⚛️ quantum physics

ViBra: Configuration Interaction for Anharmonic Vibrational Spectroscopy and Quantum-Sampled Configuration Spaces

This paper introduces ViBra, a hybrid quantum-classical workflow that integrates quantum sampling algorithms with Vibrational Configuration Interaction (VCI) methods to enable accurate anharmonic vibrational spectroscopy predictions, demonstrated through a proof-of-concept application.

Original authors: Raphael F. Ligorio, Marco Antonio Barroca, Alan Duriez, Mathias B. Steiner

Published 2026-07-28

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Raphael F. Ligorio, Marco Antonio Barroca, Alan Duriez, Mathias B. Steiner

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are trying to understand a complex machine, like a grand piano, not by looking at its wood and strings, but by listening to the notes it plays. In the world of chemistry, molecules are that piano. They vibrate, wiggle, and hum in specific patterns. Scientists call this "vibrational spectroscopy." By listening to these hums, we can figure out how strong the bonds are between atoms, how they twist and turn, and even how they react with other molecules. It's like being a detective who solves a crime just by hearing the footsteps of the suspect.

However, predicting these sounds is incredibly tricky. If you treat a molecule like a simple spring, the math is easy, but the prediction is wrong. Real molecules are messy; they have "anharmonicity," which means their springs get stiffer or looser as they stretch, and they talk to each other in complicated ways. To get the right answer, scientists have to solve a massive puzzle involving millions of possible ways the molecule can vibrate. For a long time, computers have been the only ones smart enough to try, but the puzzle gets so big that even the fastest supercomputers start to sweat. This is where the story of a new tool called ViBra begins.

The Problem: A Symphony of Too Many Notes

Think of a molecule's vibration as a song. The simplest version of this song is a perfect, mathematical melody (the "harmonic" approximation). But real life is full of jazz—improvisation, unexpected chords, and notes that bleed into one another. To capture this jazz, scientists use a method called Vibrational Configuration Interaction (VCI). Imagine VCI as a choir where every possible combination of notes is sung at once. The more notes you add to the choir, the more accurate the song becomes. But here's the catch: the number of choir members grows explosively. For a molecule with just a few atoms, the choir could have billions of singers. Trying to listen to all of them at once is like trying to drink from a firehose; it's too much data for any computer to handle efficiently.

The Solution: ViBra, the Smart Conductor

Enter ViBra (Vibrational Configuration Interaction for Anharmonic Vibrational Spectroscopy), a new software package created by researchers Raphael F. Ligorio, Marco A. Barroca, Alan Duriez, and Mathias B. Steiner. If the VCI choir is a chaotic crowd of billions, ViBra is the genius conductor who knows exactly which singers matter and which ones can go home for a coffee break.

ViBra does three main things to make this impossible task possible:

The "Selected" Choir (S-VCI): Instead of asking every single singer to perform, ViBra uses a clever screening trick called Epstein–Nesbet perturbation theory. It's like a talent scout who quickly listens to a few notes and decides, "You, you, and you are essential; the rest of you can wait." This allows the software to cut the size of the choir down to a manageable 32% of its original size (in some tests) without losing the quality of the song. The result? It finds the right notes with high accuracy but uses a fraction of the computer power.
The Symmetry Shortcut (SA-VCI): Molecules often have symmetry, like a snowflake or a perfect cube. ViBra realizes that if a molecule looks the same from different angles, the math for those angles is identical. It breaks the giant choir into smaller, separate groups based on these symmetries. Instead of one massive, scary math problem, it solves eight tiny, easy ones. In tests with the ethylene molecule, this trick made the calculation run 20 times faster than before, while still getting the exact same answer.
The "Quantum" Bridge: Perhaps the most exciting part of ViBra is its ability to talk to future technology. The researchers showed that ViBra can work in a "hybrid" mode. Imagine a futuristic quantum computer (which is currently very noisy and small) acting as a scout. It can quickly pick out a few promising singers from the billions. Then, ViBra takes that small list, organizes it, and does the heavy lifting to finish the song. In a simulation using a water molecule, ViBra successfully took a list of configurations generated by a simulated quantum algorithm and refined them to match the accuracy of a full, super-computer calculation.

What They Found and What It Means

The team didn't just build the tool; they tested it rigorously. They ran ViBra on water and ethylene molecules, comparing its results against other trusted methods and experimental data.

Accuracy: For water, ViBra matched reference data within 0.01 cm⁻¹ (a tiny unit of energy), proving it's mathematically correct.
Speed: For ethylene, using the symmetry shortcut reduced the time to solve the problem from over 4 minutes to just 13 seconds.
The Quantum Test: In their "proof-of-concept" with water, they showed that even if the quantum "scout" only found a small fraction of the necessary notes (sometimes as low as 11% of the total), ViBra could use its "Selected" method to fill in the gaps and get the energy levels right within 1 cm⁻¹ of the perfect answer.

The Verdict

ViBra is not a magic wand that solves everything instantly, nor does it claim to have beaten quantum computers yet. The quantum part of their work was a simulation, not a run on real quantum hardware. However, the paper demonstrates that ViBra is a robust, user-friendly tool that makes high-accuracy vibrational spectroscopy accessible for larger molecules. It bridges the gap between the messy reality of molecular vibrations and the limits of current computing power.

Most importantly, ViBra is designed to be the "classical brain" for future "quantum scouts." As quantum computers grow stronger, they will likely be used to pick the best starting points for calculations, and ViBra is ready to be the engine that turns those seeds into a full, accurate picture of how molecules sing. For now, it's a powerful new way for chemists to hear the music of the molecular world, loud and clear.

Technical Summary of ViBra: Configuration Interaction for Anharmonic Vibrational Spectroscopy and Quantum-Sampled Configuration Spaces

Problem Statement
Accurate prediction of anharmonic vibrational spectra is essential for interpreting experimental data and understanding molecular dynamics, yet it faces significant computational challenges. While perturbative methods like VPT2 are efficient, they fail in the presence of resonances. Variational methods, specifically Vibrational Configuration Interaction (VCI), offer systematic improvability and robustness but suffer from exponential scaling with the number of vibrational modes ( $M$ ) and the maximum number of excitation quanta ( $N_q$ ). Furthermore, existing software often lacks user-friendly interfaces, comprehensive treatment of electrical anharmonicity (essential for overtone intensities), and integration with emerging quantum-centric workflows. There is a specific gap in demonstrating computational workflows that integrate quantum sampling algorithms with classical solvers for vibrational property predictions.

Methodology
The authors introduce ViBra, a hybrid quantum-classical computational package designed to perform anharmonic vibrational structure calculations. The methodology is built upon a quartic force field (QFF) representation of the potential energy surface (PES), truncated at fourth order in normal coordinates.

VSCF and VCI Framework:
- The workflow begins with a Vibrational Self-Consistent Field (VSCF) calculation to optimize one-mode functions (modals) in the average potential of all other modes.
- These modals serve as the basis for a Vibrational Configuration Interaction (VCI) expansion. The wavefunction is expressed as a linear combination of Hartree products of these modals.
- The Hamiltonian matrix elements are constructed using precomputed modal integrals, leveraging the sparsity of the quartic potential (coupling only configurations differing by at most four modes).
Selected VCI (S-VCI):
- To mitigate the exponential scaling of full VCI, ViBra implements a Selected CI approach driven by Epstein–Nesbet second-order perturbation theory (EN-PT2).
- A reference space ( $S_{ref}$ ) is defined (either by excitation order or a user-provided list).
- External configurations are screened based on their perturbative energy contribution ( $e^{(I)}_\alpha$ ). The top- $N$ most important configurations are retained to form an active space ( $S_{active}$ ), significantly reducing the Hamiltonian dimension while maintaining high accuracy for low-lying states.
Symmetry-Adapted VCI (SA-VCI):
- For molecules in Abelian point groups ( $C_1, C_s, C_i, C_2, C_{2h}, C_{2v}, D_2, D_{2h}$ ), the code performs exact block-diagonalization of the Hamiltonian.
- Normal modes are assigned irreducible representations (irreps), and the configuration space is partitioned into symmetry blocks. This reduces the diagonalization cost from $O(|S|^3)$ to $\sum O(|B_\Gamma|^3)$ .
Transition Intensities:
- The package calculates infrared intensities by expanding the dipole operator to second order in normal coordinates. This includes both linear and quadratic dipole derivatives, enabling the accurate prediction of overtone and combination band intensities, which are zero in the double-harmonic approximation.
Quantum-Centric Integration:
- ViBra is designed to accept externally generated "seed" configuration spaces. In a proposed hybrid workflow, a quantum sampling algorithm (e.g., Sample-based Quantum Diagonalization) generates a compact set of vibrational configurations.
- ViBra then acts as the classical engine: it can either diagonalize the Hamiltonian directly in this quantum-sampled space or use the EN-PT2 screening to enlarge the space before variational rediagonalization.

Key Contributions

ViBra Software Suite: A complete toolchain comprising a Fortran 90/95 computational engine and a Python-based Graphical User Interface (GUI) for input preparation, job execution, and interactive spectral visualization (including temperature dependence, baseline correction, and 3D normal-mode animations).
Algorithmic Optimizations: Implementation of sparse pair lists, inverted potential indices, and overlap caching to handle large configuration spaces efficiently.
Hybrid Workflow Demonstration: A proof-of-concept demonstration where a quantum-sampled seed space (generated via simulated Trotterized time evolution) is processed by ViBra to recover accurate vibrational energies and wavefunction compositions.
Comprehensive Benchmarking: A four-tier validation process confirming numerical correctness against reference data (Crystal23) and internal consistency across VSCF, Full VCI, SA-VCI, and S-VCI modules.

Results

Validation: The VSCF and Full VCI modules reproduce reference data for water within $0.01$ cm $^{-1}$ and $0.001$ cm $^{-1}$ , respectively. SA-VCI reproduces Full VCI results to numerical precision for various point groups. S-VCI converges variationally, achieving sub-5 cm $^{-1}$ accuracy with significantly reduced configuration spaces.
Performance: For ethylene ( $D_{2h}$ symmetry) with 6 quanta, SA-VCI achieved a $\sim 20\times$ speedup over Full VCI (13 s vs. 265 s) by exploiting block diagonalization. S-VCI with a moderate selection threshold ($MAXSCI=50$) reduced the active space to $\sim 17\%$ of the full dimension while retaining energies within $1.6$ cm $^{-1}$ of the Full VCI limit.
Quantum-Centric Proof-of-Concept: In the water molecule test, a quantum-sampled seed space (containing $\sim 11\%$ of the full space at $N_q=12$ ) yielded ground-state energies with $\sim 0.1$ cm $^{-1}$ error. However, direct diagonalization of the seed failed to reproduce correct wavefunction compositions for higher excited states. When ViBra enlarged this seed using EN-PT2 selection, the errors dropped below $1$ cm $^{-1}$ , and the correct wavefunction mixing was recovered, demonstrating the necessity of the classical refinement step.

Significance and Claims
The paper claims that ViBra provides a robust, user-friendly, and accurate platform for anharmonic vibrational spectroscopy that bridges the gap between high-accuracy variational methods and practical computational costs. Its significance lies in:

Accessibility: Making advanced VCI methods accessible through a graphical interface and open-source availability.
Accuracy: Providing reliable intensities through the inclusion of second-order dipole derivatives and handling resonances via variational methods.
Future-Readiness: Establishing a "natural interface" for quantum-centric workflows. The authors explicitly state that their H $_2$ O demonstration is a proof-of-concept for the interface between quantum sampling and classical post-processing, not a benchmark of quantum hardware or a claim of quantum advantage. The work serves to motivate the development of future quantum-classical workflows for vibrational problems, positioning ViBra as the classical refinement engine that can process quantum-generated seeds to yield physically meaningful spectra.

The Problem: A Symphony of Too Many Notes

The Solution: ViBra, the Smart Conductor

What They Found and What It Means

The Verdict

More like this