QuanDom

Quantum Dot Color 1
Dot 2
Energy Diagram
AI
Energy Diagrams 2

Inspiration

We were inspired by the recent Nobel Prize (in chemistry) on quantum dots. Why are the properties of quantum dots based on their size? How are they able to emit such a variety of colours? What semiconductors are best at making quantum dots? However, there were no good accessible simulations online that could visualize and explain this new technology… Thus, with these questions in mind, we set out to explore the kingdom of quantum dots.

What it does

When the type of semiconductor and the radius of the dot are chosen, our program will be able to output a graphical representation of the quantum energy levels and the color of the quantum dot. Specifically, the radius can be controlled through a slider. The semiconductor material can be chosen from given buttons, or the user can type in a compound they want to test. Our program will output: A vial that lights up with a beautiful array of vibrant colors; A short animation of the electron on the valence and conduction bands Numerical values for the energy of the quantum dot and the wavelength of light The colour in words (eg.: “red”, “green”, “blue”, etc.)

How we built it

The main structure was written using Python, while the machine learning components were written using TenserFlow. The data for machine learning was acquired using DataScript. The front-end interface was implemented using the customtkinter library. The graphs were created using the matplotlib library. Calculations were implemented as Python functions.

Challenges we ran into

Our first challenge was using the customtkinter library, TenserFlow, and DataScript, since it was our first time using it, but we learned throughout the hackathon. Furthermore, we had bugs when implementing the energy diagrams from matplotlib as animations in the interface due to high memory usage; some of our computers crashed while uploading big libraries and each other’s programs. Some of our other challenges were the lack of sleep and excess of sodas and coffee.

Accomplishments that we're proud of

We are very proud of our beautiful interface with awesome graphs, front-end graphics, and machine learning. It is informative, yet easy to use, so the public can understand how quantum dots work. The interactive component can spark more interest in quantum science! Furthermore, the machine learning component, if further developed, can potentially help predict what chemical compounds can be the quantum dot development.

What we learned

This is our first time at the hackathon, so we learned how to work together and put together everyone’s codes to make the whole project function. More specifically, we learned how to work with the customtkinter library, TensorFlow, and DataScript, as well as how to be efficient with variable names and organization. We learned how to describe physics phenomena with programming.

What's next for QuanDom

We would like to… Continue training AI with more parameters to improve our machine learning model. Because of the time limits, we trained the neural network with limited data. With more training, we would be able to get more accurate results. Make our model more accurate and complete. Specifically, we would like to include exciton radius in calculations and calculate strong, intermediate, and weak confinement. Be able to vary more parameters, such as the energy of the light shining on the quantum dot, the band gap energy and effective masses. Add more accessibility futures, such as reading out the colors and values.