Liquid Metal Surfaces P. S. Pershan SEAS

1
Liquid Metal SurfacesP. S. PershanSEAS Dept
of Physics, Harvard Univ., Cambridge, MA, USA

• Colleagues

Balagurusamy, V. S. K. Berman, E. Deutsch, M. DiMasi, E. Fukuto, M. Gebhardt , J. Gog, T. Graber, T. Grigoriev, A. Huber, P. Kawamoto, E. H. Kuzmenko, I. Lin, B. H. Magnussen, O. M. Mechler, S. Meron, M. Ocko, B. M. Pontoni, D. Regan, M. J. Sellner, S. Shpyrko, O. G. Steimer, C. Stoltz, S. Streitel, R. Tostmann, H. Yahel, E
Harvard, Non-Harvard, Beam Line
2
1st Synchrotron Studies Liquid Crystal Surfaces
Reflectivity
3
Kinematics Reflectivity Flat Surface
4
No Layering for Water and Simple Liquids
A. Braslau et al. PRL (1985).
5
Free Surface of Liquid Metal Hard Wall
Metallic Liquids (DEvelyn Rice 83)
6
Capillary Waves Thermal Roughness
7
Capillary Effects H20 Ga
Water (Schwartz 90)
F(Qz) for Liquid Ga (Regan, 96)
8
Diffuse Scattering ? Surface Tension(g)
9
Simplest Surface Structure Model
DCM (Magnussen 95)
10
Elemental Liquid Metals Studied
K Ga In Sn Bi Hg
DCM DCM DCM 1 1 ?
? ? ? ? ?

• Why are 1st Layers for Bi and Sn different from
  K, Ga and In?
• Why is Hg different from all others?

11
Eutectic Alloys
J. W. Gibbs 1920Surface Adsorption A/B
AlloyIf Surface Tension ?A gt ?B Surface is Rich
in B.
AxB1-x g(A)/g(B) DH (mixing) Concentration of Surface Layers Concentration of Surface Layers Concentration of Surface Layers
AxB1-x g(A)/g(B) DH (mixing) 1st 2nd 34d
GaxBi1-x 718/3781.90 4 Liquid-Liquid Phase Sep. Liquid-Liquid Phase Sep. Liquid-Liquid Phase Sep.
Ga83.5In16.5 718/5561.29 5 97In
In78Bi22 556/3781.47 -1 35Bi
Sn57Bi43 560/3781.48 1 96Bi 25Bi 53Bi
Au71Sn29 1100/5601.96 -10 96Sn lt1Sn 24Sn
Au72Ge28 1100/6211.77 -21 No Gibbs Absorption No Gibbs Absorption No Gibbs Absorption
Au82Si18 1100/8651.27 -30 4-layers, 2DXtal (AuSi2) 4-layers, 2DXtal (AuSi2) 4-layers, 2DXtal (AuSi2)
Pd81Ge19 1500/6212.4 -44 40 Å wetting layer (No Measureable Gibbs Absorption) 40 Å wetting layer (No Measureable Gibbs Absorption) 40 Å wetting layer (No Measureable Gibbs Absorption)
(kJ/mol)Takeuchi and Inoue, Mater. Trans. 46
(2005)
12
Gibbs Surface Adsorption(BiSn)
13
Surface Freezing Au82Si18Eutectic
There is no theoretical explanation!
14
AuGe Eutectic(Should be Similar to Au-Si)
g(Au)/g(Si or Ge) DH
Au72Ge28 1100/6211.77 -21
Au82Si18 1100/8651.27 -30
Au-Si

1. Bump?higher density in 1st layer.
2. No Energy effect ? Ge in 1st layer 40atm.

• Small Gibbs (Different from Au-Sn, etc)!
• No Enhanced Layering or 2D order(Different from
  Au-Si)!

15
AuSiGe-Ternary Eutectic
Surface Frozen Ge6.5 atm
What is the physics of the cross over from Si
type to Ge type surface between 2.5 atm and 6.5
atm?
16
Pd81Ge19(Dec.08)
Au82Si18 Pd81Ge19
Glass former yes better
DH -30 -44
Expected same 2D surface order for Pd81Ge19 as
Au82Si18! Not found however, something new!
Metallic Clusters (Giant Unit Cells)
Ref Urban Feuerbacher, J.Non-Crys.Sol.(04) Que
nched Icosahedral Clusters
Others NaCd2 30Å YbCu4.5
44-49Å Al3Mg2 28Å
17
Summary

• Metal/Vapor Interface?Atomic Layering
• Surface Structure Factor - F(Qz) Measurement
  affected by thermal roughness. Requires knowledge
  of surface tension.
• Surface tension measured with diffuse
  scattering
• Surface tension effect demonstrated for Ga/In
• Subtle differences in elemental surfaces (Ga, In,
  K vs. Sn, Bi vs Hg)
• Alloys Surface tension vs. Enthalpy of
  MixingGibbs absorption is not simple. No
  reliable theory.
• Au82Si18 anomalously strong layering and 2D
  order.Why are Au82Si18, Au72Ge28 and Pd81Ge19
  all different?
• Need for THEORY!
• New Result (Preliminary) Surfaces
  Icosahedral Metallic Clusters