J. Phys. Chem. A 2010, 114, 12237–12243
12237
Observation of the Crystallization and Supersaturation of Mixed Component NaNO3-Na2SO4 Droplets by FTIR-ATR and Raman Spectroscopy Hai-Jie Tong,† Jonathan P. Reid,‡ Jin-Ling Dong,† and Yun-Hong Zhang*,† The Institute for Chemical Physics, Key Laboratory of Cluster Science, Beijing Institute of Technology, Beijing 100081, People’s Republic of China, and School of Chemistry, UniVersity of Bristol, Bristol, United Kingdom BS8 1TS ReceiVed: August 25, 2010; ReVised Manuscript ReceiVed: October 11, 2010
We present here a study of the phase behavior of mixed component NaNO3-Na2SO4 (SNS) droplets with NaNO3 to Na2SO4 molar ratios of 1:1, 3:1, and 10:1, comparing observations with thermodynamic predictions. Measurements are made by Fourier transform infrared attenuated total reflection and micro-Raman spectroscopy for SNS droplets deposited on ZnSe and quartz substrates, respectively. The conventional deliquescence/ efflorescence hysteresis in phase behavior is observed. On drying, heterogeneous crystallization leads to phase behavior that is consistent with bulk solution thermodynamics, with the formation of the mixed salt NaNO3 · Na2SO4 · H2O, Na2SO4 (s), and NaNO3 (s) all observed to form at relative humidities that coincide with predictions by the aerosol inorganics model. However, conditioning of the droplet at high relative humidity prior to drying is observed to lead to quantitative differences between the fractions of different salts formed. When substrate effects do not influence the crystallization process, supersaturated solutions are formed, and this leads to the observation of contact ion pairs. Such measurements of the phase behavior of mixed component droplets are important for testing the reliability of thermodynamic models. I. Introduction Aerosols influence the climate and chemistry of the atmosphere by scattering and absorbing solar radiation,1,2 acting as cloud condensation nuclei (CCN),2 and providing sites for heterogeneous chemical reactions.3 Inorganic salts represent a significant fraction of troposphere aerosol mass, accounting for 25-50% of fine aerosol mass.4,5 Na+, NO3-, and SO42- are important components of mineral and aged marine aerosols in addition to other ions such as Cl- and Mg2+, etc.6,7 In addition, sodium sulfate is accepted as one of the most damaging salts for the built cultural environment leading to chemical or salt weathering.8 Considering that most atmospheric aerosols are composed of more than one component, the complex phase behavior of multicomponent aerosols must be understood. Aerosol containing both NaNO3 and Na2SO4 (referred to as the SNS system below) provides a good model system for investigating the formation and properties of multicomponent tropospheric aerosols. Under conditions of varying relative humidity (RH) and temperature, Na2SO4 (s, SS) can exist as an anhydrous or hydrated crystal with varying degrees of water content:9-11
Na2SO4 + 7H2O f Na2SO4 · 7H2O + 3H2O f Na2SO4 · 10H2O (1) Na2SO4 (s) exists in five phases (I-V), and both phases III and V are observed following the evaporation of Na2SO4 solution.12,13 Phase III is metastable, while phase V is stable. However, phase III has a slow transition rate to phase V and * To whom correspondence should be addressed. Telephone: 86-1086668406. Fax: 86-10-68913596. E-mail:
[email protected]. † Beijing Institute of Technology. ‡ University of Bristol.
can be kept indefinitely at ambient temperature.13,14 In addition, a transition from Na2SO4 (V) to Na2SO4 (III) has been observed at temperatures above 32 °C.12-15 When mixed with NaNO3, the phase behavior accompanying evaporation is expected to be more complex. Tang observed the formation of the mixed salt Na2SO4 · NaNO3 in SNS aerosols by Raman spectroscopy16 and identified the presence of anhydrous Na2SO4, Na2SO4 · 10 H2O, and NaNO3 (s). Though numerous thermodynamic models have been developed for single or multicomponent inorganic aerosols,17-22 it remains a challenge to predict the crystallization process and the exact composition of supersaturated aerosols with mixed components. There are only a few studies that examine the dependence of the transitions between different metastable salts in SNS aerosol on the molar ratio of Na2SO4 and NaNO3 and relative humidity. Although experiments have been carried out on SNS droplets using the EDB technique,16,22 these have not included an identification of the crystalline phase of Na2SO4 or an investigation of the interactions of the different ionic components, which may be expected to form ion pairs in supersaturated droplets at low RH. Further, studies of crystallization of NaNO323 and Na2SO424 have been performed by Raman spectroscopy for solution droplets deposited on a substrate, although the influence of the substrate on crystallization was not fully considered. In this publication, Fourier transform infrared attenuated total reflection (FTIR-ATR) and micro-Raman spectroscopy are used to investigate the compositional and morphologic changes of SNS droplets during desiccation on different substrates. Both FTIR-ATR and micro-Raman spectroscopy are sensitive to phase transitions and intermolecular interactions between solute and solvent molecules25,26 and have successfully been used to monitor the phase transitions of NaNO3 and mixed Na2SO4MgSO4 aerosols in our recent work.24,25 In the present work, the measurements of SNS droplets with different radii on ZnSe and quartz substrates (with different contact angles27,28) are
10.1021/jp1080548 2010 American Chemical Society Published on Web 10/28/2010
12238
J. Phys. Chem. A, Vol. 114, No. 46, 2010
Tong et al.
Figure 1. Schematic diagram of the experiment setup of the FTIRATR (a) and micro-Raman (b) systems.
carried out. FTIR-ATR spectroscopy is used to investigate the crystallization of SNS droplets on the ZnSe substrate. Raman spectroscopy is used to observe the morphology and compositional changes of single SNS droplets on quartz substrate at different RHs. Based on these two spectroscopic methods, the objectives of this work are to (1) monitor the heterogeneous crystallization and identify the composition and phase of the individual crystallization products formed from SNS aerosols of varying solute molar ratio and (2) investigate the evolution in intermolecular interactions between solvent and solute molecules of single SNS aerosol droplets in a supersaturated state. II. Experimental Approach The experimental strategy used in this work combining FTIRATR and micro-Raman spectroscopy has been described in detail elsewhere23,25 and will be only briefly summarized here. Figure 1 provides the details of the two systems used in the present work. As shown in Figure 1a, the FTIR-ATR system is composed of three parts: aerosol generation, RH control, and spectroscopic detection. Aerosols 1-5 µm in diameter were produced by a medical nebulizer.25 The RH in the aerosol chamber was regulated by introducing a humidified N2 stream, and the RH and temperature were recorded by a hygrometer (CENTER 310, with an accuracy of (2.5%/RH, ( 0.7 °C/T). A MAGNA-IR 560 FTIR spectrometer equipped with a horizontal ZnSe ATR cell (refractive index 2.43, spectra-Tech Inc., U.S.) was used, with 12 total internal reflections of the IR beam at the ZnSe/air interface. A MCT/A liquid nitrogen-cooled detector was used to record the FTIR spectra. Each spectrum was recorded from the accumulation of 64 scans performed on the droplets residing on the surface to obtain a high signal-tonoise level. Spectra were acquired with a resolution of 4 cm-1 in the range of 400-4000 cm-1. The cell was carefully cleaned before each measurement. A schematic diagram of the micro-Raman system is represented in Figure 1b. Briefly, this system is equipped with a Leica DMLM microscope and a 514.5 nm Ar-ion incident laser (LS514 model, Laserphysics) with an output power of 20 mW. The Raman spectral coverage is from a Stokes shift of 200-4000 cm-1 with a resolution of about 1 cm-1. The aerosol generation
Figure 2. (a) Raman spectra of Na2SO4 crystals formed on the ZnSe substrate. (b) FTIR-ATR spectra of Na2SO4/H2O aerosols with varying RH during the desiccation process. The Raman spectra A and B are the ν3-SO42- of Na2SO4 powder for phase III (C) and V (D) observed at 200 °C by Choi.32
and RH control part for the Raman experiments are similar to those described for the FTIR measurements. The ZnSe substrate used for the FTIR and quartz substrate for Raman measurements are the same as those used in our previous work.23,25 Na2SO4 (Beijing Chemical Reagents Company, g99.0%) and NaNO3 (ACROS, g99.0%) were dissolved directly without further purification. Pure Na2SO4 droplets and SNS mixtures with NaNO3 to Na2SO4 molar ratios of 1:1, 3:1, and 10:1 were studied using FTIR-ATR spectroscopy. Molar ratio 3:1 and 10:1 SNS were studied using micro-Raman spectroscopy. Each RH was maintained for more than 20 min, and all measurements were taken at 22 ( 1 °C. III. Results and Discussion III.a. Crystallization of SS Droplets. Figure 2a shows examples of the Raman spectra of crystallized SS droplets on a quartz substrate (A and B). Figure 2b shows an example of a sequence of FTIR-ATR spectra from SS droplets on a ZnSe substrate during the desiccation process. The assignments of the infrared bands and Raman shifts are presented in Table 1 according to refs 16, 23-25, and 29-31. The Raman spectra of crystals A and B in Figure 2a are assigned to Na2SO4 (s) in phase III and a mixture of phases III and V, respectively, according to the assignments presented in refs 13, 29, 32, and 33. The Raman spectra of pure phases III (C) and V (D) observed by Choi33 at 200 °C are shown for comparison. Most crystals form on ZnSe substrate with needle morphology, similar to that observed by Rodriguez-Navarro12 and Amirthalingam.15 Figure 2b indicates that Na2SO4 (s) crystallized on the ZnSe substrate at 83% RH, which is the deliquescence point of Na2SO4 (III) reported by Steiger.11 On repeated measurements, SS droplets were observed to always crystallize to Na2SO4 (s) phase III. The infrared band at 1132 cm-1 can be assigned to the ν3-SO42- of phase III in accordance with the infrared spectra of Na2SO4 (III) reported by Durie.30 The preferential formation of phase III on ZnSe may be attributed to the increased influence of homogeneous nucleation for
Mixed Component NaNO3-Na2SO4 Droplets
J. Phys. Chem. A, Vol. 114, No. 46, 2010 12239
TABLE 1: IR Bands Position or Raman Shift of SS/SNS Aerosols (in cm-1) from the References vibration modes/Raman Shift
wavenumber (cm-1)
ν-OH (IR, NaNO3 · Na2SO4 · H2O, crystal)31 ν3-NO3- (IR, SN, aqueous)25 ν3-SO42- (IR, SS, crystal, phase V)30 ν2-SO42- (IR, NaNO3 · Na2SO4 · H2O, crystal)31 ν3-SO42- (IR, SS, crystal, phase III)30 ν3-SO42- (R, SS, aqueous)24 ν1-NO3- (R, SN, crystal)16 ν1-NO3- (R, NaNO3 · Na2SO4 · H2O, crystal)16 ν1-NO3- (R, SN, aqueous)16,23 ν1-SO42- (R, SS, crystal, phase V)24 ν1-SO42- (R/IR, SS, crystal, phase III)24,24 ν1-SO42- (IR, NaNO3 · Na2SO4 · H2O, crystal)31 ν1-SO42- (R, SS, crystal, phase II)24 ν1-SO42- (R, SS, crystal, phase I)24 ν1-SO42- (R, SS, aqueous)24 ν2- NO3- (IR, SN, crystal)25 ν2-NO3- (IR, SN, aqueous)25 ν4-SO42- (R, SS, crystal, phases I, II)24 ν4-SO42- (R, SS, crystal, phase V)24 ν4-SO42- (R, SS, crystal, phase III)24 ν4-SO42- (R, SS, aqueous)24 ν2-SO42- (R, SS, aqueous, phases I, II, III)24 ν2-SO42- (R, SS, crystal, phase V)24 ν2-SO42- (R, SS, aqueous)24
3430 1348 1113, 1131, 1145 1112, 1150 1108, 1133, 1175 1105 1067 1063 1048 993 997 995 994 989 981 836 829 627 623, 633, 647 617, 637 611 460 453, 467 450
droplets on the ZnSe substrate: the contact angle of SS droplets on the ZnSe substrate is larger than that on the quartz substrate.27,28 The formation of Na2SO4 (s) phase III on drying is accompanied by a sudden intensity decrease of ν-OH and the shifts of the ν1-SO42- band from ∼981 to 997 cm-1 and of the ν3-SO42- band from ∼1093 to 1132 cm-1. RH dependent X-ray diffraction studies of the crystallization of Na2SO4/H2O solution in a porous material have reported the dominant formation of the metastable phase III on drying,14 consistent with the formation of Na2SO4 (s) in phase III in this work. III.b. Crystallization of 1:1, 3:1, and 10:1 SNS Aerosols. Figure 3 and Figure 4b show the RH dependence of the Raman and FTIR-ATR spectra of 3:1 SNS droplets deposited on quartz and ZnSe substrates, respectively. Figure 4a and c shows the FTIR-ATR spectra of 1:1 and 10:1 SNS droplets on a ZnSe substrate during the desiccation process. Figure 5a-c shows the predicted molar evolution of different salts in 1:1, 3:1, and 10:1 SNS aerosols with varying humidity estimated from the E-AIM model.18 The original amounts of NaNO3 in 1:1, 3:1, and 10:1 SNS particles are 1, 3, and 10 mol, respectively. The amount of Na2SO4 in 1:1, 3:1, and 10:1 SNS particles is 1 mol. The two bands at 1063 and 1068 cm-1 in Figure 3 can be assigned to the ν1-NO3- band of NaNO3 · Na2SO4 · H2O and
Figure 3. Raman spectra and images versus RHs of 3:1 SNS aerosol particles on the quartz substrate during the desiccation process.
Figure 4. FTIR-ATR spectra versus RHs of 1:1 (a), 3:1 (b), and 10:1 (c) SNS aerosols on the ZnSe substrate during the desiccation process.
NaNO3 (s), respectively. This is consistent with the observations of Tang16 who suggested that the 1063 cm-1 could be attributed to the formation of NaNO3 · Na2SO4 and the 1068 cm-1 to anhydrous NaNO3 in the crystallized SNS droplet. The bands of 1117 and 1155 cm-1 in Figure 4 can be assigned to the ν3SO42- band of NaNO3 · Na2SO4 · H2O according to the FTIR assignments reported by Ericksen and references therein.31 The band of 1117 cm-1 can be assigned to the ν3-SO42- of thenardite30 too. However, Figure 2b indicates that the 1117 cm-1 band in the FTIR spectra of pure SS particles is much weaker than the 1132 cm-1. In addition, the 1132 cm-1 is not obvious in Figure 4a and b. Thus the 1117 cm-1 band in Figure 4 can be assigned to the ν3-SO42- band of NaNO3 · Na2SO4 · H2O. The contribution from insignificant Na2SO4 (s) cannot be excluded either. Both sets of spectra recorded on quartz and ZnSe substrates indicate the formation of NaNO3 · Na2SO4 · H2O at a RH between 81 and 74% and NaNO3 between 67 and 64% RH. The crystallization of NaNO3 · Na2SO4 · H2O is accompanied by the narrowing of ν-OH (3246, 3385 cm-1), the splitting of the ν3-SO42- (∼1101 cm-1) and the ν3-NO3- (∼1362 cm-1) bands, and shifts of the ν1-SO42- band from 981 and 997 cm-1 and of the ν2-NO3- band from 829 to and 831 cm-1 in the FTIR spectra. The crystallization of NaNO3 is accompanied by a shift of the ν2-NO3- band from 829 to 837 cm-1. The phase behavior identified from the Raman and FTIR measurements is consistent with the model predictions and phase diagram reported by Clegg et al.18,22 The E-AIM calculations in Figure 5b indicate that the equilibrium state for 3:1 SNS particles can be considered to go through three steps with increasing RH: the transition from a dry particle containing NaNO3 and Na2SO4 to a particle containing NaNO3 · Na2SO4 · H2O and NaNO3 at 38% RH, the deliquescence of remaining NaNO3 at 73% RH, and, finally, the gradual deliquescence of NaNO3 · Na2SO4 · H2O between 73 and 81%. Thus, the observations of the crystallization of NaNO3 ·
12240
J. Phys. Chem. A, Vol. 114, No. 46, 2010
Figure 5. Moles prediction of NaNO3 (s, empty squares), Na2SO4 (s, solid squares), Na2SO4 · 10H2O (empty circles), and NaNO3 · Na2SO4 · H2O (empty triangles) in 1:1 (a), 3:1 (b), and 10:1 (c) SNS aerosols based on the E-AIM model on line (18). The inset graph in (a) is the enlarged region from 80.25 to 81.75% RH of (a). The original amounts of NaNO3 in 1:1, 3:1, and 10:1 SNS particles are 1, 3, and 10 mol, respectively. The amount of Na2SO4 in 1:1, 3:1, and 10:1 SNS particles is 1 mol.
Na2SO4 · H2O between 81 and 74% and NaNO3 between 67 and 64% are broadly consistent with predictions from the E-AIM model. In these measurements, heterogeneous nucleation is clearly playing a role and supersaturated solutions are not formed. It can be observed from Figure 4a that NaNO3 · Na2SO4 · H2O once again crystallized at ∼80% RH in 1:1 SNS droplets. This phase transition is accompanied by the narrowing of ν-OH (3246, 3385 cm-1), the splitting of the ν3-SO42- (∼1101 cm-1) and the ν3-NO3- (∼1362 cm-1) bands, and the shifts of the ν1SO42- band from 981 to 997 cm-1 and of the ν2-NO3- band from 829 to 831 cm-1, respectively. These observations are in agreement with the predictions for the 1:1 mixture in Figure 5a. The E-AIM calculations indicate that the equilibrium RH dependence of 1:1 SNS particles shows four steps. The first step coincides with the transition from a dry particle containing NaNO3 and Na2SO4 to a particle containing NaNO3 · Na2SO4 · H2O at 38% RH. At ∼80%, the mixed salt is predicted to deliquesce, although some anhydrous sodium sulfate is expected to remain in a crystal form. At ∼81%
Tong et al. RH, this anhydrous salt becomes hydrated and then gradually deliquesces with the increase of RH. Experimentally, we observe the direct formation of NaNO3 · Na2SO4 · H2O at RHs between ∼83 and 80% in Figure 4b, marginally higher than the deliquescence point predicted by E-AIM for the mixed salt. However, we do not observe signatures from either anhydrous or hydrated SS crystalline domains on drying. It can be observed from Figure 4c that NaNO3 (s) crystallized at ∼60% RH in 10:1 SNS aerosols, accompanied by a shift of the ν2-NO3- from ∼829 to 837 cm-1. The band attributed to the symmetric stretching vibration of sulfate at frequencies in the range 1053-1155 cm-1 is observed to change in shape, with the peak shifting to higher frequency. The band components at higher frequency (>1117 cm-1) can be attributed to the formation of NaNO3 · Na2SO4 · H2O and of Na2SO4 (III), in line with assignments by Durie and Ericksen,30,31 also at 60% RH. The E-AIM calculations in Figure 5c indicate that, with increasing RH, NaNO3 · Na2SO4 · H2O is formed at 38% RH from the independent crystalline phases. This is followed by the deliquescence of the remaining NaNO3 and NaNO3 · Na2SO4 · H2O at ∼73% RH. As observed for the 3:1 SNS droplet, the deliquescence of the mixed salt is gradual with increasing RH. As a result of the low volume fraction of NaNO3 · Na2SO4 · H2O and Na2SO4 (s) formed for the 10:1 experiment shown in Figure 4c due to the dominance in mole fraction of NaNO3, the bands that provide signatures of sulfate in Na2SO4 (s) and the mixed salt, appearing at 997 and 1063 cm-1, are not as apparent as those observed in Figure 4a and b. However, during the drying process, we do not observe the formation of the mixed salt at the high RHs of ∼80% observed in the preceding measurements. It should be stressed that the trends in Figure 4a-c are representative of a large number of measurements on many droplets. In order to investigate further the phase transitions occurring, different RH conditioning cycles were performed prior to the final drying process during which FTIR spectra were recorded. These conditioning measurements were performed for SNS droplets of molar ratio 10:1. Figure 6a shows in detail the FTIR spectra around the ν3-SO42- band from Figure 4c, without any preconditioning cycle in RH and with just the gradual decrease to 20 µm in diameter). In addition, it was observed that the smaller droplets achieved a more highly supersaturated state before crystallization on the quartz substrate.22 The second crystallization pathway on the same quartz substrate is shown in Figure 8b and can be broadly described as showing supersaturated solution behavior without crystallization. The experimental conditions for the droplet in Figure 8b are the same as those for the droplet in Figure 8a. In this case, a gradual decrease in the RH by small steps to 47% was
12242
J. Phys. Chem. A, Vol. 114, No. 46, 2010
Figure 9. Intensity ratios of I1058.0/I1047.6 (a), I1052.5/I1047.6 (b), I1049.0/ I1047.6 (c), I3488/I3256 (d) for the Raman spectra in Figure 8(b) and the WSRs (e) from E-AIM model for the 10:1 SNS droplet. Open cycles correspond to the intensity ratios for the NaNO3 droplets in ref 23. Solid cycles correspond to the 10:1 SNS droplets of the present work. Standard deviations of σ as error bars were also presented.
accompanied by frequency shifts for ν1-SO42- from 982 to 986 cm-1 and for ν1-NO3- from 1052.5 to 1058.0 cm-1, but no recognizable crystallization. In our previous work, we have identified and resolved spectroscopically free solvated NO3- ions, SIPs (solventseparated ion pairs), CIPs (contact ion pairs), and complicated CIPs of NaNO3 at frequencies of 1047.6, 1049.0, 1052.5, and 1058.0 cm-1, respectively, in single NaNO3/H2O aerosol droplets on a quartz substrate using band-component analysis.23 The intensities of the Raman scattering at 1047.6, 1049.0, 1052.5, and 1058.0 cm-1 can be used in the present work to investigate the evolution of nitrate ion pairs in the 10:1 SNS aerosols. The ratios of I1058.0/I1047.6, I1052.5/I1047.6, I1049.0/I1047.6, and I3488/I3256 from the Raman spectra in Figure 8b are presented in Figure 9, along with the calculated water-to-solute ratio (WSR) for a solution of 10:1 molar ratio SNS aerosol from the E-AIM model,18 suppressing the formation of any crystalline phases. The temperature used for the calculation was 25 °C. Further, the ratio I986/I982, WSRs, and the full width at half-height (fwhh) of the ν1-SO42- band together with the area ratios between the right part and the left part (Aa/Ab) of the band (central wavenumber of 982.6 cm-1) are presented in Figure 10. The intensity ratios and fwhh values in Figures 9 and 10 are average data from three different supersaturated 10:1 SNS droplets on a quartz substrate which followed this pathway with decreasing RH. The data in Figure 9 can be separated in to three regions according to the values of I1058.0/I1047.6, I1052.5/I1047.6, and I1049.0/ I1047.6 when compared to 1 and the evolution of the slope of the intensity ratios of I1058.0/I1047.6 (a), I1052.5/I1047.6 (b), I1049.0/I1047.6 (c), and I3488/I3256 (d). In region 1 (WSR > 7.8, RH > 82%), the order of I1052.5/I1047.6 < 1 can be attributed to the dominance of SIPs of NO3-. The intensity of the band at 3256 cm-1 is dominated by the strong hydrogen bonds among water molecules and the band of 3488 cm-1 by weaker hydrogen bonds, including those between water and NO3- in the SNS aerosol droplets.34,35 The slow increase of I3488/I3256 in region I of Figure 9d indicates that the NO3- increasingly influences the strong interactions between Na+ and water molecules. In region II of Figure 9 (5.7 < WSR < 7.8, 73% I1052.5/
Tong et al.
Figure 10. I986/I982 (a), Aa/Ab (b), fwhh (full width at half-height) of ν1-SO42- (c) of the spectra in Figure 8b, and the WSRs of 10:1 SNS aerosols from the E-AIM model (d). Aa/Ab is the area ratio between the right part (Aa) and the left part (Ab) of ν1-SO42-, which is separated by the 982.6 cm-1. The area of Aa and Ab for each spectrum was integrated through the OMNIC80.iso software supplied by the Thermo Fisher Scientific Inc. Standard deviations of σ values as error bars are also presented.
I1047.6 > 1 suggests a significant increase in the amount of nitrate CIPs, accounting for the largest proportion of nitrates ion pairs.23 In region III (WSR < 5.7, RH < 73%), I1058.0/I1047.6, I1052.5/I1047.6, and I3488/I3256 increase further, indicating the significant formation of nitrate CIPs and the disruption of the hydrogen bonding network with the decrease of RH. Figure 9 also indicates that the intensity ratios of I1052.5/I1047.6 and I1058.0/I1047.6 are slightly higher than those recorded in a NaNO3/H2O droplet, suggesting that the formation of nitrate CIPs may be enhanced by the presence of SO42- in the 10:1 SNS aerosol droplets. The intensity ratios of I1049.0/I1047.6 and I3488/I3256 are quite close to those measured in NaNO3/H2O droplets, suggesting that the amount of NO3- free ion pairs and the ratio between the strong and the weak hydrogen bonds in the NaNO3/H2O droplet could be influenced by the existence of SO42-. This is in agreement with Omta’s conclusion that the addition of Na2SO4 has no significant effect on the rotational dynamics of H2O molecules outside the first hydration shell.36 Various ion pairs between Na+ and SO42- in the gas phase37 or aqueous solution have been investigated by different techniques.38-41 Figure 10 can be also separated in to two regions according to the evolution of the intensity ratios of I986/I982, Aa/Ab, and the full width at half-height (fwhh) of ν1-SO42- in the 10:1 molar ratio SNS droplet of Figure 8b. In the present work, the bands of 982 and 986 cm-1 are assigned to SO42- free ion pairs and CIPs, respectively. In region I (WSR > 2.8, RH > 55%, fwhh 11.4), the I986/I982 and fwhh of ν1-SO42- are observed to increase significantly and a ratio I986/I982 > 1 suggests that the CIPs of SO42- account for the largest proportion of sulfate ion pairs in region II. At the same time, the SO42- might be expected to enter into the first coordination layer of Na+ and replace the hydration water of Na+ in this region. It is suggested by Figures 9 and 10 that the CIPs of SO42are formed at a lower RH than those of NO3- in the 10:1 SNS aerosol droplets. The evolution of free nitrate ion pairs in the 10:1 SNS aerosol droplets is similar to that in NaNO3/H2O
Mixed Component NaNO3-Na2SO4 Droplets aerosols, but the nitrate CIPs are slightly enhanced by the existence of SO42-. The ratios between the strong and the weak hydrogen bonds including that between water and NO3- are not influenced by the existence of the SO42- significantly. As a result the hydration numbers of Na+, NO3-, and SO42- are 3-6, ∼3, and 7-12, respectively,42-44 and the SO42- is normally regarded as a “structure-making” anion, forming hydrogen bonds with water molecules around the tetrahedral anion.45-49 Thus, the sulfate CIPs formed at lower RH than the nitrate CIPs in the 10:1 SNS aerosol droplet, inducing a greater slope for I986/ I982 than that for I1052.5/I1047.6. IV. Conclusions Given that understanding the phase behavior of complex mixtures is of crucial importance for understanding the phase behavior of aerosols in the atmosphere, experiments which test the phase behavior of mixtures with varying molar ratios of solutes are of importance for benchmarking the thermodynamic models. In this work, FTIR-ATR and micro-Raman spectroscopy were used to investigate the supersaturation and crystallization process of 1:1, 3:1, and 10:1 molar ratio SNS droplets. The crystallization of metastable Na2SO4 (III), NaNO3 (s), and NaNO3 · Na2SO4 · H2O in 1:1 and 3:1 SNS aerosols on both ZnSe and quartz substrates were observed. Two distinct pathways for change in composition with decrease in RH were observed, consistent with the conventional deliquescence/efflorescence hysteresis observed in isolated droplet studies. When substrate effects can be assumed to lead to heterogeneous nucleation in the surface supported droplets during drying, the phase behavior can be confirmed to follow predictions from the E-AIM thermodynamic model. In the absence of heterogeneous nucleation for some drying events, the formation of contact ion pairs was identified with the tendency to form complicated contact ion pairs increasing with decreasing RH. Cycling the RH at high RH, prior to drying to low RH, led to conditioning of the droplets and a greater preference for formation of the more stable thermodynamic phase on complete drying. Acknowledgment. This work was supported by the NSFC (20933001 and 20873006) and by the 111 Project B07012. J.P.R. acknowledges the support of the EPSRC through the award of a Leadership Fellowship. References and Notes (1) Murphy, D. M.; Anderson, J. R.; Quinn, P. K.; McInnes, L. M.; Brechtel, F. J.; Kreidenweis, S. M.; Middlebrook, A. M.; Po´sfai, M.; Thomson, D. S.; Buseck, P. R. Nature 1998, 392, 62. (2) Andreae, M. O. Atmos. Chem. Phys. 2009, 9, 543. (3) De Haan, D. O.; Brauers, T.; Oum, K.; Stutz, J.; Nordmeyer, T.; Finlayson-Pitts, B. J. Int. ReV. Phys. Chem. 1999, 18, 343. (4) Harrison., R. M.; Plo, C. A. EnViron. Sci. Technol. 1983, 17, 160. (5) Finlayson-Pitts, B. J.; Pitts, J. N. J. Chemistry of the upper and lower atmosphere-theory, experiments, and applications; Academic Press: San Diego, CA, 2001. (6) Gibson, E. R.; Hudson, P. K.; Grassian, V. H. J. Phys. Chem. A 2006, 110, 11785. (7) Guimbaud, C.; Arens, F.; Gutzwiller, L.; Ga¨ggeler, H. W.; Ammann, M. Atmos. Chem. Phys. 2002, 2, 249.
J. Phys. Chem. A, Vol. 114, No. 46, 2010 12243 (8) Goudie, A. S.; Viles, H. Salt weathering hazards; Wiley: Chichester, U.K., 1997. (9) Linnow, K.; Zeunert, A.; Steiger, M. Anal. Chem. 2006, 78, 4683. (10) Oswald, L. D. H.; Hamilton, A.; Hall, C.; Marshall, W. G.; Prior, T. J.; Pulham, C. R. J. Am. Chem. Soc. 2008, 130, 17795. (11) Steiger., M.; Asmussen, S. Geochim. Cosmochim. Acta 2008, 72, 4291. (12) Rodriguez-Navarro, C.; Doehne, E.; Sebastian, E. Cem. Concr. Res. 2000, 30, 1527. (13) Xu, B.; Schweiger, G. J. Aerosol Sci. 1999, 30 (Suppl.), 379. (14) Linnow, K.; Zeunert, A.; Steiger, M. Anal. Chem. 2006, 78, 4683. (15) Amirthalingam, V.; Karkhanavala, M. D.; Rao, U. R. K. Acta Crystallogr. 1977, A33, 522. (16) Tang, I. N.; Fung, K. H. J. Aerosol Sci. 1989, 20, 609. (17) Clegg, S. L.; Seinfeld, J. H. J. Phys. Chem. A 2004, 108, 1008. (18) Clegg, S. L.; Brimblecombe, P.; Wexler, A. S. J Phys. Chem. A 1998, 102, 2155. (19) Topping, D. O.; McFiggans, G. B.; Coe, H. Atmos. Chem. Phys. 2005, 5, 1205. (20) Clegg, S. L.; Kleeman, M. J.; Griffin, R. J.; Seinfeld, J. H. Atmos. Chem. Phys. 2008, 8, 1057. (21) Gao, Y.; Chen, S. B.; Yu, L. Y. J. Phys. Chem. A 2006, 110, 7602. (22) Clegg, S. L.; Brimblecombe, P.; Liang, Z.; Chan, C. K. Aerosol Sci. Technol. 1997, 27, 345. (23) Li, X. H.; Wang, F.; Lu, P. D.; Dong, J. L.; Wang, L. Y.; Zhang, Y. H. J. Phys. Chem. B 2006, 110, 24993. (24) Dong, J. L.; Xiao, H. S.; Zhao, L. J.; Zhang, Y. H. J. Raman Spectrosc. 2009, 40, 338. (25) Lu, P. D.; Wang, F.; Zhao, L. J.; Li, W. X.; Li, X. H.; Dong, J. L.; Zhang, Y. H.; Lu, G. Q. J. Chem. Phys. 2008, 129, 104509. (26) Rosenoern, T.; Schlenker, J. C.; Martin, S. T. J. Phys. Chem. A 2008, 112, 2378. (27) Shephard, J. J.; Savory, D. M.; Bremer, P. J.; McQuillan, A. J. Langmuir 2010, 26, 8659. (28) Sumner, A. L.; Menke, E. J.; Dubowski, Y.; Newberg, J. T.; Penner, R. M.; Hemminger, J. C.; Wingen, L. M.; Brauers, T.; Finlayson-Pitts, B. J. Phys. Chem. Chem. Phys. 2004, 6, 604. (29) Murugan, R.; Ghule, A.; Chang, H. J. Phys.: Condens. Matter 2000, 12, 677. (30) Durie, R. A.; Milne, J. W. Spectrochim. Acta 1978, 34A, 215. (31) Ericksen, G. E.; Mrose, M. E. Am. Mineral. 1970, 55, 1500. (32) Xu, B.; Vehring, R.; Schweiger, G. J. Aerosol Sci. 1998, 29 (Suppl. 1), S865. (33) Choi, B. K.; Lockwood, D. J. J. Phys.: Condens. Matter. 2005, 17, 6095. (34) Liu, J. H.; Zhang, Y.-H.; Wang, L. Y.; Wei, Z. F. Spectrochim. Acta, Part A 2005, 61, 893. (35) Zhao, L. J.; Zeng, Q. X.; Zhang, Y. H. J. Phys. Chem A 2009, 113, 215. (36) Omta, A. W.; Kropman, M. F.; Woutersen, S.; Bakker, H. J. Science 2003, 301, 347. (37) Wang, X. B.; Woo, H. K.; Jagoda-Cwiklik, B.; Jungwirth, P.; Wang, L. S. Phys. Chem. Chem. Phys. 2006, 8, 4294. (38) Buchner, R.; Capewell, S. G.; Hefter, G.; May, P. M. J. Phys. Chem. B 1999, 103, 1185. (39) Daly, F. P.; Brown, C. W.; Kester, D. R. J. Phys. Chem. 1972, 76, 3664. (40) Weinga¨rtner, H.; Price, W. E.; Edge, A. V. J.; Mills, R. J. Phys. Chem. 1993, 97, 6289. (41) Xu, J. X.; Fang, Y.; Fang, C. H. Chin. Sci. Bull. 2009, 54, 876. (42) White, J. A.; Schwegler, E.; Galli, G.; Gygi, F. J. Chem. Phys. 2000, 113, 4668. (43) Wang, X. B.; Yang, X.; Wang, L. S.; Nicholas, J. B. J. Chem. Phys. 2002, 116, 561. (44) Ohtaki, H.; Radnai, T. Chem. ReV. 1993, 93, 1157. (45) Walrafen, G. E. J. Chem. Phys. 1971, 55, 768. (46) Walrafen, G. E. J. Chem. Phys. 1962, 36, 1035. (47) Walrafen, G. E. J. Chem. Phys. 1964, 40, 3249. (48) Pye, C. C.; Rudolph, W. W. J. Phys. Chem. A 2001, 105, 905. (49) Marcus, Y. Chem. ReV. 2009, 109, 1346.
JP1080548
