The Kitchin Research Group

In many liquid metal alloys the diffusivity and viscosity are related to each other through the Stokes–Einstein–Sutherland (SES) equation. This is useful because it is difficult to measure these properties in liquid metals, and the correlation can be used in design. Near the melting point, however, this relation often fails. In this work we use molecular dynamics to investigate deviations in the SES for molten Si-Al. We find that the viscosity changes faster than the diffusivity due to the formation of atomic clusters. These clusters cause the SES deviation in this liquid alloy system.

@article{zhan-2021-origin-stokes,
  author =       {Ni Zhan and John R. Kitchin},
  title =        {Origin of the Stokes-Einstein Deviation in Liquid Al-Si},
  journal =      {Molecular Simulation},
  volume =       {},
  number =       {},
  pages =        {1-11},
  year =         2021,
  doi =          {10.1080/08927022.2021.2012572},
  url =          {http://dx.doi.org/10.1080/08927022.2021.2012572},
  DATE_ADDED =   {Fri Dec 17 07:41:39 2021},
}

Copyright (C) 2021 by John Kitchin. See the License for information about copying.

org-mode source

Org-mode version = 9.5.1

In this new collaborative work we show how we combine a high-throughput photoreactor with a design of experiment approach to efficiently find optimal in situ synthesis compositions for making metal nanoparticle catalysts for light-driven hydrogen production. The challenge in this work is there are several components that interact with each other including metal salts, stabilizing ligands and photosensitizers, as well as some noise in the measurements. It is difficult to optimize these components one at a time, so we use a design of experiment approach. The high-throughput data enables us to explore the composition space around the optimum and to identify specific compositions for focused and expensive characterization efforts. We use this on Au, Cu, Fe and Ni and show that all of them can have high activity when they are independently optimized. It is interesting to note that Au and Cu form stable metallic nanoparticles, but Ni appears to form oxide nanoparticles and Fe appears to form sulfide nanoparticles.

@article{bhat-2022-accel-optim,
  author =       {Maya Bhat and Eric Lopato and Zoe C Simon and Jill E Millstone
                  and Stefan Bernhard and John R Kitchin},
  title =        {Accelerated Optimization of Pure Metal and Ligand Compositions
                  for Light-Driven Hydrogen Production},
  journal =      {Reaction Chemistry \& Engineering},
  volume =       {},
  number =       {},
  pages =        {},
  year =         2022,
  doi =          {10.1039/d1re00441g},
  url =          {http://dx.doi.org/10.1039/D1RE00441G},
  DATE_ADDED =   {Mon Nov 29 17:00:12 2021},
}

Copyright (C) 2021 by John Kitchin. See the License for information about copying.

org-mode source

Org-mode version = 9.5

In this paper we show that we can stabilize in situ synthesized nanoparticles by including stabilizing ligands in the synthesis. This gives the in situ particles comparable stability to pre-synthesized particles, while retaining the flexibility of the in situ synthesis. We show that not all stabilizing ligands work for all metals, although PEGSH is a reasonable starting point for the metals we considered in this work. Additionally, we show that the incorporation of PEGSH can make some metals that appear inactive without stabilization due to precipitation, become active when stabilized. Although the addition of stabilizing ligands complicates the optimization of the synthesis conditions, it leads to more reproducible, and in some cases better results than leaving them out.

@article{simon-2021-ligan-enhan,
  author =       {Zoe C. Simon and Eric M. Lopato and Maya
                  Bhat and Paige J. Moncure and Sarah M. Bernhard and John R.
                  Kitchin and Stefan Bernhard and Jill Millstone},
  title =        {Ligand Enhanced Activity of in Situ Formed Nanoparticles for
                  Photocatalytic Hydrogen Evolution},
  journal =      {ChemCatChem},
  volume =       {},
  number =       {},
  pages =        {},
  year =         2021,
  doi =          {10.1002/cctc.202101551},
  url =          {http://dx.doi.org/10.1002/cctc.202101551},
  DATE_ADDED =   {Mon Nov 29 17:01:16 2021},
}

Copyright (C) 2021 by John Kitchin. See the License for information about copying.

org-mode source

Org-mode version = 9.5

Say you have acquired some data, and fitted a nonlinear model to it. The fit looks good, but how good are predictions from the model? In high dimensional space it is tricky to tell if you are extrapolating, what should you do? First, read this paper! We illustrate a simple tool called the delta method that can help you estimate the prediction uncertainty from your model using automatic differentiation to get the required derivatives. We show some examples, and how you can use this method to refine what data you should use in fitting your models. We even show how to handle some tricky cases with models where the Hessian is not invertable! The examples are from molecular simulation, but the approach is general and should work for other models too.

@article{zhan-2021-uncer-quant,
  author =       {Ni Zhan and John R. Kitchin},
  title =        {Uncertainty Quantification in Machine Learning and Nonlinear
                  Least Squares Regression Models},
  journal =      {AIChE Journal},
  volume =       {},
  number =       {},
  pages =        {},
  year =         2021,
  doi =          {10.1002/aic.17516},
  url =          {http://dx.doi.org/10.1002/aic.17516},
  DATE_ADDED =   {Mon Nov 8 08:51:21 2021},
}

Checkout the video brief here:

Copyright (C) 2021 by John Kitchin. See the License for information about copying.

org-mode source

Org-mode version = 9.5

Over the past few months, I have been making a series of short Python videos on YouTube. You can find the playlist at https://www.youtube.com/playlist?list=PL0sMmOaE_gs2yzwy54kLZk5c1ZH-Nh-62. They are not particularly well organized there, since I make them in the order I feel like, and when I have some spare time, so today I took some time to organize them by some topics here. If you find them useful, please subscribe to the channel and tell your friends about them!

• Basic Python
  • Conditional statements https://www.youtube.com/watch?v=XymPeBMILUY&list=PL0sMmOaE_gs2yzwy54kLZk5c1ZH-Nh-62&index=2
  • Membership operators https://www.youtube.com/watch?v=CZstHHjfCHo&list=PL0sMmOaE_gs2yzwy54kLZk5c1ZH-Nh-62&index=3
  • Logical operators https://www.youtube.com/watch?v=q-uDWDSF0l8&list=PL0sMmOaE_gs2yzwy54kLZk5c1ZH-Nh-62&index=4
  • Comparison operators https://www.youtube.com/watch?v=BayqeeF_iKM&list=PL0sMmOaE_gs2yzwy54kLZk5c1ZH-Nh-62&index=5
  • Operator precedence https://www.youtube.com/watch?v=Vy4USf-UVAI&list=PL0sMmOaE_gs2yzwy54kLZk5c1ZH-Nh-62&index=19
  • Iteration https://www.youtube.com/watch?v=7rVsD9kT9RM&list=PL0sMmOaE_gs2yzwy54kLZk5c1ZH-Nh-62&index=16
  • functions https://www.youtube.com/watch?v=kidVLLHtzbc&list=PL0sMmOaE_gs2yzwy54kLZk5c1ZH-Nh-62&index=17
• Math
  • Integration https://www.youtube.com/watch?v=6xAO7te0kdA&list=PL0sMmOaE_gs2yzwy54kLZk5c1ZH-Nh-62&index=10
  • polynomials https://www.youtube.com/watch?v=8e1yOKMTP8o&list=PL0sMmOaE_gs2yzwy54kLZk5c1ZH-Nh-62&index=11
  • Numpy array broadcasting https://www.youtube.com/watch?v=slWgreHaQNE&list=PL0sMmOaE_gs2yzwy54kLZk5c1ZH-Nh-62&index=23
  • Array broadcasting example https://www.youtube.com/watch?v=FXhEKNtvoVs&list=PL0sMmOaE_gs2yzwy54kLZk5c1ZH-Nh-62&index=24
• Root finding
  • solving nonlinear functions https://www.youtube.com/watch?v=KeRNoXWs_y0&list=PL0sMmOaE_gs2yzwy54kLZk5c1ZH-Nh-62&index=15
  • Newton's method https://www.youtube.com/watch?v=spLsyP-5PF8&list=PL0sMmOaE_gs2yzwy54kLZk5c1ZH-Nh-62&index=1
  • Replacing fsolve https://www.youtube.com/watch?v=_1bOzIYcDaA&list=PL0sMmOaE_gs2yzwy54kLZk5c1ZH-Nh-62&index=6
  • root failures https://www.youtube.com/watch?v=FaOJxeVfeH4&list=PL0sMmOaE_gs2yzwy54kLZk5c1ZH-Nh-62&index=26
• Differential equations
  • first order ODEs https://www.youtube.com/watch?v=4H0Qr-gxMN4&list=PL0sMmOaE_gs2yzwy54kLZk5c1ZH-Nh-62&index=14
  • events in ODEs https://www.youtube.com/watch?v=fv_-3ZtYBEo&list=PL0sMmOaE_gs2yzwy54kLZk5c1ZH-Nh-62&index=25
  • BVPs https://www.youtube.com/watch?v=vWcRuay1tt4&list=PL0sMmOaE_gs2yzwy54kLZk5c1ZH-Nh-62&index=7
• Optimization
  • minimizing a function https://www.youtube.com/watch?v=2HMKU2nHAbE&list=PL0sMmOaE_gs2yzwy54kLZk5c1ZH-Nh-62&index=13
  • constrained optimization https://www.youtube.com/watch?v=QKiOm1iqciE&list=PL0sMmOaE_gs2yzwy54kLZk5c1ZH-Nh-62&index=12
• Data
  • Using Excel in colab https://www.youtube.com/watch?v=rfcstL5eTbs&list=PL0sMmOaE_gs2yzwy54kLZk5c1ZH-Nh-62&index=8
  • Reading data files in colab https://www.youtube.com/watch?v=xf6qprxmBaM&list=PL0sMmOaE_gs2yzwy54kLZk5c1ZH-Nh-62&index=21
  • More ways to read data in colab https://www.youtube.com/watch?v=NjRd41QtU14&list=PL0sMmOaE_gs2yzwy54kLZk5c1ZH-Nh-62&index=22
  • linear regression https://www.youtube.com/watch?v=ZXSaLcFSOsU&list=PL0sMmOaE_gs2yzwy54kLZk5c1ZH-Nh-62&index=20
  • nonlinear regression https://www.youtube.com/watch?v=hbchKAgeDcU&list=PL0sMmOaE_gs2yzwy54kLZk5c1ZH-Nh-62&index=18
• Miscellaneous
  • hand written work in Jupyter lab https://www.youtube.com/watch?v=5qCY9Eoeoyo&list=PL0sMmOaE_gs2yzwy54kLZk5c1ZH-Nh-62&index=9
  • uncertainty quantification https://www.youtube.com/watch?v=KP-km6XedVg&list=PL0sMmOaE_gs2yzwy54kLZk5c1ZH-Nh-62&index=27
  • units in python https://www.youtube.com/watch?v=au5Jwz_STXI&list=PL0sMmOaE_gs2yzwy54kLZk5c1ZH-Nh-62&index=28

Copyright (C) 2021 by John Kitchin. See the License for information about copying.

org-mode source

Org-mode version = 9.5