Brianna Greenstein

Chemistry PhD Candidate at University of Pittsburgh

[Geoff Hutchison (@[email protected]) @ghutchis]

Last year, @greensteinbri published v1 of our OPEP model for predicting the performance of organic #solar using #compchem - this is v2 and shows tandem devices can have great performance

We're working on v3 soon (even bigger!) - happy to collaborate w/ both experiment and theory

[The Journal of Physical Chemistry @JPhysChem]

Work by @greensteinbri @ghutchis @Pitt_Chemistry establishes one of the largest data sets of non-fullerene acceptors, a new ML-derived model of device PCE & uses them to derive new tandem solar cell designs with non-overlapping optical absorption spectra.
https://t.co/gTDmmTlomz

[Brianna Greenstein @greensteinbri]

Check out my newest research, first co-authored with Danielle Elsey, on best practices and hyperparameter tuning of genetic algorithms for molecular discovery!

https://t.co/4Qhae9am8P

[Geoff Hutchison (@[email protected]) @ghutchis]

Congrats to @greensteinbri for winning a @PGHQuantum Graduate Fellowship "based on her exceptional ability and promise in the field of Quantum Science and Engineering."

[ScienceInsider @ScienceInsider]

BREAKING: White House issues new policy that will require, by 2026, all federally-funded research results to be freely available to public without delay, ending longstanding ability of journals to paywall results for up to 1 year. Coverage coming on @ScienceInsider.

[Geoff Hutchison (@[email protected]) @ghutchis]

While I won't be at #ACSFall2022 in Chicago, my group will be well-represented by @OmriAb @greensteinbri and Danielle Hiener right at the start on Sunday in #compchem

@OmriAb at 9:40 am in S402a McCormick Place
Danielle at 10:40 .. same room
@greensteinbri at 2:40 .. same room

[Brianna Greenstein @greensteinbri]

Excited to share my most recent work! I curated one of the largest non-fullerene acceptor/donor datasets to date (1,225 pairs), and also combined ML and genetic algorithms to find highly efficient combinations for tandem organic solar cells.

[Geoff Hutchison (@[email protected]) @ghutchis]

This work by @greensteinbri is now on @ChemRxiv - first with a huge set of non-fullerene photovoltaics to train ML models, then using the best models (~1.5% error) to discover new *tandem* organic solar cells. #compchem

[Brianna Greenstein @greensteinbri]

2 papers in a week! Excited to share my second publication, published in @JPhysChem Letters on predicting efficiencies with data-driven models for non-fullerene organic solar cells ☀️https://t.co/YXejMR0TWC

[Brianna Greenstein @greensteinbri]

I am so excited to share my first first-author publication, published in @JChemPhys! We used a genetic algorithm to find high-performing non-fullerene acceptors for organic solar cells! https://t.co/RTr7I3lSzb

[Brianna Greenstein @greensteinbri]

@AlysiaMandato Congrats!!

[Matthew @MCeeP]

Mini #CartoonAbstract made for @greensteinbri who was this years #RSCPoster runner up in #RSCDigital

[Brianna Greenstein @greensteinbri]

Honored to have won runner-up with my #RSCposter for #RSCdigital. Thank you @RoySocChem for organizing, it was so fun to be able to share my research virtually with participants around the world!

[Royal Society of Chemistry @RoySocChem]

#RSCPoster runner up for #RSCDigital is @greensteinbri. Congratulations, Brianna on going where no one has gone before. https://t.co/BMl6CupsWm

[Brianna Greenstein @greensteinbri]

@ghutchis @Pitt_Chemistry Thank you for all of your help and guidance!

[Brianna Greenstein @greensteinbri]

@dftba_forthepod This mutation step ensures diversity within the population so we are not limited to the units selected in the initial population.

[Brianna Greenstein @greensteinbri]

Chemical Space: The Final Frontier. Join me and go where no one has gone before... to find organic solar cell materials!
📝https://t.co/0jtmxCqQZK

#RSCPoster #RSCDigital #RSCEnergy

[Brianna Greenstein @greensteinbri]

@dftba_forthepod Thanks! Once a "child" molecule has undergone crossover, it has a 40% chance to mutate. This mutation can either be a replacement of a building block unit with a new one or rearranging the units for a new sequence.

[Brianna Greenstein @greensteinbri]

@anatmilo @pschwllr Thank you! The main criteria are 1) non-fullerene small molecules as acceptors (unfused and fused backbone structures) 2) donors must be polymers 3) fill factor above 40% since this parameter is highly dependent on processing conditions and the device may not be optimized

[Brianna Greenstein @greensteinbri]

@mmwieclaw @IOC_PAS @FurChem Nice! Does it work with Orca output files or just Gaussian?

[Brianna Greenstein @greensteinbri]

@pschwllr Yes! The dataset can be found here: https://t.co/MxexluyE9s (the references can be found in the supplementary info of the paper)

All work for this paper is openly available on our groups's Github: https://t.co/C5Wf9IYGzQ

[Brianna Greenstein @greensteinbri]

@pschwllr Thanks! I curated a dataset containing ~100 non-fullerene acceptors/ polymer donor pairs with experimental PCE. My current work is expanding this dataset to 1000 pairs to update the model.

[Brianna Greenstein @greensteinbri]

@pablos_pv The preprint for creating this model can be found here: https://t.co/fFPG0l69tE

[Brianna Greenstein @greensteinbri]

@pablos_pv Thank you! We developed a data-driven model using LASSO regression to predict the PCE. A donor and acceptor each undergo QM calculations and properties from those calculations are used as descriptors in the model.