Seebeck coefficient
The Seebeck coefficient, denoted as S, is a fundamental material property in thermoelectrics that quantifies the magnitude and direction of the thermoelectric voltage generated across a material in response to an applied temperature gradient, typically measured in microvolts per kelvin (\mu \mathrm{V/K}).[1] It arises from the diffusion of charge carriers from hotter to cooler regions, driven by the greater average energy of particles at higher temperatures, and is central to the Seebeck effect discovered by physicist Thomas Johann Seebeck in 1821, who observed that joining dissimilar metals at different temperatures produces an electromotive force capable of deflecting a compass needle.[2][3] Mathematically, the Seebeck coefficient is defined as S = -\frac{\Delta V}{\Delta T}, where \Delta V is the open-circuit voltage difference and \Delta T is the temperature difference across the material, with the negative sign reflecting the conventional direction for electron flow in n-type materials.[3] Its value depends on factors such as carrier concentration, effective mass, and temperature; for instance, metals exhibit small coefficients (typically 1–70 \mu \mathrm{V/K}), while heavily doped semiconductors can reach 100–1000 \mu \mathrm{V/K} or higher, with exceptional cases like certain oxide composites exceeding 3000 \mu \mathrm{V/K}.[1] The sign of S indicates the dominant charge carrier: positive for p-type materials where holes prevail, and negative for n-type where electrons dominate, influencing the efficiency of thermoelectric devices.[4] In practice, the Seebeck coefficient is pivotal for applications in temperature measurement and energy harvesting; thermocouples, formed by pairing materials with differing S values (e.g., chromel-alumel yielding ~40 \mu \mathrm{V/^\circ C}), enable precise sensing in industrial and scientific settings.[3] Additionally, high-S materials are essential for thermoelectric generators that convert waste heat to electricity via the Seebeck effect, offering solid-state alternatives to mechanical systems with potential efficiencies tied to the figure of merit ZT = S^2 \sigma / \kappa, where \sigma is electrical conductivity and \kappa is thermal conductivity.[5] Ongoing research focuses on enhancing S through nanostructuring and doping to improve ZT values, advancing sustainable energy technologies.[6]Definition and Fundamentals
Formal Definition
The Seebeck effect is a thermoelectric phenomenon in which a temperature difference applied across a material generates an electromotive force (EMF), producing a voltage without the flow of external current.[7] This arises from the diffusion of charge carriers from hotter to cooler regions due to their thermal excitation, creating a net charge separation and an associated electric field.[8] In the framework of linear irreversible thermodynamics, thermoelectric transport is described by coupled equations for electric current density \mathbf{J} and heat flux \mathbf{Q}. The electric current is given by \mathbf{J} = \sigma \mathbf{E} - \sigma S \nabla T, where \sigma is the electrical conductivity, \mathbf{E} is the electric field, S is the Seebeck coefficient, and \nabla T is the temperature gradient; the heat flux follows \mathbf{Q} = \Pi \mathbf{J} - \kappa \nabla T, with \Pi the Peltier coefficient and \kappa the thermal conductivity.[7] In the open-circuit condition where \mathbf{J} = 0, the equation simplifies to \mathbf{E} = S \nabla T, indicating that the Seebeck coefficient relates the induced electric field to the temperature gradient. The formal definition of the Seebeck coefficient S is thus S = -\frac{\Delta V}{\Delta T} for a finite temperature difference \Delta T and corresponding open-circuit voltage difference \Delta V, or more precisely in the infinitesimal limit, S = -\left( \frac{\partial V}{\partial T} \right)_{J=0}.[8] This definition emerges directly from the transport relations, where the negative sign convention ensures positive values for p-type materials (where holes dominate) and negative values for n-type materials (where electrons dominate).[7] The units of the Seebeck coefficient are volts per kelvin (V/K), though values are typically on the order of microvolts per kelvin (μV/K) for most materials.[8]Sign Convention
The standard sign convention for the Seebeck coefficient in semiconductors assigns a positive value to p-type materials, where the hot end develops a positive voltage relative to the cold end due to the predominance of positive charge carriers (holes) diffusing from hot to cold.[1] Conversely, n-type materials exhibit a negative Seebeck coefficient, as electrons, the dominant negative charge carriers, diffuse from the hot end to the cold end, making the hot end negatively charged.[4] This convention aligns the sign of the Seebeck coefficient with the sign of the majority charge carriers, facilitating consistent interpretation in thermoelectric device design and analysis.[9] In circuit analysis for thermocouples, the sign of the Seebeck coefficient determines the direction of the induced current flow under a temperature gradient. For a thermocouple formed by two materials A and B, if the relative Seebeck coefficient S_{AB} = S_A - S_B > 0, the current flows from material A to B at the cold junction when the other junction is heated, reflecting the polarity established by the differential carrier diffusion.[4] This polarity reversal with sign change is critical for predicting emf direction and ensuring proper wiring in measurement circuits, where misinterpreting the sign could lead to inverted temperature readings.[4] A representative example is the iron-copper junction, where iron exhibits a positive Seebeck coefficient relative to copper (approximately +12.5 μV/K near room temperature, derived from absolute values of +19 μV/K for iron and +6.5 μV/K for copper).[10][2] In this setup, heating the junction results in the iron side becoming positively charged relative to copper, driving current from iron to copper at the cold end.[11] A common pitfall arises from confusing absolute Seebeck coefficients (measured relative to a reference like a superconductor at zero) with relative ones (between two materials), which can lead to sign errors in multi-material systems.[12] For instance, both iron and copper have positive absolute coefficients, but their relative sign depends on the difference, potentially inverting expected polarities if not accounted for in experimental setups.[12]Relation to Other Thermoelectric Coefficients
The Seebeck coefficient S is interconnected with the Peltier coefficient \Pi and the Thomson coefficient \tau through the Kelvin relations, which provide thermodynamic linkages between these thermoelectric transport parameters. The first Kelvin relation states that the Peltier coefficient is the product of the Seebeck coefficient and the absolute temperature: \Pi = S T. The second Kelvin relation connects the temperature derivative of the Seebeck coefficient to the Thomson coefficient: \tau = -T \frac{dS}{dT}. These relations ensure consistency across the thermoelectric effects by linking heat and charge transport phenomena.[13] Physically, the Seebeck, Peltier, and Thomson coefficients all originate from the entropy transported by charge carriers in a material under a temperature gradient or current flow. In this framework, the Seebeck coefficient represents the entropy per unit charge carried by the charge carriers, while the Peltier coefficient describes the heat absorbed or released at a junction due to this entropy flow, and the Thomson coefficient accounts for the reversible heat production within a temperature gradient. This unified interpretation underscores that the coefficients are manifestations of the same underlying carrier entropy transport mechanism.[14] The interconnections are further grounded in Onsager reciprocity, a principle from nonequilibrium thermodynamics that imposes symmetry on the transport coefficient matrix. For thermoelectric effects, this reciprocity yields S_{ij} = -S_{ji}, where S_{ij} is the Seebeck coefficient relating the electric field in direction i to the temperature gradient in direction j. This antisymmetry ensures the thermodynamic consistency of the relations between Seebeck, Peltier, and Thomson coefficients in isotropic and anisotropic materials.[15] These Kelvin relations were derived by William Thomson (later Lord Kelvin) in 1854, building on the earlier observations of Seebeck and Peltier to establish a thermodynamic foundation for thermoelectricity.[16]Historical Development
Discovery by Seebeck
In 1821, Thomas Johann Seebeck, a Baltic German physicist, conducted pioneering experiments on the interaction between heat and electrical conduction in metals. He formed a closed circuit by joining wires of dissimilar metals, specifically bismuth and copper, creating two junctions. Upon heating one junction while keeping the other at a lower temperature, Seebeck observed a significant deflection of a nearby compass needle, which he attributed to the generation of magnetism induced by the temperature difference. This observation marked the initial discovery of what would later be known as the thermoelectric effect, though Seebeck misinterpreted it as a form of thermomagnetism akin to the recently discovered electromagnetism by Hans Christian Ørsted.[17][18] Seebeck detailed his findings in a series of papers published between 1822 and 1823, titled Magnetische Polarisation der Metalle und Erze durch Temperatur-Differenz, presented to the Prussian Academy of Sciences and appearing in their Abhandlungen der physikalischen Klasse. In these works, he described extensive tests with various metal combinations, noting the consistent compass deflections and proposing that temperature gradients could polarize metals magnetically. These publications laid the groundwork for understanding the phenomenon, emphasizing its reproducibility across different materials and temperature ranges, though still framed within a magnetic paradigm. The practical implications of Seebeck's observations quickly emerged in the development of the thermopile during the 1820s. Inspired by Seebeck's observations, Italian physicist Leopoldo Nobili pioneered this device in the 1820s, often in collaboration with Macedonio Melloni, by connecting multiple thermocouples in series to amplify the voltage generated by temperature differences. The thermopile enabled sensitive detection of thermal radiation and infrared heat, serving as an early tool for radiometry and pyrometry before the advent of more advanced sensors.[19][20] Seebeck's magnetic interpretation was soon corrected through further experimentation. Shortly thereafter, Danish physicist Hans Christian Ørsted recognized that the compass deflection resulted from an electric current induced by the temperature difference, rather than direct thermomagnetism, and termed the phenomenon "thermoelectricity".[2] This clarification shifted the focus from thermomagnetism to the direct conversion of heat into electricity, paving the way for subsequent theoretical and applied advancements.Evolution of the Concept
Following Thomas Johann Seebeck's discovery of the thermoelectric effect in 1821, the conceptual understanding of the Seebeck coefficient evolved rapidly in the mid-19th century through thermodynamic linkages to related phenomena. In the 1850s, William Thomson, later Lord Kelvin, established reciprocal relations that connected the Seebeck effect to the Peltier and Thomson effects, providing a unified thermodynamic framework for thermoelectricity. These relations demonstrated that the Peltier coefficient is proportional to the product of the Seebeck coefficient and absolute temperature, while the Thomson coefficient relates to the temperature derivative of the Seebeck coefficient, enabling predictions of heat absorption or release under current flow in temperature gradients.[17] This work shifted the Seebeck coefficient from an empirical observation to a thermodynamically grounded quantity essential for analyzing irreversible processes in conductors. By the early 20th century, the Seebeck coefficient gained prominence in the emerging field of solid-state physics, where classical models were refined through quantum mechanical interpretations. Arnold Sommerfeld's application of quantum statistics to electron transport in metals during the late 1920s provided an initial quantum framework, explaining the Seebeck coefficient as arising from the asymmetric scattering of electrons near the Fermi level. This was further advanced in 1931 by Lars Onsager's reciprocal relations for non-equilibrium transport, which generalized the Seebeck coefficient's role in anisotropic media and laid the groundwork for understanding coupled heat and charge flows beyond simple metals. These developments integrated the Seebeck coefficient into band theory, highlighting its dependence on electronic density of states and carrier diffusion, thus bridging classical thermodynamics with quantum solid-state phenomena. A major milestone occurred in the mid-20th century with the recognition of the Seebeck coefficient's potential in semiconductors for thermoelectric energy conversion. In the 1950s, Abram Ioffe and collaborators emphasized semiconductors' large Seebeck coefficients—often orders of magnitude higher than in metals—due to their tunable carrier concentrations and band structures, spurring research into materials like bismuth telluride (Bi₂Te₃). This era saw experimental verification that doping could optimize the Seebeck coefficient alongside electrical conductivity, boosting interest in practical thermoelectrics and establishing it as a key parameter in the figure of merit ZT.[21] Frank J. Blatt's studies further clarified how band-edge positions and scattering mechanisms influence the coefficient in silicon and germanium, solidifying semiconductors as superior to metals for thermoelectric applications.[22] From the late 20th century to 2025, the concept has evolved through integration with nanotechnology, where low-dimensional systems exhibit enhanced Seebeck coefficients due to quantum confinement effects. Seminal theoretical work by L. D. Hicks and M. S. Dresselhaus in 1993 predicted that quantum wells and superlattices could increase the Seebeck coefficient by sharpening the density of states, reducing thermal conductivity while preserving electrical transport.[23] This spurred experimental advances, such as silicon nanowires demonstrating up to 100% enhancement in Seebeck coefficient compared to bulk silicon, attributed to boundary scattering and one-dimensional confinement. Similarly, quantum dots in Ge/Si structures have shown Seebeck enhancements of 40% or more through energy filtering of low-energy carriers, with ongoing research up to 2025 exploring hybrid nanowire-dot arrays for further optimization in nanoscale thermoelectrics.[24] These developments have redefined the Seebeck coefficient as a tunable parameter in confined geometries, extending its theoretical scope beyond bulk materials.Measurement Methods
Relative Seebeck Coefficient
The relative Seebeck coefficient, denoted as S_{AB}, quantifies the thermoelectric voltage generated by the difference in Seebeck coefficients between two dissimilar materials, A and B, in a thermocouple configuration. This setup involves forming a closed loop with the two materials joined at two junctions maintained at different temperatures, creating a temperature gradient along their lengths. When a temperature difference \Delta T is applied, an electromotive force \Delta V is induced, such that S_{AB} = \frac{\Delta V}{\Delta T} = S_A - S_B, where S_A and S_B are the absolute Seebeck coefficients of the respective materials.[25] This differential measurement offers key advantages over absolute methods, as it eliminates the requirement for a universal reference material or precise absolute temperature scale, relying instead on the relative response between the paired conductors. It is the foundational principle for thermocouple thermometry, enabling reliable temperature sensing in industrial and scientific applications without complex calibration against primary standards. The sign of the generated \Delta V follows the convention where a positive voltage indicates the hot junction drives current from material A to B in the circuit.[26][27] In practice, the measurement procedure establishes a controlled linear temperature gradient across the sample using heaters and heat sinks, with voltage probes attached at points along the materials where local temperatures are monitored via auxiliary thermocouples or sensors. The relative Seebeck coefficient at a target temperature is then determined by plotting the measured voltages against the corresponding temperature differences and extrapolating to the isotherm (where \Delta T = 0) to isolate the intrinsic response at uniform temperature. Potential error sources include contact resistance at the probe-material interfaces, which can cause voltage offsets, as well as deviations from linearity in the temperature profile due to thermal conduction losses or radiative heat transfer.[28][29] Historically, relative Seebeck coefficient measurements have been central to the standardization of thermocouple types, such as Type K (chromel-alumel) and Type J (iron-constantan), which were developed and formalized in the mid-20th century through efforts by the Instrument Society of America (now part of ISA) and adopted into ASTM and ANSI standards for consistent performance across applications.[30][31]Absolute Seebeck Coefficient
The absolute Seebeck coefficient establishes a universal scale for the thermoelectric response of a material by referencing it to a standard where the Seebeck coefficient is zero, such as a superconductor in its superconducting state at temperatures approaching 0 K, where perfect conductors exhibit S = 0 due to the absence of entropy transport by charge carriers.[32] This definition allows the direct determination of the absolute value, S_absolute, from the measured thermoelectric voltage gradient across the sample when connected via superconducting leads that contribute no voltage.[33] Key techniques for measuring the absolute Seebeck coefficient rely on cryogenic conditions to utilize superconducting references below 1 K, where the sample and leads are immersed in a controlled low-temperature environment to create a precise temperature gradient.[32] Alternative approaches include immersion in variable-temperature liquid helium baths, which adjust the reference temperature incrementally, or adiabatic demagnetization of paramagnetic materials to reach millikelvin ranges without continuous cooling.[33] These methods contrast with relative measurements, which serve as a prerequisite for extending absolute values to higher temperatures via integration.[33] Challenges in these measurements stem from the need for specialized cryogenic setups, such as dilution refrigerators or helium cryostats, to maintain stable temperatures and minimize thermal leaks, ensuring reliable voltage detection with nanovolt sensitivity.[32] Achieving high accuracy, typically on the order of 0.1 μV/K or better, requires careful control of contact resistances, lead contributions, and thermal gradients, with uncertainty analyses often revealing contributions from finite temperature resolutions and material inhomogeneities.[32] Advancements in the 1970s and 1980s included standardization efforts by the National Institute of Standards and Technology (NIST), which developed reference tables and calibration protocols for thermocouple materials, laying the groundwork for absolute scales through low-temperature benchmarks.[34] More recent progress, extending into the 2020s, encompasses non-cryogenic optical techniques, such as laser-heating methods that enable absolute Seebeck determinations at elevated temperatures up to several hundred kelvin by modulating thermal gradients without physical contacts.[35] These innovations, including graphene-based transistor references for room-temperature absolutes, reduce reliance on cryogenics while preserving precision.[36]Material Properties
Values for Common Metals
The Seebeck coefficient for metals is generally small in magnitude, typically on the order of a few microvolts per kelvin (μV/K), reflecting the dominance of free electron diffusion with minimal contributions from other mechanisms. At room temperature (around 300 K), pure metals exhibit values that are often positive for transition metals and negative for noble metals, with a nearly linear dependence on temperature due to the Fermi gas model of electron transport. These values are compiled from standardized measurements and are crucial for applications in precision thermometry and alloy design.| Metal/Alloy | Seebeck Coefficient (μV/K at 300 K) | Notes |
|---|---|---|
| Copper (Cu) | +1.8 | Pure annealed; increases slightly with temperature. [1] |
| Iron (Fe) | +15.0 | Alpha phase; shows stronger temperature dependence due to magnetic effects. [1] |
| Nickel (Ni) | -15.0 | Negative sign indicates electron-like carriers; sensitive to impurities. [1] |
| Aluminum (Al) | -1.5 | Low value typical of free-electron metals; linear up to 500 K. [1] |
| Platinum (Pt) | -5.0 | Standard reference; stable and linear over wide temperature range. [1] |
| Gold (Au) | -1.8 | Similar to copper; used in high-purity thermocouples. [1] |
| Constantan (Cu-Ni alloy) | -35.0 | Engineered for low thermal expansion; nearly temperature-independent. [1] |
| Type S Thermocouple (Pt-10%Rh vs. Pt) | +6.4 to +11.0 | Differential value; increases with Rh content for higher sensitivity. [37] |
Values for Semiconductors and Other Materials
Semiconductors exhibit significantly higher Seebeck coefficients than metals due to the presence of a band gap, which breaks the symmetry between electrons and holes, leading to an asymmetric density of states near the Fermi level that enhances the thermoelectric response.[39] This asymmetry results in Seebeck coefficients typically two orders of magnitude larger in semiconductors compared to metals, making them preferable for thermoelectric applications where a large voltage response to temperature gradients is required.[40] In semiconductors, the sign of the Seebeck coefficient indicates the dominant charge carrier type: positive values for p-type materials (hole conduction) and negative values for n-type materials (electron conduction).[41] The magnitude depends strongly on doping level, with lower carrier concentrations yielding higher absolute values due to increased entropy per carrier, while higher doping reduces it by bringing the Fermi level closer to the band edge. Temperature dependence often shows a peak in the Seebeck coefficient near temperatures related to the band gap energy, where thermal excitation balances diffusive and drag contributions, followed by a decrease at higher temperatures due to increased carrier excitation across the gap.[42] Representative examples illustrate these trends across various semiconductor classes. Bismuth telluride (Bi₂Te₃), a benchmark thermoelectric material, achieves a Seebeck coefficient of approximately 200 μV/K for p-type variants at room temperature, benefiting from its narrow band gap and high carrier mobility.[43] Silicon at low doping levels (around 10¹⁸ cm⁻³) can reach up to 660 μV/K, highlighting the potential for enhanced values in lightly doped intrinsic-like regimes, though practical applications often involve higher doping for balanced conductivity.[44] Organic semiconductors like poly(3,4-ethylenedioxythiophene) polystyrene sulfonate (PEDOT:PSS) typically exhibit lower values around 50 μV/K, tunable via post-treatments or additives to improve power factors in flexible devices.[45] Emerging materials further expand the range of high Seebeck coefficients. Halide perovskites, such as methylammonium lead iodide (MAPbI₃), demonstrate exceptional values up to 920 μV/K, attributed to their soft lattice and defect-tolerant band structures that minimize thermal conductivity while maintaining decent electrical transport.[46] Two-dimensional materials like graphene show Seebeck coefficients up to 100 μV/K in doped or nanostructured forms, leveraging Dirac cone band structures for tunable transport, though pristine graphene remains limited by its semimetallic nature.[47] Half-Heusler alloys, such as ZrNiSn-based compounds, achieve around 150-165 μV/K for n-type compositions at intermediate temperatures, owing to their robust half-filled band gaps and resistance to thermal degradation.[48] Exceptional cases include certain oxide composites, which can exceed 3000 μV/K due to enhanced phonon drag and low carrier concentrations.[1]| Material | Type | Seebeck Coefficient (μV/K at ~300 K) | Key Notes |
|---|---|---|---|
| Bi₂Te₃ | p-type | ~200 | Narrow band gap enhances near-room-temperature performance.[43] |
| Si | Low n- or p-doping | ~660-1000 | Peaks at low carrier concentrations; decreases with doping.[44][49] |
| PEDOT:PSS | p-type | ~50 | Organic; tunable via solvents or additives for flexibility.[45] |
| MAPbI₃ (perovskite) | n- or p-type | ~920 | High due to defect states; solution-processable.[46] |
| Graphene (doped) | Ambipolar | Up to ~100 | 2D Dirac bands; substrate effects influence value.[47] |
| Half-Heusler (e.g., ZrNiSn) | n-type | ~150-165 | Stable at high temperatures; band engineering optimizes.[48] |