Lecture 4 - Bulk applications of SCs
Lecture 4 - Bulk applications of SCs
Superconductivity - Lecture 4
(Bulk applications)
High field superconducting magnets
Principle is to make a superconducting solenoid, which will be able to carry high currents (in the order of few 100A to few 1000A)
𝐵 = 𝜇0 𝑛𝐼 (n = number of turns/metre)
For 1T field, if you use normal copper wire of 1mm diameter wire (n= 1000), current required = 795A!
Total length of copper wire required for a 1m long solenoid (assume solenoid diameter is 1m) = 1000 ∗ 2𝜋 ∗ 0.5 = 3141.6 𝑚
2
Resistivity of copper = 0.017 𝑂ℎ𝑚 𝑚𝑚 Τ𝑚
Total resistance of this wire = 0.017 ∗ 3141.6/ 𝜋 ∗ 0.52 = 68 ohms
Power required = 𝐼 2𝑅 ≅ 43𝑀𝑊
• For superconducting solenoid, power required is only to cool the solenoid below the 𝑇𝑐 of the wire material.
• For low 𝑇𝑐 materials, liquid Helium is needed. For high 𝑇𝑐 materials (most ceramic superconductors), liquid nitrogen is good
enough to cool down below 𝑇𝑐 .
• Typical cost of liquid nitrogen is Rs 2/litre. Typical cost of liquid Helium is Rs. 5000/litre
• Helium is a constantly dwindling resource as it escapes from earth’s atmosphere to outer space!
Lets consider a spatially homogeneous superconductor. (This can happen if the wire diameter is less than both 𝜆𝑙 𝑎𝑛𝑑 𝜉𝐺𝐿)
Lets assume for a spatially homogeneous order parameter Ψ = Ψ0 𝑒 𝑖𝑘𝑥 𝑘 𝑖𝑠 𝑡ℎ𝑒 𝑤𝑎𝑣𝑒𝑛𝑢𝑚𝑏𝑒𝑟 𝑎𝑙𝑜𝑛𝑔 𝑡ℎ𝑒 𝑤𝑖𝑟𝑒
𝑞ℏ 𝑞2
Substitute the above in 𝐽𝑠 = Ψ ∗ ∇Ψ − Ψ∇Ψ ∗ − Ψ 2 𝐴 𝑎𝑛𝑑 𝑡𝑎𝑘𝑒 𝑜𝑛𝑙𝑦 𝑥 𝑔𝑟𝑎𝑑𝑖𝑒𝑛𝑡 𝑓𝑜𝑟 𝑠𝑖𝑚𝑝𝑙𝑖𝑐𝑖𝑡𝑦
2𝑚𝑖 𝑚
ℏ𝑘−𝑞𝐴𝑥
𝐽𝑠,𝑥 = 𝑞Ψ02 = 𝑞Ψ02 𝑣𝑥 = 𝑞𝑛𝑠 𝑣𝑥 -- (1)
𝑚
Hence Cooper pair density decreases with increasing 𝑣𝑥 , and ultimately will go to 0. Substitute (2) in (1), then you find that the
𝑑𝐽𝑠,𝑥
supercurrent density is also dependent on 𝑣𝑥 . After substitution, for maximum supercurrent equate =0
𝑑𝑣𝑥
2𝛼 2 2𝛼
For max 𝐽𝑠,𝑥 , 𝑣𝑥 = . And maximum supercurrent is 𝐽𝑠,𝑥 𝑚𝑎𝑥
= 𝑞 𝑛𝑠
3𝑚 3 3𝑚
𝐵𝑐𝑡ℎ 2 𝑚 𝟐 𝟐 𝑩𝒄𝒕𝒉
Use 𝛼 = −
𝜇0 𝑛𝑠
, and 𝜆𝑙 =
𝜇0 𝑛𝑠 𝑞2
, you will get 𝑱𝒔,𝒙 𝒎𝒂𝒙
=
𝟑 𝟑 𝝁𝟎 𝝀𝒍
High field superconducting magnets (maximum current in Type I superconductors)
Any current in the wire will create a magnetic field around it. But a Type I Superconductor will expel all magnetic fields, expect for
a small shell cross section at the edges corresponding to 𝜆𝑙 . This therefore implies that the bulk of the wire will not be able to
carry any supercurrent, and the supercurrent carrying region should only be confined to 𝜆𝑙 at the wire edge.
For a Type II SC, it is not necessary that the bulk needs to be completely free of magnetic fields.
In the region 𝐵𝑐1 < 𝐵𝑐𝑡ℎ < 𝐵𝑐2 (𝑆ℎ𝑢𝑏𝑛𝑖𝑘𝑜𝑣 𝑃ℎ𝑎𝑠𝑒), quantized magnetic flux (vortices) can be present in the bulk.
Lets assume a Type II SC tape, within such a field range. The black circles are vortices which are essentially ‘normal metallic’
regions. The vortices therefore feel a Lorentz force = 𝐵𝐼𝐿, and they start to move along the width of the tape, therefore making
SC regions normal and giving rise to electrical resistance and generating heat.
Even a small current can cause such vortex motion, and this therefore makes these TypeII SC materials useless for high current
applications, despite having a a high 𝑩𝒄𝟐.
Solution – vortices have to be somehow pinned to their locations so that they do not move around (Flux pinning)
High field superconducting magnets (Flux pinning – reaping advantages of high 𝑩𝒄𝟐 )
Vortex is a normal state region with currents circulating outside. Vortex formation starts when the magnetic field exceeds 𝐵𝑐1. At
this point, if there are 𝑛 flux lines (vortex) per unit area, each flux line carries exactly 1 flux quantum.
Vortex formation needs some energy. Let that energy be 𝑒𝑣 per unit length. Let the Area of the current carrying cylindrical wire
be 𝑆, and let its length be 𝐿. Let’s try to see what 𝑒𝑣 depends on.
𝐸𝑓 = energy needed to induce magnetization in the flux containing region = 𝐸𝑚 = 𝐵𝑐1 ∆𝑀 𝑆 𝐿 where. ∆𝑀 = 𝑛Φ0 /𝜇0
𝐵𝑐1 Φ0
Hence 𝑒𝑣 = (Vortex formation energy per unit length is a material property)
𝜇0
Put some ‘metallic inclusions’ in a TypeII SC (vortex a), the energy there is reduced by amount 𝑒𝑣 𝑙.
Hence, vortex ‘a’ needs external energy to move (is otherwise ‘pinned’), while vortex ‘b’ is free.
Such inclusions are therefore called ‘pinning centers’ and are helpful in achieving high currents in SCs.
Ceramic (High 𝑇𝑐 ) SCs have low coherence lengths, and hence tiniest of defects can act as pinning centers.
𝜋ℏ𝑣𝑓
𝜉0 = 0.18
𝑘𝐵 𝑇𝑐
High field superconducting magnets (Preventing flux jumps)
It can sometimes happen that due to mechanical shock or thermal fluctuations in the environment, vortices can ‘tear off’ from
their ‘pinned’ sites (Flux Jumping event). This movement of vortices, as discussed before, causes heat generation and if not
addressed immediately, can in turn heat the whole wire and turn it metallic (‘quenching’ of superconducting magnet). This would
be a catastrophic event, as huge currents in the wire will probably melt it and cause fires.
Solution to the above is to encapsulate multiple superconducting wire filaments within metallic regions with low resistance. If
’Flux jump’ happens, the vortices will move to the normal conducting regions, will heat it negligibly, but the SC filaments will be
saved and a ‘Quench’ event can be avoided.
Non reversible magnetization curves can be obtained for Type II SCs with widespread defect sites (pinning centers)
Apply a magnetic field and make such a material non superconducting above 𝐵𝑐2. Then reduce the field to zero. Flux pinning will
happen as you reduce the field. Ultimately at 0 field, there will be a lot of trapped flux. Remove the external field, and this flux
will still remain pinned as the materials is still in the SC state, and will remain in such a state forever.
(Click the link and watch from 10:50 onwards for how SC maglev trains work)
Superconducting microscopic devices (the future of technology)
Superconducting Qubits – JJs used as anharmonic oscillator to create a well defined quantum 2 level system
Superconducting tunnel (SIS) junctions – for inter galactic radiation detections (discovery of Cosmic Microwave background)
RSFQ technology – Superconducting processors operating at clock speeds close to 800GHz (State of the art Si technology – 4GHz)