Materials Science · Electrical transport and electronic materials

Superconductivity and its operating conditions

Useful background: Atoms, bonding, structure and defects; Measuring and checking material claims. Each worked case explains the quantities it uses.

Back to the chapter

After reading, you can

State superconductivity through temperature, magnetic field and current constraints.
Distinguish a transport resistance bound from magnetic evidence for a superconducting state.
Explain Meissner flux expulsion, the type-II mixed state and why vortex motion can produce a voltage.
Use a supplied temperature–field current map and a stated electric-field criterion within their given ranges.
Account separately for conductor, joint and refrigeration contributions without turning a specimen result into a loss-free cable claim.

A special state has an operating region

A superconductor can exhibit zero DC resistance and characteristic magnetic behaviour in an appropriate state. Its useful region is constrained by temperature, magnetic field and current, with microstructure and geometry affecting performance. A transition temperature alone does not specify a practical conductor.

A wire’s voltage becomes too small to resolve when it is cooled. That is useful transport evidence, but how small is the remaining resistance, and what magnetic state was observed? Even after a superconducting state is established, a current that meets one criterion at one temperature and field may fail after either condition changes. Start with those separate questions rather than treating a transition temperature as a cable rating.

In conventional superconductors, correlated pairs of electrons, called Cooper pairs, participate in a collective quantum state. Its response differs from a normal metal with merely fewer collisions. The normal lattice, defects and possible excitations have not vanished. This lesson develops the measurable behaviour and its limits; it does not derive a microscopic pairing theory or assume that every superconductor has the same pairing mechanism. Charge carriers and electronic energy bands supplies the earlier electronic-state context.

A resistance bound and flux expulsion answer different questions

For steady positive current I and gauge voltage drop V, an ohmic interpretation gives R = V/I. If a validated observation only bounds the true magnitude of the gauge drop by |V| < V_lim, then it bounds |R| < V_lim/I. An instrument displaying zero is not a proof of an exactly zero physical quantity. Specify offsets, sensitivity, current path, gauge length and contact contribution. Conductors, insulators and semiconductors develops the required geometry and voltage definitions.

For example, stipulate a corrected steady gauge measurement with I = 2.00 A and a valid bound |V| < 20.0 nV. Then |R| < 10.0 nΩ. If the gauge is uniform, length L = 20.0 mm and section A = 0.0200 mm², the corresponding bulk-resistivity bound is |ρₑ| = |R|A/L < 1.00 × 10⁻¹⁴ Ω m. This is an invented conditional measurement record, not a new material result or a complete uncertainty assessment. A lower voltage limit tightens the bound; it does not alone identify the state.

The Meissner effect supplies a different property. In an appropriate equilibrium Meissner state, screening currents near the surface exclude magnetic induction from the interior of a sufficiently large, simply connected specimen, apart from a finite penetration layer. B denotes magnetic induction, measured in tesla. The penetration depth is a material- and temperature-dependent length; it is not zero. Sample shape and surface barriers can affect the observed magnetic response.

Why is this more than perfect conduction? Imagine a solid long cylinder in a weak steady field along its axis while it is normal. The field initially penetrates it. Magnetic flux measures the field passing through a chosen surface. In the classical idealisation where only resistivity becomes zero, induced currents oppose later changes of flux; the field already inside need not disappear. The equilibrium Meissner state instead expels that initially present bulk field on entering the state. Maintaining an existing flux and expelling it are different predictions. The diagram assumes a solid cylinder much wider than its penetration depth, with negligible end effects and a field below the relevant Meissner limit.

A real state assessment compares transport, magnetic response and specimen identity under consistent conditions. An extremely conductive path can dominate a voltage measurement while occupying a small part of the object. Magnetic signals also need background, shape, volume and history interpretation; a shielding signal alone is not a universal phase-fraction meter. A resistance drop and an unqualified photograph of levitation do not replace that evidence.

Field penetration need not mean the whole specimen is normal

The simple type-I equilibrium picture has a Meissner region below a critical field at the stated temperature and a normal state above it. Finite shapes can introduce intermediate arrangements, so a single ideal curve is not a complete shape-independent testing rule. Increasing temperature towards the transition generally reduces the available field range.

Type-II materials have a lower critical field and an upper critical field. Between them, bulk superconductivity can coexist with quantised magnetic vortices: narrow regions of suppressed superconductivity carrying flux, surrounded by circulating superconducting currents. This mixed state is different from both complete bulk flux exclusion and the normal state. Here “upper critical field” refers to the bulk transition; possible special surface superconductivity is outside this core description.

A transport current can exert a force on vortices. If they move across the specimen, a voltage and dissipation can result even while superconducting regions remain. Defects and inclusions can impede that motion, a process called pinning. Pinning can improve a useful low-voltage current range, while boundaries or damage can also interrupt current paths. More defects therefore do not imply universally better superconductivity.

Current capacity depends on temperature, field magnitude and orientation, strain, microstructure and connections. A finite-voltage current criterion is not the same as the thermodynamic boundary that removes superconducting order. Trapped flux and magnetic hysteresis in a pinned type-II specimen also mean that absence of complete field expulsion after one history does not by itself disprove superconductivity. State the regime and history rather than applying the ideal Meissner cartoon to every observation.

Original schematic cross-sections of solid long cylinders in a weak axial field. Dotted circles denote field out of the page; their count is an illustration, not a measured flux or vortex count. The upper two initially normal cylinders both contain the field. After the left hypothetical material only becomes perfectly conducting, the previously present bulk field can remain. In the right equilibrium Meissner comparison, the bulk field is expelled and a clockwise surface screening current opposes the applied outward field. The cylinder radius is assumed much greater than its finite penetration depth, with negligible end effects. A violet surface layer is exaggerated and is not a measured length. This right panel is below the relevant Meissner field limit and is not a cartoon of all type-II histories. The separate lower strip depicts a type-II mixed state: three schematic vortex cores with suppressed superconductivity and circulating currents, rather than complete bulk field exclusion or a wholly normal specimen. Vortex dimensions, separation and currents are not to scale; pinning can impede vortex motion, and current-driven motion can cause voltage. No actual material or experimental observation is represented.

Read the supplied current map at the actual temperature and field

Use a fictional conductor C whose superconducting state has already been independently established. Its zero-applied-field transition is T_c = 12.0 K. The table supplies criterion currents I_c at fixed specimen geometry, pressure, strain and field orientation. B_a is the externally applied field magnitude in tesla. The field generated by transport current is implicitly included for this geometry in the supplied response; B_a is not the internal field everywhere. These records are not a universal critical surface or a safe operating rating.

The map’s I_c is defined by a steady electric-field criterion E_c = 1.00 µV/cm = 1.00 × 10⁻⁴ V/m along the gauge. At gauge length L = 0.100 m, this is V_c = E_c L = 10.0 µV. It is a specified small finite voltage, not exactly zero. Changing the criterion or gauge interpretation changes the reported current; using the same raw voltage criterion at another L would change E_c.

Only within 4.00 ≤ T ≤ 8.00 K and 1.00 ≤ B_a ≤ 3.00 T, stipulate interpolation by I_c(T,B_a) = 200 A [1 − (T − 4 K)/(8 K)] [1 − (B_a − 1 T)/(5 T)]. Each bracket is dimensionless. This product reproduces the four corner records and provides the chosen interpolation between them; four measurements alone would not establish its validity in a real conductor. Do not extend it to T_c or another field orientation.

For one declared comparison, require operating current I ≤ 0.800 I_c. The factor 0.800 is an invented screening margin, not a failure probability, reliability guarantee or standard. At 6.00 K and 2.00 T, the factors are 0.750 and 0.800, giving I_c = 120 A and screen limit 96.0 A. At 8.00 K and 1.00 T, I_c = 100 A and the screen limit is 80.0 A. A 96.0 A current therefore meets the first screen but misses the second, although both temperatures are below 12 K and 96 A is still below the second criterion current.

For fixed I = 96.0 A, the screen requires I_c ≥ 120 A. In the stated rectangle, its boundary is B_a = 1 T + 5 T {1 − 0.600/[1 − (T − 4 K)/(8 K)]}. It joins 4 K, 3 T to 7.20 K, 1 T. Allowed points lie at the lower-field side, including the boundary. At higher T within the rectangle, no B_a ≥ 1 T meets this chosen screen. This is a current-margin map, not a Meissner/normal phase diagram.

Invented criterion-current records for the same conductor and fixed field orientation, strain and geometry. E_c = 1.00 µV/cm; interpolation is separately stipulated.
Temperature T	I_c at B_a = 1.00 T	I_c at B_a = 3.00 T
4.00 K	200 A	120 A
8.00 K	100 A	60.0 A

A current criterion does not eliminate the voltage curve

For conductor C at each condition in the map, additionally supply a local empirical relation E = E_c(I/I_c)²⁰ over 0.600 ≤ I/I_c ≤ 1.10 for positive steady current, stable temperature and the same gauge. It describes a finite dissipative response near the criterion. It is not the ideal zero-resistance law, a microscopic pairing calculation or a model outside that stated current interval. In a real sample, noise, drift, transfer lengths, vortex dynamics and heating can limit such a fit.

At 6 K and 2 T, I_c = 120 A. With I = 96.0 A, I/I_c = 0.800, so E = 1.15292 × 10⁻⁶ V/m and V = EL = 0.115292 µV across 0.100 m. The corresponding gauge power is P_gauge = IV = 11.0680 µW. A small voltage is still finite; passing the chosen 80% current screen does not mean this empirical curve has zero loss.

The same curve also shows criterion dependence. If E_c is changed while the underlying E–I relation and conditions remain unchanged, solve E_new/E_c = (I_new/I_c)²⁰. A twofold change gives I_c,new = 120 × 2^(1/20) = 124.232 A, ratio 1.03526, inside the supplied range. A tenfold change would require I_c,new = 134.642 A, ratio 1.12202, beyond its upper limit 1.10; that prediction is therefore not validated by this local fit. A different reporting convention has not improved the material.

An ordinary low-resistance ohmic conductor can also cross a finite E_c: E = RI/L gives I_cross = E_c L/R. For R = 1.00 µΩ and L = 0.100 m, it crosses 1.00 µV/cm at 10.0 A. This mathematical crossing is not evidence that it is superconducting. The curve’s shape and independent magnetic/material observations matter. A criterion current is useful only with its definition and physical context.

Original invented conductor map, not a universal phase diagram. The upper axes show temperature four to eight kelvin and applied field one to three tesla, with fixed geometry, strain and field orientation. Current is fixed at 96 amperes and the chosen screen is I less than or equal to 0.800 times the interpolated criterion current. The hatched region meets this screen. Its boundary runs from four kelvin and three tesla through point A at six kelvin and two tesla to 7.20 kelvin and one tesla. Equality is included. Point B at eight kelvin and one tesla misses the screen even though its criterion current 100 amperes exceeds 96 amperes. The lower graph shows the separately stipulated local relation E/E_c=(I/I_c) to power 20 only for current ratios 0.600 to 1.10. Its vertical axis is logarithmic: equal intervals multiply electric field by ten. Point A at ratio 0.800 has E/E_c=0.0115292; B at 0.960 has 0.442002. The dashed horizontal line at one is the original finite criterion. Doubling that criterion moves the crossing to ratio 1.03526 on the same curve, not a better material. The map is an explicitly supplied interpolation; its criterion, local fit, operating screen and superconducting phase boundary are distinct quantities.

The complete system still dissipates energy

Joints, AC effects, flux motion and refrigeration can contribute losses or operating costs. Distinguish a material measurement from a complete magnet or cable. Extraordinary transition claims require magnetic, transport and structural evidence appropriate to the claimed state.

Separate the tested gauge from the assembled system. A series joint of resistance R_joint produces P_joint = I²R_joint. In the invented 96.0 A case, total joint resistance 2.00 µΩ gives 18.4320 mW, much larger than the model’s 11.0680 µW gauge loss. Other normal-metal paths, leads and connections need their own accounts.

Cooling removes heat at the cold temperature; a refrigerator needs input work to do that. Supply a partial cold-stage budget of 0.400 W background heat leak, the two gauge/joint terms above, and no other included cold loads. The sum is 0.418443 W. If the supplied refrigerator requires 250 W of electrical input per watt removed at that operating point, the required input for this budget is 104.611 W. That factor is a given device ratio, not a thermodynamic constant or a complete cable efficiency.

Changing currents or fields can produce additional losses, including vortex hysteresis and losses in surrounding normal metal. A local transition into a dissipative region can increase heating and spread a loss of the intended superconducting operating state, often called a quench. Stability and protection require thermal, electrical and mechanical evidence beyond this steady model. This comparison supplies no prediction of recovery after a disturbance.

A persuasive conductor or cable claim therefore retains the specimen state, continuity, electric-field criterion, temperature, field orientation, strain, current and relevant duration. It also measures the losses and cooling appropriate to the complete assembly and waveform. From measured transport to device behaviour continues that material-to-system distinction; a finite specimen observation alone supplies neither a loss-free system nor safe-service qualification.

Worked examples

Work through the case

A sample shows a resistance drop below a stated temperature. Does that single observation certify a loss-free power cable?

Check whether the superconducting state is established.
Assess current and field limits and conductor continuity.
Include joints, AC operation and cooling.

No. A component needs substantially broader evidence.

Translate an unresolved gauge drop into a bounded statement

A corrected steady gauge record at I = 2.00 A establishes a true-drop bound |V| < 20.0 nV. With uniform gauge L = 20.0 mm and A = 0.0200 mm², find the resistance and resistivity bounds. Has exact zero resistance or a bulk superconducting state been established?

Convert 20.0 nV = 20.0 × 10⁻⁹ V. Divide by 2.00 A to give |R| < 1.00 × 10⁻⁸ Ω = 10.0 nΩ.
Convert A = 2.00 × 10⁻⁸ m² and L = 0.0200 m. Then |ρₑ| < (1.00 × 10⁻⁸)(2.00 × 10⁻⁸)/0.0200 = 1.00 × 10⁻¹⁴ Ω m under the supplied uniform-gauge assumptions.
The bound depends on valid correction/sensitivity and the defined current path. It has not resolved exactly zero or supplied magnetic phase evidence. Repeating a displayed zero does not remove an unrecognised common offset.

Report |R| < 10.0 nΩ and conditional |ρₑ| < 1.00 × 10⁻¹⁴ Ω m. State what the method resolved and seek complementary magnetic and specimen evidence.

Apply the map before interpreting the voltage

For fictional conductor C, compare I = 96.0 A at point A: 6.00 K, 2.00 T and point B: 8.00 K, 1.00 T. Use I ≤ 0.800 I_c and the stated local E–I relation. Which screen passes? Find each gauge voltage for L = 0.100 m.

At A, I_c = 200(0.750)(0.800) = 120 A; its screen limit is 96.0 A, so equality is permitted and the screen passes.
At B, I_c = 200(0.500)(1.00) = 100 A; its screen limit is 80.0 A, so 96.0 A misses it even though T < T_c and I < I_c. Missing a declared margin alone is not proof that the entire specimen is normal.
The current ratios are 0.800 and 0.960, both inside the supplied [0.600,1.10] fit domain. At A, V = (1.00 × 10⁻⁴)(0.800)²⁰(0.100) = 0.115292 µV. At B, V = (1.00 × 10⁻⁴)(0.960)²⁰(0.100) = 4.42002 µV.
The map and the near-criterion voltage relation answer specified operating comparisons. Neither proves zero loss, a probability of failure, or validity at another field orientation or temperature outside the map.

A meets the chosen current screen; B misses it. Their conditional gauge voltages are 0.115292 and 4.42002 µV; both are finite.

Keep the conductor result separate from the cooling bill

At point A with 96.0 A, use gauge power 11.0680 µW, total cold joint resistance 2.00 µΩ and background cold heat leak 0.400 W. The cooler requires 250 W input per watt removed at this point. Find the included cold load and input work rate. These categories are stipulated separate and complete only for this partial budget.

Joint heating is I²R_joint = 96.0² × 2.00 × 10⁻⁶ = 0.0184320 W. The finite gauge loss is 0.0000110680 W.
Add the separate included cold loads: Q̇_cold = 0.400 + 0.0184320 + 0.0000110680 = 0.418443 W.
The given refrigerator ratio requires Ẇ_input = 250 Q̇_cold = 104.611 W. This input is not another cold heat leak and should not be added to Q̇_cold.
A complete cable comparison also needs delivered power, leads, converter losses and waveform/thermal records where relevant. This partial calculation establishes none of those missing quantities.

The included cold-stage load is about 0.418443 W and its refrigerator input about 104.611 W. A very small gauge voltage does not remove the joint or cooling requirements.

Try the reasoning

Can a magnetic field change the operating state?
Show the answer and reasoning
Yes.
Superconductivity has field-dependent limits.
Does zero DC resistance imply zero refrigeration demand?
Show the answer and reasoning
No.
Maintaining temperature remains a system requirement.
At 6.00 K, change only B_a from 2.00 to 3.00 T in the supplied map while I remains 96.0 A. Find I_c and the 80% screen limit. May the local E–I fit be used at that current?
Show the answer and reasoning
I_c = 90.0 A, the screen limit is 72.0 A, and 96.0 A misses it. The ratio 96/90 = 1.06667 remains inside the stipulated fit domain.
The temperature factor stays 0.750, while the field factor becomes 0.600. Multiplying gives 200 × 0.750 × 0.600 = 90.0 A. The fit can calculate the stated finite response only if its other steady-state assumptions hold; passing its input range is not an operating approval.
Keep the 6 K, 2 T E–I curve unchanged, but double the field criterion from E_c to 2E_c. Find the new reported criterion current, check the supplied fit domain, and explain whether the material improved.
Show the answer and reasoning
I_c,new = 120 × 2^(1/20) ≈ 124.232 A; the ratio 1.03526 is inside [0.600,1.10]. No material improvement follows.
Solve 2E_c = E_c(I_new/120)²⁰. Taking a twentieth root changes the reporting point along the same curve. Geometry, state and underlying E–I relation were held fixed. A tenfold criterion change would require ratio 10^(1/20) ≈ 1.12202, outside the stated fit range, so that extrapolation is not supplied.
An ohmic normal conductor of R = 2.00 µΩ spans L = 0.0500 m. At what current does its E = RI/L reach 1.00 µV/cm? Does that finite threshold crossing establish superconductivity?
Show the answer and reasoning
I_cross = 2.50 A. It does not establish superconductivity.
Convert the criterion to 1.00 × 10⁻⁴ V/m, so V = EL = 5.00 µV and I = V/R = 2.50 A. Any suitable finite ohmic resistance has a finite crossing. Magnetic and other state evidence cannot be replaced by assigning that crossing a critical-current label.
A field was present before a solid specimen entered its candidate state. One model predicts only that existing bulk flux stays unchanged; another predicts equilibrium expulsion from the interior apart from a thin penetration layer. Which observation distinguishes ideal perfect conduction from a Meissner state? Does trapped flux in a pinned type-II sample automatically refute superconductivity?
Show the answer and reasoning
Expulsion of previously present interior flux distinguishes the ideal Meissner prediction. Trapped flux in a pinned type-II specimen does not automatically refute superconductivity.
The first model opposes changes without requiring its initial flux to vanish. The ideal Meissner comparison assumes the appropriate field, shape and equilibrium state. A type-II mixed state and pinning introduce different magnetic histories; interpret them in their actual regime rather than forcing every observation into the simple cylinder cartoon.
In the point-A partial budget, halve total joint resistance to 1.00 µΩ without changing current, gauge response or background load. Find the new included cold load and required input at the supplied cooler ratio. Has refrigeration demand vanished?
Show the answer and reasoning
The included cold load becomes 0.409227 W and input about 102.307 W. Refrigeration remains required.
Joint heat becomes 96² × 10⁻⁶ = 0.009216 W. Add 0.400 W and 0.0000110680 W gauge loss to obtain 0.409227068 W; multiply by 250. Improving a joint helps this budget but does not remove other heat or establish overall cable efficiency.
The fictional conductor has zero-field T_c = 12 K. May the supplied current-map formula certify 10 K, 2 T operation or predict that changing field orientation preserves I_c?
Show the answer and reasoning
Neither inference is supplied.
The interpolation domain stops at 8 K, and field orientation was fixed. T < T_c alone neither locates a finite-field current criterion nor validates an extension of an interpolation. Obtain appropriate state/current and thermal evidence for the changed conditions.

What to carry forward

Use an operating region, not one transition number.
Separate intrinsic observations from system performance.
An unresolved transport voltage supplies a conditional bound; Meissner magnetic evidence asks a different question about the state.
Type-II vortex penetration and motion distinguish bulk superconducting order from a practically useful low-voltage current range.
Report the current criterion and all map conditions; neither a finite E threshold nor a chosen current margin is a universal phase boundary.
Include joints, waveform-dependent losses and cooling before making a complete conductor or cable claim.

References

Teaching resources and research supporting this lesson’s explanations and examples.