Simulated Non-Abelian Statistics of Majorana Zero Modes from a Kitaev Lattice
Foster Sabatino 1,2, Matthew Brooks 3, Charles Tahan 3, Silas Hoffman 1,4,5,6 · 2025
Read Simulated Non-Abelian Statistics of Majorana Zero Modes from a Kitaev Lattice by Foster Sabatino 1,2, Matthew Brooks 3, Charles Tahan 3, Silas Hoffman 1,4,5,6 (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-2503-15405.txt.
Introduction
Foster Sabatino, Matthew Brooks, Charles Tahan, and Silas Hoffman present a 2025 paper on simulating exotic particle statistics on a quantum computer. Majorana zero modes are prized because swapping them stores information in a way that resists noise. Usually they sit at the ends of quantum wires, which makes them awkward to represent.
A Kitaev lattice model gives a local picture instead. The team spells out a braiding protocol for two and four modes as a spin Hamiltonian. Projecting onto a chosen subspace reveals the non-Abelian phase the braid should build. A gate set mimicking that phase ran on a real quantum computer. Braiding fidelity reached about eighty-five percent for two modes and forty-seven percent for four.
Catalog record
| CREATOR | Foster Sabatino 1,2, Matthew Brooks 3, Charles Tahan 3, Silas Hoffman 1,4,5,6 |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 14 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-15405
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Simulated Non-Abelian Statistics of Majorana Zero Modes from a Kitaev Lattice Foster Sabatino 1,2, Matthew Brooks 3, Charles Tahan 3, Silas Hoffman 1,4,5,6 Abstract We simulate the non-Abelian exchange of Majorana zero modes (MZMs) on a quantum computer. Rather than utilizing MZMs at the boundaries of quantum Ising chains, which are typically represented as nonlocal operators on a quantum computer, using a Kitaev lattice allows us to exploit a local representation of MZMs. We detail the protocol for braiding two and four MZMs in terms of a spin Hamiltonian, i.e. physical qubit Hamiltonian. Projecting this onto a subspace of states, we extract an effective Hamiltonian which drives a non-Abelian Berry’s phase. Using several approximations, we construct a set of gates which mimics this accumulation of non-Abelian phase and process this construction on a quantum computer. For two and four MZMs, we realize braiding fidelities of approximately 85% and 47%, respectively. I Introduction The many-body wavefunction of electrons acquires an overall minus sign upon exchanging any two constituent electrons. In contrast, anyons are so-named because an analogous many-body wavefunction can acquire a fixed root of negative one upon exchange nayakRMP08 . Moreover, some anyons, known as non-Abelian anyons, are equal to their original many-body wavefunction by a matrix multiplication. One example of non-Abelian anyons are Majorana zero modes (MZMs). These particles were predicted to exist at the ends of one-dimensional topological superconductors kitaevPU01 . While the natural occurrence of such materials is unknown, several proposals have shown that such topological superconductors could be engineered using conventional materials lutchynPRL10 ; oregPRL10 ; aliceaPRB10 . While many experiments have shown signatures consistent with the presence of MZMs nadj-pergeSCI14 ; agheePRB23 ; agheeNAT25 , no definitive proof exists. Measurement of the non-Abelian statistics of MZMs would provide compelling evidence of their existence. This is, however, in general experimentally difficult. In a completely different setup, MZMs are known to be emulated in quantum spin chains liebAoP61 . Taking advantage of the Jordan-Wigner mapping, one can show that MZMs are supported in spin-1/2 quantum Ising chains. However, owing to the nonlocal nature of the mapping, the MZMs are similarly nonlocal objects when mapped back onto the spin degrees of freedom. Nonetheless, some of the properties for which MZMs are well-known, e.g. perfect Andreev reflection hoffmanPRB23 , fractional Josephson junctions, and braiding, are present in these spin-emulated MZMs backensPRB17 ; stengerPRR21 . In this work, we emulate Majorana fermions (MFs) inspired by the spin-1/2 Kitaev lattice kitaev2006anyons . In contrast to the Ising chain, within this representation of MFs, the interactions between any pair of MFs can be emulated by local two-spin interactions. Consequently, to demonstrate braiding on a quantum computer emulating the spins, only two-qubit gates are necessary. In particular, we show that adiabatic manipulation of two-qubit interactions in a system of four qubits effectively braids two MZMs. Generalizing this system to ten qubits, we show a similar manipulation can braid four MZMs. This paper is organized as follows: in Sec. II we give the basic setup of the four and ten qubit systems, map it onto the equivalent MZM system, and calculate the non-Abelian gauge fields in both the reduced MZM space and spin space. In Sec. III we map this protocol onto a set of quantum gates and run the protocal on a quantum computer. We discuss our results and provide an outlook in Sec. . II Theoretical Model Consider a Y -junction of four Majorana fermions (MFs), \gamma_{j} for j=0,\ldots 3 , where \{\gamma_{i},\gamma_{j}\}=2\delta_{ij} described by the Hamiltonian \displaystyle\tilde{H} \displaystyle=i[\Delta_{z}(\gamma_{0}\gamma_{1})+\Delta_{y}(\gamma_{0}\gamma_{2})+\Delta_{x}(\gamma_{0}\gamma_{3})] \displaystyle=i\gamma_{0}(\vec{\Delta}\cdot\vec{\gamma})\,. (1) These four MFs can be operated as a logical qubit by controlling the coupling of the MFs which we parameterize by the coordinates on a sphere, \vec{\Delta}=\Delta(\sin\theta\cos\phi,\sin\theta\sin\phi,\cos\theta) , fixing |\vec{\Delta}|=\Delta . When \theta=0 , the system is idle and the MFs \gamma_{2} and \gamma_{3} commute with the Hamiltonian and are therefore MZMs. The Hamiltonian also commutes with the parity operator \tilde{n}=i\gamma_{2}\gamma_{3} which has eigenvalues of \pm 1 , defining the logical state of the qubit. Moreover, there exists low- and high-energy subspaces according to the eigenvalue of \tilde{h}=i\gamma_{0}\gamma_{1}=\mp 1 , respectively; for fixed \Delta , Eq. ( 1 ) cannot mix low- and high-energy states. We choose to work in a subset of states that reside in the low energy subspace without loss of generality. One can perform a rotation about the z axis of the Bloch sphere, using qubit states defined by the eigenvalues of \tilde{n} , by adiabatically changing \theta and \phi to trace a closed contour, \Omega_{c} , over the surface of a sphere. The resulting unitary operation is \mathcal{R}_{z}(\Omega_{c})=e^{-i\Omega_{c}\eta_{z}/2} with \eta_{z} the Pauli matrix acting in the space of eigenvectors of \tilde{n} . When \vec{\Delta} traces out an octant of the unit sphere with corners along the x , y , and z axes, then \Omega_{c}=\pi/2 and one can show that this corresponds to braiding of \gamma_{2} and \gamma_{3} . We refer to the case when \Omega_{c}\neq\pi/2 as ‘partially braiding’ the MZMs. Figure 1 : 10 MZM, triple Y-junction system described in Eq. ( 2 ). Each blue node houses the labeled MZM which are connected to the relevant neighbors by the red arms. Each arm is labeled by the relevant parameter of the clock arm vectors \vec{\Delta} , \vec{\Delta}^{\prime} and \vec{\Lambda} . The black dashed box highlights the Y-junction containing the first of the two topological qubits within this system and is equivalent to the system described in Eq. ( 1 ) Let us generalize the Y -junction to ten MFs wherein groups of three MFs are coupled through a central MF (Fig. 1 ), \tilde{H}_{10}=i\gamma_{0}(\vec{\Delta}\cdot\vec{\gamma})+i\gamma^{\prime}_{0}(\vec{\Delta}^{\prime}\cdot\vec{\gamma}^{\prime})+i\zeta_{0}(\vec{\Lambda}\cdot\vec{\zeta})\,. (2) Here, \vec{\gamma}^{\prime}=(\gamma_{1}^{\prime},\gamma_{2}^{\prime},\gamma_{3}^{\prime}) and \vec{\zeta}=(\zeta_{1},\gamma_{3}^{\prime},\gamma_{2}) with \gamma^{\prime}_{j} and \zeta^{\prime}_{j} MFs obeying the usual \{\zeta_{i},\zeta_{j}\}=\delta_{ij} , \{\gamma^{\prime}_{i},\gamma^{\prime}_{j}\}=\delta_{ij} and \{\gamma_{i},\gamma_{j}^{\prime}\}=\{\gamma_{i},\zeta_{j}\}=\{\gamma_{i}^{\prime},\zeta_{j}^{\prime}\}=0 . The parameters \vec{\Delta}^{\prime}=\Delta^{\prime}(\sin\theta^{\prime}\cos\phi^{\prime},\sin\theta^{\prime}\sin\phi^{\prime},\cos\theta^{\prime}) and \vec{\Lambda}=\Lambda(\sin\alpha\cos\beta,\sin\alpha\sin\beta,\cos\alpha) parameterize three clock arms. When \theta=\theta^{\prime}=\alpha=0 , the system is idle and \gamma_{2},\gamma_{3},\gamma^{\prime}_{2} and \gamma_{3}^{\prime} commute with the Hamiltonian and the energy is characterized by the eignevalues of \bar{h} , \bar{h}^{\prime}=i\gamma_{2}^{\prime}\gamma_{3}^{\prime} , and \bar{h}^{a}=i\zeta_{0}\zeta_{1} . Analogous to the case of four MFs, one can perform unitary transformations on n and n^{\prime}=i\gamma_{2}^{\prime}\gamma_{3}^{\prime} by moving the clock arms. It is easy to see that adiabatically moving \vec{\Delta}^{\prime} over the unit sphere, tracing a closed contour of area \Omega_{c} generates the operation \mathcal{R}_{z}^{\prime}(\Omega_{c})=\exp(-i\Omega_c\eta'_z/2) where \eta^{\prime\alpha} are the Pauli matrices acting on the space of states with i\gamma_{2}^{\prime}\gamma_{3}^{\prime}=\pm 1 . Adiabatically tracing a contour of area \Omega_{c} on the unit sphere with the vector \vec{\Lambda} performs the operation \mathcal{R}_{xx}(\Omega_{c})=\exp(-i \Omega_c\eta_x\eta_x'/2) . By moving \vec{\Delta}^{\prime} ( \vec{\Lambda} ) in a loop enclosing an area of \Omega_{c}=\pi/2 on the unit sphere, \gamma_{2}^{\prime} and \gamma_{3}^{\prime} ( \gamma_{2} and \gamma_{3}^{\prime} ) are braided. In order to emulate the MFs in Eq. ( 1 ), consider three qubits whose coupling to a central qubit are noncommuting, H=\Delta_{z}\sigma_{0}^{z}\sigma_{1}^{z}+\Delta_{y}\sigma_{0}^{y}\sigma_{2}^{y}+\Delta_{x}\sigma_{0}^{x}\sigma_{3}^{x}\,. (3) We recognize this as a piece of the so-called Kitaev lattice which itself is known to support dispersing Majorana edge modes. Inspired by the latter, we extend the dimension of the space by writing the Pauli matrices as products of Majoranas, \sigma^{\alpha}_{i}=i\gamma^{\alpha}_{i}\gamma_{i} , where \alpha=x,y,z , i=1,2,3,4 . Here, \{\gamma^{\alpha}_{i},\gamma^{\beta}_{j}\}=\delta_{ij}\delta_{\alpha\beta} , \{\gamma_{i},\gamma_{j}\}=\delta_{ij} , and \{\gamma_{i}^{\alpha},\gamma_{j}\}=0 . Upon appropriate substitution (see appendix), one can show that H , h=\sigma_{0}^{z}\sigma_{1}^{z} , n=\sigma_{2}^{z}\sigma_{3}^{z} can be mapped into \tilde{H} , \tilde{h}