The effect of organic compounds on the hygroscopic properties of inorganic aerosols

(2003-2005)

E. Weingartner, U. Krieger, S. Sjögren, A. Zardini, U. Baltensperger, T. Peter

SNF-project 200021-100280/1

1. Goals of the project

The aim of this project is an improved understanding of the hygroscopic properties of mixed atmospheric aerosol particles. Special emphasis is put on the investigation of the influence of organic species on the hygroscopic behaviour of inorganic salts, because atmospheric aerosols consist of mixtures of both substances. The project has the unique advantage that the complementary core capabilities of the two institutes will be used: The Electrodynamic Balance (EDB) at the IAC allows for the investigation of the hygroscopic properties of larger particles (diameter d > 1 µm) within a timescale of hours or even days, whereas the Hygroscopicity Tandem Differential Mobility Analyzer (H-TDMA) at the LAC investigates the same properties for smaller particles (d < 300 nm) and for timescales smaller than a few minutes. According to the original proposal, defined model substances as well as particles extracted from real atmospheric samples will be investigated with the two methods.

2. Activity report

2.1. Measurements on model substances

2.1.1 Ammonium sulfate (AS) and mixtures of AS with adipic acid

In a first phase the students became familiar with the instruments and performed measurements with defined laboratory generated test aerosols. Such experiments were carried out during an H-TDMA intercomparison workshop at PSI where two H-TDMAs (one from PSI and one from the University of Manchester, UMIST) were involved. The goals of this workshop were to i) intercompare the instruments and the subsequent data analysis and ii) to study the water vapor equilibration times of mixed aerosols. Therefore, chambers of different volumes were installed downstream of the humidifiers, which allowed studying the effect of residence time in a higher relative humidity environment on the hygroscopic growth. The investigated residence times ranged from 1 s to 2 min.
Initially, the UMIST and a PSI instrument were intercompared using pure ammonium sulfate (AS) particles. Figure 1 shows that the two H-TDMA instruments obtained very similar results, which agree well with theory. As expected, no influence of different residence times at higher relative humidity (RH) was noted because the equilibration times for pure inorganic particles are very short (a few milliseconds).
Figure 2 shows the hygroscopic growth of a super micrometer sized pure ammonium sulfate particle measured with the EDB at ETH. Here the residence time of the particle in the instrument was on the order of a few hours and the result also compares very well with the same thermodynamic model. It has to be noted that measured growth factors presented in Figure 1 and 2 can not be directly compared, since the H-TDMA measures the particles hygroscopic growth in terms of a relative diameter change whereas the EDB data is presented as a relative change of the particle mass. For a direct comparison, the different growth factors can be transferred into each other by knowledge of the particles density and the assumption of a spherical particle shape.

Figure 1: Hygroscopic growth (in terms of relative diameter change) of ammonium sulfate particles measured with the UMIST resp. the PSI H-TDMA. The solid line is a theoretical curve.

Figure 2: Hygroscopic growth of a pure ammonium sulfate particle (in terms of relative mass change) measured with the ETH EDB instrument. The solid line is a theoretical curve.

Subsequently, internally mixed particles composed of AS and adipic acid (AA, which is much less soluble than AS) were analyzed. Figure 3 shows that for these mixtures a reduction of the hygroscopic growth of AS with increasing amount of AA is observed with the H-TDMA.

Figure 3: Hygroscopic growth of mixtures of ammonium sulfate and adipic acid (both H-TDMAs employed). All points were measured with residence time > 4 s. The solid line is the same theoretical curve for pure AS as in Figure 1, and the dashed lines are fitted curves which were added to guide the eye.

Figure 4 shows that the residence time in a higher RH environment influences the hygroscopic growth of the mixed deliquesced particles: Mixed particles which experienced a residence time < 4 s showed a lower growth factor compared to particles which have experienced the high RH for more than 4 s. These measurements clearly indicate that the equilibrium times are an important issue for the investigated mixed organic-inorganic particles. EDB measurements of the same particles with even longer timescales are planned which will complement these measurements.
These results are new and relevant for processes in the atmosphere. They will be presented at the European Aerosol Conference in Budapest this year and it is planned to submit a paper with these results by the end of this year to the Journal of Aerosol Science.

Figure 4: Influence of residence time on the hygroscopic growth of mixed particles of ammonium sulfate and adipic acid (50/50 %wt left graph; 24/76 %wt right graph).

2.1.2 Mixtures of ammonium sulfate with citric acid and glutaric acid

Presently, simultaneous measurements with the H-TDMA and with the EDB are performed with particles composed of mixtures of AS and citric as well as AS and glutaric acid. These substances were chosen since we expected to see no significant differences in the result obtained with the two methods for AS - citric acid particles. However, we expect to see significant differences for AS - glutaric acid particles since it is speculated that this mixture requires much longer water vapor equilibrium times. This difference is clearly seen in Figure 5 (all data taken from literature), where the hygroscopic growth of pure glutaric acid particles was analyzed using different methods. This is interesting, as the two substances have similar properties such as solubility and hygroscopicity, so it is speculated that the difference in results come from mass transfer limitations for glutaric acid.

Figure 5: Hygroscopic growth of glutaric acid particles measured with the EDB,
the H-TDMA and from bulk solutions.

The hygroscopic growth of pure citric acid particles was measured with the EDB at ETH (Figure 6) and with the H-TDMA at PSI (Figure 7). A good agreement with literature data is found. Both humidograms are characterized by no distinct efflorescence and deliquescence points. Also the magnitude of the water uptake observed with both methods is similar: Assuming spherical particles, a citric acid density of 1.542 g cm-3 and supposing volume additivity, the measured H-TDMA growth factor of D/Do = 1.24 at RH= 85% (see Figure 7) can be converted into a mass change of m/mo = 1.54. This value compares well with the corresponding EDB measurement (m/mo = 1.65 ± 0.1 at RH= 85%, as seen in Figure 6).
In other experiments, a solution of AS and citric acid (1:1 molar ratio) was nebulized and internally mixed particles with Do = 100 nm were analyzed with the H-TDMA (Figure 8). Again, no distinct efflorescence and deliquescence points were measured for these mixed particles. Starting at dry conditions, there is a small decrease in size at around 17% RH. This small restructuring is attributed to a mobility effect, i.e. a reduction of the dynamic shape factor due to microstructural rearrangement. Similar effects were observed by Gysel et al. (2004) for the hygroscopic behaviour of mixed organic/inorganic particles that were renebulized from atmospheric samples. Such distinct restructuring is not observed for pure inorganic particles at this low RH. It is therefore speculated that the presence of such organics (with low deliquescence RH) might induce such a microstructural rearrangement at lower RH.

Figure 6: Hygroscopic growth of a citric acid particle measured with the EDB at ETH
and comparison with data by Choi et al., 2002.

Figure 7: Hygroscopic growth of citric acid measured with the H-TDMA at PSI and comparison with
data by Peng et al., 2001 and Joutsensaari et al., 2001. Residence time = ~10s.

Figure 8: Hygroscopic growth of AS-citric acid mixture (molar ratio 1:1) measured with the
H-TDMA at PSI. Included is also AS pure and the CA pure (dotted lines). Residence time = ~10s.
First EDB measurements performed at ETH on pure glutaric acid particles are shown in Figure 9. The measured humidogram is characterized by a distinct hysteresis, with well defined deliquescence and efflorescence points. Apart from the different absolute value for the deliquescence RH (within 5% and which is probably due to different ambient conditions and subject of further studies) the data shows good agreement with the literature data.

Figure 9: Hygroscopic growth of a glutaric acid particle measured with the EDB at ETH
and comparison with the only data available from literature.

Other measured humidograms of pure glutaric acid particles were very similar to the one shown in Figure 9. Often the efflorescence point could not be detected in form of an abrupt mass change as function of RH. However, the phase transition of organic particles from the liquid to solid state could always be clearly identified in a distinct change of the particles optical properties (i.e. by an analysis of the variability of the scattered light signal). This points to the complexity of organic particles. In contrast to inorganic particles they do not necessarily experience a well defined transition from the solid to liquid state - they can keep the uptaken water in a not well defined intermediate state.
First H-TDMA measurements of pure glutaric acid particles also indicate that the submicron particles tend to be sensitive to evaporation and/or mass transfer limitation effects.

2.2. Measurements at the PSI smog chamber

During this project it is also planned to investigate the hygroscopic properties of secondary organic aerosol (SOA) particles which were generated in controlled laboratory experiments in the PSI smog chamber. First hygroscopicity measurements have already been performed at this facility where a gaseous precursor (1,3,5-trimethylbenzene) was exposed to photo-oxidation in an air/NOx mixture. Figure 10 shows the increase of the hygroscopic growth factor for such an experiment where the photo-oxidation starts at 00:00. The different sizes available for measurements change over time; smaller particles (dry diameter d0 < 100 nm) grow rapidly by condensation to larger particles and thus disappear successively during the first hours of the experiment.

Figure 10: Evolution of the hygroscopic growth factor with time (for RH = 83-87%). The different symbols denote the investigated particle sizes in the diameter range 30 - 200 nm.

After steady state is reached (ca. 9 h after lights on) the H-TDMA was operated in the scanning mode and humidograms were measured. Such a humidogram (see Figure 11) shows a continuous increase of water uptake with increasing humidity, i.e. no deliquescence or efflorescence was observed and the particles stay in liquid form down to an RH of about 15%.

Figure 11: Humidogram for d0 = 100 nm particles, obtained during 9-14 h after lights on.

For the next year it is planned to take filter samples at various stages of such smog chamber experiments. Then, the aerosol particles will be separated from the filter for off-line analysis of the hygroscopicity with the EDB. A fraction of the material will be nebulized and re-analyzed with the H-TDMA as a control experiment. It will be highly interesting to see if these H-TDMA results can be reproduced with the EDB.

2.3. Field measurements

According to the original proposal, the influence of organic compounds on the hygroscopic behaviour of real atmospheric aerosols will also be investigated. Therefore, as part of extensive field campaigns at the high alpine research station Jungfraujoch, high-volume filter samples were taken in summer 2002 (30 samples, 1 day time resolution) and as well in winter 2004 (5 samples, 7 day time resolution). These samples are presently stored in a freezer. In a later phase of this project the water soluble particulate material will be extracted from these samples, renebulized and analyzed with the H-TDMA and EDB. The obtained results will be compared with the in-situ H-TDMA data that were continuously measured during the campaigns with a time resolution of 1 hour. Together with other in-situ measured aerosol parameters (such as number size distributions, light scattering and absorption coefficients as well as chemical composition) the effects of the organic fraction on atmospheric processes of relevance to climate will be estimated.

References:

Choi, M.Y., and C.K. Chan, The effects of organic species on the hygroscopic behaviors of inorganic aerosols, Environ. Sci. Technol., 36 (11), 2422-2428, 2002.
Cruz, C.N., and S.N. Pandis, Deliquescence and hygroscopic growth of mixed inorganic-organic atmospheric aerosol, Environ. Sci. Technol., 34 (20), 4313-4319, 2000.
Joutsensaari, J., P. Vaattovaara, M. Vesterinen, K. Hameri and A. Laaksonen, A novel tandem differential mobility analyzer with organic vapor treatment of aerosol particles, Atmos. Chem. Phys., 1, 1-22, 2001.
Peng, C., M.N. Chan and C.K. Chan, The hygroscopic properties of dicarboxylic and multifunctional acids: Measurements and UNIFAC predictions, Environ. Sci. Technol., 35 (22), 4495-4501, 2001.
Gysel, M., E. Weingartner, S. Nyeki, D. Paulsen, U. Baltensperger, I. Galambos, and G. Kiss, Hygroscopic properties of water-soluble matter and humic-like organics in atmospheric fine aerosol, Atmos. Chem. Phys., 4, 35-50, 2004.

Contacts

At PSI: Staffan Sjögren At ETH: Alessandro Zardini

Links

Aerosol Physics Group - Laboratory of Atmospheric Chemistry - PSI

to top of page

last updated: 2005-02-11