Learning goal: Explain how electron spin, magnetic order and spin-dependent transport create information-processing devices that use more than electric charge alone, progressing from magnetoresistance to spin currents, magnetic tunnel junctions, spin-transfer torque, spin–orbit torque, spin waves, skyrmions, antiferromagnets and altermagnets.
Scope boundary: Magnetism and Electromagnetism remains the owner of classical magnetic fields; Semiconductors and Transistors owns charge-based electronic devices; Superconductivity and Quantum Materials owns superconducting/quantum-material phases; Quantum Measurement owns quantum-state measurement fundamentals. This article owns the specialist job of how spin-polarised transport and magnetic state are coupled to electrical read/write operations in spintronic devices.
Reader-safety boundary: Educational condensed-matter/device physics only.
Wait, What? A Current Can Carry Angular Momentum Even When Net Charge Flow Is Zero
Ordinary electronics focuses on charge.
Spintronics adds another carrier of information:
electron spin
A material can support:
- charge current;
- spin-polarised charge current;
- pure spin current.
The device can therefore move angular momentum even when little or no net charge is transferred in the same direction.
That is the conceptual jump.
The One-Sentence Answer
Learn spintronics by first separating charge current from spin polarisation, then use magnetoresistance and spin-dependent tunnelling to understand magnetic readout before learning how spin-transfer and spin–orbit torques can write magnetic states and how new magnetic orders extend the design space beyond ferromagnets.
Stage 1: Spin Is an Intrinsic Quantum Degree of Freedom
Electron spin behaves like intrinsic angular momentum.
It is not literally a tiny classical sphere rotating in space.
Spin has:
- quantised projection;
- magnetic moment;
- quantum statistics.
Stage 2: Magnetisation Is a Collective State
In a ferromagnet, exchange interactions favour alignment of many microscopic magnetic moments.
The material develops macroscopic magnetisation.
Spintronics uses that ordered state as an information-bearing variable.
Stage 3: Majority and Minority Spin Channels Can Conduct Differently
Electronic states for opposite spin orientations need not be equivalent inside a ferromagnet.
The two spin channels can therefore have different:
- density of states;
- scattering;
- conductivity.
This is the foundation of spin-polarised transport.
Stage 4: Spin Polarisation Measures Imbalance
A useful transport concept is how unequal the spin-up and spin-down populations or currents are.
A fully spin-polarised current would use one spin channel only.
Real materials usually sit between zero and perfect polarisation.
Stage 5: Spin Relaxation Erases Information
A nonequilibrium spin population does not persist forever.
Scattering and spin–orbit coupling randomise spin orientation.
Important scales include:
- spin lifetime;
- spin diffusion length.
Spin information has a range.
Stage 6: Spin Diffusion Length Is a Device Constraint
If a spin current must cross a layer much thicker than its spin diffusion length, most of the spin information is lost.
Device geometry must therefore be matched to spin-relaxation physics.
Stage 7: Giant Magnetoresistance Turns Relative Magnetisation Into Resistance
In a multilayer containing ferromagnets separated by nonmagnetic material, electrical resistance can depend strongly on whether magnetic layers are:
- parallel;
- antiparallel.
This is giant magnetoresistance (GMR).
Magnetic configuration becomes an electrical signal.
Stage 8: GMR Is Not the Same as Ordinary Magnetoresistance
The “giant” effect arises from spin-dependent scattering across engineered multilayers.
It is not simply a conductor’s resistance changing slightly in a magnetic field.
Stage 9: GMR Made Magnetic Readout Extremely Sensitive
Hard-disk read heads used GMR to detect small magnetic domains.
The scientific lesson is broader:
a nanoscale spin configuration can be converted into a macroscopic voltage
Stage 10: Magnetic Tunnel Junctions Use Quantum Tunnelling
A magnetic tunnel junction contains:
- ferromagnetic layer;
- ultrathin insulating barrier;
- second ferromagnetic layer.
Electrons tunnel through the barrier.
The tunnelling probability depends on spin and magnetic alignment.
Stage 11: Tunnelling Magnetoresistance Reads Magnetic State
The resistance differs between parallel and antiparallel configurations.
A May 2026 Nature Reviews Methods Primers article describes MgO-based magnetic tunnel junctions as benchmark spintronic devices because coherent spin-dependent tunnelling produces large room-temperature TMR.
Stage 12: MgO Is More Than a Passive Insulator
Crystalline MgO filters electronic wavefunctions by symmetry.
This can strongly favour selected spin states during tunnelling.
The barrier’s crystal structure therefore helps create the readout.
Stage 13: One Magnetic Layer Is Usually Fixed, One Is Switchable
A memory junction typically contains:
- reference layer;
- free layer.
The relative orientation stores the bit.
The read operation measures resistance without intentionally switching the free layer.
Stage 14: Thermal Stability Protects Stored Information
A nanoscale magnet experiences thermal fluctuations.
If the energy barrier between magnetic states is too small:
- the bit flips spontaneously.
A memory cell therefore needs sufficient thermal stability for retention.
Stage 15: Smaller Memory Cells Become Harder to Stabilise
Shrinking a magnet reduces magnetic volume.
That lowers the energy barrier.
Scaling therefore creates a conflict between:
- density;
- retention;
- write energy.
Stage 16: Spin-Transfer Torque Uses the Current to Write the Magnet
A spin-polarised current entering a magnetic layer transfers angular momentum.
That torque can reorient magnetisation.
The current becomes both:
- information carrier;
- actuator.
Stage 17: STT-MRAM Separates Non-Volatility From Charge Storage
Conventional volatile memory loses state when power disappears.
MRAM stores information in magnetic orientation.
The state survives without continuously maintaining stored charge.
Stage 18: STT-MRAM Is Already a Practical Technology
A 2024 Nature Reviews Electrical Engineering review described STT-MRAM as an established non-volatile memory platform with strong speed, endurance and embedded-memory relevance.
The topic is not only future physics.
It is semiconductor manufacturing.
Stage 19: Read and Write Share a Path in STT-MRAM
The same tunnel junction often carries:
- small read current;
- larger write current.
This simplifies structure.
It also creates read-disturb and endurance design constraints.
Stage 20: Spin–Orbit Coupling Connects Charge Motion to Spin
When electron motion interacts with relativistic spin–orbit effects, charge current can generate transverse spin accumulation or spin current.
This opens a second writing strategy.
Stage 21: The Spin Hall Effect Produces Transverse Spin Current
In selected heavy metals and related materials, charge current along one direction creates spin current in a perpendicular direction.
The conversion efficiency is often described through a spin Hall angle or related parameter.
Stage 22: Spin–Orbit Torque Can Switch a Nearby Magnet
A spin current generated by spin–orbit coupling can exert torque on an adjacent ferromagnet.
This is spin–orbit torque (SOT).
The current need not flow through the tunnel barrier used for reading.
Stage 23: SOT-MRAM Separates Read and Write Paths
A March 2026 review highlights SOT-MRAM’s ability to separate the write path from the magnetic tunnel junction read path.
Potential advantages include:
- high endurance;
- fast switching.
The price can be larger cell area or extra terminals.
Stage 24: Field-Free SOT Switching Is an Engineering Challenge
Simple SOT switching may require a symmetry-breaking field.
Researchers seek built-in structural or magnetic asymmetry that removes this requirement.
The device must break symmetry somehow.
Stage 25: Rashba–Edelstein Effects Offer Another Conversion Route
At interfaces lacking inversion symmetry, spin–orbit coupling can convert charge current into interfacial spin accumulation.
Interface engineering becomes a spin-current source.
Stage 26: Spin Pumping Runs Conversion in Reverse
A precessing magnet can inject spin angular momentum into an adjacent material.
This is spin pumping.
The resulting spin current can be detected through inverse spin Hall conversion.
Stage 27: Thermal Gradients Can Generate Spin Signals
The spin Seebeck effect links temperature gradients to spin transport.
This connects spintronics with thermoelectric-style nonequilibrium transport.
But thermoelectric voltages can also mimic spin signals, so measurement controls matter.
Stage 28: Spin Caloritronics Adds Heat as a Control Variable
Spin caloritronics studies coupled:
- heat;
- charge;
- spin transport.
The challenge is separating true spin conversion from ordinary thermal artefacts.
Stage 29: Spin Waves Can Carry Information Without Net Electron Flow
A local disturbance in an ordered magnet can propagate as a collective spin wave or magnon.
Magnonics uses these waves for information transport and processing.
The information carrier becomes a collective excitation.
Stage 30: Spin-Wave Devices Trade Joule Heating for Other Losses
Reducing charge motion can reduce some resistive losses.
But magnons still suffer:
- damping;
- scattering;
- transduction losses.
“No charge current” does not mean “no energy cost”.
Stage 31: Magnetic Skyrmions Are Topological Spin Textures
A skyrmion is a nanoscale swirling magnetic configuration with nontrivial topology.
Its stability arises from exchange, anisotropy, spin–orbit effects and magnetic interactions.
Topology helps but does not make the object indestructible.
Stage 32: Skyrmions Can Move Under Small Currents
Spin torques can drive skyrmions along tracks.
This inspired racetrack-memory concepts.
But device use must manage:
- pinning;
- edge annihilation;
- skyrmion Hall motion.
Stage 33: Antiferromagnets Have Ordered Spins but Little Net Magnetisation
Neighbouring moments align oppositely.
The net magnetisation can be near zero.
Advantages can include:
- weak stray fields;
- ultrafast dynamics.
Reading and switching the order parameter is harder than for ferromagnets.
Stage 34: Antiferromagnetic Spintronics Expands the State Space
Information can be encoded in the Néel vector rather than net magnetisation.
Spin–orbit effects can manipulate that order.
This moves spintronics beyond conventional ferromagnetic memory.
Stage 35: Altermagnets Add a New Symmetry Class
Altermagnets can have zero net magnetisation while electronic bands show strong spin splitting.
A July 2026 Nature Physics review describes altermagnetism as a new platform combining selected advantages of ferromagnetic spin splitting with antiferromagnetic compensation.
Stage 36: Zero Net Magnetisation Does Not Mean Spin Degeneracy
In ordinary intuition, no net magnetisation suggests spin-up and spin-down bands should coincide.
Altermagnets show that crystal and magnetic symmetry can produce momentum-dependent spin splitting without macroscopic magnetisation.
This is a symmetry lesson, not merely a materials label.
Stage 37: 2D Magnets Add Interface-Controlled Spin Physics
Van der Waals magnets can be stacked with:
- graphene;
- semiconductors;
- topological materials.
Atomically sharp interfaces allow strong tuning through:
- gating;
- twist;
- proximity.
Stage 38: Probabilistic Spintronics Uses Thermal Fluctuation Instead of Fighting It
A thermally unstable magnetic element can fluctuate between states.
Instead of treating that as a defect, probabilistic computing can use the stochastic state as a p-bit.
A 2026 SOT-MRAM review highlights this as an emerging application.
Stage 39: Neuromorphic Devices Use Magnetic Dynamics as Computation
Spin-torque oscillators and magnetic junction networks can mimic selected neural-network operations.
The physics supplies:
- oscillation;
- nonlinearity;
- memory.
It does not reproduce biological neurons literally.
Stage 40: Measurement Artefacts Are a Major Professional Problem
Spin-current experiments often measure tiny voltages.
Ordinary effects can imitate spin signals:
- thermoelectric voltages;
- anomalous Hall effects;
- rectification;
- contact asymmetry.
One voltage peak is not enough.
Stage 41: Magnetoresistance Needs a Geometric Control
Changing magnetic field can rotate magnetisation and also alter contacts, heating or orbital motion.
Strong experiments vary:
- angle;
- thickness;
- temperature.
Mechanism should survive those controls.
Stage 42: Spin Hall Measurements Are Model-Dependent
Extracting a “spin Hall angle” often requires assumptions about:
- current distribution;
- interface transparency;
- spin diffusion;
- damping.
The parameter is inferred, not directly observed.
Stage 43: Professional Spintronics Is a Conversion-and-Lifetime Problem
Which microscopic degree of freedom carries angular momentum, how efficiently is charge converted to spin or spin to charge, how far and how long does that spin information survive, and which control experiment rules out an ordinary electrical or thermal artefact?
Evidence: How Do We Know a Magnetic Tunnel Junction Stores State Magnetically?
Strong evidence combines:
- hysteretic resistance switching;
- independent magnetic measurement;
- angle dependence;
- retention;
- switching-current dependence.
The electrical resistance is the readout; the magnetic state is the proposed stored variable.
Misconceptions Worth Hunting
- Electron spin means a tiny ball literally spinning.
- Spin current always means charge current.
- GMR and TMR are the same mechanism.
- A magnetic field must be applied to read MRAM.
- Non-volatile means zero energy for every operation.
- Smaller MRAM cells are automatically better.
- Spin–orbit torque eliminates all write losses.
- Skyrmions are indestructible because they are topological.
- Antiferromagnets have no useful spin order.
- Zero net magnetisation means no spin splitting.
- Any transverse voltage proves a spin Hall effect.
Transfer Check
A current transports equal numbers of opposite charges in the same direction but opposite spin populations move differently. Can spin current exist without large net charge current? Yes.
Two magnetic tunnel junction states show different resistance. Does that alone prove which layer switched? No.
A nanomagnet becomes smaller and its retention drops. Which energy scale likely weakened? The magnetic thermal-stability barrier.
A spin-pumping experiment shows a voltage that reverses with a temperature gradient. Should thermal artefacts be checked? Yes.
How We Know the Learning Has Held
A learner should be able to:
- define spin polarisation;
- explain spin relaxation and spin diffusion length;
- distinguish GMR and TMR;
- explain magnetic tunnel junctions;
- explain STT and SOT conceptually;
- explain spin Hall and inverse spin Hall conversion;
- explain spin pumping and spin caloritronics;
- explain magnons and skyrmions;
- distinguish ferromagnetic, antiferromagnetic and altermagnetic order;
- explain MRAM retention/write trade-offs;
- identify common spintronic measurement artefacts.
Model Limits
Two-current models simplify band structure. Jullière-type TMR models omit coherent symmetry filtering. Macrospin switching can fail for nonuniform nanomagnets. Spin diffusion models compress interface scattering. Extracted spin Hall angles depend on device models.
Professional spintronics therefore keeps magnetic order + band symmetry + spin lifetime + interface transparency + device geometry + thermal stability + artefact controls visible together.
Teaching Guide
Teach in this order:
spin → magnetic order → spin polarisation → relaxation → GMR → tunnelling → TMR → magnetic memory → STT → spin Hall → SOT → spin pumping → magnons → skyrmions → antiferromagnets → altermagnets → metrology.
Begin with:
“Can a device move spin angular momentum even when almost no net charge is moving in the same direction?”
Connect This to the eduKate Learning Estate
- https://edukatesengkang.com/2026/08/28/how-to-learn-semiconductors-transistors-energy-bands-modern-electronics/
- https://edukatesengkang.com/2026/08/29/how-to-learn-superconductivity-quantum-materials/
- https://edukatesengkang.com/2026/08/29/how-to-learn-quantum-measurement-superposition-tunnelling-state-reasoning/
- https://edukatesengkang.com/2026/08/29/how-to-learn-thermoelectricity-thermoelectric-materials/
Research Foundations and Further Learning
- Magnetic tunnel junctions — Nature Reviews Methods Primers, 28 May 2026.
- Altermagnetic spintronics — Nature Physics, 6 July 2026.
- Spin–Orbit Torque MRAM: Fundamentals, Device Engineering, and Applications in Probabilistic Computing — Advanced Quantum Technologies, 4 March 2026.
- Spin-transfer torque magnetoresistive random access memory technology status and future directions — Nature Reviews Electrical Engineering, 2024.
- MRAM: A Versatile Non-Volatile Memory for Next-Generation Computing — 2026 review.
- Spintronics technology: A comprehensive review of materials, applications, and future trends — Journal of Science: Advanced Materials and Devices, 2026.
The Quiet Ending
The beginner asks, “Why does electron spin matter in electronics?”
The developing physicist asks, “How does magnetic alignment change resistance?”
The advanced learner asks, “How is angular momentum written into the free layer?”
And the professional asks:
Which spin-conversion mechanism, lifetime scale and artefact control make the measured voltage a defensible signal of spin transport rather than an ordinary electrical effect wearing a magnetic label?
