Nernst Effect¶

Measure the ordinary off-diagonal thermoelectric response in which a longitudinal temperature gradient and perpendicular applied magnetic field generate a transverse open-circuit electric field, while keeping the response geometry distinct from its several microscopic mechanisms.

Version: v1 · 2026-08-30 · History
Domain-specific #: 2361
Origin domain: condensed matter physics
Subdomain: thermoelectric transport

Core Idea¶

The ordinary Nernst effect is the off-diagonal thermoelectric response in which a longitudinal temperature gradient in an electrically conducting sample, together with an applied magnetic field perpendicular to that gradient, produces an electric field transverse to both. Fix a common right-handed geometry: the thermal drive is \(-\partial_x T\), the applied field is \(B_z\), and the readout is \(E_y\). One widely used convention defines the Nernst signal and its weak-field coefficient by

\[ N(B_z)=\frac{E_y}{-\partial_xT}, \qquad \nu=\lim_{B_z\to0}\frac{N(B_z)}{B_z}. \]

Authors sometimes reverse axes or absorb a minus sign into \(N\) or \(\nu\), so a reported sign has no meaning until the axis, field, thermal-gradient, and voltage conventions are stated. The ratio \(N/B\) is a field-independent coefficient only in a suitable low-field linear-response regime; at higher field the signal can be nonlinear and \(N/B\) can vary strongly.^[1]

The ordinary effect is identified by response geometry and tensor structure, not by one universal picture of individual charge carriers being bent by a Lorentz force. In linear transport,

\[ \mathbf J_e=\boldsymbol\sigma\mathbf E- \boldsymbol\alpha\boldsymbol\nabla T. \]

An open-circuit voltage experiment sets \(\mathbf J_e=0\), hence \(\mathbf E=\boldsymbol\sigma^{-1}\boldsymbol\alpha\boldsymbol\nabla T\). The measured transverse field therefore depends on both the electrical-conductivity tensor \(\boldsymbol\sigma\) and the thermoelectric tensor \(\boldsymbol\alpha\); it is not simply a direct reading of \(\alpha_{xy}\). For an in-plane isotropic tensor under the convention above,

\[ N=\frac{\sigma_{xx}\alpha_{xy}-\sigma_{xy}\alpha_{xx}} {\sigma_{xx}^{2}+\sigma_{xy}^{2}}. \]

Here the displayed scalar formula uses the common in-plane antisymmetric convention \(\boldsymbol\sigma=\bigl(\begin{smallmatrix}\sigma_{xx}&\sigma_{xy}\\-\sigma_{xy}&\sigma_{xx}\end{smallmatrix}\bigr)\) and the analogous convention for \(\boldsymbol\alpha\); changing tensor-index or axis conventions changes the displayed signs, not the underlying relation \(\mathbf E=\boldsymbol\sigma^{-1}\boldsymbol\alpha\nabla T\). This combination explains why the carrier charge sign alone does not determine the Nernst sign and why cancellations, multiple bands, or different scattering-energy dependences matter. In other regimes the same transverse observable can receive contributions from superconducting vortices or fluctuations, Landau quantization, and Berry curvature.^[1] The mechanism must be inferred from field, temperature, symmetry, and material dependence rather than built into the effect's definition.

The canonical node remains the ordinary applied-field effect. The anomalous Nernst effect is a qualified variant in which spontaneous magnetization or magnetic order can supply time-reversal breaking even at zero applied field; Berry curvature can dominate its off-diagonal thermoelectric conductivity.^[2]^[3] The planar Nernst effect is another qualified variant: its field lies in the plane defined by the temperature gradient and magnetization and is typically quadratic in magnetization, with a different angular and reversal symmetry from the cross-product anomalous response.^[4] These forms belong in the vocabulary neighborhood, but neither is a lossless alias for the ordinary effect because constitutive roles and diagnostics change.

Structural Signature¶

Sig role-phrases:

the electrically conducting medium — a metal, semiconductor, or superconducting regime in which coupled electrical and thermal transport can be measured
the longitudinal temperature gradient — the spatially directed thermal drive, conventionally \(-\partial_xT\), whose magnitude and direction must be calibrated
the perpendicular applied magnetic field — the ordinary effect's constitutive time-reversal-breaking control, conventionally \(B_z\); anomalous variants replace this role with magnetization or magnetic order rather than preserving it literally
the off-diagonal thermoelectric coupling — the tensor channel that converts a longitudinal thermal drive into a transverse electrical response instead of a collinear Seebeck response
the transverse electrical readout axis — the axis orthogonal to both the ordinary thermal-gradient and magnetic-field directions on which \(E_y\) or a voltage is detected
the electrical boundary condition — commonly open circuit, \(\mathbf J_e=0\), which converts the coupled \(\boldsymbol\sigma\) and \(\boldsymbol\alpha\) tensors into the measured electric field
the reversal and contamination protocol — comparison of opposite applied fields to isolate the ordinary field-odd signal and diagnose longitudinal Seebeck pickup from imperfectly aligned contacts
the convention-bound response measure — \(N=E_y/(-\partial_xT)\) and, only in the weak-field limit, \(\nu=N/B\), reported with axes, signs, units, field range, and temperature

These roles yield a strict recognition test for the ordinary effect. There must be a calibrated longitudinal temperature gradient, a perpendicular applied magnetic field, an electrically transverse response, and an explicit electrical boundary condition. Reversing the applied field should reverse the ordinary field-odd contribution after any magnetic hysteresis or background is handled. A transverse voltage that persists because of contact pickup, a Hall voltage driven by current, or a magnetization-dependent planar voltage with different symmetry does not pass merely because it is measured on transverse contacts.

The signature deliberately separates invariant roles from microscopic occupants. Quasiparticle velocities and scattering times can occupy the response mechanism in a normal metal; moving flux-carrying vortices can occupy it in a mixed-state superconductor; Berry curvature can occupy it in a magnetic material. What remains invariant is the controlled thermal drive, time-reversal-breaking structure, off-diagonal coupling, transverse electrical readout, and coefficient protocol. What does not remain invariant across qualified variants is specifically the ordinary node's applied-field role or its simple field-odd diagnostic.

What It Is Not¶

Not a universal Lorentz-deflection story. Semiclassical quasiparticle deflection is important but cannot define the whole response. Vortex motion, superconducting fluctuations, and intrinsic Berry-curvature terms can generate Nernst signals with different microscopic carriers and field dependences.
Not just \(\alpha_{xy}\). Under open circuit, \(\mathbf E=\boldsymbol\sigma^{-1}\boldsymbol\alpha\nabla T\); longitudinal thermopower mixed through Hall conductivity can add to or cancel the direct off-diagonal thermoelectric term.
Not any transverse voltage in a magnetic field. Hall pickup, misaligned longitudinal contacts, thermal Hall effects, and magnetoresistive artifacts can all appear on transverse leads. The imposed drive, boundary condition, field parity, and tensor channel decide the attribution.
Not a charge-current flow requirement. The standard electrical readout is commonly made at zero net electrical current. Heat and entropy transport remain present, but Flow is not a universal genus for the measured open-circuit effect.
Not universally linear in field. \(N\propto B\) and \(\nu=N/B\) are low-field approximations. Quantum oscillations, superconducting crossovers, saturation, and other nonlinearities make a finite-field ratio regime-dependent.
Not sign-equivalent to the carrier charge. In a simple one-band metal, competing contributions can nearly cancel; multiband conduction, energy-dependent scattering, vortices, and Berry curvature further break any one-sign inference.
Not identical to anomalous or planar Nernst response. The anomalous form can survive at zero applied field through magnetic order; the planar form has different coplanar geometry and reversal symmetry. Both require qualified labels.
Not the Ettingshausen effect. Ettingshausen exchanges the experimental drive and response: an electrical current in a magnetic field produces a transverse temperature gradient or heat flow. Reciprocal relations connect transport coefficients under field reversal but do not collapse the two experiments into one effect.^[5]

Scope of Application¶

The effect applies wherever off-diagonal thermoelectric transport can be established and separated from longitudinal and other transverse channels.

Ordinary metals and semimetals. The Nernst signal probes Fermi-surface geometry, carrier mobility, scattering-energy dependence, compensation, and multiband transport. A small signal can reflect cancellation rather than an absence of thermoelectric response; a large signal can reflect high mobility or a small Fermi energy, but neither inference is mechanism-free.^[1]
Semiconductors. Tunable carrier density and mobility make transverse thermoelectric measurements sensitive to band structure and scattering, provided bipolar conduction and contact effects are modeled.
Superconductors. In a type-II mixed state, thermally driven vortices carry magnetic flux and entropy and can generate a strong transverse electric field. Above the transition, superconducting fluctuations can also contribute. The Nernst response is therefore widely used to study vortex matter and fluctuation regimes without redefining the effect as quasiparticle deflection.^[1]
Quantum and topological materials. Landau quantization, small Fermi surfaces, and Berry curvature can strongly restructure the response. Here field and temperature sweeps, quantum oscillations, and comparison with Hall and Seebeck data become part of mechanism diagnosis.
Magnetic conductors. The anomalous Nernst variant uses magnetization or magnetic order as the time-reversal-breaking structure and can produce a transverse thermovoltage at zero applied field.^[2]^[3] It is useful for magnetic-state readout and transverse thermoelectric conversion, but must remain qualified because it replaces an ordinary mandatory role.
Thin films and nanodevices. Transverse geometry can support heat-flux sensing and energy conversion, but three-dimensional temperature gradients, magnetic anisotropy, contact placement, and planar/anomalous admixtures become unusually consequential.^[4]

The abstraction does not apply merely because a material is conducting, thermoelectric, or magnetic. A usable attribution requires the complete geometry, a stated electrical boundary condition, and enough reversal or angular data to separate the desired off-diagonal channel. In insulators a spin Nernst current may be definable, but a pure spin-current phenomenon is distinct from this charge-electric-field node unless an electrical conversion layer and its own mechanism are explicitly included.

Clarity¶

The Nernst effect clarifies a phrase such as “heat made a sideways voltage” by forcing six questions: Which direction is the calibrated temperature gradient? What supplies time-reversal breaking? Which direction is called transverse? Is the electrical condition open circuit or current biased? Which parity under field or magnetization reversal is expected? Which quantity—raw voltage, \(N\), low-field \(\nu\), or \(\alpha_{xy}\)—is actually being reported?

Those questions prevent three common category errors. First, they keep a voltage ratio from being confused with an intrinsic tensor component: the measured open-circuit field combines \(\boldsymbol\sigma\) and \(\boldsymbol\alpha\). Second, they keep response identity from being confused with microscopic explanation: the same geometry can be produced by quasiparticles, vortices, fluctuations, or Berry curvature. Third, they keep neighboring effects separate by drive and response. Seebeck is longitudinal thermal-to-electric response; Hall is electrical-to-transverse-electric response; Ettingshausen is electrical-to-transverse-thermal response; Righi–Leduc is longitudinal-thermal-to-transverse-thermal response.

The sign convention is part of the claim, not editorial decoration. “Positive Nernst signal” is incomplete unless the directions of \(-\nabla T\), \(B\), and positive transverse voltage are fixed. Similarly, “Nernst coefficient” should say whether it means the finite-field ratio \(N/B\), the differential low-field slope, or another convention. Naming the roles turns an apparently simple scalar into an auditable transport statement.

Manages Complexity¶

Thermoelectric experiments contain many simultaneously active channels: longitudinal electrical and thermal conductivities, Hall conductivity, Seebeck response, transverse thermoelectric conductivity, contact offsets, magnetoresistance, and, in magnetic or superconducting materials, hysteresis, vortices, fluctuations, and Berry curvature. The Nernst framework compresses this sprawl into a controlled off-diagonal measurement: impose one thermal drive, choose an orthogonal field, read one transverse voltage, and use tensor inversion plus reversal symmetry to identify what combination was observed.

That compression is conditional. It works only when the analyst preserves the boundary condition and parity information. Field antisymmetrization,

\[ E_y^{\mathrm{odd}}(B)=\frac{E_y(+B)-E_y(-B)}{2}, \]

removes a field-even offset and much of the longitudinal Seebeck pickup caused by transverse-contact misalignment in an ordinary nonmagnetic measurement.^[6] The complementary even component is not waste: it diagnoses misalignment, thermopower background, or a different symmetry channel. In magnetic materials, hysteresis and magnetization reversal must be analyzed explicitly rather than assuming that every unwanted term is even in applied field.

The framework also manages model proliferation by separating levels. The first level is observational: geometry, boundary condition, parity, and coefficient. The second is constitutive: \(\boldsymbol\sigma\) and \(\boldsymbol\alpha\). The third is microscopic: bands, scattering, vortices, fluctuations, or Berry curvature. Evidence may support a level-three mechanism only after the first two have been fixed. This ordering lets different microscopic models compete without changing what experiment they are trying to explain.

Abstract Reasoning¶

The Nernst effect licenses a repeatable set of reasoning moves.

Freeze the oriented geometry. Assign \(x\) to the thermal drive, \(z\) to the ordinary applied field, and \(y\) to the readout. Reverse one control at a time and predict the sign change before looking at data.
State the boundary condition. Under open circuit solve \(\mathbf J_e=0\); under current bias use the full transport equations instead. Never move between those conditions as if the same scalar were measured.
Separate raw signal from response quantity. Convert voltage to electric field using contact separation, temperature difference to gradient using thermometer spacing, and then define \(N\) and the applicable low-field slope.
Decompose by symmetry. Form field-odd and field-even parts, inspect magnetization hysteresis where relevant, and rotate the field or gradient when planar and anomalous channels may coexist.
Distinguish observable from coefficient tensor. Use \(\mathbf E=\boldsymbol\sigma^{-1}\boldsymbol\alpha\nabla T\) to see how \(\alpha_{xy}\), \(\alpha_{xx}\), \(\sigma_{xy}\), and \(\sigma_{xx}\) combine.
Test regimes rather than assuming linearity. Determine the range where \(N(B)\) is linear; treat \(N/B\) outside that range as a finite-field ratio, not a material constant.
Infer mechanism comparatively. Ask whether field, temperature, angular, thickness, and material dependence favor quasiparticles, vortices, superconducting fluctuations, or Berry curvature. No sign or magnitude alone settles the issue.
Reclassify changed roles. If the signal survives at zero applied field because magnetization supplies broken time-reversal symmetry, call it anomalous Nernst. If the electric field is coplanar and has quadratic magnetization symmetry, test the planar Nernst form. Do not silently stretch the ordinary signature.

These moves make the effect a reasoning tool, not only a name. They specify what can be inferred from a reversal, what cannot be inferred from a sign, when a coefficient is meaningful, and when a superficially similar sideways voltage belongs to another transport channel.

Knowledge Transfer¶

Within condensed-matter and materials physics, the Nernst framework transfers as a common diagnostic across mechanisms without losing its role vocabulary.

Normal-metal transport: thermal drive, applied field, transverse contacts, and tensor inversion expose mobility, compensation, and scattering-energy dependence.
Superconductivity: the same readout geometry is retained, while vortex motion or superconducting fluctuations replace the normal-quasiparticle mechanism. The distinction between effect and cause is what makes this transfer legitimate.
Topological materials: the off-diagonal thermoelectric tensor is retained, while Berry curvature supplies an intrinsic contribution. Joint Hall, Seebeck, and Nernst measurements constrain the electronic structure more strongly than any one scalar does.
Magnetic materials: the anomalous variant retains longitudinal thermal drive and transverse electric readout but replaces the ordinary applied-field role with magnetization or order. Reversal follows the magnetic state and may be hysteretic.
Sensor and energy-conversion design: transverse geometry permits voltage contacts to extend along a heat source without lying along the heat-flow direction. The abstraction transfers as a design grammar—thermal drive, symmetry-breaking control, orthogonal readout—but materials optimization still depends on conductivity, geometry, and heat leakage.
Experimental metrology: parity decomposition and angular scans transfer to any transport experiment where contact misalignment mixes a large longitudinal channel into a smaller transverse one. What transfers is the diagnostic method, not the claim that every odd voltage is Nernst.

The transfer has a hard boundary. Using “Nernst” for the electrochemical Nernst equation, the Nernst heat theorem, or a generic cross-domain metaphor imports only a name, not the transport roles. Even within physics, “spin Nernst” changes the transported/read-out quantity from transverse electric field to transverse spin current. The abstraction travels robustly across material substrates because the transport geometry and tensor reasoning recur literally; it does not become substrate-neutral.

Examples¶

Canonical¶

Illustrative measurement; the numbers are chosen to expose the procedure rather than report a particular material. A rectangular nonmagnetic conductor is held with \(-\partial_xT=100\ \mathrm{K\,m^{-1}}\). A perpendicular field is applied along \(z\), with \(B_z=\pm0.50\ \mathrm T\), and transverse contacts separated along \(y\) yield raw fields \(E_y(+B)=+7\ \mu\mathrm{V\,m^{-1}}\) and \(E_y(-B)=+3\ \mu\mathrm{V\,m^{-1}}\). The field-even component is \(5\ \mu\mathrm{V\,m^{-1}}\), consistent with longitudinal Seebeck pickup or offset. The field-odd ordinary component is

\[ E_y^{\mathrm{odd}}=\frac{7-3}{2}=2\ \mu\mathrm{V\,m^{-1}}. \]

Thus \(N=2\times10^{-8}\ \mathrm{V\,K^{-1}}=20\ \mathrm{nV\,K^{-1}}\), and, if \(\pm0.50\ \mathrm T\) lies in the verified linear regime, \(\nu=40\ \mathrm{nV\,K^{-1}\,T^{-1}}\). A second field magnitude is required to verify that linear-regime premise; one symmetric pair alone establishes parity, not field linearity.

Mapped back: The rectangular conductor is the electrically conducting medium; \(-\partial_xT=100\ \mathrm{K\,m^{-1}}\) is the longitudinal temperature gradient; \(B_z=\pm0.50\ \mathrm T\) is the perpendicular applied magnetic field; the odd conversion from that thermal drive to \(E_y\) is the off-diagonal thermoelectric coupling; the \(y\)-separated contacts provide the transverse electrical readout axis; the high-impedance voltmeter enforces the electrical boundary condition \(\mathbf J_e=0\); the \(+B/-B\) difference is the reversal and contamination protocol; and the explicitly signed \(N\) and conditionally low-field \(\nu\) are the convention-bound response measure.

Applied / In Practice¶

In the ferromagnetic Weyl semimetal \(\mathrm{Co_3Sn_2S_2}\), Yang and colleagues observed a large anomalous Nernst response: a longitudinal temperature gradient generated transverse voltage even without an external magnetic field, and the signal showed hysteretic behavior tied to the magnetic state. Their measurements and calculations attributed the large anomalous transverse thermoelectric conductivity to Berry curvature near Weyl points.^[3] This is a boundary-demonstrating application, not a reason to erase “applied magnetic field” from the ordinary node. The qualified anomalous variant replaces that role with spontaneous magnetization and magnetic order.

Mapped back: \(\mathrm{Co_3Sn_2S_2}\) is the electrically conducting medium and the imposed longitudinal heating is the longitudinal temperature gradient. The ordinary role the perpendicular applied magnetic field is deliberately changed: spontaneous magnetization and magnetic order supply time-reversal breaking at zero external field. Berry-curvature-mediated \(\alpha_{yx}^{A}\) supplies the off-diagonal thermoelectric coupling; the orthogonal voltage contacts supply the transverse electrical readout axis; the voltage measurement supplies the electrical boundary condition; magnetization reversal and its hysteresis replace simple \(+B/-B\) antisymmetrization as the reversal and contamination protocol; and the anomalous Nernst signal or anomalous transverse thermoelectric conductivity, with its axes and magnetic history stated, is the convention-bound response measure. The mapped role replacement is exactly why “anomalous Nernst effect” is a recognized qualified variant rather than an exact alias.

Structural Tensions¶

T1: Response identity versus microscopic explanation. A mechanism-neutral tensor definition lets normal quasiparticles, vortices, fluctuations, and Berry curvature be compared as explanations of the same transverse observable. But if the definition is made too neutral, any sideways thermovoltage can be mislabeled Nernst; if it is made Lorentz-specific, major legitimate regimes are excluded. Diagnostic: Are geometry, boundary condition, and symmetry fixed before a microscopic mechanism is inferred, and does the proposed mechanism explain the observed field and temperature dependence without being smuggled into the definition?

T2: Ordinary identity versus qualified variants. A broad family resemblance connects ordinary, anomalous, and planar Nernst effects. Yet anomalous response can remove the external-field role, while planar response changes the orientation and reversal law; treating all three as role-identical destroys the ordinary recognition test. Diagnostic: Does the case preserve a perpendicular applied field and field-odd transverse geometry, or must it be labeled anomalous or planar because the time-reversal source, geometry, or parity has changed?

T3: Open-circuit observable versus intrinsic tensor component. Measuring \(E_y\) at \(\mathbf J_e=0\) is experimentally direct, while theory often discusses \(\alpha_{xy}\). The former mixes thermoelectric and conductivity tensors; the latter is not what the voltmeter reads without inversion. Diagnostic: Is the claim about raw \(E_y\), Nernst signal \(N\), low-field \(\nu\), thermopower tensor \(S=\sigma^{-1}\alpha\), or \(\alpha_{xy}\), and were the conversions performed with compatible sign conventions?

T4: Weak-field compression versus nonlinear regime. The coefficient \(\nu=N/B\) neatly compresses a linear response, enabling comparisons among samples. The same ratio obscures crossovers, maxima, quantum oscillations, or saturation when used beyond the range in which \(N\) is proportional to \(B\). Diagnostic: Has linearity been established over a finite low-field interval, or is a finite-field ratio being mislabeled a field-independent material coefficient?

T5: Symmetry isolation versus experimental contamination. Field antisymmetrization powerfully removes field-even Seebeck leakage and offsets from an ordinary measurement. But magnetic hysteresis, anomalous terms, thermometry asymmetry, or field-dependent contacts can violate the simple even/odd model. Diagnostic: Do raw \(+B\) and \(-B\) traces, repeated thermal reversals, angular scans, and magnetic history support the parity decomposition rather than merely assume it?

T6: Reciprocal constraint versus distinct reciprocal experiment. Onsager–Casimir relations constrain thermomagnetic coefficients under magnetic-field reversal, and historically tie Nernst and Ettingshausen measurements together.^[5] Yet exchanging force and response changes the laboratory experiment and named effect. Diagnostic: Is reciprocity being used to relate coefficients with the required field reversal, or to incorrectly call the current-driven transverse thermal response “the Nernst effect”?

T7: Cancellation versus absence. In a simple quasiparticle picture, terms contributing to the measured Nernst signal may nearly cancel; multiband or energy-dependent scattering can lift that cancellation. A small signal therefore need not mean weak thermal-electric coupling, and a large signal does not identify one cause. Diagnostic: Have \(\sigma_{xx}\), \(\sigma_{xy}\), \(\alpha_{xx}\), and \(\alpha_{xy}\), or experimentally equivalent Hall and Seebeck data, been considered before interpreting magnitude and sign?

T8: Autonomy versus reduction. Nernst Effect is autonomous enough to own a stable experimental grammar, tensor relation, coefficient family, reversal protocol, and mechanism-neutral inferential closure. It nevertheless reduces structurally to a temperature Gradient feeding a specialized Coupling; if its thermomagnetic roles are stripped away, nothing uniquely Nernst remains. Diagnostic: Resolve toward the named domain abstraction when the oriented thermoelectric geometry and diagnostic package are in play, and toward Gradient plus Coupling when only the substrate-neutral thermal drive and cross-dependency are being discussed.

Structural–Framed Character¶

The Nernst effect is mixed-structural, strongly structural-leaning. On evaluative weight, it is neutral: a transverse thermoelectric coefficient can be large, small, positive, or negative without becoming morally or epistemically good or bad. Engineering may value a large response, but that goal is external to the definition. On human-practice-boundness, measurement choices matter—axes, sign convention, open-circuit wiring, and antisymmetrization are human protocols—yet they reveal a material response rather than constitute it. Removing laboratories would remove our measurements and names, not the coupled transport relation in the material.

On institutional origin, the abstraction arose within experimental and theoretical physics, but no institution grants a sample its Nernst response. Standardized symbols and conventions vary, which is why the entry requires them to be stated; this conventional layer does not make the underlying response institution-created. On whether the vocabulary travels, the complete vocabulary—\(\boldsymbol\sigma\), \(\boldsymbol\alpha\), open-circuit electric field, perpendicular magnetic field, vortices, Berry curvature—travels literally across metals, semiconductors, superconductors, and magnetic/topological materials. It does not travel intact into organizational, social, or biological substrates. Only the thinner language of a directed gradient feeding an off-diagonal coupling escapes condensed-matter transport.

On import versus recognition, physicists recognize the same mechanism-level response grammar when they move from a metal to a superconductor or topological material: the substrate and microscopic occupant change while the oriented drive/readout relation remains. Outside that orbit, calling a policy response or social effect “Nernst-like” would be metaphorical import. Even close names impose boundaries: the electrochemical Nernst equation and Nernst heat theorem are namesakes, while spin Nernst changes the transported quantity. Its character: an evaluatively neutral and non-institutional physical response whose tensor geometry is structural across several material regimes, but whose magnetic, electrical, and thermoelectric roles remain too domain-bound for prime status.

Structural Core vs. Domain Accent¶

What is skeletal. The portable core is a directed Gradient serving as an input to a cross-variable Coupling: variation of one quantity over space drives a response in another channel, and the response is read along a direction different from the drive. That skeleton can be recognized in many substrates. The gradient establishes an oriented local difference; coupling establishes that one subsystem or variable becomes input to another through a specified channel. Neither prime contributes a magnetic field, a conductivity tensor, an open-circuit voltage, or a Nernst coefficient.

What is domain-bound. The accent is almost the whole working apparatus: an electrically conducting material, coupled charge-and-heat transport tensors, a longitudinal temperature gradient, a perpendicular applied magnetic field for the ordinary effect, a transverse electric field, the \(\mathbf J_e=0\) boundary condition, field-reversal antisymmetry, and convention-bound quantities \(N\) and \(\nu\). The mechanism family—quasiparticles, vortices, superconducting fluctuations, Landau levels, and Berry curvature—and the qualified anomalous and planar variants are likewise condensed-matter cargo. Strip those roles and the Nernst identity disappears.

Why this does not clear the prime bar. A prime must survive free substitution with essentially the same vocabulary and reasoning leverage. “Gradient feeding an off-diagonal coupling” survives, but “Nernst effect” does not: outside thermoelectric transport there is no applied magnetic field/thermal-gradient/transverse-voltage triad, no \(\boldsymbol\sigma^{-1}\boldsymbol\alpha\) inference, no field-odd isolation, and no Nernst coefficient. The candidate therefore cannot replace Gradient or Coupling, nor is it entailed by their conjunction. It owns a proprietary domain-level recognition test and diagnostic closure, making it a genuine domain-specific abstraction rather than a mere composite or a new prime.

The Nernst effect strictly presupposes Gradient by composition. Its ordinary thermal drive is not a finite temperature difference alone but an oriented spatial temperature gradient. Removing that drive leaves no Nernst experiment. Gradient can exist without any magnetic field or transverse electrical response, so the direction is asymmetric.

The effect is also a strict subsumption of Coupling: it is a specific off-diagonal thermal-to-electrical cross-coupling whose channel, direction, boundary condition, and response measure are fixed by thermomagnetic transport. Generic coupling lacks these commitments, while every ordinary Nernst case contains the dependency by which a longitudinal thermal drive changes a transverse electrical variable. The anomalous and planar qualified variants retain a thermoelectric cross-coupling even though their time-reversal source or geometry changes.

Flow is a useful neighbor but not a necessary parent. Heat and entropy transport underlie the experiment, and vortices or quasiparticles may move, yet the defining electrical observation is commonly made at zero net charge current; flow vocabulary does not distinguish the effect. The live Reciprocity prime owns social mutual exchange, not Onsager–Casimir transport reciprocity, so it is a wrong-sense parent. The live domain-specific Field node is algebraic field, not magnetic field, and is likewise a wrong-sense endpoint. Hall, Seebeck, Ettingshausen, Righi–Leduc, anomalous Nernst, and planar Nernst are neighboring response channels or qualified variants, not required live parents.

Relationships to Other Abstractions¶

Current abstraction Nernst Effect Domain-specific

Parents (2) — more general patterns this builds on

Nernst Effect is a kind of Coupling Prime

The Nernst effect strictly presupposes Gradient by composition.
Nernst Effect presupposes Gradient Prime

The Nernst effect strictly presupposes Gradient by composition.

Hierarchy paths (2) — routes to 2 parentless roots

Nernst Effect → Coupling

Show alternative path (1)

Neighborhood in Abstraction Space¶

Nernst Effect sits in a sparse region of the domain-specific corpus (88^th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Unclustered & Miscellaneous (1565 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-09-08

Not to Be Confused With¶

Seebeck effect. A temperature gradient produces an electric field primarily along the gradient, with no perpendicular magnetic-field geometry required. Tell: Is the electrical readout longitudinal to \(\nabla T\) (Seebeck) or transverse to it and field-conditioned (ordinary Nernst)?
Hall effect. An imposed electrical current and magnetic field produce a transverse electric field. Tell: Is the drive electrical current (Hall) or a temperature gradient under a stated electrical boundary condition (Nernst)?
Ettingshausen effect. Electrical current in a magnetic field produces a transverse temperature gradient or heat response. Tell: Is the response thermal and the drive electrical (Ettingshausen), or the response electrical and the drive thermal (Nernst)?
Righi–Leduc or thermal Hall effect. A longitudinal heat flow produces a transverse temperature gradient. Tell: Are the transverse contacts thermometers measuring temperature (Righi–Leduc) or electrodes measuring electric field (Nernst)?
Anomalous Nernst effect. Magnetization or magnetic order can produce transverse thermoelectric response at zero applied field and with hysteretic reversal. Tell: Is a perpendicular applied field constitutive and the ordinary field-odd part sought, or does the signal track an internal magnetic order parameter even at zero field?
Planar Nernst effect. A magnetization-dependent transverse thermopower lies in the plane of \(\nabla T\) and \(M\), typically with a \(\sin 2\phi\)-type angular dependence and evenness under 180° magnetization reversal. Tell: Is the response coplanar and quadratic in magnetization (planar), or orthogonal to the \(\nabla T\)-field plane and odd under reversal (ordinary/anomalous cross-product form)?
Spin Nernst effect. A longitudinal temperature gradient generates a transverse spin current, which may need another layer or inverse spin Hall process for electrical detection. Tell: Is the primary transported/read-out quantity spin current (spin Nernst) or an open-circuit transverse electric field (this node)?
Nernst signal versus Nernst coefficient. \(N\) is the transverse field per temperature gradient; \(\nu\) is commonly its weak-field value per magnetic field. Tell: Is the reported number in \(\mathrm{V/K}\) for \(N\), in \(\mathrm{V/(K\,T)}\) for \(\nu\), or a raw voltage that still requires geometry and gradient normalization?
Electrochemical Nernst equation. This relates an electrode or cell potential to activities, temperature, and reaction quotient. Tell: Is the claim about chemical equilibrium voltage and ionic activities (Nernst equation) or a transverse thermomagnetic transport response (Nernst effect)?
Nernst heat theorem / third law. This is a thermodynamic statement about entropy changes approaching zero near absolute zero. Tell: Is the subject low-temperature entropy and thermodynamic limits (heat theorem) or transverse voltage under a temperature gradient and magnetic field (effect)?
Raw transverse thermovoltage. Contact misalignment can project a much larger longitudinal Seebeck signal onto transverse leads. Tell: Does the signal have the expected field or magnetization reversal, angular geometry, and repeatable gradient dependence after background separation, or is it simply voltage on transverse contacts?

References¶

[1] Kamran Behnia and Hervé Aubin, “Nernst effect in metals and superconductors: a review of concepts and experiments,” Reports on Progress in Physics 79 (2016), 046502. DOI; open preprint. registry ↩a ↩b ↩c ↩d

[2] Di Xiao, Yugui Yao, Zhong Fang, and Qian Niu, “Berry-Phase Effect in Anomalous Thermoelectric Transport,” Physical Review Letters 97 (2006), 026603. DOI; open preprint. registry ↩a ↩b

[3] Haiyang Yang et al., “Giant anomalous Nernst effect in the magnetic Weyl semimetal \(\mathrm{Co_3Sn_2S_2}\),” Physical Review Materials 4 (2020), 024202. DOI. registry ↩a ↩b ↩c

[4] Ulrike Martens et al., “Anomalous Nernst effect and three-dimensional temperature gradients in magnetic tunnel junctions,” Communications Physics 1 (2018), 65. DOI. registry ↩a ↩b

[5] H. B. Callen, “The Application of Onsager's Reciprocal Relations to Thermoelectric, Thermomagnetic, and Galvanomagnetic Effects,” Physical Review 73 (1948), 1349–1358. DOI. registry ↩a ↩b

[6] N. Doiron-Leyraud et al., “Evidence for a small hole pocket in the Fermi surface of underdoped \(\mathrm{YBa_2Cu_3O_y}\),” Nature Communications 6 (2015), 6034. Its methods give the field-antisymmetrized Nernst extraction and longitudinal-background rejection. DOI. registry ↩