Design of Experiments for Nanocrystal Syntheses: A How-To Guide for Proper Implementation

Abstract

The understanding and control of colloidal nanocrystal syntheses are essential for discovery and optimization of desired properties and therefore play a key role in the applications of these materials. Typical one variable at a time (OVAT) methods limit the ability of researchers to achieve such goals by providing one-dimensional insight into a complex, multidimensional experimental domain, wasting precious resources in the process. Design of experiments (DoE) in conjunction with response surface methodology (RSM) offers an accelerated route for multivariate investigation and optimization of nanocrystal syntheses. The method enables systematic analysis and multidimensional modeling of the independent and dependent effects that any number of factors have on chosen responses, resulting in easy optimization of a large synthetic space in a fraction of the experiments. Herein, we will outline the general steps to follow when utilizing DoE and RSM for screening and optimization of nanocrystal syntheses, as well as the background needed to appropriately design an investigation and understand the results.

Predictive Synthesis of Copper Selenides Using a Multidimensional Phase Map Constructed with a Data-Driven Classifier

“…DoE screening and optimization matrices were utilized to construct the surrogate model. Despite the inability of the data to be modeled via regression because of its discrete nature, DoE provides an orthogonally balanced experimental sampling of the design space. , This prevents the over- and under-sampling of any one region in the n -dimensional domain, which can occur when other techniques like random sampling are used …”

“…One solution is to utilize data-driven learning to render synthetic phase maps that allow for rational targeting of materials within the high-dimensional variable space. Because phase is a categorical (or discrete) outcome, regression-based multivariate techniques like design of experiments (DoE) and response surface methodology (RSM) cannot be used because they require continuous outcomes. , Deep learning techniques like convolutional neural networks can map the multidimensional variable space, but they require large data sets that are not feasible when novel chemistry is being employed and/or done in a low-throughput manner. , On the other hand, a trained classification algorithm can handle both smaller data sets and categorical variables, making it a tractable solution to this problem …”

Artificial Scanning Electron Microscopy Images Created by Generative Adversarial Networks from Simulated Particle Assemblies

“…[ 26–29 ] Here, AI can help to analyze data and to predict suitable synthetic routes, taking experimental data (usually from electron microscopy) into account. [ 30–33 ]…”

“…[26][27][28][29] Here, AI can help to analyze data and to predict suitable synthetic routes, taking experimental data (usually from electron microscopy) into account. [30][31][32][33] Here, we present a considerable enhancement of the GAN procedure by introducing image depth information in the form of simulated height maps. This gives the GAN additional information on the particle positions and improves the simulation quality.…”

Electronic Lab Notebooks for Materials Synthesis

“…Thanks to ELNs, new researchers can easily access current synthesis protocols and also learn what has already been tried within the lab. Last but not least, ELNs represent a huge resource for data-driven discovery, which many materials chemists are pursuing in collaboration with computational scientists. , Using reliable data generated within the same laboratory and under standardized conditions is invaluable for comparison with larger data sets extracted via data mining from the literature, which is challenged by the lack of completeness and reporting standardization across different laboratories . Because ELNs provide the primary interface to research data, an opportunity exists to directly integrate computational approaches in order to automatically derive insights into the data and suggest the next step in the synthesis development.…”