Hot carrier transport in silicon can depart from ordinary diffusion when a semiconductor is driven far from equilibrium. Ultrafast imaging of photoexcited silicon has revealed that, at high laser excitation, carrier movement slows and develops an oscillatory component before settling into longer-term diffusion. The observations point to electrostatic interactions between spatially separated electrons and holes as an important influence on transport.
For engineers and technical decision-makers working with semiconductor materials and devices, the result underscores a practical consideration: carrier behavior under intense excitation may not be predicted adequately by a simple diffusion model. Understanding how charge distributions evolve in both space and time can help clarify the physical processes relevant to optoelectronic, photovoltaic and thermoelectric systems.
Why spatially resolved carrier measurements matter
Optical pump-probe spectroscopy is commonly used to examine how a material responds after a short optical pulse. However, its spatial resolution is limited by diffraction, and the optical signal can include a substantial contribution from the bulk or substrate. Those limitations can make it difficult to separate surface behavior from the response of the wider sample, particularly in reduced-dimension materials.
Scanning ultrafast electron microscopy (SUEM) offers a different view. It combines a femtosecond optical pump with an electron probe that scans the sample surface. The method provides snapshots of changes in carrier density with spatial and temporal resolution. In the reported work, the electron probe detected changes in secondary-electron emission associated with changes in surface potential. The resulting contrast images therefore mapped laser-induced carrier dynamics at the surface rather than directly counting every carrier.
The experiment used a heavily doped p-type silicon wafer excited with a 515 nm laser pulse. Images were collected over delays ranging from before excitation to several nanoseconds afterward. The measurement had approximately 1 ps temporal resolution; spatial resolution of 200 nm was used for this study. Because secondary electrons emerge from only the upper 5–10 nm of the sample, SUEM provided surface-sensitive information.
Hot carrier transport in silicon changes at high fluence
After excitation, the images showed a strong signal near the laser-exposed region. The carrier distribution broadened as it spread laterally, with faster expansion at early times and slower evolution at later times. At higher excitation, an alternating, ring-like pattern also appeared away from the center and gradually faded.
The researchers quantified the distribution’s width using its second spatial moment. At a lower laser fluence of 0.16 mJ/cm², the distribution expanded rapidly at first and then followed a more conventional diffusion pattern beyond roughly 500 ps. At the higher fluence of 3.0 mJ/cm², the total expansion was smaller despite the stronger excitation. The second moment and calculated transport rate also showed oscillations at early times. These oscillations were not observed in the lower-fluence measurement.
This contrast matters because increasing excitation does not simply produce a proportionally faster spread. At high carrier density, the measured dynamics include processes that oppose or redirect the initial expansion. The oscillatory response is temporary: as the system evolves, it fades and normal diffusion becomes more prominent.
How charge separation can oppose diffusion
The proposed explanation centers on the electric field created as electrons and holes separate. Following optical excitation, the built-in field in silicon’s subsurface space-charge region drives electrons toward the surface and holes into the bulk. The carriers also diffuse outward from the excited region. Because electrons diffuse faster than holes in the model, the two populations become spatially offset, producing a changing charge distribution and a transient electric field.
That field can act against further separation and slow carrier movement. Its influence becomes more apparent when the carrier density is high enough for electrostatic interactions to compete with diffusion. As carriers lose excess energy to the lattice and recombination reduces the free-carrier population, the initial rapid transport and the electrostatic effect both diminish.
To examine this interpretation, the researchers developed an advection-diffusion model coupled to Poisson’s equation. The model included carrier diffusion, motion driven by electric fields and a time-dependent carrier temperature. Simulations reproduced key qualitative features, including slowed expansion and subtle oscillations. The simulated oscillations were smaller than those measured experimentally, so the model does not capture every aspect of the observed response. The authors noted that simplifying assumptions, including the treatment of recombination, could contribute to the difference.
Implications for semiconductor device engineering
The findings show why a single diffusion coefficient may be insufficient to describe carriers under intense, transient excitation. In the high-fluence regime, measured transport reflects a combination of energy-dependent diffusion, charge separation and electrostatic feedback. A distribution may briefly broaden, partially contract or show oscillatory changes before approaching ordinary diffusion.
For device development, the broader lesson is methodological as well as physical: measurement approaches should match the scale and location of the process under investigation. Surface-sensitive imaging can reveal behavior that spatially averaged measurements may obscure. At the same time, interpreting contrast requires attention to what the instrument measures—in this case, changes in surface potential inferred through secondary-electron emission.
The study does not establish a universal transport rule for all silicon devices or operating conditions. It examines a specific, highly excited p-type silicon sample. Its value is in demonstrating directly that carrier transport in this regime can be nonlinear and oscillatory, and in linking that behavior to evolving electric fields generated by electron-hole separation.
Conclusion
Ultrafast electron microscopy captured a high-excitation regime in which hot carrier transport in silicon slowed and displayed transient oscillations. The observations, together with advection-diffusion modeling, indicate that electrostatic interactions between separated electrons and holes can hinder diffusion. For semiconductor engineers and technical teams, the practical implication is clear: under strong excitation, carrier dynamics may require spatially resolved analysis and models that account for coupled transport and electric fields—not diffusion alone.