skip to main content NIST Center for Neutron Research NIST Center for Neutron Research National Institute of Standards and Technology
Home Live Data Instruments CHRNS Proposals

Probing the Structure of Aerosol Nanodroplets

Nanodroplet aerosols form readily in the supersonic expansions that occur, for example, in turbomachinery, jet exhausts, and volcanic eruptions. Thus, understanding particle formation and growth when cooling rates approach 106 K / s is of broad scientific interest.

From a fundamental point of view, particles with radii < 10 nm are important because they lie in the critical transition zone between large molecular clusters and bulk material. In multicomponent droplets, there is also strong theoretical evidence for surface enrichment, i.e., that the surface and interior compositions can differ significantly. Surface enrichment is important because it affects nucleation, growth and evaporation kinetics, and the heterogeneous chemistry of aerosol droplets. A major goal of our work is to use small angle neutron scattering to find direct evidence for surface enrichment in nanodroplets.

We produce nanodroplets with radii between 5 nm and 12 nm and number densities in the range of 1011 cm-3 to 1012 cm-3 by rapidly expanding a dilute vapor mixture of D2O (or other condensible) and N2 in a supersonic nozzle apparatus (Refer to Figure 1) (Refer to Reference 1). The volume fraction of droplets is always less than 10-5 and, thus, the scattering signals are close to the instrument background. Furthermore, the high velocity of the droplets (400 m/s to 500 m/s) is comparable to the speed of the neutrons. This leads to a Doppler shift in the momentum of the scattered neutrons (Refer to References 2 and 3), and the 2-D scattering patterns are anisotropic (Refer to Figure 2) even though spherical droplets at rest scatter isotropically. This Doppler shift can be used to directly measure the velocity of the particles, and, in a test case, this velocity was within 2 % of the average velocity derived from the pressure trace measurements (Refer to Reference 2).

Graphics Caption FIGURE 1. The nozzle is placed in the sample chamber at right angles to the neutron beam.

Graphics Caption FIGURE 2. The observed 2-D scattering pattern for an aerosol with an average particle velocity vp = 435 m / s measured using a neutron wavelength of λ = 1.0 nm (vn = 400 m / s). The contour levels correspond to absolute intensities of 0.08, 0.03, 0.008 and 0.003 cm-1, respectively.

Graphics Caption FIGURE 3. Scattering from a D2O – h-butanol aerosol is compared to that from a H2O – d-butanol aerosol formed under identical conditions in the nozzle. Both aerosols contain approximately 6 % molecular fraction butanol. In the high q region, the intensity falls off as q-4 for the D2O-rich droplets, but only as q-2 for the H2O-rich droplets. Our preliminary modeling shows that the feature in the region 0.4 nm-1 < q < 0.5 nm-1 of the H2O – d - butanol spectrum cannot be reproduced by a well mixed droplet model. S D D is the sample-to-detector distance.

Rather than simply examine nanodroplets for which segregation of the components is severe, we use different materials to observe the transition between well-mixed and fully segregated droplets. Thus, our experiments include binary mixtures of H2O, D2O, ethanol, and n - butanol, or its fully deuterated analog d - butanol. Almost degenerate mixtures, such as D2O – H2O, appear to form well-mixed droplets (Refer to Reference 4). In contrast, binary nanodroplets containing H2O (or D2O) and a low molar concentration of d - butanol (or h - butanol) should consist of a water-rich core and an alcohol-rich shell. Our recent S A N S experiments with H2O – d - butanol nanodroplets clearly demonstrate for the first time the existence of the alcohol-rich shell. As illustrated in Figure 3, for these droplets the scattered intensity falls off as q-2 in the high q region, as is characteristic for scattering by a shell. Furthermore, the feature in the region 0.4 nm-1 < q < 0.5 nm-1 cannot be reproduced by a well-mixed droplet model that matches the low q data. Also shown in Figure 3 is the scattering spectrum from the complementary D2O – h-butanol experiment. As expected, the scattered intensity is much higher, and the signal intensity decreases as q-4 in the high q region. In this case, the D2O-rich core dominates the scattering.

In summary, S A N S provides us with a powerful new way to study the properties of nanometer sized liquid droplets in the environment in which they form. To date, it is the only technique to directly probe the microstructure of aerosol nanodroplets. Combined with pressure trace measurements and modeling, S A N S provides information critical to our understanding of droplet formation and growth in the nanometer size regime.


[1] B. E. Wyslouzil, J. L. Cheung, G. Wilemski, and R. Strey, Phys. Rev. Lett. 79, 431 (1997).

[2] B. E. Wyslouzil, G. Wilemski, J. L. Cheung, R. Strey, and J. Barker, Phys. Rev. E 60, 4330 (1999).

[3] G. Wilemski, Phys. Rev. E 61, 557 (2000).

[4] C. H. Heath, K. A. Streletzky, J. Wolk, B. E. Wyslouzil, and R. Strey,

in Nucleation and Atmospheric Aerosols, 2000, ed. by B. N. Hale and M. Kulmala, American Institute of Physics, New York (2000) p. 59.


B. E. Wyslouzil, C. H. Heath, U. M. Dieregsweiler, and K. A. Streletzky
Department of Chemical Engineering
Worcester Polytechnic Institute
Worcester, MA 01609

R. Strey, J. Wölk
Institut für Physikalische Chemie
Universität zu Köln
D-50939 Cologne, Germany

G. Wilemski
Physics Department
University of Missouri – Rolla
Rolla, MO 65409

J. Barker
N I S T Center for Neutron Research
National Institute of Standards and Technology
Gaithersburg, MD 20899-8562

Back to CHRNS Scientific Highlights

Last modified 29-May-2002 by website owner: NCNR (attn: Jeff Krzywon)