Applications of the Quantum Phase Difference Estimation Algorithm to the Excitation Energies in Spin Systems on a NISQ Device

This paper demonstrates the practical viability of a novel, constant-depth Quantum Phase Difference Estimation (QPDE) algorithm for accurately calculating excitation energy gaps in diverse spin systems on noisy intermediate-scale quantum (NISQ) devices, achieving 85% to 93% accuracy through advanced noise suppression techniques.

The Gist

The Big Picture: Finding the "Price Tag" of Quantum Energy

Imagine you are trying to figure out the price difference between two rare collectibles. In the world of quantum physics, these "collectibles" are energy states of atoms or molecules. Scientists often need to know the energy gap (the difference in energy) between a system's calm, resting state (ground state) and a more excited, energetic state.

Usually, calculating this is like trying to weigh a feather on a scale that is shaking violently. The current tools (quantum computers) are powerful but "noisy"—they make mistakes easily, like a radio with static.

This paper introduces a new, smarter way to measure that energy gap using a noisy quantum computer. They call it the QPDE (Quantum Phase Difference Estimation) algorithm. Think of it as a "noise-canceling headphone" for quantum math.

---

The Problem: The Old Way Was Too Heavy

Previously, scientists used a method called QPE (Quantum Phase Estimation).

• The Analogy: Imagine trying to measure the height of a building by stacking thousands of heavy bricks one by one. If you drop one brick (a computer error), the whole tower falls.
• The Issue: This method requires very deep, complex circuits (too many steps). On today's "Noisy Intermediate-Scale Quantum" (NISQ) devices, the computer gets tired and makes mistakes long before it finishes the stack.

The Solution: The "Twin" Trick (QPDE)

The authors propose a new method, QPDE, which is lighter and faster.

• The Analogy: Instead of stacking bricks, imagine you have two identical twins. You put them on a seesaw. You don't need to know exactly how heavy each twin is individually; you just need to know the difference in their weight.
• How it works: The algorithm creates a "superposition" (a quantum mix) of the ground state and the excited state simultaneously. It lets them "dance" together. By watching how their dance steps get out of sync over time, the computer can calculate the energy difference directly.
• The Benefit: This method skips the heavy "controlled" operations that cause errors. It's like using a feather-light balance beam instead of a heavy crane.

The Magic Ingredient: The "Match Gate"

The researchers focused on Spin Systems (tiny magnets acting like little compass needles). They realized that the math describing how these magnets move (the Heisenberg Hamiltonian) has a special secret structure.

• The Analogy: Imagine a complex dance routine. Usually, adding more dancers (more time steps) makes the choreography infinitely harder. But in this specific dance, the steps are so repetitive and symmetrical that no matter how long the music plays, the dance routine stays the same length.
• The Result: They found a way to compress the quantum circuit so that even if they simulate the system for a long time, the computer doesn't get overwhelmed. The "circuit depth" (the number of steps) stays constant. This is a huge deal for current noisy hardware.

The Experiment: Testing on Real Hardware

The team took this algorithm to IBM's quantum computers (real, physical machines that currently suffer from noise). They tested it on different arrangements of "spins" (magnets):

1. Two spins: A simple pair.
2. Three spins in a line: A chain.
3. Three spins in a triangle: Some were "frustrated" (like a triangle of friends where everyone wants to be friends with everyone else, but physics says they can't, creating tension).

The Results:
Despite the "static" and errors in the hardware, their method was incredibly accurate.

• They achieved 85% to 93% accuracy.
• The Metaphor: It's like trying to hit a bullseye on a target while riding a bumpy horse. Most people would miss, but this team managed to hit the bullseye almost every time by using a special saddle (noise suppression) and a steady hand (the algorithm).

The Secret Sauce: Noise Suppression

To get these high scores, they didn't just rely on the algorithm; they also used "noise suppression" techniques.

• Pauli Twirling & Dynamical Decoupling: Think of these as "shaking the dice" or "tapping the table" to reset the computer's memory before it gets confused by the noise. It's like a drummer tapping a rhythm to keep the band from drifting out of time.

Why Does This Matter?

1. Real-World Chemistry: In real life, we can't always measure the total energy of a molecule easily. But we can measure the energy difference (like the color of light an atom emits). This algorithm is built specifically to find those differences.
2. Future-Proofing: It proves that we don't need to wait for perfect, error-free quantum computers to do useful science. We can get great results right now on imperfect machines if we use the right tricks.

Summary

The paper shows that by using a clever "twin" algorithm (QPDE) and exploiting the symmetrical nature of magnetic spins, scientists can calculate energy gaps on today's noisy quantum computers with high accuracy. It's a major step toward using quantum computers to discover new drugs and materials, even before the technology is perfect.

Technical Summary

1. Problem Statement

Quantum computing holds promise for solving quantum many-body problems that are intractable for classical computers, particularly in calculating energy eigenvalues and gaps. However, current hardware is limited to the Noisy Intermediate-Scale Quantum (NISQ) era, characterized by noise, limited qubit connectivity, and short coherence times.

• Limitations of Existing Algorithms:
  • Variational Quantum Eigensolver (VQE): While suitable for NISQ, it suffers from "barren plateaus" (optimization difficulties), high shot counts for statistical accuracy, and reliance on classical optimizers.
  • Quantum Phase Estimation (QPE): Offers high accuracy but requires deep, complex circuits with controlled-unitary operations. These are resource-intensive and prone to errors on current hardware with limited connectivity and crosstalk.
• Specific Challenge: Many experimental observables (e.g., spectroscopy) measure energy differences (excitation energies) rather than total energies. Existing methods often require multiple separate executions to calculate these differences, amplifying computational overhead and error accumulation.

2. Methodology

The authors propose and implement the Quantum Phase Difference Estimation (QPDE) algorithm, an extension of QPE designed specifically for NISQ devices.

A. The QPDE Algorithm

• Core Concept: Instead of estimating the absolute phase of a single eigenstate, QPDE computes the difference between two eigenvalues ($\Delta E$) of a unitary operator directly.
• Mechanism:
  1. Superposition: It creates a superposition of an approximate ground state ($|\Phi_0\rangle$) and an excited state ($|\Phi_1\rangle$) using a controlled-Excit gate (which does not require controlled-unitary operations).
  2. Time Evolution: A control-free time evolution operator ($e^{-iHt}$) is applied.
  3. Phase Shift & Measurement: A phase rotation gate $T_z(\Delta\epsilon t)$ is applied, followed by a Hadamard gate and measurement of an ancilla qubit.
  4. Probability Function: The probability of measuring $|0\rangle$ is given by:
     $$P(0) = \frac{1}{2} \left( 1 + \sum_{j,k} |c_j|^2 |d_k|^2 \cos \{(\Delta E_{jk} - \Delta\epsilon)t\} \right)$$
     The energy gap $\Delta E$ is found by identifying the phase shift $\Delta\epsilon$ that maximizes $P(0)$.
• Adaptive Framework: The algorithm uses an iterative process with Gaussian fitting to refine the estimate of the energy gap, narrowing the search range until convergence.

B. System Model: Heisenberg Spin Systems

The study focuses on 2-spin and 3-spin systems governed by the Heisenberg Hamiltonian:
$$H_H = -2 \sum_{\langle i,j \rangle} J_{ij} S_i \cdot S_j$$
Configurations tested include:

• Symmetric (Linear, Triangular).
• Asymmetric (Linear, Triangular with varying coupling strengths).
• Frustrated vs. Non-frustrated geometries.

C. Circuit Optimization for NISQ

A critical innovation is the handling of the time evolution operator ($e^{-iHt}$):

• Match-Gate Structure: The authors exploit the specific structure of the Heisenberg Hamiltonian's time evolution, which resembles match gates.
• Constant Depth: Through multi-step pipeline optimization (Qiskit $\to$ Pytket $\to$ Qiskit), they achieve constant-depth quantum circuits. This means the circuit depth and two-qubit gate count remain invariant regardless of the number of Trotter steps or evolution time $t$. This is crucial for running long-time simulations on noisy hardware.
• Noise Suppression: Techniques such as Pauli Twirling and Dynamical Decoupling are employed to mitigate hardware noise.

3. Key Contributions

1. First NISQ Implementation of QPDE on Spin Systems: The paper demonstrates the first practical application of QPDE to determine excitation energies in spin systems on real IBM Quantum processors (Kyoto, Sherbrooke, Kyiv).
2. Control-Free Architecture: By eliminating controlled-unitary operations, the algorithm reduces circuit complexity and crosstalk sensitivity, making it highly suitable for devices with limited qubit connectivity.
3. Constant-Depth Circuits: The derivation of constant-depth circuits for Heisenberg time evolution allows for accurate simulations even with high Trotter steps, overcoming the typical depth limitations of NISQ devices.
4. Adaptive Convergence Strategy: The implementation of an adaptive Gaussian fitting framework ensures reliable convergence to the correct energy gap even in the presence of significant hardware noise.

4. Results

The algorithm was tested on various configurations using IBM's 127-qubit Eagle processors.

• Accuracy: The results showed remarkable agreement with classical exact diagonalization, achieving accuracy rates between 85% and 93%.
  • Two-Spin System: Achieved 87% accuracy (Result: $1.74 \pm 0.28$ vs. Exact: $2.0$).
  • Three-Spin Linear Chain: Achieved 88% accuracy (Result: $0.88 \pm 0.32$ vs. Exact: $1.0$).
  • Three-Spin Frustrated Triangle: Achieved 88% accuracy (Result: $2.64 \pm 0.34$ vs. Exact: $3.0$).
  • Three-Spin Non-Frustrated Triangle: Achieved up to 93.4% accuracy (Result: $4.67 \pm 0.31$ vs. Exact: $5.0$).
  • Asymmetric Chain: Achieved high accuracy ($3.04 \pm 0.18$ vs. Exact: $3.15$) even when spin eigenfunctions were not simultaneous eigenfunctions of the Hamiltonian.
• Robustness: The method successfully handled geometric frustration and asymmetric coupling strengths. The use of noise suppression techniques was critical in maintaining these high accuracy levels despite hardware imperfections.
• Circuit Efficiency: Optimization reduced circuit depths significantly (e.g., a 3-spin chain with 620 Trotter steps was reduced from a depth of ~7448 to a constant depth of ~54-184 depending on the architecture mapping).

5. Significance

• Practical Viability: This work establishes that QPDE is a viable, robust framework for quantum many-body simulations on current NISQ hardware, bridging the gap between theoretical algorithms and experimental reality.
• Efficiency for Energy Gaps: By directly computing energy differences without needing separate total energy calculations, QPDE offers a more resource-efficient path to obtaining spectroscopic data, which is directly relevant to chemistry and materials science.
• Scalability: The constant-depth circuit property suggests that this approach can scale to larger systems and longer evolution times without the exponential error accumulation typical of standard Trotterization on NISQ devices.
• Future Outlook: The study paves the way for more complex quantum simulations, including larger spin lattices and eventually fault-tolerant quantum computing applications, by providing a blueprint for noise-resilient, control-free quantum algorithms.