycine zwitterion. Sulfuric acid is found to be able to catalyze the spontaneous formation of the zwitterion and subsequently stabilize the formed ion. The formation of the glycine zwitterion occurs with a low Gibbs free energy barrier of 2.10 kcal/mol, indicating that this formation could occur rapidly in the atmosphere.
1. INTRODUCTION
Since organic acids have been shown to enhance nucleation and amines have been shown to interact strongly with sulfuric acid, the geminate nature of the amino acids poses as an interesting possibility for involvement in aerosol formation and growth. New particle formation in marine environments has earlier been correlated with oxidation products of the amino acid L-methionine.49 Recently, it was shown that dicarboxylic acids exhibited the intriguing trend of enhancing nucleation in two directions in geometrical space due to having two acid moieties.30 Similar effects could be possible with amino acids. The intrinsically amphiprotic nature of amino acids allows for protonation states which might enhance the interaction with aerosol precursors through the formation of ionic species. Considering the properties of amino acids and the known presence in aerosols, we here wish to computationally investigate the interaction between common aerosol nucleation precursors (H2O, NH3, H2SO4) and amino acids, using the simplest amino acid glycine (NH2CH2COOH) in various states of protonation as a model system.
The formation and growth of tropospheric aerosols via nucleation plays an important role in Earth’s climate and atmospheric chemistry.1 The assessment of the impact of new particle formation and growth is not possible without a fundamental understanding of the mechanism behind nucleation. Sulfuric acid has been accepted as a key molecule in atmospheric nucleation, but the participation of other nucleating vapors remains unknown. The participation of an atmospheric stabilizer such as ammonia,2−6 amines,7−12 ionic species,13−21 and organic compounds22−36 in the nucleation process has previously been examined. While omnipresent in atmospheric aerosol particles, their role in nucleation is not yet defined. Zhang et al. (2004) found considerable enhancement in nucleation rates due to the presence of organic acids.24 Direct experimental evidence of the thermodynamic advantage of involving organic acids on bisulfate clusters has been shown,37 and recently, Kulmala et al. affirmed the crucial role of organic compounds in atmospheric aerosol formation.38 Studies have shown that water-soluble organic nitrogen compounds comprise an important portion of aerosols and rainwater.39,40 Water-soluble organic nitrogen compounds contribute about 18% to the total mass of fine aerosol particles.41 A major portion of identified nitrogen compounds are amino acids, which have been detected in atmopsheric aerosols and fog droplets.42−47 In particular, glycine, threonine, serine, and alanine are identified as the most abundant free amino acids detected.42 32 different amino acids have been identified in the atmosphere, and amino acids are recognized as significant components in atmospheric aerosols.48 © 2013 American Chemical Society
2. COMPUTATIONAL METHODOLOGY All density functional theory calculations have been performed in Gaussian 09,50 and all explicitly correlated coupled cluster calculations have been performed in Molpro, version 2012.1.51 Geometry optimization and frequency calculations have been performed using M06-2X52/6-311++G(3df,3pd) which has Received: September 6, 2013 Revised: November 5, 2013 Published: November 5, 2013 12990
dx.doi.org/10.1021/jp408962c | J. Phys. Chem. A 2013, 117, 12990−12997
The Journal of Physical Chemistry A
Article
Table 1. Mean Absolute Errors in Geometrical Parameters of the DFT Functionals B3LYP, PW91, M06-2X, CAM-B3LYP, and ωB97X-D Using DF-LCCSD(T)-F12a/VDZ-F12 as the Reference B3LYP
PW91
M06-2X
CAM-B3LYP
ωB97X-D
compound
bond (Å)
angle (deg)
bond (Å)
angle (deg)
bond (Å)
angle (deg)
bond (Å)
angle (deg)
bond (Å)
angle (deg)
Gly Gly⊖ Gly⊕ (Gly)(H2O) (Gly⊖)(H2O) (Gly⊕)(H2O) (Gly)(NH3) (Gly⊖)(NH3) (Gly⊕)(NH3)
0.0015 0.0018 0.0021 0.0022 0.0034 0.0037 0.0019 0.0022 0.0025
0.53 0.38 0.30 0.71 0.87 1.91 0.84 2.34 0.66
0.0074 0.0083 0.0083 0.0101 0.0098 0.0095 0.0094 0.0087 0.0109
0.21 0.68 0.67 0.54 1.34 0.82 0.53 2.01 0.63
0.0027 0.0034 0.0026 0.0026 0.0034 0.0034 0.0028 0.0028 0.0039
0.39 0.35 0.30 0.39 0.87 0.78 0.40 0.93 0.60
0.0035 0.0039 0.0033 0.0041 0.0042 0.0034 0.0033 0.0039 0.0034
0.60 0.51 0.45 0.80 1.03 1.02 0.81 1.19 0.72
0.0031 0.0034 0.0028 0.0028 0.0030 0.0026 0.0025 0.0030 0.0030
0.45 0.35 0.38 0.56 0.67 0.78 0.64 0.67 0.61
total WMAE
0.0024
0.99
0.0093
0.83
0.0031
0.57
0.0036
0.81
0.0029
0.58
weighted mean absolute errors (WMAE). In the case of the ωB97X-D functional, the (Gly)(H2O) and (Gly)(NH3) structures required an ultrafine grid and tight convergence criteria to converge without any low lying imaginary frequencies. It is observed that all five functionals perform well in describing the DF-LCCSD(T)-F12a/VDZ-F12 structure with a total WMAE of
