Accepted Articles

Abstract

In this project, I delineate a method for addressing the scarcity of energy and cutting-edge microchip production supplies that were cut off from the US. Experts have been indicating peak oil has passed, and several recent global crises, including the war in Ukraine and the COVID-19 pandemic, have exacerbated an era of energy poverty, which in turn produced global increases in energy and food prices by 50% and 20%, respectively. According to experts, there can be further similar changes because of the war in the Middle East. The White House has demonstrated commitment to bringing novel nanotechnology, specifically semiconductor chips, which are ingrained in the automotive, aerospace, and technology industries. As they are currently only produced in Taiwan, the economic safety and diminishing effects of the current economic downturn depend on bringing production home through the novel technology I outline. State-of-the-art production of microchips only exists overseas, which caused steep price increases by 50% of cars to the point, for example, of being completely absent in New York State. I demonstrate a novel methodology in nanocatalyst and nanotechnology real-time characterization using a novel mathematical framework for machine learning algorithms. Prior approaches failed to satisfy the requirements of nanoscale, real-time analysis needed for lights-out and smart manufacturing, as prioritized by the government. Previous research in machine learning applications to materials science did not have adapt-on-the-fly models to exploit the hidden patterns of the particular material dataset, which are necessary to operate novel flexible manufacturing smart technology moves in the nanoscale. Past models failed to capture the crucial features of nanomaterial structure and properties, which my multi-task algorithm for variational auto-encoding (MAVEN) does. MAVEN creates a disentangled, interpretable latent space through my novel mathematical framework by performing multi-objective optimization for tasks tailored to the data, including novel loss functions and evaluation metrics. I demonstrate the power of this method through studying palladium nanoparticles, which are potent materials used prominently in industrial catalysis, batteries, and fuel cells and essential to greener, optimized systems of energy. New smart manufacturing, including cutting-edge microchips, requires palladium catalysts. Results demonstrate algorithmic independence and real-time structural analysis, which are essential for efficient production. Such information and processing were not previously available for nanocatalyst design and analysis. Furthermore, MAVEN’s interpretable capabilities create insight into the nature of fine structure spectra relationships, which are used as feedback in nanomanufacturing. MAVEN demonstrates efficacy in promoting a greener, more efficient energy model, which will bring advanced scientific and computational production back to the US.

Reference

[1] Olesya S. Ablyasova, Meiyuan Guo, Vicente Zamudio-Bayer, Markus Kubin, Tim Gitzinger, Mayara da Silva San tos, Max Flach, Martin Timm, Marcus Lundberg, J. Tobias Lau, and et al. Electronic structure of the complete series of gas-phase manganese acetylacetonates by x-ray absorption spectroscopy. The Journal of Physical Chemistry A, 127(34):7121–7131, 2023. 

[2] M. Ahmadi, J. Timoshenko, F. Behafarid, and B. Roldan Cuenya. Tuning the structure of pt nanoparticles through support interactions: An in situ polarized x-ray absorption study coupled with atomistic simulations. The Journal of Physical Chemistry C, 123(16):10666–10676, 2019. 

[3] Carlos Alejandre, Ortzi Akizu-Gardoki, and Erlantz Lizundia. Optimum operational lifespan of household appliances considering manufacturing and use stage improvements via life cycle assessment. Sustainable Production and Consumption, 32:52–65, Jul 2022. 

[4] Md. Eaqub Ali, Md. Motiar Rahman, Shaheen M. Sarkar, and Sharifah Bee Hamid. Heterogeneous metal catalysts for oxidation reactions. Journal of Nanomaterials, 2014:1–23, Dec 2014. 

[5] Morris Argyle and Calvin Bartholomew. Heterogeneous catalyst deactivation and regeneration: A review. Catalysts, 5(1):145–269, Feb 2015. 

[6] John N. Armor. A history of industrial catalysis. Catalysis Today, 163(1):3–9, Apr 2011. 

[7] Michael L. Baker, Michael W. Mara, James J. Yan, Keith O. Hodgson, Britt Hedman, and Edward I. Solomon. K- and l-edge x-ray absorption spectroscopy (xas) and resonant inelastic x-ray scattering (rixs) determination of differential orbital covalency (doc) of transition metal sites. Coordination Chemistry Reviews, 345:182–208, 2017. 

[8] Jon Bateman. U.s.-china technological “decoupling”: A strategy and policy framework ..., 2022. 

[9] Tarek Ben Hassen and Hamid El Bilali. Impacts of the russia-ukraine war on global food security: Towards more sustainable and resilient food systems? Foods, 11(15):2301, 2022. 

[10] Matthew R. Carbone, Shinjae Yoo, Mehmet Topsakal, and Deyu Lu. Classification of local chemical environments from x-ray absorption spectra using supervised machine learning. Phys. Rev. Mater., 3:033604, Mar 2019. 

[11] Danilo S. Carvalho, Giangiacomo Mercatali, Yingji Zhang, and Andre Freitas. Learning disentangled representa tions for natural language definitions, 2023. 

[12] Tian Qi Chen, Xuechen Li, Roger B. Grosse, and David Duvenaud. Isolating sources of disentanglement in variational autoencoders. CoRR, abs/1802.04942, 2018. 

[13] Raphaël Dang-Nhu. Evaluating disentanglement of structured representations. 2021. [14] Harry Dent and Gordon Long. Financial repression authority, Apr 2016. 

[15] Kien Do and Truyen Tran. Theory and evaluation metrics for learning disentangled representations. CoRR, abs/1908.09961, 2019. 

[16] Kien Do and Truyen Tran. Theory and evaluation metrics for learning disentangled representations. 02 2020. [17] Camila Domonoske. Why car prices are still so high - and why they are unlikely to fall anytime soon, Mar 2023. 

[18] Lin Gao, Jie Yang, Tong Wu, Yu-Jie Yuan, Hongbo Fu, Yu-Kun Lai, and Hao Zhang. SDM-NET: deep generative network for structured deformable mesh. CoRR, abs/1908.04520, 2019. 

[19] Unai Garciarena, Alexander Mendiburu, and Roberto Santana. Towards automatic construction of multi-network models for heterogeneous multi-task learning. ACM Trans. Knowl. Discov. Data, 15(2), mar 2021. 

[20] Juan García-Serna, Raúl Piñero-Hernanz, and Desirée Durán-Martín. Inspirational perspectives and principles on the use of catalysts to create sustainability. Catalysis Today, 387:237–243, Mar 2022. 

[21] Alejandro de la Garza. Gaza’s energy crisis could turn hospitals into “morgues”, Oct 2023. 

[22] Spyros Gidaris and Nikos Komodakis. Generating classification weights with GNN denoising autoencoders for few-shot learning. CoRR, abs/1905.01102, 2019. 

[23] Spyros Gidaris and Nikos Komodakis. Generating classification weights with GNN denoising autoencoders for few-shot learning. CoRR, abs/1905.01102, 2019.