Understanding hot-hole dynamics at metal–semiconductor interfaces is important for designing devices that use light to generate and separate electrical charge. In a gold nanoparticle and p-type gallium nitride (p-GaN) system, ultrafast measurements show that light-driven hot holes move from the gold into the semiconductor in less than 200 femtoseconds. The transfer does more than create separated charge: it also changes the subsequent cooling behavior of hot electrons in the gold.
These findings add a time-resolved view of hole collection to a field that has more often focused on hot electrons. For engineers and technical decision-makers following plasmonic materials, the central implication is that the two sides of a photo-excited electron–hole pair cannot always be treated as independent. Charge extraction at an interface can influence carrier behavior within the metal itself.
Why hot-hole dynamics matter at an interface
When a metallic nanostructure absorbs light, optical transitions can produce energetic, non-equilibrium electrons and holes. These carriers are called “hot” because their energies differ from those of carriers in thermal equilibrium. They initially occupy a distribution of energy states, then undergo electron–electron scattering and approach an elevated electronic temperature. Later, energy transfers from the electron population to the lattice through electron–phonon coupling.
This sequence unfolds quickly. If a device is intended to collect a hot carrier, transfer across the interface must compete with the carrier’s relaxation. A rectifying contact, often described by a Schottky barrier, can help selectively collect electrons or holes depending on the semiconductor and its doping type. While many plasmonic metal–semiconductor studies have examined hot-electron collection, fewer have measured how quickly hot holes transfer into a p-type semiconductor or how their removal affects the remaining carriers.
Using p-GaN to observe hot-hole injection
The study used gold nanoparticles on p-GaN as a platform for examining this process. The interface has a reported Schottky barrier of about 1.1 eV, so only sufficiently energetic holes can cross from the gold into the semiconductor. The p-GaN band gap is 3.4 eV, which prevents the visible pump light used in the experiment from directly exciting free carriers in the semiconductor. In this arrangement, p-GaN serves as a destination for hot holes generated in the gold.
Researchers excited the gold nanoparticles with a 530 nm laser pulse and monitored the response using ultrafast transient absorption spectroscopy. Infrared probing tracked the injected holes in the p-GaN valence band, while a separate visible-light probe followed changes in the gold nanoparticles’ hot-electron response. This distinction allowed the two carrier populations to be examined through different spectral regions.
The infrared signal indicated that hot-hole injection occurred within the experiment’s temporal resolution of 200 femtoseconds. The signal then decayed through multiple timescales: a fast component on the order of a picosecond, followed by slower components on the order of tens of picoseconds and several nanoseconds. No corresponding transient signal appeared on bare p-GaN under the reported control conditions. The result establishes rapid hole transfer in this particular gold–p-GaN structure; it should not be taken as a universal transfer time for other materials or interfaces.
How hole collection changes electron cooling
To assess the effect on hot electrons, the researchers compared gold nanoparticles on p-GaN with gold nanoparticles on insulating glass (SiO2), where charge separation across the interface is prohibited. The visible transient absorption response differed between the two structures. On p-GaN, the transient plasmon bleach was narrower and showed little apparent spectral shift as it decayed. On SiO2, the bleach shifted over several picoseconds as the electron population cooled.
The width of the transient bleach provides an indication of the elevated electronic temperature. The narrower feature on p-GaN was consistent with a lower peak electronic temperature than on SiO2. Measurements also showed shorter electron–phonon coupling times for gold on p-GaN across the tested pump conditions. Because differences in substrate thermal conductivity alone did not explain the observed trend, the results pointed to hot-hole collection as a factor in the changed electron relaxation dynamics.
A mechanism linked to the gold d-band
First-principles calculations supported this interpretation. In gold, the electronic heat capacity depends on the electronic structure, including the density of available d-band states. Removing hot holes from below the Fermi level changes the electron population and, in turn, the conditions under which the remaining hot electrons thermalize. The calculations predicted that greater hot-hole removal limits the peak electronic temperature and shortens the electron–phonon relaxation time, consistent with the experimental trends.
The study’s comparison between experiment and theory suggested that a substantial fraction of holes energetic enough to overcome the interfacial barrier were injected into p-GaN. This interpretation links interfacial charge transfer to changes in the metal’s electronic response—not merely to the presence of charge in the semiconductor.
Implications for plasmonic device design
For device developers, the practical lesson is that a metal–semiconductor interface should be evaluated as a coupled carrier system. Selecting a p-type support to collect holes may also affect the energy distribution and cooling rate of electrons in the metal. That interaction could matter when designing plasmonic optoelectronics or photocatalytic systems that depend on carrier extraction before thermalization.
The results do not establish device-level performance or demonstrate a particular industrial application. Instead, they provide a mechanistic basis for considering hot-hole collection as a design variable. Future assessments of candidate materials and interfaces can benefit from measuring both the transferred carrier and the carrier population left behind in the metal, rather than treating either process in isolation.
Conclusion
In gold nanoparticles supported on p-GaN, hot-hole injection takes place in less than 200 femtoseconds and is accompanied by measurable changes in hot-electron thermalization. The combination of infrared and visible transient absorption measurements, supported by electronic-structure calculations, shows how interfacial hole removal can lower the peak electron temperature and shorten electron–phonon coupling time. For researchers developing plasmonic devices, the key takeaway is to account for both carrier collection and its effect on the metal’s remaining carriers when evaluating interface design.