Sabin Baral

Project Overview

This research focuses on developing efficient methodologies for fabricating and characterizing block copolymer thin films at scale. Block copolymers self-assemble into nanoscale structures that have applications in next-generation semiconductor manufacturing, filtration membranes, and energy storage devices. The high-throughput approach allows for rapid screening of processing conditions to optimize film morphology and properties.

My Role & Responsibilities

As the lead undergraduate researcher on this project, I designed and executed the experimental workflow including:

• Sample preparation using spin-coating and thermal annealing techniques
• Conducting GIWAXS and GISAXS measurements at synchrotron facilities
• Data analysis and pattern interpretation using IgorPro and Python
• Developing automated processing scripts to handle large datasets

Techniques & Methods

• Grazing-Incidence Wide-Angle X-ray Scattering (GIWAXS) - Crystalline structure analysis
• Grazing-Incidence Small-Angle X-ray Scattering (GISAXS) - Morphology characterization
• Atomic Force Microscopy (AFM) - Surface topography imaging
• Thermal Analysis - DSC and TGA for phase behavior studies
• Computational Tools - MATLAB and Python for data processing and visualization

Results & Outcomes

Developed a streamlined workflow that reduced characterization time by 40% while maintaining data quality. Identified key processing parameters that influence domain spacing and orientation. Results contributed to understanding structure-property relationships in block copolymer systems. Currently preparing 300+ thin film samples for our next beamtime at ALS or NSLS.

Workflow Video

Links & Resources

Lab Group Website

Project Overview

Organic photovoltaics (OPVs) offer a promising pathway to sustainable energy with advantages like flexibility, low-cost manufacturing, and tunable optical properties. This project investigated the thermal stability of polymer-based solar cells, addressing one of the key challenges limiting their commercial viability. Understanding degradation mechanisms under thermal stress is critical for extending device lifetime and improving performance reliability.

My Role & Responsibilities

I was responsible for the complete device lifecycle from fabrication to characterization:

• Fabricating organic solar cells using solution processing and vacuum deposition techniques
• Conducting accelerated aging studies under controlled temperature conditions
• Measuring current-voltage characteristics and power conversion efficiency
• Performing morphological analysis using AFM to track structural changes
• Analyzing thermal stability using TGA and DSC and X ray techniques.
• Correlating performance degradation with structural and compositional changes

Techniques & Methods

• Device Fabrication - Spin coating, thermal evaporation, and encapsulation
• J-V Characterization - Solar simulator and source meter measurements
• Atomic Force Microscopy (AFM) - Surface morphology and phase separation analysis
• Thermal Analysis - TGA for decomposition temperature, DSC for glass transition
• UV-Vis Spectroscopy - Absorption spectra and optical bandgap determination
• Data Analysis - OriginPro and Python for statistical analysis and visualization

Results & Outcomes

Identified critical temperature thresholds beyond which device performance degrades significantly. Discovered that morphological stability of the active layer is a primary factor in thermal degradation. Proposed optimized annealing protocols that improve initial efficiency while maintaining better long-term stability.

Links & Resources

Research Poster | Lab Website

Project Overview

This project has evolved from solving everyday laboratory inconveniences to engineering critical infrastructure for automated research workflows. By leveraging additive manufacturing, I design and fabricate custom interfaces that bridge the gap between standard laboratory equipment and advanced robotic systems. Recent work focuses on high-throughput sample handling and automated characterization, enabling the integration of North Robotics arms and motorized stages into closed-loop experimental cycles. This initiative demonstrates how rapid prototyping can unlock new capabilities in materials science research.

My Role & Responsibilities

End-to-end ownership of the design and fabrication process:

• Identifying problems and needs in laboratory and practical settings
• Conceptualizing and designing solutions using CAD software (AutoCAD, OpenSCAD)
• Optimizing designs for printability, structural integrity, and material efficiency
• 3D printing using FDM technology with mostly PLA
• Post-processing and quality testing of printed components
• Iterating designs based on lab members feedback and performance testing

Techniques & Methods

• CAD Design - AutoCAD for complex geometries, OpenSCAD for parametric modeling
• 3D Printing - FDM (Fused Deposition Modeling) with PLA, PETG, and ABS materials
• Material Selection - Choosing appropriate polymers based on mechanical requirements
• Process Optimization - Print parameter tuning for quality and efficiency
• Iterative Design - Rapid prototyping and testing cycles

Results & Outcomes

Successfully designed and fabricated over 20 unique solutions including custom sample holders for characterization equipment, modular storage systems for laboratory consumables, and specialized fixtures for experimental setups. These tools have improved lab efficiency and reduced equipment costs. The project has also enhanced my skills in design thinking, materials selection, and practical engineering problem-solving.

Example Applications

• Ossila Four-Point Probe Automated Mount
  The Design: A precision bracket that mounts the Ossila probe head directly onto a Thorlabs KVS30/M motorized vertical stage.
  Impact: Transitioned the lab from manual to fully automated conductivity measurements. By using Python scripts to control the stage, the probe now contacts samples at the exact same pressure and height every time, completely eliminating human error.
• North Robotics N9 Drop Cast Holder
  The Design: A custom platform that replaces the standard "wafer garage" on the N9 system, utilizing existing alignment pins for a perfect fit.
  Impact: This "hack" repurposed a high-end robot for specific thin-film applications, allowing it to directly deposit solutions into drop-cast structures with high repeatability.
• High-Capacity Thin-Film Sample Holders
  The Design: Custom-engineered holders with optimized spatial geometry to secure thin-film samples during processing.
  Impact: Redesigned the standard architecture to fit 100 samples within the same footprint that originally held only 20. This 500% increase in capacity significantly reduced cost-per-sample and maximized throughput.
• Universal Testing Machine (UTM) Protective Cover
  The Design: A custom-fitted housing for the Universal Testing Machine's sensitive mechanics.
  Impact: Protects internal components from dust and accidental splashes during wet tensile testing, preventing corrosion and debris buildup on the drive system.
• Lab Hood Edge-Mount Pipette Holder
  The Design: A space-saving C-port clip that attaches directly to the metal lip of a fume hood.
  Impact: Cleared the deck of clutter by moving expensive tools to eye level, improving workflow organization and maintaining better airflow and sterility within the hood.
• Robosoccer Competition Attachments
  The Design: A two-part system featuring a front-end "pusher/plow" for ball control and a protective top housing.
  Impact: Turned a base-model robot into a championship winner by providing superior ball handling and protecting the electronics from high-speed collisions.

Laying the "Walking Blocks"

A true closed-loop system requires mastering individual instrument control first. By developing custom Python drivers, I have created a Hardware Abstraction Layer (HAL) that allows a central computer to coordinate disparate hardware—originally designed as standalone units—into a single, unified laboratory ecosystem.

Developed Hardware Infrastructure

Access the specific drivers developed for each laboratory "block":

The Integrated Autonomous Loop

I am currently integrating these drivers with our N9 North Robotics platform to achieve a fully automated cycle:

Project Overview

As part of the U.S. DOE-funded US CMS PURSUE Summer Research Internship, I contributed to the High Luminosity LHC (HL-LHC) upgrade. The program began with intensive computational training at Fermilab—covering the UNIX shell, PyROOT, and machine learning—before I relocated to Kansas State University. At K-State, I joined the US Midwest 2 testing site to focus on the electrical validation of bare modules (silicon sensors bump-bonded to CMS Readout Chips) for the CMS Inner Tracker upgrade.

My Role & Responsibilities

I transitioned from executing standard testing protocols to actively optimizing the laboratory's workflow:

• Conducted rigorous electrical testing and optical inspections of highly sensitive bare modules inside a Class 10,000 cleanroom.
• Identified critical efficiency bottlenecks in the terminal-heavy DAQ testing procedures.
• Designed and developed a centralized Python-based Graphical User Interface (GUI) from the ground up to streamline the testing workflow.
• Integrated hardware controls directly into the GUI, automating WPAC initialization, contact resistance checks, IREF trim-bit measurements, and Sensor I-V scans.

Techniques & Methods

• Software Engineering - Python, PyROOT, RDataFrame, and custom GUI development for hardware abstraction.
• Hardware Integration - Ph2_ACF framework, Keithley source-meters, and MPI-TS3000 probe station operation.
• Diagnostic Analysis - High-Charge scans and coupled/uncoupled crosstalk scans for dead pixel and open-bump identification.
• Cleanroom Protocol - ESD-safe handling, ISO 7 cleanroom standards, and microscopy for physical defect inspection.

Results & Outcomes

Successfully abstracted complex, terminal-level commands into an intuitive GUI, drastically improving the team's productivity and testing efficiency. The automated testing suite includes built-in exception handling and safe-exit procedures that automatically cut high-voltage and low-voltage power during faults, securing delicate hardware. Actively processed and graded multiple fragile modules, ensuring the early detection of bump-bonding defects before expensive downstream assembly steps.

Links & Resources

US CMS Operations at Fermilab

Project Overview

Built for the Mississippi State Board of Nursing (MSBN), VeriScript is a serverless, AI-assisted transcript review platform designed to combat nursing credential fraud. The system automatically ingests applicant PDFs, extracts critical academic data, runs a battery of deterministic fraud-detection rules, and routes flagged applications to a human reviewer. Crucially, the system is strictly advisory—it generates a defensible, immutable audit trail to augment staff judgment, making zero final licensing decisions on its own.

My Role & Responsibilities

I contributed across the entire full-stack architecture, from the infrastructure pipeline to the reviewer dashboard:

• Engineered a resilient, serverless AWS pipeline using Python and CDK, orchestrated by AWS Step Functions to ensure fully logged state transitions.
• Designed a dual-pass AI extraction engine combining Amazon Textract for visual forensics (detecting tampered text) and Amazon Bedrock Nova Pro for semantic academic extraction.
• Programmed a robust deterministic rules engine featuring 14 specific fraud checks across physical authenticity, educational chronology, and MSBN curriculum compliance.
• Developed a responsive, lightweight frontend dashboard in React 19 to translate complex AI outputs and flagged issues into an intuitive UI for MSBN staff.

Techniques & Methods

• Cloud & Infrastructure - AWS CDK, Lambda, Step Functions, DynamoDB (Single-Table Design), S3, API Gateway, CloudFront, Cognito.
• Frontend Development - React 19, TypeScript, Vite, custom JWT authorization.
• AI & Optical Character Recognition - Amazon Bedrock (Nova Pro), Amazon Textract, smart aggregation algorithms for multi-page flattening.
• System Architecture - Serverless orchestration, idempotent event handling, exponential backoff strategies, and warm Lambda client reuse.

Results & Outcomes

The platform fundamentally modernized the MSBN review process, cutting average manual review times from 45 minutes down to 15 minutes per application. By leveraging highly optimized AWS serverless architecture (costing approximately $0.35 per application), the solution delivered a ~65% cost reduction compared to purely manual reviews. The resulting system successfully protects public health by identifying sophisticated credential fraud while maintaining the transparent, appeal-ready documentation required of state regulatory boards.