Majorana nanowire devices are being investigated as a possible platform for topological quantum computing. In semiconductor nanowires coupled to a superconductor, a magnetic field can drive the system into a topological superconducting regime, where Majorana zero modes (MZMs) are expected to form near the wire ends. A zero-bias peak in tunneling conductance can be an important signal, but it is not conclusive on its own: ordinary Andreev bound states and other effects can produce similar features.
The next step is therefore to combine better-controlled devices with measurements that probe more than a local conductance peak. Several proposed transport experiments use gate control, multiple terminals, spatially separated probes or interference to test distinct aspects of Majorana behavior. Together, they offer a more rigorous route to evaluating candidate devices.
Why device quality matters
Early semiconductor–superconductor nanowire experiments reported zero-bias peaks at finite magnetic field. Their gate and field dependence was consistent with expectations for MZMs, but alternative, topologically trivial explanations were also proposed. A further concern was the soft induced superconducting gap: appreciable sub-gap conductance can obscure signatures and undermine the conditions needed for topological protection.
Disorder at the superconductor–nanowire interface can contribute to both problems. Improved fabrication and cleaner interfaces have enabled ballistic transport signatures, including quantized conductance plateaus. Epitaxial growth, in which the superconductor is grown directly on the semiconductor nanowire, has also produced cleaner interfaces and hard superconducting gaps in the devices described in the research. These advances strengthen the foundation for more discriminating measurements, but they do not by themselves establish that a zero-bias feature is a Majorana mode.
Majorana nanowire devices: five transport tests
1. Look for a peak-to-dip transition
In a normal-lead–nanowire–superconductor device, a gate can tune the tunnel barrier between the normal lead and the wire. With only one spin-resolved channel transmitting, an MZM can produce a zero-bias conductance peak quantized at 2e2/h. The proposed next test is to lower the barrier further, allowing a second spin-resolved channel to participate. For the Majorana case, a spin-selection effect can keep the zero-bias conductance at 2e2/h, turning the peak into a zero-bias dip relative to the surrounding conductance.
This behavior may help distinguish MZMs from partially separated Andreev bound states, also called quasi-Majorana states, which can imitate a quantized peak under some conditions. The comparison depends on how the additional channel couples to the states, so it calls for controlled barrier tuning rather than relying on peak height alone. A more strongly coupled tunnel gate, such as a wrap-around gate, is one possible device-design approach discussed for reaching the relevant regime.
2. Test whether a remote gate affects the signal
Tunneling spectroscopy mainly probes the local density of states near the tunnel barrier. Adding a gate farther along the nanowire creates a way to test non-local behavior: changing the remote electrochemical potential can move the distant MZM toward the local one, causing the zero-bias peak to split as the two modes hybridize.
A response to this non-local gate would be harder to reconcile with trivial states confined near the tunnel barrier. However, gate cross-capacitance could make a nominally remote gate affect the local region. Experiments therefore need controls that assess this coupling, for example by checking whether the remote gate influences known local trivial states.
3. Measure correlations and non-local conductance
A three-terminal normal–superconductor–normal device can measure conductance at both wire ends. Because MZMs are expected to occur as a pair, zero-bias peaks at the two ends—and changes such as peak splitting—should be correlated as magnetic field and gate voltages are varied. Their peak heights and widths need not match, since local tunnel barriers affect each measurement independently.
The superconducting section must be long enough to reduce the risk that both probes are detecting the same ordinary Andreev bound state. Correlations should also persist across systematic changes in fields and gate settings; a single coincident signal is not a sufficient test.
The same general device can be configured to measure crossed Andreev reflection through non-local conductance. In a sufficiently long wire, this measurement can reveal information about the induced gap across the proximitized region. Near a topological phase transition, the predicted non-local signal has a bias-asymmetric, rectifying form, including negative differential conductance on one side. Pairing this measurement with end-to-end correlation tests can connect a gap closing and reopening to the appearance of end-localized peaks.
4. Probe the wire interior
A T-shaped device adds a side electrode that samples the wire bulk, alongside an end electrode that measures near an expected MZM location. The side probe is expected to detect gap closing and reopening as the magnetic field changes, rather than a zero-bias peak, because the MZMs are localized at the ends.
This layout has practical challenges. Side probes can introduce potential inhomogeneity, which may create trivial Andreev bound states or disrupt the topological region. The tunnel barrier may also be difficult to tune independently. Device geometry and electrostatic control are therefore central to interpreting spatially resolved measurements.
5. Test phase coherence and Kondo behavior
A Majorana interferometer uses a superconducting island with finite charging energy and a separate coherent reference path. In the proposed Fu-interferometer scheme, magnetic flux produces Aharonov–Bohm oscillations. A change in island parity is expected to shift the interference phase by π. Since similar phase shifts can occur in devices without MZMs, a sufficiently long island is important for suppressing trivial tunneling processes while retaining coherent transport associated with spatially separated Majorana modes.
A different test is the topological Kondo effect. In a device with multiple MZMs coupled to normal leads, charging energy can produce Coulomb blockade, while the low-energy conductance is predicted to evolve with temperature toward a non-Fermi-liquid regime. Finite overlap between Majorana modes changes the expected low-temperature conductance. Observing the characteristic temperature-dependent crossover would probe non-local Majorana dynamics, although it requires a device with multiple modes and carefully controlled coupling.
Building a stronger case from complementary measurements
No single transport signature resolves every ambiguity in Majorana nanowire devices. Quantized peaks, gate response, correlations, non-local conductance and interference each test different properties—and each has its own device-specific confounders. For experimental teams, the practical implication is to design for independent control of tunnel barriers, electrostatic gates, wire length and measurement terminals from the outset. Combining complementary measurements, with controls for disorder and unintended gate coupling, offers a more reliable way to evaluate Majorana physics and guide progress toward braiding-based quantum operations.