Fragmentation of Virtual Orbitals for Quantum Computing: Reducing Qubit Requirements through Many-Body Expansion

Federico Zahariev, Mark S. Gordon · 2025

Read Fragmentation of Virtual Orbitals for Quantum Computing: Reducing Qubit Requirements through Many-Body Expansion by Federico Zahariev, Mark S. Gordon (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2510-20950.txt.

Introduction

Federico Zahariev leads two co-authors on cutting qubit needs for quantum chemistry by fragmenting the virtual orbital space. Correlated calculations demand large active spaces, and qubits are the scarcest resource on current quantum hardware.

Their Q-FVO method keeps the full occupied space, splits the localized virtual space into chemical fragments, and recovers correlation energy through a many body expansion with inclusion exclusion counting. Across six molecular benchmarks, one body calculations cut qubit needs by roughly half, while two body versions cut less but recover most correlation energy with small errors. The trade offers a practical dial between hardware cost and accuracy.

Catalog record

CREATORFederico Zahariev, Mark S. Gordon
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE30 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2510-20950

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Fragmentation of Virtual Orbitals for Quantum Computing: Reducing Qubit Requirements through Many-Body Expansion Federico Zahariev, Mark S. Gordon Abstract The development of quantum computing for molecular simulations is constrained by the limited number of qubits available on current Noisy Intermediate-Scale Quantum (NISQ) devices. The present work introduces the Virtual Orbital Fragmentation (FVO) method, a systematic approach that reduces qubit requirements by 40–66% while maintaining chemical accuracy. The method partitions the virtual orbital space into chemically intuitive fragments and employs many-body expansion techniques analogous to spatial fragmentation methods. Applications to six molecular systems demonstrate that the 2-body FVO expansion achieves errors below 3 kcal/mol, while the 3-body expansion provides sub-kcal/mol accuracy. When integrated with the Variational Quantum Eigensolver (VQE) and combined with the Effective Fragment Molecular Orbital (EFMO) method for multi-molecular systems, the hierarchical Q-EFMO-FVO approach achieves 96–100% accuracy relative to full calculations. The method provides a practical pathway for quantum chemical calculations on current 50–100 qubit processors and establishes virtual orbital fragmentation as a complementary strategy to spatial fragmentation for managing quantum computational complexity. Keywords: Quantum computing, Virtual orbitals, Fragmentation methods, VQE, EFMO, NISQ, Many-body expansion 1 Introduction Quantum computing offers transformative potential for molecular simulations by providing polynomial rather than exponential scaling for solving the electronic Schrödinger equation [ 1 , 2 , 3 ] . The Variational Quantum Eigensolver (VQE) has emerged as the leading algorithm for molecular electronic structure calculations on near-term quantum devices [ 4 , 5 ] , offering a hybrid quantum-classical approach particularly suited to current hardware limitations. However, Noisy Intermediate-Scale Quantum (NISQ) devices [ 6 ] remain severely constrained, with typical processors containing 50–100 qubits, significant noise levels, limited gate fidelities, and short coherence times. Such hardware limitations create a substantial gap between the theoretical promise of quantum algorithms and practical implementation for chemical systems of interest. The qubit requirements for molecular calculations present the most immediate bottleneck. Using standard Jordan-Wigner or Bravyi-Kitaev transformations [ 7 , 8 ] , the number of qubits scales as twice the number of spatial molecular orbitals. For even modest molecular systems, such an arrangement creates severe limitations: a single water molecule in cc-pVDZ basis requires 52 qubits, while methanol requires 96 qubits—already exceeding most current quantum processors. Since virtual (unoccupied) orbitals typically comprise 70–90% of the total orbital space, they dominate qubit requirements despite contributing primarily through electron correlation effects. The observation motivates the development of methods to systematically reduce the virtual orbital space while maintaining chemical accuracy. Recent progress in quantum chemistry on NISQ devices has focused on several complementary strategies: active space selection [ 9 , 10 ] , orbital optimization [ 11 , 12 ] , problem decomposition [ 13 , 14 ] , and error mitigation techniques [ 15 , 16 ] . Such approaches have enabled calculations on systems with up to 12 qubits with chemical accuracy. However, scaling to the 50–100 qubit systems needed for molecules of practical interest requires additional algorithmic innovations. Fragmentation methods, which decompose large quantum chemistry problems into smaller subproblems, offer a promising path forward by reducing both qubit counts and circuit depths. 1.1 Background: Fragmentation Methods in Quantum Chemistry Fragmentation methods in quantum chemistry fall into two complementary categories: spatial fragmentation and orbital space fragmentation. Spatial fragmentation methods, such as the Fragment Molecular Orbital (FMO) method [ 17 ] and its derivative, the Effective Fragment Molecular Orbital (EFMO) method [ 18 , 19 , 20 ] , partition large molecular systems into smaller spatial regions (monomers), then recover the total energy through a many-body expansion of monomer, dimer, and potentially trimer contributions. Such methods have demonstrated exceptional accuracy and computational efficiency for systems ranging from small molecular clusters to large biomolecules and materials [ 21 ] . The FMO energy expansion is expressed as E_{\text{FMO}}=\sum_{i}E_{i}+\sum_{i<j}(E_{ij}-E_{i}-E_{j}) , where E_{i} represents the energy of spatial fragment i and E_{ij} represents the dimer energy. EFMO enhances FMO by incorporating the Effective Fragment Potential (EFP) method [ 22 , 23 , 24 ] to describe long-range interactions (Coulomb, polarization, dispersion, exchange-repulsion, and charge transfer), enabling many-body polarization effects without expensive trimer calculations. Recent work combining quantum Monte Carlo (QMC) with EFMO [ 25 , 26 , 27 ] has demonstrated that spatial fragmentation can reduce computational costs while maintaining sub-kcal/mol accuracy for both ground and excited states. For water clusters, spatial fragmentation with cutoff distances of 1.4 van der Waals radii achieves errors below 1%, demonstrating that fragmentation methods based on many-body expansions maintain high accuracy when applied to demanding quantum calculations [ 25 ] . The QMC-EFMO method has been further extended to handle fragmentation across covalent bonds [ 28 ] , enabling application to systems where fragments are connected through chemical bonds rather than only intermolecular interactions. The extension demonstrated that spatial fragmentation maintains accuracy even when cutting through covalent bonds, achieving correlation energy errors below 2 kcal/mol for systems including glycine oligomers, dipeptide formation reactions, silica-based materials, and polyalanine chains. The successful application to both intermolecular and intramolecular fragmentation validates the robustness of the EFMO approach and establishes that many-body expansion fragmentation is a general strategy applicable across different bonding situations. 1.2 Virtual Orbital Fragmentation: A Complementary Approach While spatial fragmentation reduces the number of calculations needed, the method does not reduce qubit requirements per calculation. The Virtual Orbital Fragmentation (FVO) method presented here represents an orthogonal fragmentation strategy: rather than partitioning physical space, the method partitions the virtual orbital space. The approach directly addresses the qubit bottleneck for NISQ devices by reducing the orbital space included in each quantum calculation. The physical justification for FVO rests on three key observations. First, electron correlation effects involving virtual orbitals are predominantly local in character—excitations from occupied orbitals typically involve only nearby virtual orbitals in real space. Second, the virtual space, especially with large basis sets, exhibits considerable redundancy, with multiple virtual orbitals describing similar regions of space and energy scales. Third, virtual orbital fragmentation is complementary to spatial fragmentation, enabling hierarchical decomposition strategies that fragment along multiple orthogonal dimensions of the quantum chemistry problem. The FVO method systematically reduces the virtual space through many-body expansion analogous to spatial fragmentation. Unlike spatial fragmentation, which must carefully manage fragment boundaries cutting through bonds, virtual orbital fragmentation maintains the complete occupied orbital space in all calculations. Such an approach avoids complications with chemical bonding and electron counting while achieving substantial computational savings. All fragment calculations remain embarrassingly parallel, providing excellent computational scalability on both classical and quantum hardware. The results demonstrate that FVO reduces maximum qubit requirements by 40–66% across diverse molecular systems while maintaining chemical accuracy (errors below 1 kcal/mol). For quantum algorithms like VQE, such qubit reduction is transformative, bringing systems which would require 96–128 qubits down to 48–74 qubits—well within the range of current quantum processors. When combined with EFMO for multi-molecular systems, the hierarchical Q-EFMO-FVO approach enables accurate calculations on molecular clusters which would otherwise far exceed current quantum hardware capabilities. Figure 1 provides a schematic overview of the FVO concept. Figure 1: Schematic illustration of the Virtual Orbital Fragmentation (FVO) approach. Left: Full molecular orbital space showing occupied orbitals (OO, blue) and complete set of virtual orbitals (VO 1 –VO N , gray). Center: Fragmentation

Read the complete text