A scanning tunneling microscope (STM) is a microscope that investigates a surface by moving an extremely sharp conducting tip across it and measuring the current that flows through the narrow gap between tip and sample. Its operation depends on quantum tunneling, which allows electrons to cross a barrier that would prevent passage in classical physics. Under suitable conditions, STM resolves individual atoms and probes electronic properties on atomic length scales. Unlike an optical microscope, it does not form an image with lenses: its images are reconstructed from electrical measurements made during scanning.
Operating principle
When a conducting tip approaches a conducting or semiconducting surface to within a distance typically of a fraction of a nanometre, their electronic wave functions overlap. An applied bias voltage creates conditions for a net flow of electrons between occupied states on one side and available unoccupied states on the other. This produces a small tunneling current, commonly measured in the picoampere-to-nanoampere range.
The essential sensitivity comes from the approximately exponential dependence of tunneling current on tip–sample separation. In a simplified barrier model, at fixed bias and with other factors held constant,
where is the current, is the separation, is the electron mass, is an effective barrier height, and is the reduced Planck constant. The effective barrier is related to the materials’ work functions and the applied voltage; it is not generally a perfectly rectangular barrier.
For a representative barrier height of several electronvolts, increasing the separation by approximately one ångström can reduce the current by about an order of magnitude. This strong distance dependence makes extremely small vertical displacements detectable. It also concentrates the signal near the closest part of the tip, enabling atomic lateral resolution when the tip apex and sample are suitable.
The signal nevertheless depends on more than distance. It also reflects the electronic states of both tip and sample, their orbital character, and the tunneling probability between them.
Instrument design
A typical STM contains:
- A sharp conducting tip, often made from tungsten or a platinum–iridium alloy. Its microscopic apex strongly influences image quality.
- A positioning and scanning system, usually based on piezoelectricity, which converts applied voltages into very small mechanical displacements.
- A coarse approach mechanism, which brings the tip close enough for tunneling without crashing it into the surface.
- A current amplifier, which converts the very small tunneling current into a measurable voltage.
- Control electronics and data acquisition, which regulate scanning, record signals, and construct images.
- Mechanical and environmental isolation, which suppresses vibration, acoustic disturbance, electrical interference, and thermal drift.
Piezoelectric scanners provide motion in two lateral directions and in the vertical direction. Their displacement is not perfectly linear or independent of previous motion, so calibration and correction of scanner behavior are important for quantitative measurements.
STMs can operate in air, liquids, or ultrahigh vacuum, and at temperatures ranging from cryogenic conditions to elevated temperatures in specialized instruments. Vacuum helps preserve clean surfaces, while low temperature reduces thermal motion and improves the energy resolution of electronic measurements. Neither vacuum nor cryogenic cooling is an absolute requirement for STM.
Imaging modes
Constant-current mode
In constant-current imaging, a feedback controller adjusts the vertical tip position to maintain a chosen tunneling current as the tip scans laterally. The recorded vertical displacement forms the main image.
This mode accommodates surface steps and moderate roughness because the tip follows changes in the tunneling signal. However, its speed is limited by the response of the feedback system. Incorrect settings can produce distortion or oscillation, or allow a tip–sample collision.
The resulting image is often called a topographic image, but the apparent height includes electronic contributions. Two locations at the same physical height can appear different if they have different electronic states within the energy range selected by the bias.
Constant-height mode
In constant-height imaging, the tip scans with little or no feedback adjustment, and changes in current provide image contrast. Because large vertical corrections are unnecessary, this mode can be faster than constant-current imaging.
It is mainly suitable for sufficiently flat regions. Unexpected protrusions, drift, or surface steps can bring the tip into contact with the sample. In practice, some implementations retain slow feedback to compensate for long-term drift.
Interpreting atomic contrast
STM does not directly show the positions of atomic nuclei. Bright features represent locations with greater tunneling probability under the selected conditions. They may correspond to atoms, bonds, particular orbitals, or spatial variations in electronic states.
For example, images of graphite often emphasize only one of the two carbon sublattices in its surface layer because the two sites have different electronic environments. Changing the bias or tip condition can change the observed pattern without changing the underlying atomic structure.
Scanning tunneling spectroscopy
The same tip–sample junction can perform scanning tunneling spectroscopy (STS), which measures electronic behavior as a function of voltage rather than recording only a surface image.
A common procedure positions the tip, stabilizes the junction at a selected current and voltage, temporarily disables feedback, and records the current–voltage relation . Its differential conductance, , is often measured using a small alternating voltage and phase-sensitive detection.
Under commonly used assumptions—including a weakly varying tip density of states, a sufficiently simple tunneling matrix element, and low temperature—the differential conductance approximately follows the sample’s local density of states:
where denotes position, is the sample local density of states, is the equilibrium Fermi energy, and is the positive elementary charge. Here the sample voltage is defined relative to the tip.
With that convention, positive sample bias generally probes unoccupied sample states, while negative bias probes occupied states. This interpretation is approximate: finite temperature, the tip’s electronic structure, voltage-dependent tunneling transmission, and local electrostatic effects can alter the spectrum.
STS can reveal a semiconductor’s band gap, localized electronic states, and the energy gap associated with superconductivity. Spatially resolved spectroscopy maps how these features vary around impurities, defects, interfaces, or magnetic vortices.
Development and historical significance
Gerd Binnig and Heinrich Rohrer developed STM at IBM’s Zurich research laboratory in 1981. Their work established that a controllable vacuum tunneling junction could be used to obtain atomic-scale surface information. Early studies included investigations of reconstructed silicon surfaces, helping establish STM as an important tool in surface science.
Binnig and Rohrer received half of the 1986 Nobel Prize in Physics for their design of the scanning tunneling microscope. The other half went to Ernst Ruska for fundamental work in electron optics and the design of the first electron microscope.
STM became a foundational technique of scanning probe microscopy, in which a localized probe is scanned across a surface to measure an interaction. The later atomic force microscope extended this general approach to force measurements, allowing investigation of many insulating materials that conventional STM cannot readily image.
Applications and specialized forms
Surface structure and reactions
STM reveals atomic arrangements, surface reconstructions, steps, vacancies, and other defects. It is widely used to investigate adsorption, molecular organization, and reactions on surfaces.
Repeated imaging can follow diffusion or structural changes, although processes faster than the acquisition rate may be missed or appear distorted. The tip can also influence the process being observed through its electric field, current, or mechanical interaction.
Atomic and molecular manipulation
Carefully controlled changes in tip position, voltage, and current can move individual atoms or molecules, transfer them between tip and surface, or induce selected changes in bonding. A well-known early demonstration arranged xenon atoms on a nickel surface to spell “IBM”; the work was published in 1990.
Such experiments enable the construction of artificial atomic structures and the study of confined electronic states. They generally require carefully prepared surfaces and tightly controlled conditions, rather than providing a general-purpose method for assembling arbitrary materials.
Electronic and magnetic materials
STM and STS are important tools in condensed-matter physics. They investigate electron scattering, superconducting states, charge ordering, and other spatially varying electronic phenomena.
In spin-polarized STM, a magnetically sensitive tip makes the tunneling signal depend on the relative spin polarization of tip and sample. This can resolve magnetic structures, but separating magnetic contrast from ordinary electronic and structural contrast requires controlled measurements.
Electrochemical interfaces
Electrochemical STM examines electrode surfaces in liquids while controlling electrode potentials. It can follow processes such as metal deposition, dissolution, and molecular adsorption. Insulation of most of the tip helps suppress unwanted electrochemical currents, leaving the exposed apex to provide the localized tunneling signal.
Limitations and measurement reliability
Electrical access is essential. Conventional STM needs a sufficiently conductive path through the sample. Bulk insulating specimens usually cannot be examined directly, although ultrathin insulating films on conducting substrates may permit tunneling.
Tip structure can dominate the image. An unstable or contaminated apex can change contrast abruptly. Multiple active protrusions may produce duplicated features. Atomic resolution therefore does not, by itself, guarantee an unambiguous structural interpretation.
Apparent height is not purely geometric. Bias-dependent electronic structure can make a feature appear taller, shorter, or even reverse its contrast. Structural conclusions are stronger when supported by measurements at several biases, spectroscopy, calculations, or complementary techniques.
Environmental stability limits performance. Vibration, thermal drift, electrical noise, and scanner distortion can obscure or displace atomic features. Quantitative work must account for these effects and report appropriate measurement uncertainty.
The measurement can disturb the sample. High electric fields, electron injection, local heating, or accidental contact can move adsorbates or change the surface. This interaction is useful for manipulation but can complicate observation.
Finally, STM generally examines a small, accessible surface region rather than the interior or bulk average of a specimen. Its exceptionally local information complements, rather than replaces, diffraction, spectroscopy, and other structural methods.