Spin-Polarized Scanning Tunneling Microscopy (SP-STM)

While Scanning Tunneling Microscopy (STM) revolutionized our ability to see individual atoms, it remains blind to one of their most fundamental quantum properties: spin. This leaves an entire world of magnetic phenomena, from the data stored on our hard drives to the exotic behavior of quantum materials, invisible at the atomic scale. The challenge, then, is how to supplement our view of atomic topography with a map of their magnetic landscape. Spin-Polarized STM (SP-STM) provides the answer. This advanced technique transforms the STM from a simple altimeter into a quantum compass, capable of resolving magnetic spin orientation atom by atom.

This article delves into the principles and powerful applications of SP-STM. In the chapter on ​​Principles and Mechanisms​​, we will explore the quantum mechanical rules that govern spin-dependent tunneling, from the foundational two-channel model to the subtle effects of temperature and quantum resonance. Subsequently, in ​​Applications and Interdisciplinary Connections​​, we will journey through the vast landscapes that SP-STM has unveiled, from mapping magnetic bits and exotic skyrmions to probing the quantum frontier of topological matter and manipulating individual spins.

Imagine you're in a crowded room, trying to follow two different conversations at once. If all you can hear is the total volume of sound, you'll have a hard time distinguishing what's being said in each. A conventional Scanning Tunneling Microscope (STM) is in a similar situation; it measures the total flow of electrons from a surface, which is like hearing only the overall volume. Spin-Polarized STM (SP-STM), however, is a revelation. It's like having a magical set of headphones that can tune into each conversation separately. It doesn't just measure the total electron flow; it can distinguish between two special "conversations" happening at the atomic scale—those carried by electrons with "spin-up" and those with "spin-down". This ability to listen in on the spin-polarized chatter of electrons is what unlocks the rich, magnetic world hidden on the surfaces of materials.

So, what are these "spin-up" and "spin-down" electrons? In the quantum world, electrons possess an intrinsic property called ​​spin​​, which makes them behave like tiny magnets. When we place these electrons in a magnetic field, they can align either with the field (spin-up) or against it (spin-down). In a non-magnetic material like copper, there's an equal number of up and down electrons everywhere; the magnetic chatter is a balanced, uniform hum.

However, in a ferromagnetic material like iron, this balance is broken. The interactions between electrons cause more of them to align in one direction. This creates two distinct populations: a majority-spin group and a minority-spin group. To a physicist, the number of available electronic states at a given energy is called the ​​density of states (DOS)​​. In a ferromagnet, we therefore speak of a ​​spin-resolved density of states​​—a separate count for spin-up states, $N^{\uparrow}$, and spin-down states, $N^{\downarrow}$. The imbalance between them is captured by a simple, powerful parameter: the ​​spin polarization​​, $P$.

This value tells us the net magnetic character of the electron sea at the energy we are probing. If $P=0$, the material is non-magnetic. If $P=1$, all electrons have the same spin, a rare and special state called a "half-metal". For a typical ferromagnet, $P$ is somewhere in between.

The central idea of SP-STM, as laid out in the beautifully simple ​​Jullière model​​, is that the tunneling process respects a fundamental rule: ​​spin is conserved​​. A spin-up electron from the microscope's tip must find an empty spin-up state in the sample to tunnel into; it can't just flip its spin mid-flight. This means the total tunneling current, $I$, isn't one monolithic flow. It's the sum of two independent streams, or channels: one for spin-up electrons and one for spin-down electrons. The size of each stream is determined by the availability of states on both sides—the tip (T) and the sample (S). So, the total conductance, $G = I/V$, is given by:

This elegant formula is the key to everything. It tells us that the quantum conversation isn't just a monologue from the sample; it's a dialogue, depending on the spin states of both the tip and the sample.

Let's see what this two-channel model predicts. We use a sharp, magnetic tip (with its own polarization, $P_T$) and scan it over a magnetic sample (with polarization $P_S$).

First, consider the ​​parallel (P) configuration​​, where the tip's magnetization is aligned with the sample's magnetization. The tip's majority-spin electrons see a wealth of available majority-spin states in the sample. The tip's minority-spin electrons see the few available minority-spin states in the sample. Both channels are open for business, and electrons flow relatively easily. A detailed calculation shows the conductance is proportional to $G_P \propto (1 + P_T P_S)$.

Now, imagine we move the tip over a domain where the sample's magnetization is flipped, pointing in the opposite direction. This is the ​​antiparallel (AP) configuration​​. The situation changes dramatically. The tip's abundant majority-spin electrons now face the sample's scarce minority-spin states. It's like a four-lane highway suddenly narrowing to a single lane. Likewise, the tip's minority-spin electrons face the sample's majority-spin states. Both channels face a mismatch, creating a bottleneck that impedes the flow of electrons. The conductance drops, and the math tells us it is now $G_{AP} \propto (1 - P_T P_S)$.

This difference in conductance between the parallel and antiparallel cases is a famous effect called ​​Tunnel Magnetoresistance (TMR)​​, and it's the source of magnetic contrast in SP-STM. We can quantify this contrast with a value called the magnetic asymmetry, $\mathcal{C}$:

This result is wonderfully simple and intuitive. The magnetic signal we measure is directly proportional to the product of the tip and sample polarizations! Flipping the tip's magnetic polarization, for instance, inverts the measured contrast, providing a definitive signature that the signal is truly magnetic in origin. In a real experiment, these polarizations might be something like $P_T = 0.3$ and $P_S = 0.4$, giving a contrast of only $P_T P_S = 0.12$, or 12%. This means the magnetic signal is often a tiny ripple on top of a large background current, a fact that makes the experimental achievement of these atomic-scale magnetic images all the more remarkable.

Treating spins as only "parallel" or "antiparallel" is a bit like describing color in just black and white. Nature has a full palette. What happens if the sample's magnetization isn't perfectly aligned with the tip, but sits at some arbitrary angle $\theta$?

Here, the full beauty of quantum mechanics unfolds. The sample's spin states can be viewed as a combination (a "superposition") of the tip's up and down states. When we do the math, a simple and profound relationship emerges: the conductance varies as the cosine of the angle between the magnetizations.

This single equation elegantly encapsulates all our previous findings. When the magnets are parallel ($\theta=0$), $\cos\theta=1$, and we recover the high-conductance state. When they are antiparallel ($\theta=\pi$), $\cos\theta=-1$, and we get the low-conductance state. But this formula tells us more. It describes a smooth, continuous variation between these extremes, like a magnetic dimmer switch. And it predicts something truly strange: when the tip and sample magnetizations are perpendicular to each other ($\theta=\pi/2$), $\cos\theta=0$. The magnetic term vanishes entirely! The magnetic tip becomes blind to the sample's magnetism. It's a purely quantum mechanical effect, a direct consequence of the geometry of spin.

With these principles, SP-STM becomes far more than just a camera for magnets. It becomes a miniature physics laboratory, capable of probing the fundamental properties of materials.

One powerful extension is ​​spin-polarized spectroscopy​​. Our simple model assumed we only care about the DOS right at the energy level of the last electron, the Fermi level. But by applying a bias voltage $V$ between the tip and sample, we open up a window of energy states that can participate in tunneling. If there is a peak or a valley in the spin-resolved DOS within this energy window, it will cause a change in the current at that specific voltage. By sweeping the voltage and measuring the current, we can map out the energy landscape of the spin-up and spin-down electrons separately. This allows us to see not just that atoms are magnetic, but why—what specific electronic orbitals are giving rise to their magnetism.

We can also explore how magnetism changes with temperature. We all have an intuition for this: if you heat a permanent magnet, it gets weaker. At a critical temperature, the ​​Curie temperature ($T_C$)​​, the thermal jiggling of the atoms becomes so violent that it completely destroys the ordered magnetic state, and the material becomes non-magnetic. Since the SP-STM signal $C(T)$ is proportional to the sample's polarization, which is in turn proportional to its overall magnetization, we should see the magnetic contrast disappear as we approach $T_C$. Landau's powerful theory of phase transitions predicts that near this critical point, the magnetization should vanish in a very specific way. By tracking the SP-STM contrast as a function of temperature, we can test this prediction. The result is a thing of beauty: the contrast follows the elegant relation:

where $C_0$ is the contrast at zero temperature. How remarkable! A measurement of single electrons quantum tunneling across a nanometer-scale gap precisely reveals a property of a macroscopic thermodynamic phase transition. It's a stunning example of the unity of physics, connecting the quantum and classical worlds.

The simple model of $G \propto (1 + P_T P_S \cos\theta)$ is incredibly powerful, but the universe is always more subtle and interesting upon closer inspection. The true power of a scientific tool often lies in discovering where the simple picture breaks down.

For one, we've assumed the vacuum gap is a passive stage for the tunneling dance. But the barrier itself can have opinions. The probability for an electron to tunnel can depend on its spin, a feature captured by a "matrix element asymmetry" $\beta$. This adds another layer of complexity, modifying the simple TMR formulas and showing that the barrier itself can act as a weak spin filter.

Even more dramatically, what if we deliberately place a single magnetic atom on the surface? This atom can act as a stepping stone for the tunneling electrons. If we tune our voltage just right to match a resonant energy level of this adatom, something amazing can happen. The adatom can act as an active ​​spin filter​​, resonantly enhancing the transmission of one spin type—say, spin-down—while suppressing the other. This effect can be so strong that it completely overwhelms the normal TMR effect. It can cause the entire magnetic contrast to invert, making parallel alignment look like antiparallel, and vice-versa. It's a stunning reminder that the signal we measure is a complex story written by the tip, the sample, and the precise path the electrons take between them. Discovering and understanding these exceptions to the rule is where the real frontier of the science lies, pushing our understanding of the quantum dance of electrons to an ever-deeper level.

Now that we have tinkered with the principles of spin-polarized tunneling, we might be tempted to put our feet up, content with the cleverness of it all. We’ve built a machine that can sense the orientation of a single atomic magnet! But the real adventure is just beginning. To a physicist, a new tool isn't an end in itself; it's a new pair of eyes to look at the universe. And what do we see with these new eyes? We see a world that was previously invisible, a bustling, dynamic landscape of magnetism that underpins the technology of today and the quantum dreams of tomorrow. So, let’s go on a tour and see what our spin-polarized scanning tunneling microscope (SP-STM) can do.

The most direct use of our new tool is simply to look. Imagine flying over a landscape in the dark. With a regular STM, which senses topography, it’s like having an altimeter—we can map out the mountains and valleys of atoms. But with SP-STM, it's as if we've been given a special night-vision camera that sees not height, but the direction of tiny, invisible compass needles on the ground.

Where do we find such landscapes? Look no further than the hard drive in a computer, which stores data in microscopic magnetic domains. With SP-STM, we can fly over the surface of such a material and see these domains directly. As the magnetized tip scans across the surface, the tunneling current rises when the tip's spin is parallel to the local surface spin and falls when it is antiparallel. By translating this fluctuating current into a color map, we can paint a direct picture of the magnetic data bits written on the surface. We can even go beyond a qualitative picture; by carefully analyzing the average current over a region, we can deduce the precise fractional areas of "spin-up" versus "spin-down" domains, giving us a quantitative map of the material's magnetic state.

This capability is even more profound when we look at materials that are invisible to conventional magnetic probes. Consider an antiferromagnet. Here, the atomic spins are arranged in a perfect checkerboard pattern, alternating "up, down, up, down" from one atom to the next. Macroscopically, their magnetic fields cancel out completely. It's a perfectly ordered magnetic world, but one that is completely "dark" to a standard magnet. But with SP-STM, we can bring our tip close enough to sense each atom individually. As we scan from one atom to its neighbor, we see the tunneling current oscillate—high, then low, then high again—revealing the alternating spin pattern with stunning clarity. For the first time, we can directly witness the fundamental order of an antiferromagnet, atom by atom.

Nature, of course, is more interesting than simple black-and-white domains. The boundaries between domains, known as domain walls, are not sharp lines but smooth, transitional regions where the spins gradually rotate from one orientation to another. With SP-STM, we can "zoom in" on these walls and map out their internal structure. For a typical wall, the magnetization component perpendicular to the surface might follow a smooth profile, like a hyperbolic tangent function, $m_z(x) = \tanh(x/w)$, where $w$ is the wall's width. By measuring the gradual change in our spin-polarized current as we cross the wall, we can directly measure this profile and determine the wall's characteristic width, a fundamental parameter that governs its behavior.

The world of spin textures can get even more exotic. In certain materials, spins can arrange themselves into beautiful, stable whirlpools called magnetic skyrmions. These are like tiny magnetic tornadoes, only a few nanometers across, where the spins artfully twist from pointing down at the center, outwards in a radial pattern, and up at the edges. These skyrmions are not just pretty; they are robust, particle-like objects that can be moved around with tiny currents, making them prime candidates for future high-density memory and logic devices. How do we see such a complex, non-collinear object? Our SP-STM is the perfect tool. As the tip scans across the skyrmion, its current is modulated by the dot product of its own spin polarization with the local, ever-changing spin direction of the skyrmion, producing a unique and identifiable signature that reveals the skyrmion's presence and structure.

In fact, we can do even better. By using a "vector magnet" to orient the tip's magnetization not just up or down, but in any direction in 3D space, we can perform a full magnetic reconstruction. This allows us to distinguish between different types of domain walls, such as a Bloch wall (where spins rotate like a corkscrew) and a Néel wall (where spins rotate like a wheel). By parking the tip at the center of a wall and rotating its in-plane magnetization, the signal will oscillate sinusoidally. The phase of this sine wave directly tells us if the spins in the wall's core point along the wall (Bloch) or perpendicular to it (Néel). This isn't just an academic curiosity; these different wall types have different energies and dynamics, which is crucial for technologies like "racetrack memory" that aim to shuttle domain walls around.

So far, we have used our microscope to take snapshots of static magnetic structures. But what if we want to understand their dynamics? Like any physical system, collections of spins can vibrate, resonate, and be excited. SP-STM is more than just a camera; it is also a spectroscope that can listen to the "music" of spins.

The technique is called inelastic electron tunneling spectroscopy (IETS). The idea is simple: most electrons tunnel across the gap elastically, without losing energy. But a small fraction can give a little "kick" of energy to the surface as they tunnel, for instance by flipping a local spin or exciting a molecular vibration (a phonon). This inelastic channel only opens up when the bias voltage $V$ is large enough to provide the required energy, $eV \ge \Delta$, where $\Delta$ is the excitation energy. When this happens, a new pathway for current opens, and we see a small step in the conductance $dI/dV$. The voltage at which these steps appear tells us the characteristic energies of the system's excitations.

A wonderful puzzle then presents itself: if we see a step at, say, 10 millivolts, did our tunneling electron excite a spin-flip or a phonon? Here, SP-STM offers a beautiful set of tools to play detective. First, we can apply an external magnetic field, $B$. A spin has a magnetic moment, so its energy levels will shift and split in a magnetic field (the Zeeman effect). A spin-flip excitation energy will therefore change linearly with $B$. A phonon, being a quantum of mechanical vibration, is largely indifferent to the magnetic field. If the position of our conductance step moves with the field, it’s a spin; if it stays put, it’s a vibration. Second, we can use our spin-polarized tip. The interaction that flips a spin depends on the electron's spin, so the size of the inelastic step will change if we use a spin-polarized tip and will even show an asymmetry depending on bias voltage. Vibrational excitations, being non-magnetic, are largely insensitive to the tip's polarization. Finally, we can use temperature as a knob. For low-energy spin excitations, a modest increase in temperature can thermally populate the excited state. This allows a tunneling electron to gain energy by de-exciting the spin, producing a step at the opposite bias voltage. This "anti-Stokes" signal, whose intensity grows with temperature, is a hallmark of a low-energy spin system. Through this systematic investigation, we can unambiguously separate the magnetic from the mechanical world at the atomic scale.

Armed with these powerful capabilities, we can venture into the deepest and most intriguing territories of modern condensed matter physics. SP-STM has become an indispensable tool for exploring exotic states of quantum matter.

Take, for example, a topological insulator. These are remarkable materials that are electrically insulating in their bulk but host a perfectly conducting surface. The surface electrons have a strange property called "spin-momentum locking": an electron moving to the right must be spin-up, while one moving to the left must be spin-down. What happens if we place a single magnetic impurity on this bizarre surface? The impurity's magnetic field scatters the surface electrons and induces a complex, spatially varying spin texture in the surrounding electron sea. SP-STM allows us to directly visualize this texture. As we move the tip away from the impurity, we see beautiful ripples in the spin polarization, which oscillate as a function of distance and direction. These are quantum interference patterns—not of charge, as in ordinary materials, but of spin—revealing the unique physics of the topological surface state.

The stakes get even higher when we turn our attention to the hunt for one of the most sought-after particles in physics: the Majorana fermion. These are exotic entities that are their own antiparticles, predicted decades ago but never definitively observed as fundamental particles. However, it's believed they can emerge as "quasi-particle" excitations in certain condensed matter systems. One of the most promising platforms is a chain of magnetic atoms on the surface of a superconductor. Theory predicts that a special zero-energy state, a Majorana zero mode (MZM), should appear at each end of the chain. These MZMs are the holy grail for building a fault-tolerant quantum computer. But how do you prove you’ve found one? The predicted signature is a sharp peak in the conductance precisely at zero bias voltage. But other, more mundane effects can cause such peaks. The definitive proof requires looking at its spin structure. A Majorana mode has a unique, non-trivial spin texture. By probing the end of the chain with our SP-STM, we can measure the conductance with the tip polarized parallel and anti-parallel to the chain's magnetic axis. The ratio of these two conductances gives a direct measure of the MZM's internal spin character, providing a crucial fingerprint to distinguish it from impostors. SP-STM thus places itself at the very heart of the quest for the future of computation.

Throughout our journey, we have treated our STM as a passive probe, a tool for seeing and understanding. But the interaction between the tip and the sample is a two-way street. The spin-polarized current is not just an observer; it can be an actor. It can exert a force—or rather, a torque—on the local magnetic moments.

This phenomenon, known as spin-transfer torque (STT), occurs because the electrons tunneling from the tip carry spin angular momentum. When they interact with a magnetic moment on the surface, they can transfer some of this angular momentum, giving the moment a tiny rotational "kick". By carefully controlling the current and its polarization, we can use this effect to manipulate magnetism at the ultimate, single-atom limit. We can pump energy into a single magnetic moment to make it precess, or with a strong enough current pulse, we can even force it to flip its orientation entirely. This transforms our microscope from a device that only "reads" magnetic information into one that can also "write" it, atom by atom.

The creative power of the tunneling electron doesn't stop there. In a fascinating marriage of spintronics and quantum optics, the SP-STM junction can be made to emit light. When a tunneling electron loses energy through an inelastic spin-flip process, that energy can be released not as heat or a phonon, but as a photon. What's more, because the process involves a change in the electron's spin, angular momentum must be conserved. The emitted photon carries away this angular momentum, which means the light is circularly polarized. A spin-up electron flipping to spin-down might emit a right-circularly polarized photon, while a spin-down to spin-up flip would emit a left-circularly polarized one. By analyzing the degree of circular polarization of the light coming from the junction, we can learn about the spin-dependent electronic structure of the tip and sample in remarkable detail. We have, in essence, created the world's smallest source of polarized light, a device where magnetism, quantum tunneling, and light interact at the single-atom, single-photon level.

So, SP-STM is far more than one instrument; it is a versatile, atomic-scale laboratory. It is our eye for seeing the invisible magnetic world, our spectroscope for hearing its rhythms, our probe for exploring the frontiers of quantum matter, and our hand for manipulating and creating at the ultimate limit of smallness. It is a beautiful testament to the idea that by asking a simple question—"what if we could see spin?"—we unlock a universe of discovery.