Resistance vs. Temperature theory

The temperature dependence of electrical resistance in pure metals is fundamentally explained by the classical Drude-Lorentz model and quantum electron-phonon scattering theory. In a metallic crystal lattice such as copper (face-centered cubic), valence electrons are delocalized into a free electron Fermi gas with high electron density n ≈ 8.5 × 10²⁸ m⁻³. According to the Drude conductivity equation, electrical resistivity is given by ρ = m_e / (n e² τ), where m_e is the electron effective mass, e is elementary charge, and τ is the mean collision time (relaxation time) between scattering events. At temperatures well above the Debye temperature (for copper, Θ_D = 343 K ≈ 70 °C), the dominant scattering mechanism for conduction electrons is not collisions with other electrons, but collisions with quantized thermal lattice vibrations known as phonons. As temperature T increases, the thermal energy of the lattice ions increases according to equipartition, causing the mean-square vibrational displacement of copper nuclei ⟨u²⟩ to increase in direct proportion to absolute temperature T: ⟨u²⟩ ∝ T. Because the scattering cross-section σ_phonon presented by vibrating ion cores is proportional to ⟨u²⟩, the electron scattering rate 1/τ increases linearly with temperature (1/τ ∝ T), causing electrical resistivity to rise linearly with temperature: ρ(T) ∝ T.

Over moderate temperature intervals (such as 0 °C to 100 °C), the variation of electrical resistance for most pure metals is exceptionally well represented by a first-order Taylor expansion: R(T) = R₀ [1 + α₀ (T − T₀)], where R₀ is the baseline resistance at reference temperature T₀ (conventionally chosen as 0.0 °C), T is the operating temperature, and α₀ is the fractional temperature coefficient of resistance referenced to 0 °C. Rearranging this relationship into standard slope-intercept form gives R(T) = (R₀ α₀) · T + R₀, which represents a straight line y = m · x + b with independent variable x = T (in °C), dependent variable y = R (in Ω), slope m = R₀ α₀ (in Ω °C⁻¹), and vertical intercept b = R₀ (in Ω). The temperature coefficient α₀ is defined physically as the fractional change in resistance per unit temperature rise: α₀ = (1 / R₀) · (dR / dT). From the linear fit coefficients, α₀ is extracted directly as the ratio of slope to vertical intercept: α₀ = m / b. For commercial annealed standard copper (IACS), α₀ at 0 °C is 0.00427 °C⁻¹, while referenced to 20.0 °C room temperature, α₂₀ is 0.00393 °C⁻¹.

If the linear resistance characteristic R(T) = R₀(1 + α₀ T) is extrapolated downward to the temperature where theoretical resistance would drop to zero (R = 0), we find 0 = R₀(1 + α₀ T_abs) ⟹ T_abs = −1 / α₀ = −b / m. For pure electrolytic copper with α₀ ≈ 0.00427 °C⁻¹ (or α₂₀ = 0.00393 °C⁻¹ referenced to 0 °C line fit), this linear extrapolation points to an inferred absolute zero of approximately −234.5 °C to −254.5 °C. In the historical development of 19th-century thermodynamics and electrical science, this extrapolation provided early empirical evidence that thermal motion governs electrical resistance and corroborated the thermodynamic Kelvin scale absolute zero (−273.15 °C). However, real metals deviate from this simple linear extrapolation at cryogenic temperatures due to Matthiessen's rule: as thermal phonon vibrations freeze out (ρ_phonon ∝ T⁵ at low T via the Bloch-Grüneisen law), the resistance levels off to a non-zero constant known as the residual resistance ratio (RRR), governed strictly by static crystal lattice defects and chemical impurities.

Measuring small resistance changes (such as ΔR ≈ 0.39 Ω per °C in a 100 Ω coil) requires rigorous elimination of systematic lead resistance errors. In a conventional two-terminal ohmmeter or standard Wheatstone bridge, the measured resistance includes not only the coil winding R_coil, but also the resistance of the connecting test leads R_lead and the contact resistance of the terminal binding posts R_contact: R_meas = R_coil + 2R_lead + 2R_contact. Because ordinary copper hookup wires possess resistances between 0.05 Ω and 0.20 Ω, two-wire measurements introduce a systematic positive zero-offset of up to 0.4 Ω (equivalent to a 1 °C temperature error). To eliminate this artifact, precision laboratories utilize a Kelvin 4-wire sensing configuration (or Callendar-Griffiths bridge with compensating dummy leads). In the Kelvin technique, two outer leads (Current Force leads) supply a stable, low excitation current, while two inner leads (Voltage Sense leads) connect directly to the coil winding to sense the potential drop across high-impedance inputs. Because virtually zero current flows through the sensing leads, no IR drop occurs along them, ensuring that the measured voltage represents solely the true coil winding resistance.

The linear dependence of electrical resistance on temperature forms the operational basis of industrial Resistance Temperature Detectors (RTDs), the global standard for precision thermometry (IEC 60751 / ITS-90). While copper is cost-effective and highly linear between 0 °C and 100 °C, its low chemical oxidation threshold limits its use above 150 °C. Consequently, primary metrology and industrial standards employ platinum resistance thermometers (such as the standard PT100 sensor, having R₀ = 100.00 Ω at 0.0 °C and α = 0.003851 °C⁻¹). Over broader temperature ranges (−200 °C to +850 °C), subtle non-linear lattice effects require the Callendar-Van Dusen equation: R(T) = R₀ [1 + A T + B T² + C (T − 100) T³], where the quadratic coefficient B ≈ −5.775 × 10⁻⁷ °C⁻² accounts for slight downward curvature caused by higher-order electron scattering and lattice anharmonicity at elevated temperatures. In the 0 °C to 100 °C range examined in this laboratory, the quadratic term contributes less than 0.06% deviation, confirming the exceptional validity of the linear model.

While pure metals exhibit large positive temperature coefficients (α ≈ +0.004 °C⁻¹), electrical engineering frequently requires standard resistors and instrument shunts whose resistance remains completely invariant during temperature fluctuations. This is achieved by creating specialized solid-solution alloys such as Constantan (55% Cu, 45% Ni) and Manganin (84% Cu, 12% Mn, 4% Ni). According to Nordheim's rule and quantum scattering theory, introducing a high concentration of foreign solute atoms into a solvent crystal lattice creates severe localized electrostatic potential perturbations that shatter lattice periodicity. This static impurity scattering (ρ_impurity) becomes so massive that it completely dominates the total resistivity (ρ_impurity ≫ ρ_phonon). Consequently, thermal phonon vibrations add only an insignificant fractional change to the total scattering rate, reducing the temperature coefficient of resistance by more than two orders of magnitude to α ≈ ±0.00001 °C⁻¹ (±10 ppm/°C). Manganin is intentionally engineered to have its resistance peak precisely at room temperature (20 °C to 25 °C), ensuring virtually zero drift in precision resistance boxes.

Two critical experimental systematic errors must be mitigated when evaluating temperature coefficients: Joule self-heating and thermal lag. Joule self-heating occurs because the measuring current I passed through the coil by the bridge generates internal heat at rate P = I² R. If the excitation current is excessive (e.g. I > 10 mA in a 100 Ω coil, dissipating > 10 mW), the internal temperature of the fine copper wire will rise significantly above the surrounding oil bath temperature, creating a systematic positive resistance offset. To prevent self-heating, measuring currents are kept strictly below 1.0 mA (dissipating < 0.1 mW, corresponding to self-heating < 0.002 °C). Thermal lag occurs because the protective brass casing, electrical insulation varnish, and ceramic former possess finite thermal resistance and heat capacity. If the bath is heated too rapidly, the internal coil temperature lags behind the external mercury thermometer. This lag is detected by taking readings during both heating and cooling cycles; any separation between the heating and cooling curves indicates thermal hysteresis caused by premature measurement before steady state.