Quantum computing, with its promise to revolutionize fields from drug discovery and materials science to breaking current encryption methods and developing unhackable communication, is currently hampered by a fundamental fragility inherent in its core components: qubits. Unlike the bits in classical computers that reliably represent either a 0 or a 1, qubits exist in a superposition of both states simultaneously. This quantum state, while powerful, is incredibly susceptible to disruption from environmental noise, a phenomenon known as decoherence. Decoherence manifests as errors that can corrupt the delicate quantum information. The two primary types of errors are bit flips, where a qubit abruptly switches from its ‘0’ or ‘1’ state, and phase flips, where the relative phase of a qubit’s superposition unexpectedly reverses. These errors, occurring randomly and unpredictably, can easily derail intricate quantum calculations, even with a single incident. Mitigating these errors is therefore a paramount challenge for quantum engineers striving to build reliable quantum machines.

To combat decoherence, the prevailing strategy involves the concept of logical qubits. Instead of relying on a single, highly vulnerable physical qubit, researchers encode the information of one logical qubit across multiple physical qubits. This redundancy, coupled with sophisticated error correction protocols, allows for the detection and correction of errors, effectively creating a more stable and resilient quantum information storage unit. While this approach has proven effective for preserving quantum information over time, quantum computers are not just about storage; they are about computation. To execute quantum algorithms, qubits must be actively manipulated using quantum gates, the fundamental building blocks of quantum operations. The true hurdle lies in performing these active operations – the very essence of computation – without introducing new errors, a task far more challenging than simply keeping qubits stable when idle.

Now, a team spearheaded by Professor Andreas Wallraff from the Department of Physics at ETH Zurich, in collaboration with scientists from the Paul Scherrer Institute (PSI) and theorists led by Professor Markus Müller at RWTH Aachen University and Forschungszentrum Jülich, has demonstrated a novel technique that directly addresses this critical challenge. Their work, recently published in the prestigious journal Nature Physics, showcases a method to execute quantum operations between superconducting logical qubits while simultaneously correcting errors. This breakthrough represents a significant stride towards achieving fault-tolerant quantum computing, a paradigm where quantum computations can proceed reliably and accurately, impervious to the constant threat of errors.

The distinction between classical and quantum error correction is profound. Classical computers can readily employ redundancy by copying information. If a bit flips, multiple identical copies allow for a simple majority vote to identify and correct the error. This fundamental principle, however, cannot be applied to quantum systems. "With qubits, things are a lot more complicated," explains Dr. Ilya Besedin, a postdoctoral researcher in Wallraff’s group and a co-leading author of the study. The no-cloning theorem, a cornerstone of quantum mechanics, strictly forbids the perfect copying of an arbitrary quantum state. Instead, quantum information must be distributed across entangled qubits. Furthermore, quantum systems are plagued by phase flip errors, a type of error with no direct analogue in classical computing, necessitating unique and specialized correction mechanisms.

A leading candidate for implementing quantum error correction is the use of surface codes. In this architecture, the quantum information of a single logical qubit is dispersed across an array of physical data qubits. Error detection is achieved through the periodic measurement of "stabilizers," auxiliary qubits that work in concert with the data qubits to define the logical qubit. These stabilizers are interconnected with the data qubits, and their measurement outcomes provide crucial information about whether a bit flip or a phase flip has occurred since the last check. Specifically, Z-type stabilizers are sensitive to bit flips, while X-type stabilizers detect phase flips. Crucially, the data qubits themselves are never directly measured during this process, ensuring that the encoded quantum state remains intact and undisturbed while errors are identified and corrected.

The complexity escalates dramatically when researchers aim to perform logical operations, such as a controlled-NOT (CNOT) gate, between two logical qubits. Errors can arise not only during the storage of information but also during the execution of these gate operations, and these operational errors must also be corrected in real-time. "Performing a logical operation in this fault-tolerant way would be relatively easy if we could move our qubits around and connect them arbitrarily to each other," notes Michael Kerschbaum, a PhD student in Wallraff’s group and the other co-leading author. However, in superconducting quantum processors, qubits are typically fixed in position, with interactions limited to neighboring qubits. This physical constraint poses a significant limitation on how complex logical operations can be implemented.

To circumvent these architectural limitations, the team ingeniously employed a technique known as "lattice surgery." Their experiment began with a single logical qubit encoded within seventeen physical qubits, arranged in a roughly square configuration that included both data qubits and stabilizers. Over several carefully timed cycles, stabilizers were measured every 1.66 microseconds to continuously correct for both bit flips and phase flips. The pivotal moment in their demonstration involved the measurement of three central data qubits running through the heart of the square. This targeted measurement effectively "split" the surface code into two independent halves. Concurrently, the measurement of the X-type stabilizers was temporarily suspended.

"The end result of this operation was that we had two logical qubits entangled with each other," explains Besedin. During this splitting process, bit flip errors continued to be addressed. Following the split, bit flip error correction resumed independently on each of the two resulting logical qubits. While this specific lattice surgery operation, as demonstrated, doesn’t directly produce a CNOT gate, it forms a fundamental building block. By combining multiple splitting and merging operations, it is possible to construct arbitrary logical gates, including the essential CNOT gate.

"One could say that the lattice surgery operation is the operation, and all the others can be constructed from it," emphasizes Besedin. He further highlights the novelty of their achievement: "To the best of our knowledge, this is the first time lattice surgery has been performed on superconducting qubits." While acknowledging that further development is necessary – for instance, approximately 41 physical qubits would be required to render the splitting operation stable against phase flips as well – this successful demonstration of lattice surgery on superconducting qubits marks a crucial milestone. It brings the ambitious goal of constructing useful quantum computers with thousands of highly stable and controllable qubits significantly closer to reality, paving the way for an era of unprecedented computational power.