Scanning Tunneling Spectroscopy (STS)

While Scanning Tunneling Microscopy (STM) revolutionized our ability to "see" the atomic world, its power lies in mapping the topography of a surface. But what if we could not only see where atoms are but also understand their electronic character? This is the fundamental gap bridged by ​​Scanning Tunneling Spectroscopy (STS)​​, a powerful extension of STM that transforms it from a mere microscope into a quantum spectrometer for individual atoms and molecules.

This article delves into the world of STS, exploring its foundational principles and diverse applications. In the first chapter, ​​Principles and Mechanisms​​, we will explore how STS works at the quantum level, defining the concept of the Local Density of States (LDOS) and explaining how measuring the change in tunneling current with voltage provides a direct map of this crucial electronic property. We will also examine the practical considerations and ultimate physical limits that govern the technique's precision. Following this, the chapter on ​​Applications and Interdisciplinary Connections​​ will showcase the remarkable power of STS in action. We will journey from measuring the fundamental band gaps of semiconductors and imaging the frontier orbitals of single molecules to investigating exotic quantum phenomena in advanced materials. Through this exploration, we will uncover how STS provides an unparalleled window into the electronic music of matter.

Imagine you are in a completely dark room with a collection of musical instruments. If you reach out and touch them, you can feel their shapes, their sizes, their textures. This is what a standard Scanning Tunneling Microscope (STM) does; it "feels" the topography of a surface, atom by atom, giving us a breathtakingly detailed map of the atomic landscape.

But what if you could do more? What if, instead of just feeling the shape, you could gently tap each instrument and listen to the specific notes it produces? A tiny drum would sound different from a long violin string. This is the essence of ​​Scanning Tunneling Spectroscopy (STS)​​. It goes beyond just seeing where atoms are; it allows us to "listen" to their electronic music. Each peak in an STS spectrum is like a musical note, a resonant frequency corresponding to an allowed energy level for an electron at that precise location. By sweeping through a range of "notes," we can record the unique electronic symphony of a single atom or molecule.

So, what is this "music" that we are listening to? In the language of quantum mechanics, it’s a property called the ​​Local Density of States (LDOS)​​. This might sound intimidating, but the concept is wonderfully intuitive. Think of a giant, multi-story parking garage. The total number of parking spots in the entire garage is the "Total Density of States." But the LDOS is much more specific. It answers the question: "If I stand at this exact spot ($\mathbf{r}$) on this specific floor (energy $E$), how many parking spots are available right here?"

More formally, for a single electron in a material, its behavior is described by wavefunctions, $\psi_n(\mathbf{r})$, each with a corresponding energy, $E_n$. The probability of finding that electron at a specific point $\mathbf{r}$ is given by $|\psi_n(\mathbf{r})|^2$. The LDOS at a location $\mathbf{r}$ and energy $E$, often written as $g(\mathbf{r}, E)$, is the sum of the probabilities of all possible states that have that exact energy $E$. Mathematically, this looks like:

This quantity is local because of the $|\psi_n(\mathbf{r})|^2$ term—it can be very large in one place (e.g., on top of an atom) and very small in another (e.g., between atoms). This is why a map of the LDOS is not uniform; it reveals the intricate electronic patterns that define chemical bonds, defects, and surface phenomena. It's the fundamental canvas upon which chemistry happens.

The remarkable trick of STS is to measure this LDOS. The setup is elegantly simple. We bring the sharp metallic tip of the STM to a halt above a point of interest on the sample. Then, we apply a small DC voltage, $V$, between the tip and the sample.

This voltage acts like a small step or ramp, shifting the energy levels of the sample relative to the tip. Let's set the "sea level" of electrons in the tip—its ​​Fermi level​​—to zero energy. If we apply a positive voltage $V$ to the sample, we are effectively lowering the sample's energy landscape by an amount $eV$. This opens a window of energy, from $0$ to $eV$, into which electrons from the tip can tunnel. Conversely, a negative voltage on the sample raises its energy levels, allowing electrons to tunnel from the sample's occupied states into the tip.

The flow of these tunneling electrons creates a tiny electrical current, $I$, which is the signal we measure. This current depends on two main things: the number of filled "starting" states and the number of empty "destination" states available within the energy window created by the voltage $V$.

Simply measuring the total current $I$ for a given voltage $V$ isn't enough. The total current is an accumulation of all tunneling possibilities within the energy window from the Fermi level up to $eV$. It’s like knowing the total volume of water in a bucket, but not the rate at which the tap is flowing.

To find the density of states at a specific energy, we need a more subtle approach. Imagine we slowly increase the voltage from $V$ to $V+dV$. This opens up a tiny, additional sliver of energy, $dE = e dV$, for tunneling to occur. The extra current we measure, $dI$, must be coming from electrons tunneling into this new energy sliver. Therefore, this additional current $dI$ is directly proportional to the number of available states within that sliver—which is precisely the LDOS at that energy!

This gives us the central, beautiful relationship in STS:

The differential conductance, $dI/dV$, is directly proportional to the sample's Local Density of States at the position of the tip ($\mathbf{r}_0$) and at an energy $E$ equal to the bias energy $eV$. By sweeping the voltage $V$ and recording $dI/dV$ at each step, we trace out the LDOS spectrum, note by note.

A positive sample voltage ($V \gt 0$) means electrons from the tip tunnel into the sample. This probes the empty, or ​​unoccupied states​​ of the sample at energies above its Fermi level. A peak in the $dI/dV$ spectrum at, say, $V = +0.5$ V indicates a high density of available electronic states $0.5$ eV above the sample's Fermi level. Conversely, a negative sample voltage ($V \lt 0$) pulls electrons out of the sample into the tip, thereby probing the ​​occupied states​​ below the Fermi level.

This beautifully simple proportionality is, like many things in physics, an elegant approximation. It holds true under a specific set of ideal conditions. Understanding these conditions, and what happens when they are not met, is where the real art of spectroscopy begins.

The Ideal Probe

For the magic of $dI/dV \propto \text{LDOS}$ to work perfectly, our experimental setup should be ideal. This means:

1. ​​A "Boring" Tip:​​ The tip itself must be electronically "flat." Its own density of states should be constant over the energy range we are measuring. If the tip has its own peaks and valleys in its DOS, they will get mixed in with the sample's signal, like trying to listen to a violin over a noisy radio. We want a featureless, metallic tip.

2. ​​Transparent Tunneling:​​ The probability of an electron tunneling through the vacuum barrier must be the same for all electrons, regardless of their energy. In reality, this probability (the tunneling matrix element) has some energy dependence, but for small bias voltages, we can often ignore this.

3. ​​An Absolute Zero World:​​ The derivation assumes the experiment is done at a temperature of absolute zero ($T=0$ K). At this temperature, the electron energy levels in a metal are filled neatly up to the Fermi level and are completely empty above it, creating a perfectly sharp "sea level."

The Reality of a Warm World: Thermal Broadening

Of course, we don't live at absolute zero. At any finite temperature $T$, the electronic "sea level" is not perfectly calm; it's fuzzy. The electrons' energies are smeared out by thermal energy, following the ​​Fermi-Dirac distribution​​. This means some states just below the Fermi level are empty, and some just above are filled.

This thermal fuzziness acts like a blurry lens on our spectrometer. Instead of measuring the true, sharp LDOS, we measure the LDOS convolved (or smeared) with a thermal broadening function. The width of this blur is determined by the temperature. The full-width at half-maximum (FWHM) of this broadening is found to be:

where $k_B$ is the Boltzmann constant. For a typical experiment conducted at liquid helium temperature ($T=4.2$ K), this thermal broadening sets a fundamental resolution limit of about $1.3$ meV. Any features in the true LDOS that are sharper than this will be blurred out in our measurement.

Other Wrinkles in the Measurement

The real world introduces other fascinating complications. When studying semiconductors, for instance, the strong electric field from the tip can penetrate the sample and bend the energy bands, shifting the very states we want to measure. This effect, known as ​​tip-induced band bending​​, means the simple relation $E=eV$ no longer holds, and careful modeling is needed to interpret the spectra.

Clever experimentalists have developed techniques to counteract some of these issues. For example, sometimes the tip-sample distance can fluctuate. Since the tunneling current depends exponentially on this distance, this can distort the spectrum. One trick is to measure a ​​normalized conductance​​, typically $(dI/dV)/(I/V)$. In this quantity, the exponential dependence on distance largely cancels out, giving a cleaner look at the electronic structure.

What is the absolute, final limit on how sharp our energy measurement can be? It is tempting to think about the time it takes for an electron to "traverse" the vacuum gap. An electron tunneling is a quantum event, and one might think the time-energy uncertainty principle, $\Delta E \Delta t \gtrsim \hbar/2$, connects a short traversal time to a large energy uncertainty. This, however, is a common but misleading idea. STS is a steady-state measurement of a continuous current, not a time-of-flight experiment for a single electron. The traversal time does not define the energy resolution.

The uncertainty principle does, however, play two crucial—and much more subtle—roles.

First, there is the experimental integration time. At each voltage point, we measure the current for some duration, $\tau_{\text{int}}$, which might be a few milliseconds. The uncertainty principle states that this finite measurement time imposes a limit on our energy resolution of $\Delta E \gtrsim \hbar/(2\tau_{\text{int}})$. For a typical $\tau_{\text{int}} = 10$ ms, this limit is an incredibly tiny $\sim 10^{-14}$ eV. This is astronomically smaller than any other broadening effect, like thermal broadening, and is therefore completely negligible.

The second role is far more profound. The electronic states we are measuring within the material are not immortal. They are ​​quasiparticles​​—excitations that exist only for a finite ​​lifetime​​, $\tau_{\text{life}}$, before they are scattered by other electrons or vibrations. And here, the uncertainty principle strikes with full force. A state that lives for only a short time $\tau_{\text{life}}$ cannot have a perfectly defined energy. It must have an intrinsic energy width, or ​​lifetime broadening​​, of $\Delta E_{\text{life}} \approx \hbar/\tau_{\text{life}}$.

For a typical quasiparticle lifetime of 10 femtoseconds ($10^{-14}$ s), the intrinsic energy broadening is about 66 meV. This is an inherent quantum fuzziness of the state itself. No instrument, no matter how perfect or cold, can ever measure this state with a better resolution. This is not a failure of our experiment; it is a fundamental property of the universe, and STS allows us to measure it directly. We are not just listening to the notes; we are hearing how long they last before they fade away.

In the previous chapter, we dissected the marvelous quantum trick that allows an electron to leap across a vacuum—tunneling. We saw that the rate of this leap, the tunneling current, is exquisitely sensitive to the landscape it is leaping into. In particular, by measuring how the current changes with voltage, we get a direct line to the local density of states (LDOS), which is nothing less than a map of the available energy levels for an electron at a specific point in space.

Now, with this powerful tool in hand, let's go on an adventure. We are no longer just observing atoms as tiny bumps on a surface; we are becoming quantum mechanics, poised at the tip, ready to tunnel. We are going to probe the very soul of materials—their electronic structure—and see how this simple principle unlocks secrets across physics, chemistry, and materials science. This is not just a microscope; it is a spectroscope for the quantum world.

Let's begin with the fundamentals. Every material, be it a semiconductor in your phone or a lump of metal, has a set of rules that governs its electrons. These rules are encoded in its electronic band structure. The most important rule for a semiconductor is the existence of a "forbidden zone" of energy—the band gap. States in the valence band are filled with electrons, while states in the conduction band are empty. The gap is the energy cost to move an electron from the filled band to the empty one, making it available to conduct electricity.

How do we measure this gap? With tunneling spectroscopy, it's astonishingly direct. We place our tip above the semiconductor and slowly ramp up the voltage. For a range of small voltages, nothing much happens; the current is nearly zero. This is because there are no available states in the gap for our tunneling electrons to jump into. But, as we apply a sufficient positive bias, electrons from our tip suddenly find open arms in the sample's empty conduction band, and the current surges. If we apply a negative bias, we eventually reach a point where electrons from the sample's filled valence band can tunnel out into our tip. The voltage range between these two onsets directly gives us the size of the band gap. We are, quite literally, measuring the boundaries of this forbidden zone, eV by eV.

But the electronic landscape is not always a simple story of filled, empty, and forbidden zones. The density of states is often not flat; it can have dramatic peaks and valleys. Just as the shape of a concert hall creates acoustic resonances, the geometric arrangement of atoms in a crystal creates electronic resonances—energies at which there is a huge number of available states. These sharp peaks in the DOS are known as van Hove singularities. With our trusty $dI/dV$ measurement, which directly maps the DOS, these singularities appear as prominent peaks in our spectrum. We are seeing how the crystal's symmetry and geometry choreograph the behavior of its electrons, a beautiful link between structure and property.

Let's zoom in now, from the vast landscape of a crystal to the intimate world of single atoms and molecules. Suppose we have a surface with different kinds of atoms scattered upon it. How can we tell them apart? We could try to measure their size, but a far more elegant way is to take their electronic fingerprint. We can position our STM tip directly above an unknown atom and record its $dI/dV$ spectrum. Each type of atom has its own unique set of orbital energies, which appear as peaks in the LDOS at characteristic voltages. By simply looking at the energy of these peaks, we can identify an atom with certainty, performing chemical analysis at the ultimate limit of a single atom.

This principle takes on a breathtaking dimension when we look at molecules. The pictures of molecular orbitals—the famous s, p, and d clouds—that you see in chemistry textbooks are not just cartoons. They are real! With tunneling spectroscopy, we can make them visible. The Highest Occupied Molecular Orbital (HOMO) is the outermost orbital holding electrons, while the Lowest Unoccupied Molecular Orbital (LUMO) is the first empty one. To "see" the HOMO, we apply a negative bias, coaxing electrons to tunnel out of this specific orbital. By scanning the tip across the molecule while holding the voltage steady, we map the spatial probability of that orbital. To see the LUMO, we do the opposite: apply a positive bias so our tunneling electrons can jump into it. The result is a pair of images that show the exact shape and location of the molecule's frontier orbitals, the very orbitals that dictate all of its chemical reactivity. We are no longer just inferring quantum mechanics; we are looking right at it.

But there is more to a molecule than just where its electrons are. Molecules also vibrate; their bonds stretch, bend, and twist. It turns out our tunneling electrons can feel these vibrations! This is the magic of Inelastic Electron Tunneling Spectroscopy (IETS). Most electrons tunnel elastically, arriving with the same energy they left with. But occasionally, a tunneling electron can give up a tiny, specific amount of energy to excite a molecular vibration—to ring the molecule like a bell. This opens up a new "inelastic" channel for tunneling, causing a tiny step-like increase in the current. This step appears as a sharp peak in the second derivative of the current, $d^2I/dV^2$. The voltage at which this peak appears, $V$, is directly related to the vibrational energy of the bond, $\hbar\omega = eV$.

This technique is so precise that we can, for example, distinguish between carbon monoxide molecules made with the normal $^{12}\text{C}$ isotope and those made with the heavier $^{13}\text{C}$ isotope. The heavier isotope vibrates just a little more slowly, shifting the IETS peak to a slightly lower voltage—a direct consequence of the simple harmonic oscillator model, where frequency depends on mass. We are literally listening to the hum of a single chemical bond.

Armed with this versatile tool, we can now venture into the wild and wonderful frontier of modern condensed matter physics, where the collective behavior of electrons gives rise to exotic quantum phenomena.

The materials we engineer are never perfect, and often it's their imperfections—defects like a missing atom or an extra one—that define their most useful properties. Tunneling spectroscopy is an unparalleled tool for hunting down these defects and diagnosing their electronic character. In a 2D material like $\text{MoS}_2$, for example, a missing selenium atom leaves behind "dangling bonds" that create an electronic state within the material's band gap. An extra molybdenum atom, on the other hand, acts as an electron donor, creating a different kind of state. Each defect has a unique spectral signature—a peak at a specific energy—allowing us to not only find them but also understand how they will affect the material's electronic behavior.

We can even go beyond studying existing materials and start designing new "quantum landscapes." By taking two sheets of a 2D material like graphene and stacking them with a slight twist, a beautiful aperiodic pattern called a Moiré superlattice emerges. This is not just a geometric curiosity; this new, larger-scale pattern imposes a spatially varying potential on the electrons. Using STS, we can scan across a single Moiré unit cell and watch the electronic properties, like the local band gap, change dramatically from one spot to the next. This new field of "twistronics" allows us to engineer electronic behavior in ways never before possible.

Some of the most exciting quantum materials are topological insulators. These are strange materials that are insulators in their bulk but are forced by the laws of topology to have conducting states on their surfaces. These surface states are special: they are robust and protected from scattering off simple, non-magnetic defects. An STS experiment demonstrates this beautifully. The spectrum of a topological surface state has a characteristic V-shape centered at its "Dirac point." If we measure this spectrum on a perfectly flat terrace and then measure it again right on top of a messy atomic step edge, we find something remarkable: the V-shape is almost completely unchanged. The electrons simply don't see the non-magnetic obstacle because back-scattering is forbidden by the fundamental physics of their spin-momentum locking. It's a striking visualization of quantum protection in action.

Finally, tunneling spectroscopy gives us a window into the most profound aspects of quantum mechanics: the cooperative dance of many electrons acting in concert. In a superconductor, electrons pair up and condense into a collective quantum state that can carry current with zero resistance. This pairing opens up an energy gap, $\Delta$, in the electronic spectrum, which STS can measure with exquisite precision. By mapping this gap, we have discovered that in many high-temperature superconductors, the gap is not uniform. Instead, the material is a complex patchwork of regions with strong and weak superconductivity. The global properties we measure are an average over this messy, inhomogeneous landscape, a crucial insight provided only by the local eye of STS.

Perhaps the most delicate dance of all is the Kondo effect. If you place a single magnetic atom on the surface of a non-magnetic metal, you might expect it to behave like a tiny compass needle. But at low temperatures, the sea of conduction electrons surrounding the atom engages in a collective conspiracy to "screen" its magnetic moment, forming a complex, many-body singlet state. The signature of this remarkable phenomenon in an STS experiment is a single, sharp resonance that appears precisely at zero bias voltage—the Fermi level. The width of this "Kondo resonance" is directly proportional to the energy scale of the many-body state, known as the Kondo temperature, $T_K$. Seeing this peak is to witness the quantum entanglement of one atom with an entire sea of electrons.

From the band gap of a transistor to the entangled state of a single magnetic atom, the journey of a tunneling electron tells a rich and varied story. Tunneling spectroscopy is the language through which we can read that story, revealing the unity and beauty of the quantum laws that govern our world, one electron at a time.