A sputtering target is a shaped solid of controlled purity, density, and grain structure, mounted inside a vacuum chamber and used as the source material for physical vapor deposition of thin films on substrates including silicon wafers, glass, and polymer web [S3][S4][S5].
Inside the chamber, a low-pressure gas, almost always argon, is ionised into a plasma, and the positively charged Ar+ ions are accelerated onto the negatively biased target; momentum transfer knocks target atoms off the surface, and those atoms travel across the vacuum to condense on the substrate as a continuous, often nanometre-thin layer [S1][S4][S5]. For a definition of the broader process family, see the sputtering target reference page.
How the Sputtering Process Deposits a Thin Film
Sputtering is a physical vapor deposition (PVD) technique: the target never melts, atoms are removed purely by kinetic momentum exchange between inert gas ions and the target surface, and the ejected atoms transit the chamber at thermal energies before condensing on the substrate [S1][S3][S4][S5].
The chamber is first pumped to a base vacuum, typically in the 10^-6 to 10^-8 Torr range, and then back-filled with argon to a working pressure on the order of a few millitorr; a high voltage between the target (cathode) and the substrate table (anode) ignites the plasma, sustaining a glow discharge that continuously bombards the target [S1][S5]. For thicker or wear-resistant coatings on tooling and machine parts, the same physical process is also used in the wider construction machinery and equipment supply chain, where PVD-coated inserts and dies are now standard.
Each incoming Ar+ ion at typical energies of a few hundred to about 1000 eV transfers momentum to near-surface atoms; a small fraction, characterised by the sputter yield S (atoms ejected per incident ion, usually 0.5 to 2 for most metals under Ar+ at 500 eV), is ejected and travels ballistically across the chamber to the substrate [S5][S8]. The deposited film thickness is controlled by time, target power, and throw distance, with deposition rates for common metals on the order of a few nanometres per second to several tens of nanometres per second at high power densities.
Sputtering Target Materials and Purity Grades
Targets are produced from metals, alloys, ceramics, and composites, with the material grade and purity selected to match the film stoichiometry the device requires [S1][S5].
Common target families and the purity bands used in production volume are: tantalum at 5N8 (≥99.9998%), copper and copper alloys at 6N (≥99.9999%), aluminium and aluminium alloys at 5N (99.999%), tungsten and tungsten alloys at 4N5 to 5N (99.995% to 99.999%), titanium at 5N (99.999%), and nickel and nickel alloys at 4N (99.99%) [S5]. Ceramic targets such as silicon dioxide, aluminium oxide, and titanium dioxide are used where the film must be electrically insulating, wear-resistant, or optically functional, and are typically manufactured by hot pressing or sintering rather than casting [S1][S5].
Alloy and composite targets, for example nickel-chromium heater films or copper-aluminium-oxide composites, let the process engineer deposit a multi-element film in a single chamber step, avoiding the run-to-run drift that two sequential targets would introduce [S1]. A well-known composite example is a copper plus aluminium-oxide target, which yields a film that is simultaneously highly conductive and well-bonded to the substrate [S1].
Sputtering Method Comparison: DC, RF, Reactive, and Magnetron
Each sputtering variant trades off target conductivity, deposition rate, film stress, and step coverage, so the choice of method is as much a part of the process spec as the target material itself [S1][S5][S8].
A direct comparison of the four principal methods, against the four criteria that most often drive selection, looks like this:
DC (direct current) diode sputtering: works only on electrically conductive targets, delivers the highest deposition rates for metals, and is the lowest-cost option, but it cannot sputter dielectrics and produces relatively high substrate heating. Reactive sputtering: a small partial pressure of O2, N2, or a fluorocarbon is bled into the argon stream so the film becomes an oxide, nitride, or fluoride of the target metal, which is the standard route to TiO2, AlN, and ITO films, but it risks target poisoning and process drift. Magnetron sputtering: permanent or electromagnets behind the target confine secondary electrons, raising plasma density and therefore deposition rate (commonly 5 to 10× over DC diode), at the price of target erosion confined to a narrow race-track zone that lowers material utilisation to about 20% to 40% [S1][S5][S8].
In practice, more than 80% of modern semiconductor, display, and optical-coating production lines run some form of magnetron configuration, often pulsed-DC for reactive modes, because the throughput gain is decisive [S5][S8].
Typical Industrial Applications
Sputtering targets are the consumable heart of a very wide range of manufactured parts, and the application drives the target spec as much as the material class [S1][S2][S5].
In semiconductor fabrication, targets deposit barrier layers, copper seed layers, tungsten contacts, and aluminium or cobalt interconnects on 300 mm silicon wafers through repeated PVD modules integrated into cluster tools. In hard-disk manufacturing, targets lay down nanometre-thick magnetic multilayers on glass or aluminium substrates where grain orientation and film smoothness directly control areal density. In optics, dielectric stacks of SiO2, Ta2O5, and TiO2 are reactively sputtered onto lenses and mirrors to make anti-reflection, beam-splitter, and bandpass coatings [S1][S5]. In solar cells, both crystalline silicon and thin-film CdTe/CIGS lines use sputtering for back contacts and Mo or TCO layers, and the same PVD process is also used to deposit functional coatings on parts that ultimately end up in lamps and light fittings and the broader lighting equipment and electric lamps category.
Failure Modes, Limitations, and Sourcing Risks
Process engineers need to know not only what a target does but also what it cannot do, because most thin-film defects originate at the target or the target/plasma interface [S2][S5][S7].
Key failure modes and constraints to specify around: nodule formation (cone-like tantalum or tungsten protrusions) that sheds particles onto the wafer, requiring periodic conditioning runs; target poisoning in reactive modes, where the target surface converts to a dielectric and the sputter rate drops sharply until the process is rebalanced; non-uniform erosion leading to race-track thinning and end-of-life target changes that waste 60% to 80% of the original material; and micro-arcing on ceramic or contaminated targets that splatters droplets and ruins the film [S5][S7]. On the supply side, refractory metals such as tantalum, tungsten, and ruthenium have historically been exposed to single-source or geopolitically concentrated supply chains, so many fabs now dual-qualify two vendors and require 5N8 or better purity with a published lot-to-lot analysis certificate [S5]. Precious-metal targets such as gold, silver, platinum, and palladium are routinely returned to refiners for recycling rather than landfilled, since reclaim rates above 95% are typical and the recovered metal is fed back into new target production [S2][S9].
Selection Criteria for a Sputtering Target
A practical target specification captures five things in writing: composition, purity, density, grain structure, and bonding geometry [S1][S5].
The decision points that most often determine whether a target is fit for a given tool and film are: required film purity versus available target purity (the target must be at least one purity grade tighter than the film spec, because gettering, chamber background, and resputtering all degrade the deposited film); minimum density (typically ≥95% to 99% of theoretical, since porosity traps gas and arcs); grain size and orientation (a fine, uniform grain structure reduces nodule and pinhole density); bond-back quality (the target is normally indium- or elastomer-bonded to a copper backing plate, and a voided bond is a frequent cause of thermal cracking); and the form factor, which must match the tool's cathode, whether planar, rotary, or conical, with the as-bonded assembly typically supplied in diameters from 100 mm for R&D up to 300 mm or larger for production [S1][S5]. For process engineers who also work on adjacent metrology, a related pressure sensor 4 mA and 20 mA setpoints reference explains how the same kind of current-loop discipline is applied when characterising the vacuum and gas-flow systems that feed a sputtering chamber.
Trackable signals to watch over the next procurement cycle: vendor disclosures on lot-by-lot purity certificates for 5N8 tantalum and 6N copper, the publication of recycled-content fractions by major precious-metal refiners, and the next generation of rotary-magnetron target designs that lift material utilisation above the current 20% to 40% band.