Metal–Semiconductor Contacts

Introduction

A semiconductor device needs metal leads. A poorly chosen lead can form its own rectifying junction and dominate the measured response, even when the semiconductor interior is designed correctly. Metal–semiconductor contacts are therefore active interfaces, not neutral wire attachments. Work functions and electron affinities suggest ideal energy alignment, while real surface chemistry, interface states and tunnelling determine how closely a contact follows that simple picture.

Core explanation

The work function Φ m of a metal is the energy from its Fermi level to a vacuum reference just outside the surface. The semiconductor electron affinity χ is the energy from its conduction-band edge to that vacuum reference under a specified clean-surface convention. In an ideal unpinned n-type contact, an electron Schottky barrier can be written Φ Bn ≈ Φ m − χ. This Schottky–Mott expression is a starting model, not a universal measurement law. Surface dipoles, chemical reactions, oxide layers and gap states can change the actual alignment. MIT's metal–semiconductor junction lecture treats ideal barriers alongside Fermi-level pinning and practical ohmic contacts.

When metal and semiconductor touch, carriers redistribute until equilibrium electrochemical potential is common. Band edges bend near the semiconductor surface. A substantial barrier for majority carriers can make current respond asymmetrically to voltage: a Schottky contact . It often conducts by majority-carrier thermionic emission in a suitable regime, which can make its switching behaviour differ from a minority-carrier p–n diode. A contact with approximately linear I–V around zero bias and sufficiently low resistance for its application is called ohmic . “Ohmic” is a behavioural description; it does not prove that no barrier exists. Heavy doping can make a barrier thin enough for tunnelling, yielding nearly linear low-resistance transport even with a finite barrier height.

For p-type material, the relevant majority-carrier barrier is for holes, so a work-function trend stated for n-type must not be copied without reversing the carrier reasoning. Interface states can pin the semiconductor Fermi level near an energy set by the interface, reducing sensitivity of measured barrier height to the chosen metal. Native oxide, interdiffusion and microscopic inhomogeneity further complicate measurements. The voltage across an entire two-terminal sample includes both contacts and bulk material; measuring a nonlinear I–V curve does not by itself locate the barrier. MIT's photovoltaics lecture on charge extraction explicitly compares Schottky and ohmic contacts.

Contact quality matters for LEDs, transistors and solar cells. A high-resistance contact wastes electrical power and can create local heating. An unintended Schottky barrier can impede collection or injection. Conversely, a designed Schottky diode uses that barrier purposefully for rectification or sensing. Contact selection therefore begins with the required function, not a rule that every interface must be ohmic.

Step-by-step reasoning

1. Identify whether electrons or holes are the majority carriers. 2. Draw vacuum level, metal Fermi level and semiconductor band edges before contact. 3. Estimate an ideal barrier only if clean unpinned alignment is justified. 4. After contact, show charge transfer and band bending at common equilibrium E F. 5. Check interface states, oxide, doping and tunnelling before predicting measured I–V.

Visual explanation

Draw a metal block next to n-type semiconductor bands. Mark Φ m from metal E F to vacuum and χ from E c to vacuum. After contact, align E F and bend E c near the interface. Beside it, sketch a second heavily doped case with a thinner barrier and tunnelling arrow. Pair a rectifying I–V curve with a nearly linear ohmic curve.

Real-world analogy

At a border checkpoint, a tall but narrow fence may be crossed by a tunnelling-like shortcut in the physics picture, while a tall wide barrier strongly impedes passage. The exact electronic effect is quantum mechanical, not a literal tunnel under a fence. The analogy highlights that barrier width can matter as much as height for contacts.

Real-world example

An engineer measuring a newly grown semiconductor film first checks that both metal contacts are ohmic over the measurement range. Otherwise, the apparent sample resistance may largely reflect contact barriers. In a solar cell, a selective contact may intentionally favour one carrier type and block the other; this can improve charge extraction when designed carefully.

Why?

Why can heavy doping turn a rectifying contact into a useful ohmic contact without removing its nominal barrier? More doping narrows the depletion region at the interface. Carriers may tunnel through this thin barrier, lowering contact resistance and making the small-signal I–V response nearly linear. This is a transport-mechanism change, not necessarily a disappearance of all band bending.

Common misconception

“A high-work-function metal always makes a good ohmic contact” ignores semiconductor type and interface pinning. “The work function is the same as band gap” confuses a vacuum-referenced removal energy with a valence-to-conduction separation. “Ohmic means zero resistance” is false; practical contacts have finite contact resistivity.

Worked example

For an ideal unpinned n-type interface, suppose Φ m = 4.8 eV and χ = 4.0 eV. The Schottky–Mott estimate gives Φ Bn = 0.8 eV . If a second metal has Φ m = 4.4 eV under the same ideal assumptions, its estimated electron barrier is 0.4 eV. In a real interface with strong Fermi-level pinning, the measured barriers could be much closer together. The calculation demonstrates model use, not a guaranteed device result.

Quick check

1. Does a nearly linear low-resistance contact necessarily have no energy barrier at all? Answer: No. A thin barrier in a heavily doped region can allow tunnelling and yield ohmic behaviour.

Exam focus

State carrier type and assumptions before applying Φ Bn = Φ m − χ. Label all energies relative to the same vacuum reference. Distinguish rectifying behaviour from barrier height alone, and mention interface states or oxide layers when explaining deviations. A two-terminal I–V curve includes contacts as well as bulk.

Advanced insight

Barrier inhomogeneity can make different regions of one contact carry disproportionate current, so a single fitted barrier height may depend on temperature and measurement method. Specific contact resistivity, in Ω m², is more useful than a casual “ohmic” label when assessing a high-current device. Surface preparation and annealing can alter reaction layers and therefore both electrical behaviour and long-term stability.

Summary

Metal–semiconductor contacts align electrochemical potentials and produce local band bending. An ideal work-function model estimates a Schottky barrier, while real interfaces may be pinned or chemically modified. Rectifying and ohmic describe observed transport, and heavy doping can make tunnelling through a finite barrier useful for low-resistance contacts.

Practice questions

1. Define a metal work function. Answer: It is the energy to move an electron from the metal Fermi level to a vacuum reference just outside the surface. 2. In the ideal unpinned n-type model, what barrier follows from Φ m = 5.0 eV and χ = 4.2 eV? Answer: Φ Bn ≈ 5.0 − 4.2 = 0.8 eV. 3. Why can a two-terminal nonlinear I–V measurement fail to identify the semiconductor bulk response? Answer: One or both metal contacts may have a rectifying barrier and dominate the measured voltage dependence. 4. What does Fermi-level pinning change? Answer: Interface states make the actual barrier less sensitive to the selected metal work function than the ideal alignment predicts.