Ultrafast terahertz conductivity in epitaxial graphene nanoribbons: an interplay between photoexcited and secondary hot carriers

This study utilizes optical pump-terahertz probe spectroscopy to reveal that ultrafast photoconductivity in epitaxial graphene nanoribbons exhibits a non-monotonic fluence dependence driven by an interplay between negative-sign conductivity from secondary hot carriers at low fluences and positive-sign conductivity from excess carriers at high fluences, resulting in complex variations in mobility and plasmonic resonance.

The Gist

The Big Picture: A Dance of Hot Particles

Imagine a sheet of graphene (a material made of a single layer of carbon atoms) not as a flat sheet, but as a giant, crowded dance floor. The people on this floor are electrons (the dancers).

In this experiment, scientists shined a powerful, ultra-fast laser pulse (the "pump") onto these dancers. This laser gave them a sudden burst of energy, making them "hot" and hyperactive. Then, they used a special "THz probe" (like a high-speed camera) to watch how the dancers moved and interacted with each other in the trillionths of a second that followed.

The goal was to understand how this "dance" changes depending on how hard they hit the floor with the laser.

The Setup: A Narrow Hallway

The researchers didn't use a giant open ballroom. They cut the graphene into very narrow strips called nanoribbons (about 3.4 micrometers wide).

• Think of it like this: Instead of an open field, the dancers are in a long, narrow hallway with walls on both sides.
• The Twist: They shined the laser light either parallel to the hallway (along the length) or perpendicular to it (across the width). This helped them see how the walls affected the dancers' movement.

The Two Regimes: Low Energy vs. High Energy

The paper describes two very different behaviors depending on how much energy (fluence) the laser delivered.

1. The "Warm-Up" Phase (Low Laser Power)

The Scenario: You give the dancers a gentle nudge.
What happens:

• The "Secondary" Effect: The few dancers who get hit by the laser don't stay hot for long. They immediately bump into their neighbors, sharing their energy.
• The Analogy: Imagine a few people in a crowd start clapping loudly. They don't just clap; they get everyone else around them to clap too. The original clappers cool down, but the whole crowd gets a little warmer and more energetic.
• The Result: The scientists call these the "secondary hot carriers." Because the original dancers cooled down by sharing energy, the electrical conductivity actually went down (negative signal). The dancers were just getting warmer, not necessarily moving faster in a specific direction.
• The Walls: At this low energy, some dancers got stuck in "corners" or bumps on the floor (defects in the material). They were localized, meaning they couldn't move freely down the hallway.

2. The "Overload" Phase (High Laser Power)

The Scenario: You hit the dance floor with a massive, high-energy laser blast.
What happens:

• The "Excess" Effect: There are so many new dancers (electrons) created by the laser that there aren't enough "regular" dancers left to absorb all the energy. The new dancers stay hot and energetic on their own.
• The Analogy: It's like a mosh pit where everyone is jumping so high they can't be stopped. The energy is so intense that the dancers can now jump over the small bumps and obstacles on the floor that used to trap them.
• The Result: The "secondary" effect fades, and the "excess" carriers (the new, hot dancers) take over. They conduct electricity very well (positive signal).
• The "Hot Phonon" Bottleneck: The dancers are moving so fast they are constantly bumping into the floor (the substrate), creating heat waves (phonons). But the floor can't cool down fast enough. It's like a traffic jam of heat. This makes the dancers stay hot for longer, slowing down their overall relaxation.

The Surprising Twist: The "Lift"

One of the coolest findings is about the localization (the dancers getting stuck).

• At low power: The dancers are easily trapped by small bumps in the hallway.
• At high power: The dancers are so hot and energetic that they simply jump over the traps. The "stuck" behavior disappears completely. It's as if the heat gave them super-jump boots, allowing them to ignore the obstacles that were blocking them before.

The "Goldilocks" Zone

The researchers found a sweet spot in the middle.

• If you hit it too lightly, the dancers are just warming up and getting stuck.
• If you hit it too hard, the system gets clogged with too much heat (the bottleneck), and the dancers start crashing into each other too much, slowing them down again.
• The Sweet Spot: There is a specific amount of laser power where the dancers move the most efficiently. This is where the "mobility" (how fast they can travel) peaks.

Why Does This Matter?

This isn't just about watching carbon atoms dance. It helps engineers design super-fast electronics that work with light (optoelectronics).

• By understanding how these "hot" electrons behave, we can build devices that process information at Terahertz speeds (trillions of times per second), which is much faster than the computers we use today.
• It shows us that by simply turning up the "volume" (laser intensity) on a material, we can switch its behavior from "stuck and slow" to "free and fast," and then back to "clogged and slow."

Summary in One Sentence

The paper reveals that when you zap graphene with a laser, the electrons act like a crowd of people: at low energy, they share heat and get stuck in corners; at high energy, they get so hot they jump over obstacles and move freely, until they get too hot and start jamming up the system.

Technical Summary

1. Problem Statement

Graphene exhibits unique hot-carrier dynamics due to its massless Dirac fermion nature, low electronic heat capacity, and strong carrier-carrier interactions. While ultrafast carrier relaxation (thermalization, cooling, recombination) is well-studied, the nonlinear scaling of terahertz (THz) photoconductivity across a broad range of optical pump fluences remains complex. Specifically, there is a need to distinguish between the contributions of:

• Secondary hot carriers: Equilibrium carriers heated by carrier-carrier scattering (dominant at low fluences).
• Directly photoexcited (excess) carriers: New electron-hole pairs generated by the pump (dominant at high fluences).

Furthermore, the interplay between these mechanisms and structural inhomogeneities (such as lithography-induced defects) in graphene nanoribbons (GNRs) requires quantitative analysis to understand how carrier mobility, scattering times, and plasmonic resonances evolve under intense excitation.

2. Methodology

The researchers employed Optical Pump-Terahertz Probe (OPTP) spectroscopy to investigate the ultrafast photo-induced charge transport in epitaxial graphene nanoribbons.

• Sample Preparation:
  • Graphene was grown epitaxially on a 6H-SiC substrate via thermal decomposition.
  • Hydrogen intercalation was used to decouple the graphene, creating quasi-free-standing single-layer graphene.
  • Arrays of nanoribbons were fabricated using electron-beam lithography with a width ($w$) of 3.4 µm and a gap ($g$) of 0.5 µm (period $L$ = 3.9 µm).
• Experimental Setup:
  • Pump: 40 fs, 800 nm optical pulses from a Ti:sapphire amplifier (5 kHz repetition rate).
  • Probe: THz pulses generated and detected via 1 mm thick ZnTe crystals.
  • Geometry: Collinear pump-probe geometry with THz polarization measured both parallel and perpendicular to the ribbon axis.
  • Fluence Range: Experiments covered a broad range of absorbed pump fluences ($\phi$), from low ($\sim 0.25 \times 10^{12}$ photons/cm²) to high ($\sim 10 \times 10^{12}$ photons/cm²).
• Data Analysis:
  • Transient THz waveforms were analyzed at a 2 ps delay to extract photoconductivity spectra.
  • A global fitting strategy was applied using a modified Drude-Smith model (for parallel polarization, accounting for localization) and a Lorentz model (for perpendicular polarization, accounting for plasmonic resonances).
  • Key parameters extracted included scattering time ($\tau$), carrier temperature ($T_c$), chemical potential ($\mu$), Drude weight ($D$), and localization parameter ($c$).

3. Key Contributions

• Mechanism Identification: The study definitively attributes the nonlinear THz conductivity response not to Pauli blocking of interband transitions, but to the interplay between secondary hot carriers (heated equilibrium carriers) and directly photoexcited excess carriers.
• Regime Differentiation: The authors quantitatively distinguish three distinct operational regimes based on pump fluence, revealing non-monotonic behaviors in mobility and plasmonic frequency.
• Localization Dynamics: The paper provides evidence that weak carrier localization (caused by lithography contamination) is fluence-dependent; it is lifted at high fluences as increased carrier kinetic energy allows charges to overcome potential barriers.
• Quantitative Modeling: A comprehensive model linking carrier temperature, chemical potential, and scattering time to the observed conductivity saturation and plasmonic shifts is established.

4. Key Results

A. Non-Monotonic Photoconductivity Scaling

• Low Fluence ($\phi \lesssim 1 \times 10^{12}$ photons/cm²):
  • The photoconductivity is negative (increased transparency) and scales linearly with fluence.
  • Mechanism: The pump energy heats the existing equilibrium carriers (secondary hot carriers). The rise in carrier temperature reduces the scattering time slightly but increases mobility slightly due to the specific temperature dependence of the scattering mechanism.
  • Plasmonics: The plasmonic resonance exhibits a red shift due to a decrease in the Drude weight (caused by the drop in chemical potential as carriers are heated).
• High Fluence ($\phi \gtrsim 4 \times 10^{12}$ photons/cm²):
  • The photoconductivity saturates and eventually shows a positive contribution.
  • Mechanism: The density of equilibrium carriers is insufficient to absorb all deposited energy. Directly generated excess carriers dominate, increasing the Drude weight.
  • Mobility: Carrier mobility decreases significantly (by a factor of ~4) due to a sharp reduction in scattering time (increased scattering rate) despite the rising Drude weight.
  • Plasmonics: The plasmonic resonance undergoes a significant blue shift due to the increased carrier density (Drude weight) from excess carriers.

B. Carrier Dynamics and Hot Phonon Bottleneck

• Carrier Cooling: At high fluences, the carrier relaxation time (decay time) increases and does not saturate. This is attributed to the hot phonon bottleneck effect, where optical phonons cannot decay into acoustic phonons fast enough, leading to re-absorption of energy by carriers and slowed relaxation.
• Localization: At low fluences, a weak localization effect ($c < 0$) is observed, particularly for parallel polarization. As fluence increases, the carrier temperature rises (reaching ~1400 K), allowing carriers to overcome the low-energy potential barriers caused by lithography defects. Consequently, the localization parameter $c$ approaches 0 (Drude behavior) at high fluences.

C. Parameter Evolution

• Carrier Temperature: Increases linearly at low fluences, then saturates around 1400 K (corresponding to the SiC optical phonon energy of 120 meV).
• Chemical Potential: Decreases at low fluences (due to heating) but increases at high fluences (due to excess carrier generation).
• Scattering Time: Decreases monotonically with increasing fluence due to enhanced carrier-phonon and carrier-carrier scattering.

5. Significance

This work provides a critical quantitative framework for understanding nonlinear THz responses in graphene nanostructures. The findings are significant for:

1. Ultrafast Optoelectronics: Understanding the saturation limits and switching speeds of graphene-based THz modulators and detectors.
2. Hot Carrier Devices: Clarifying the transition between secondary heating and direct excitation regimes is essential for designing efficient hot-carrier solar cells and photodetectors.
3. Plasmonic Applications: The ability to tune plasmonic resonance frequencies (red-shift to blue-shift) via pump fluence offers a mechanism for dynamic, all-optical control of plasmonic devices.
4. Material Characterization: The study demonstrates how ultrafast spectroscopy can distinguish between intrinsic carrier dynamics and extrinsic effects like defect-induced localization, providing a diagnostic tool for sample quality.

In summary, the paper elucidates the complex competition between carrier heating and carrier generation in graphene nanoribbons, revealing that the macroscopic THz response is a direct result of the microscopic balance between secondary hot carriers and excess photoexcited carriers.