3D molecular resolution at solvated Fluorite (111) surface: theory meets experiment

Matt Watkins,

School of Mathematics and Physics, University of Lincoln, UK

Scanning probes

Atomic force microscopy and Scanning Tunnelling microscopy.

Local measurements.

Invasive?

Ambient or liquid conditions?

3D Solvation Structures on CaF$_2$

UHV measurements at 4K are all very well, but some applications require a more fluid environment

  • solar cells
  • batteries
  • photocathode / anodes

and of course study of material processes like weathering, corrosion, catalysis

3D Solvation Structures on CaF$_2$

UHV measurements at 4K are all very well, but some applications require a more fluid environment

  • solar cells
  • batteries
  • photocathode / anodes

and of course study of material processes like weathering, corrosion, catalysis

How do we see water density above a surface?

How do we see water density above a surface?

AFM - we'll see that this allows 3D mapping of the interface structure.

2 sided strategy: manipulate both experiment and simulation data for comparison

K Miyazawa, N Kobayashi, MW, AL Shluger, K Amano, T Fukuma, Nanoscale 8, 7334 (2016)

Theory: Solvent tip model

We don't know the actual tip apex - historical problem with AFM.

Hypothesize that there are likely strongly bound water molecules at the tip apex - and take them as the tip

They feed force back onto the cantilever.

M Watkins, B Reischl, The Journal of chemical physics 138, 154703 (2013)

Statistical mechanical model

pure water model leads to

$$ \Delta\Delta G(\mathbf{r}) = -k_BT \ln \frac{\rho(\mathbf{r})}{\rho_{\rm{bulk}}} $$

for the free energy change of bringing the tip model (water molecule) from the bulk liquid to $\mathbf{r}$

$$ F(\mathbf{r}) = \frac{\partial \Delta\Delta G(\mathbf{r})}{\partial z} = \frac{k_bT}{\rho(\mathbf{r})}\frac{ \partial \rho(\mathbf{r})}{\partial z} $$

this is the 'short range' force exerted on the tip apex.

M Watkins, B Reischl, The Journal of chemical physics 138, 154703 (2013)

Macro vs Nano

cantilever is macroscopic, tip apex is nanoscopic

experiment measures frequency change due to all interactions - macro + nano

Experiment: Macroscopic effects

removed empirically by subtracting force from averaged long range scan data over several surface locations

K Miyazawa, N Kobayashi, MW, AL Shluger, K Amano, T Fukuma, Nanoscale 8, 7334 (2016)

Experiment: fast AFM

$\textbf{3D AFM technique}$ and $\textbf{pattern matching}$ routines for massive speed up in image collection efficiency.

  • Allows image collection within few minutes of exposure of surface to liquid

    • (We also use it on the simulation data)

  • Enables data collection in pure water.

  • No longer true atomic resolution

K Miyazawa, N Kobayashi, MW, AL Shluger, K Amano, T Fukuma, Nanoscale 8, 7334 (2016)

Qualitative comparison of data

K Miyazawa, N Kobayashi, MW, AL Shluger, K Amano, T Fukuma, Nanoscale 8, 7334 (2016)

Can also compare to MD free energy calculation with explicit tips

caf2_md

STA does better than an incorrect tip model

K Miyazawa, N Kobayashi, MW, AL Shluger, K Amano, T Fukuma, Nanoscale 8, 7334 (2016)

Realistic environment

The above were all obtained in pure water.

  • requires complicated experimental protocols

  • not a realistic environment for many of the motivations

Realistic environment

The above were all obtained in pure water.

  • requires complicated experimental protocols

  • not a realistic environment for many of the motivations

What changes in electrolyte solution?

Comparison to supersaturated solution

  • The supersaturated solution ( s = 100 ) was prepared by mixing the same amounts of 38 mM CaCl$_2$ and 76 mM KF solutions

caf2_large

K Miyazawa, MW, AL Shluger, T Fukuma, Nanotechnology 28, 245701 (2017)

Adsorbed ions?

  • we use the same protocol as before

caf2_large

K Miyazawa, MW, AL Shluger, T Fukuma, Nanotechnology 28, 245701 (2017)

Adsorbed ions?

caf2_large

Possible specific cation adorption.

K Miyazawa, MW, AL Shluger, T Fukuma, Nanotechnology 28, 245701 (2017)

Challenge for simulation

here is a ~20 ns simulation of the supersaturated solution

  • of the order of 1 CaF$_2$ ion complex per 50 nm$^2$
In [4]:
%%HTML
<video width="520" height="500" controls>
  <source src="images/F2.mp4" type="video/mp4">
</video>

Conclusions

  • Evidence suggests here that stable tips are not very invasive – measurement of equilibrium water density
  • Theory vs experiment: short range forces above Ca.
  • Better agreement between implicit tip model and experiment than explicit MD
    • better tip models are needed!
  • Towards mapping specific ion adsorption sites and exploring electrolyte solutions
    • need to be cleverer than brute force to simulate

Thanks to

  • Leverhulme Trust for initial funding
  • Alex Shluger, University College London
  • Takeshi Fukuma, Kanazawa University
  • Adam Foster, Aalto
  • Bernhard Reichl, Curtin University

Lincoln is here