Efficient Excited-State Calculations for Molecules Based on Contextual Subspace Method and Symmetry Optimizations
Qianjun Yao, He Li · 2025
Read Efficient Excited-State Calculations for Molecules Based on Contextual Subspace Method and Symmetry Optimizations by Qianjun Yao, He Li (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-2502-17932.txt.
Introduction
Qianjun Yao and He Li propose, in a 2025 paper, a leaner way to compute molecules' excited states on today's small quantum computers. Excited states drive photochemistry and materials behavior, yet quantum methods for them lag behind ground-state work because qubits are scarce.
Their framework shrinks the problem with a contextual subspace method, then peels off excited states one by one with a deflation algorithm. A spin-respecting circuit design exploits spin symmetry inside the shrunken space, trimming resources further. Against a common circuit design, theirs needed up to three times fewer optimization rounds at the same circuit depth, easing near-term quantum chemistry.
Catalog record
| CREATOR | Qianjun Yao, He Li |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 27 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2502-17932
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Efficient Excited-State Calculations for Molecules Based on Contextual Subspace Method and Symmetry Optimizations Qianjun Yao, He Li Abstract Quantum computing methods for excited-state calculations remain underexplored in Noisy Intermediate-Scale Quantum (NISQ) hardware, despite their critical role in photochemistry and material science. Herein, we propose a resource-efficient framework that integrates the contextual subspace (CS) method with the Variational Quantum Deflation (VQD) algorithm to enable systematic excited-state calculations for molecules while reducing qubit requirements. On the basis of the numerical results, we find that it is unproblematic to utilize this combination in calculating the excited state to reduce qubits. Furthermore, we demonstrate that the implementation of a spin-conserving hardware-efficient ansatz, namely the 𝒩 ( θ x , θ y , θ z ) \mathcal{N}(\theta_{x},\theta_{y},\theta_{z}) block ansatz, allows exploitation of spin symmetry within the projected subspace, thereby achieving further reductions in computational resource demands. Compared to the commonly used R y R z R_{y}R_{z} ansatz, using the 𝒩 ( θ x , θ y , θ z ) \mathcal{N}(\theta_{x},\theta_{y},\theta_{z}) ansatz can reduce the number of optimization iterations by up to 3 times at a similar circuit depth. Preprint: APS/123-QED I Introduction Quantum chemistry is a promising area where NISQ [ 1 ] computers can first show their potential power and may also benefit a lot from the more interesting fault-tolerant quantum computing era due to resource-required algorithms such as quantum phase estimation [ 2 ] . The emergence of the Willow chip [ 3 ] is undoubtedly a significant milestone in moving toward the next quantum stage. By being able to handle more complex electronic structures and larger systems, it will surely provide a greater impetus to the fields of quantum chemistry [ 4 ] , materials [ 5 ] , drug development [ 6 ] , and biology [ 7 ] . Although being largely underexplored in the context of resource-constrained quantum hardware, excited state plays a vital role in the realm of quantum chemistry, specifically in chemical reactions and photochemical processes. By learning the energy distribution and transition mechanisms between the energy eigenstates, one can form the basis for chemical kinetics and spectroscopy. The properties of the excited state can lead to applications related to luminescence, novel functional materials, and so on. Therefore, in-depth studies of excited state can provide a theoretical foundation for quantum chemistry and material science [ 8 ] . Recent years have witnessed active debate in the quantum computing community regarding how to leverage existing quantum devices to achieve practical utility. Reducing the resources for the problems that we are interested in, e.g., tapering qubits [ 9 ] using Z_{2} symmetry and point group symmetry [ 10 ] , is one way to help from the original point. Although techniques like qubit tapering and contextual subspace methods [ 11 , 12 ] have successfully reduced qubit counts for ground-state calculations, their extension to excited states remains nascent. Herein, we address this gap by integrating the contextual subspace framework with the VQD [ 13 ] method (namely CS-VQD), enabling systematic excited-state calculations with reduced resource demands. Variational Quantum Eigensolver (VQE) [ 14 , 15 , 16 ] is an algorithm dedicated to the NISQ era which faces a critical challenge in designing the ansatz, a parameterized quantum circuit. The real quantum advantage of such an algorithm is the implementation of the state in a high dimension vector space, which is classically hard without the resource of superposition and entanglement properties. The tradeoff between expressibility and accessibility of the ground state often makes it a struggle to choose a hardware-efficient or a problem-inspired one. After showing the performance of CS-VQD using a problem-inspired ansatz, we also demonstrate the feasibility of exploiting the spin symmetry after the Hamiltonian is projected into the contextual subspace using a spin-preserved ansatz. Numerical simulations demonstrate that the use of the spin-preserved ansatz achieves a significant reduction in required optimization iterations compared to the conventional R_{y}R_{z} ansatz [ 17 ] . II Contextual subspace for excited state calculation We first give a brief introduction to the Contextual Subspace Variational Quantum Eigensolver (CS-VQE) [ 12 , 18 , 19 ] based on some original papers. For a more detailed and rigorous explanation, see the systematic establishment by Weaving et al. [ 20 ] . After giving the foundation knowledge, we show the performance of combining the contextual subspace method with the variational quantum deflation method to calculate the excited state. From a stabilizer perspective, the contextual subspace projection has some kind of similarity to the qubit-tapering technique [ 9 ] . It is shown that qubit tapering can be integrated by a contextual subspace projection. Thus we also present our results with the qubit tapering routine in Sec. III . The concept of quantum contextuality originates from the Bell-Kochen-Specker [ 21 , 22 ] theorem. Kirby and Love first built the condition test for an arbitrary Hamiltonian to determine noncontextuality. Contextuality subspace VQE aims at solving the ground state of a qubit Hamiltonian. This qubit Hamiltonian is composed of Pauli strings which can be defined as, \displaystyle P_{N}:=\otimes_{i=0}^{N-1}\sigma_{p}^{(i)},\,\sigma_{p}\in\{\sigma_{I},\,\sigma_{x},\,\sigma_{y},\,\sigma_{z}\}, (1) where N represents the total qubit number and \sigma_{p}^{(i)} means \sigma_{p} acting on i^{th} qubit. CS-VQE can reduce the number of qubits in a VQE routine but is different from the tapering qubits technique, which exploits the real physical symmetry in the full Hamiltonian. The contextual subspace method has been used to find a molecule’s electronic ground-state energy through a quantum-classical hybrid algorithm, VQE, and thus has been called CS-VQE. This CS-VQE method divides the electronic Hamiltonian into two disjoint parts, which are named the contextual part and the noncontextual parts. The latter one can be effectively solved using a classical computer which indicates a rough result for the original problem. To gain an accurate enough result we would need to solve the contextual part as a quantum correction. In VQE, the Hamiltonian can be represented as a sum of weighted Pauli strings after fermion-qubit transformation and as described above can then be turned into two components, \displaystyle H_{\text{qubit}}=H_{\text{c}}+H_{\text{nc}}=\sum_{p_{1}}h_{p_{\text{c}}}P_{\text{c}}+\sum_{p_{2}}h_{p_{\text{nc}}}P_{\text{nc}}, (2) where p_{1},\,p_{2} are the number of Pauli terms in the contexual part and noncontextual part, respectively. h_{p_{\text{c}}} and h_{p_{\text{nc}}} are real coefficient to guarantee H_{\text{qubit}} is hermitian. The set of Pauli terms in H_{\text{nc}} is required to be closed under inference within all the Pauli terms in H_{\text{qubit}} . By solving the non-contextual part H_{\text{nc}} , we can first obtain a noncontextual ground state energy E_{\text{nc}}^{\text{g}} by minimizing a classical objective function \eta(\mathbf{\nu},\,\mathbf{r}) . The former variable is an assignment vector over some selected qubits and can only be valued \pm 1 over these sites. The latter variable \mathbf{r} is also a vector but is restricted to unit length. This unit vector relates to an constructed observable R(\mathbf{r}):=\sum_{i=1}^{M}r_{i}C_{i} , where C_{i} is a representative operator in all the equivalence classes of commutation divided by the symmetry \mathcal{S}_{\text{nc}} of H_{\text{nc}} . A symmetry of a H_{\text{nc}} means that \displaystyle S:=\{P\in\{P_{\text{nc}}\,|\,[P,\,P_{\text{nc}}]=0\}\}. (3) Additionally, we can extend this set to a symmetry group \bar{\mathcal{S}} . Minimizing \eta(\mathbf{\nu},\,\mathbf{r}) to obtain a ground state turns out to be a constrained optimization problem that is classically tractable. The E_{\text{nc}}^{\text{g}} is shown to be close to the Hartree-Fock energy, demonstrated by