This is a collection of codes developed for Paulson et al. (2021).

1. Run uptake_curve.py to generate uptake curve figures for each ALD system and some .csv files needed by ald_opt.py (for plotting purposes only). This will produce Figure 1a and Figures S1a, S2a, and S3a.
2. Run uptake_curve_trad.py to generate traditional uptake curve figures for each ALD system. This will produce Fig. 1b, and Figures S1b, S2b, and S3b.
3. Run ald_cost_sensitivity.py to explore the sensitivity of the C_var1 and C_var2 cost function components for different gas timings and imposed measurement noise levels. This will generate Fig. 4.
4. Run ald_opt_sensitivity_initcond.py to explore the sensitivity of the expert systems optimization to initial gas timing guesses. Change noise_imposed on line 210 to explore different levels of imposed measurement noise. This will generate Fig. 5.
5. Run ald_opt_sensitivity_nrep.py to explore optimization sensitivity to different numbers of repeated cycles with the same gas timings under different imposed measurement noise. Change noise_imposed on line 207 to explore different levels of imposed measurement noise, and acq on line 201 to explore each optimization algorithm ('random' for random optimization, 'phys' for expert systems optimization, and 'EI' for Bayesian optimization). This will generate Fig. 6 and 7.
6. Run ald_opt.py to explore optimization performance for the 4 ALD systems and 3 optimization strategies with changing imposed measurement noise. This will generate Figures 8-10 and S5-S13.

• core.py and sopt.py contain common functions used for plotting, output, surrogate modeling, and optimization.
• ALDmodel.py contains the basic ALD model considering the interaction of two precursors with a growing surface.
• physicsmodel.py contains the expert systems optimization algorithm.
• maximin_lhs_l2_10_4d.csv, and maximin_lhs_l2_20_4d.csv are 4-dimensional optimized latin-hypercube designs for 10 and 20 samples, respectively.

• python/3.7.6
• matplotlib/3.1.1
• numpy/1.19.2
• pandas/1.0.3
• scikit-learn/0.24.1
• scipy/1.4.1
• seaborn/0.9.0

Paulson, N.H., Yanguas-Gil, A., Abuomar, O.Y., Elam, J.W. “Intelligent agents for the optimization of atomic layer deposition,” ACS Applied Materials and Interfaces. (2021) Accepted