About

Kunal Mukherjee is an assistant professor in Materials Science and Engineering at Stanford University. He previously served as an assistant professor at UC Santa Barbara from 2016 to 2020, and held postdoctoral appointments at IBM TJ Watson Research Center in 2016 and MIT in 2015. His professional background also includes work as a transceiver engineer at Finisar between 2009 and 2010. Mukherjee's research group specializes in semiconductors that emit and detect light in the infrared spectrum. His work focuses on enabling better materials for data transmission, sensing, manufacturing, and environmental monitoring by making high-quality thin films with IV-VI (PbSnSe) and III-V (GaAs-InAs/GaSb) material systems. A significant aspect of his research involves understanding how imperfections in the crystalline structure, such as dislocations and point defects, impact the electronic and optical properties of these materials. This understanding is key to integrating these semiconductors with silicon and germanium substrates for the development of hybrid circuits that combine infrared photonics and conventional electronics. Mukherjee holds a Ph.D. from MIT in Materials Science and Engineering, obtained in 2014, a Master's degree from the National University of Singapore in Advanced Materials for micro- and nano-systems in 2009, a Master of Engineering from MIT in 2008, and a Bachelor's degree in Electrical and Electronics from Nanyang Technological University, Singapore, in 2007.

Research topics

• Optoelectronics
• Materials science
• Physics
• Optics
• Computer Science
• Nanotechnology
• Composite material
• Condensed matter physics
• Engineering physics
• Electronic engineering
• Geography
• Electrical engineering
• Engineering

Selected publications

• Epitaxial Growth of Anisotropic SnSe on GaAs(001) via Step-Edge Orientation Control

  arXiv (Cornell University) · 2026-02-09

  articleOpen accessSenior author

  Epitaxial growth of orthorhombic SnSe on cubic substrates is challenging due to lattice-symmetry mismatch and anisotropic bonding. Here we demonstrate that epitaxial films with sharp interfaces can be achieved for layered SnSe grown directly on on-axis and 4 degree miscut GaAs(001) substrates. The substrate miscut strongly influences the growth morphology, evolving from spirals on on-axis GaAs to a terraced structure on miscut GaAs. X-ray diffraction reveals that on-axis GaAs supports SnSe with two in-plane orientation variants, whereas the miscut substrate stabilizes a single orientation and introduces a small out-of-plane tilt. Accordingly, in-plane optical anisotropy is enhanced in the single variant film compared to the double variant, as determined by cross-polar reflectance. High-resolution TEM shows that the SnSe/GaAs interface is atomically abrupt and incoherent, characteristic of quasi-van der Waals epitaxy. We find a pronounced tendency for the zigzag edges of SnSe to align parallel to step edges on both substrates, and we show that step-skipping nucleation and layer growth on the miscut substrate leads to the additional tilt. These results establish direct SnSe/GaAs heteroepitaxy as a route to integrate anisotropic layered semiconductors with cubic platforms, and show that miscut substrates provide additional control over in-plane anisotropy.

  PublisherOA PDF

• Erratum: Strongly Interacting, Two-Dimensional, Dipolar Spin Ensembles in (111)-Oriented Diamond [Phys. Rev. X <b>15</b> , 021035 (2025)]

  Physical Review X · 2026-02-05

  articleOpen access
  PublisherDOI

• Red-Teaming Claude Opus and ChatGPT-based Security Advisors for Trusted Execution Environments

  arXiv (Cornell University) · 2026-02-23

  preprintOpen access1st authorCorresponding

  Trusted Execution Environments (TEEs) (e.g., Intel SGX and ArmTrustZone) aim to protect sensitive computation from a compromised operating system, yet real deployments remain vulnerable to microarchitectural leakage, side-channel attacks, and fault injection. In parallel, security teams increasingly rely on Large Language Model (LLM) assistants as security advisors for TEE architecture review, mitigation planning, and vulnerability triage. This creates a socio-technical risk surface: assistants may hallucinate TEE mechanisms, overclaim guarantees (e.g., what attestation does and does not establish), or behave unsafely under adversarial prompting. We present a red-teaming study of two prevalently deployed LLM assistants in the role of TEE security advisors: ChatGPT-5.2 and Claude Opus-4.6, focusing on the inherent limitations and transferability of prompt-induced failures across LLMs. We introduce TEE-RedBench, a TEE-grounded evaluation methodology comprising (i) a TEE-specific threat model for LLM-mediated security work, (ii) a structured prompt suite spanning SGX and TrustZone architecture, attestation and key management, threat modeling, and non-operational mitigation guidance, along with policy-bound misuse probes, and (iii) an annotation rubric that jointly measures technical correctness, groundedness, uncertainty calibration, refusal quality, and safe helpfulness. We find that some failures are not purely idiosyncratic, transferring up to 12.02% across LLM assistants, and we connect these outcomes to secure architecture by outlining an "LLM-in-the-loop" evaluation pipeline: policy gating, retrieval grounding, structured templates, and lightweight verification checks that, when combined, reduce failures by 80.62%.

  PublisherDOI

• Solid-phase heteroepitaxy of oriented Sb2Se3 on GaAs for birefringent thin films

  Journal of Vacuum Science & Technology A Vacuum Surfaces and Films · 2026-01-13

  articleSenior author

  We investigate the amorphous-to-crystalline transformation of antimony selenide (Sb2Se3) on UHV-prepared GaAs (001) substrates. In the bulk orthorhombic form, Sb2Se3 is a layered quasi-1D semiconductor with highly anisotropic properties of interest for optical and electronic devices. We find that an amorphous layer deposited by molecular beam epitaxy annealed at or above 230 °C yields a textured-epitaxial structure among some randomly oriented domains. The textured-epitaxial Sb2Se3 grains are oriented with the covalently bonded “1D axis” constrained in-plane to GaAs [110] and with multiple van der Waals (hk0) orientations out-of-plane. The same texture was achieved exclusively without randomly oriented grains using continuous-wave laser radiation, highlighting the use of thermal and optical methods to yield anisotropic crystalline Sb2Se3 films directly from the amorphous phase. Polarized reflectance and polarized microscopy confirm the unique state of in-plane birefringence in the crystallized thin film. Overall, we show that solid-phase heteroepitaxy provides additional pathways to the integration of low-symmetry chalcogenide semiconductors for demanding applications where the inherent anisotropy needs to be preserved.

  PublisherDOI

• Optimal Transport-Guided Adversarial Attacks on Graph Neural Network-Based Bot Detection

  ArXiv.org · 2026-01-30

  articleOpen access1st authorCorresponding

  The rise of bot accounts on social media poses significant risks to public discourse. To address this threat, modern bot detectors increasingly rely on Graph Neural Networks (GNNs). However, the effectiveness of these GNN-based detectors in real-world settings remains poorly understood. In practice, attackers continuously adapt their strategies as well as must operate under domain-specific and temporal constraints, which can fundamentally limit the applicability of existing attack methods. As a result, there is a critical need for robust GNN-based bot detection methods under realistic, constraint-aware attack scenarios. To address this gap, we introduce BOCLOAK to systematically evaluate the robustness of GNN-based social bot detection via both edge editing and node injection adversarial attacks under realistic constraints. BOCLOAK constructs a probability measure over spatio-temporal neighbor features and learns an optimal transport geometry that separates human and bot behaviors. It then decodes transport plans into sparse, plausible edge edits that evade detection while obeying real-world constraints. We evaluate BOCLOAK across three social bot datasets, five state-of-the-art bot detectors, three adversarial defenses, and compare it against four leading graph adversarial attack baselines. BOCLOAK achieves up to 80.13% higher attack success rates while using 99.80% less GPU memory under realistic real-world constraints. Most importantly, BOCLOAK shows that optimal transport provides a lightweight, principled framework for bridging the gap between adversarial attacks and real-world bot detection.

  PublisherOA PDF

• MoltGraph: A Longitudinal Temporal Graph Dataset of Moltbook for Coordinated-Agent Detection

  arXiv (Cornell University) · 2026-02-28

  preprintOpen access1st authorCorresponding

  Agent-native social platforms such as Moltbook are rapidly emerging, yet they inherit and amplify classical influence and abuse attacks, where coordinated agents strategically comment and upvote to manipulate visibility and propagate narratives across communities. However, rigorous measurement and learning-based monitoring remain constrained by the absence of longitudinal, graph-native datasets for agentic social networks that jointly capture heterogeneous interactions, temporal drift, and visibility signals needed to connect coordination behavior to downstream exposure. We introduce MoltGraph as a realistic longitudinal agentic social-network graph dataset for studying how agents behave, coordinate, and evolve in the wild, enabling reproducible measurement on emerging multi-agent social ecosystems. Using MoltGraph, we provide the first graph-centric characterization of Moltbook as a dynamic network: (i) heavy-tailed connectivity with power-law exponents in the range alpha in [1.86, 2.72], (ii) accelerating hub formation and attention centralization where the top 1% agents account for 29.00% of engagements, (iii) bursty, short-lived coordination episodes, 98.33% last under 24 hours, and (iv) measurable exposure effects across submolts. In matched analyses, posts receiving coordinated engagement exhibit 506.35% higher early interaction rates (within H=5 days) and 242.63% higher downstream exposure in feeds than non-coordinated controls.

  PublisherDOI

• Mid‐Infrared LEDs Based on Lattice‐Mismatched Hybrid IV–VI/III–V Heterojunctions

  Advanced Optical Materials · 2026-02-16

  articleOpen accessSenior authorCorresponding

  ABSTRACT Light‐emitting diodes (LEDs) can bridge the gap between narrow linewidth, expensive lasers and broadband, inefficient thermal globars for low‐cost chemical sensing in the mid‐infrared (mid‐IR). However, the efficiency of III–V‐based mid‐IR LEDs at room temperature is low, primarily limited by strong nonradiative Auger‐Meitner recombination that is only partially overcome with complex quantum‐engineered active regions. Here, we exploit the intrinsically low Auger‐Meitner recombination rates of the IV–VI semiconductors PbSe and PbSnSe, while leveraging the mature III–V platform through the fabrication of hybrid heterojunctions that mediate the ∼8% lattice mismatch to GaAs. Electrically injected n‐PbSe/p‐GaAs LEDs emit at 3.8 µm with output powers up to 400 µW under pulsed operation and a peak wall plug efficiency of 0.08% at room temperature, approaching the performance of commercial III–V LEDs at similar wavelengths. Incorporating 7% Sn extends the emission to 5 µm in GeSe/PbSnSe/GaAs LEDs with output powers up to 45 µW. Notably, both devices operate despite threading dislocation densities on the order of 10 9 cm −2 , underscoring the potential of hybrid IV–VI/III–V heterojunction architectures. We show that combining the complementary advantages of IV–VI and III–V semiconductors offers a simple and efficient mid‐IR optoelectronic platform for a rapidly expanding set of applications.

  PublisherOA PDFDOI

• Creation of depth-confined, shallow nitrogen-vacancy centers in diamond with tunable density

  Applied Physics Letters · 2026-05-18

  article

  Engineering shallow nitrogen-vacancy (NV) centers in diamond holds the key to unlocking new advances in nanoscale quantum sensing. We find that the creation of near-surface NVs through delta doping during diamond growth allows for tunable control over both NV depth confinement (with a twofold improvement relative to low-energy ion implantation) and NV density, ultimately resulting in highly sensitive single defects and ensembles with coherence limited by NV–NV interactions. Additionally, we demonstrate the utility of our shallow delta-doped NVs by imaging magnetism in few-layer CrSBr, a two-dimensional magnet. We anticipate that the control afforded by near-surface delta doping will enable new developments in NV quantum sensing from nanoscale nuclear magnetic resonance to entanglement-enhanced metrology.

  PublisherDOI

• Shining light on short-range atomic ordering in semiconductors alloys

  arXiv (Cornell University) · 2026-03-29

  preprintOpen access

  The functional properties of semiconductors are typically controlled by tailoring their chemical composition and their state of strain, and by controlling their long-range structural order, including the presence of extended defects such as dislocations. In addition to these approaches, theoretical predictions suggest that short-range order (SRO) of atoms in group-IV semiconductor alloys can modify the bandgap, a defining property of any semiconductor. Herein, a new machine learning enabled, computation-guided methodology for extended X-ray absorption fine structure (EXAFS) analysis of SRO is used to quantify the effects of local atomic order on the bandgap of germanium-tin (GeSn) alloy single crystal nanostructures with well-controlled strain and composition. Correlative analysis of EXAFS and photoluminescence (PL) establishes the relationship between bandgap and the Warren-Cowley short-range order (WC-SRO) parameter of the GeSn alloys. It is further demonstrated that SRO can be tuned over a broad range by post-deposition annealing of the alloy crystals. This work establishes control of SRO as an important design parameter for semiconducting properties and suggests the potential for quantitative measurement and tuning of SRO in other semiconductor alloy systems.

  PublisherDOI

• Epitaxial growth of anisotropic SnSe on GaAs(001) via step-edge orientation control

  Journal of Applied Physics · 2026-05-05

  articleSenior author

  Epitaxial growth of orthorhombic SnSe on cubic substrates is challenging due to lattice symmetry mismatch and anisotropic bonding. Here, we demonstrate that epitaxial films with sharp interfaces can be achieved for layered SnSe grown directly on both on-axis and 4° miscut GaAs(001) substrates. The substrate miscut strongly influences the growth morphology, evolving from spirals on on-axis GaAs to a terraced structure on miscut GaAs. X-ray diffraction reveals that on-axis GaAs supports SnSe with two in-plane orientation variants, whereas the miscut substrate stabilizes a single orientation and introduces a small out-of-plane tilt. Accordingly, in-plane optical anisotropy is enhanced in the single-variant film compared to the double variant, as determined by cross-polar reflectance. High-resolution TEM shows that the SnSe/GaAs interface is atomically abrupt and incoherent, characteristic of quasi-van der Waals epitaxy. We find a pronounced tendency for the zigzag edges of SnSe to align parallel to step edges on both substrates, and we show that step skipping nucleation and layer growth on the miscut substrate lead to the additional tilt. These results establish direct SnSe/GaAs heteroepitaxy as a route to integrate anisotropic layered semiconductors with cubic platforms and show that miscut substrates provide additional control over in-plane anisotropy.

  PublisherDOI

Recent grants

• CAREER: Mixed-bonded IV-VI semiconductors for hybrid heterostructures

  NSF · $112k · 2020–2020

• CAREER: Mixed-bonded IV-VI semiconductors for hybrid heterostructures

  NSF · $593k · 2020–2025

Frequent coauthors

• Eamonn T. Hughes

  University of California, Santa Barbara

  50 shared

• Brian B. Haidet

  University of California, Santa Barbara

  36 shared

• Leland Nordin

  University of Central Florida

  34 shared

• Jarod Meyer

  33 shared

• Jennifer Selvidge

  31 shared

• Paul J. Simmonds

  28 shared

• Daniel Wasserman

  27 shared

• Aaron J. Muhowski

  27 shared

Education

• Ph.D., Materials Science and Engineering

  Stanford University

  2015

• M.S., Materials Science and Engineering

  Stanford University

  2011

• B.S., Materials Science and Engineering

  University of California, Berkeley

  2009