R&D cycles are expensive.
Unreliable computational models make them worse.
We fix that upstream.
Before battery developers, drug discovery teams, and materials scientists commit resources to synthesis and testing, Quantum-Clarity audits whether the electronic model they are building on is stable, ambiguous, or too unreliable to support confident decisions.
completed
validated
per system
per condition
Conventional workflows often return a number. ELSD returns a classification — whether the underlying electronic model is stable enough to trust, sensitive to perturbation, open-shell coherent, or too truncated to support decision-making.
Six domains, one classification engine
The same ensemble VQE diagnostic framework is applied across chemically distinct systems. Each domain has its own named platform and published or in-preparation dataset, at different stages of validation maturity.
Energy storage — NMC811 battery cathodes
Ni-rich NMC811 cathodes degrade fastest near 50% state of charge. ELSD maps the electronic energy landscape under symmetry-breaking perturbation, directly revealing bifurcation, multi-basin structure, and dopant-response mechanisms that are not accessible from a single ground-state picture.
Drug discovery — metalloenzyme active sites
Most computational drug discovery tools optimize on top of target models that were never validated for electronic reliability. Prometheus classifies whether a metal-centered active-site model is stable, coherent, multi-basin, or model-pathologic — before a single ligand screen begins.
Catalysis — nitrogen fixation and FeMoCo
Ensemble VQE diagnostics applied to Fe/Mo catalytic active-site models reveal structurally distinct electronic regimes across the redox series. Results suggest a stabilising electronic role for Mo across the catalytic landscape, supporting a more mechanistic interpretation of cofactor selection.
Rocket propulsion — energetic materials & engine alloys
Solid rocket propellants and liquid-fueled engines depend on materials behaving predictably under extreme thermal, pressure, and combustion conditions. ELSD classifies whether the electronic models underpinning energetic additives and metal–fuel interfaces are reproducible — an upstream triage layer before costly synthesis or engine testing.
CO₂ reduction — iron porphyrin redox ladder
Iron porphyrins are the leading molecular electrocatalysts for converting CO₂ into useful fuels. Despite decades of study, the electronic structure of the catalytically active reduced states has remained contested. ELSD provides the first reproducibility-certified classification of the Fe porphyrin redox ladder from resting state through CO₂ binding.
Green hydrogen — PEM electrolysis & storage materials
Proton exchange membrane electrolysers are a leading pathway for green hydrogen production. Storage materials must release hydrogen reliably under real-world conditions. ELSD classifies whether the electronic models underpinning both branches are reproducible or pathological — providing an upstream quality check before catalyst or material investment.
Four electronic landscape regimes, across all domains
ELSD produces a regime classification, not just an energy value. The same four-class framework applies whether the system is a battery cathode, a metalloenzyme active site, a catalytic cluster, or a strongly correlated material — making results comparable across domains.
DFT is the workhorse of computational materials science — fast, well-understood, and broadly reliable for stable systems. ELSD does not replace it. It adds the one layer DFT lacks: a reproducibility check that asks whether the answer DFT returned is actually trustworthy.
| DFT | ELSD | |
|---|---|---|
| Output | Single converged energy | Regime classification — Rigid Stability, Rugged, Multi-Basin, or Pathological |
| Runs per system | One | 15–35 independent optimizer trajectories |
| Reliability signal | None — convergence is assumed, not tested | Explicit — σ, dominant determinant weight, sector audit |
| Multi-reference systems | Can be unreliable; no built-in self-diagnostic | Classifies whether multi-reference character is structured or pathological |
| Active-space adequacy | Not assessed | Flagged via controlled fragment extension tests |
| Decision readiness | Implicit — assumed from convergence | Explicit — decision-grade classification or flagged constraint |
Perturbation finds nothing to split. Single basin retained across all ensemble seeds. The model is reliable enough to support downstream decision-making — ligand screening, dopant selection, or synthesis.
Multi-reference character present but well-structured. The ensemble converges within a single electronic family. Results are reproducible and the model is usable with appropriate care.
Two or more distinct electronic basins coexist under the same scaffold. The system cannot reliably resolve its electronic state. Results depend on starting conditions and should not be trusted without landscape diagnosis.
The active space or scaffold is too truncated or underconstrained to produce reliable results. Ensemble seeds diverge in ways that reflect model failure rather than physical electronic structure. Not safe to optimize against.
How ELSD is built
Four layers connecting industry problems to validated quantum computation. Each layer is purpose-built: the diagnostic engine, the GPU-native simulation infrastructure, and the QuantaCore™ hardware-validation architecture that underpins the simulation stack.
1 classification engine
~3–4 GPU hrs per target
UCCSD · 20-qubit operating point
Patent pending
Partner with us on electronically difficult targets
We work with battery developers, pharmaceutical R&D teams, catalyst designers, and materials scientists who need reliable electronic landscape classification before committing resources to synthesis, screening, or clinical-stage decisions.
Submit a candidate composition, active-site model, or dopant strategy. We return a full ELS classification report: regime, basin structure, trapped fraction, and design recommendations. No source code shared.
A defined evaluation campaign against your internal targets, with results benchmarked against your existing computational workflows. Designed for R&D teams assessing ELSD fit before a broader deployment decision.
Full platform deployed in your environment, running across your candidate pipeline on your hardware. For organizations screening many compositions or targets on an ongoing basis.
For organizations interested in platform rights, IP integration, or long-term strategic access. Details available under mutual NDA.
Bring us your electronically difficult target.
Whether it is a cathode composition, a metalloenzyme active site, a catalytic cluster, or a strongly correlated material — if conventional workflows are not giving you a reliable picture, ELSD is designed for exactly that problem.
Contact Us
Interested in working together? Fill out some info and we will be in touch shortly. We can’t wait to hear from you!