Heterostructures formed by stacking layered materials require atomically clean interfaces. However, contaminants are usually trapped between the layers, aggregating into randomly located blisters, incompatible with scalable fabrication processes. Here we report a process to remove blisters from fully formed heterostructures. Our method is over an order of magnitude faster than those previously reported and allows multiple interfaces to be cleaned simultaneously. We fabricate blister-free regions of graphene encapsulated in hexagonal boron nitride with an area ~ 5000 μm2, achieving mobilities up to 180,000 cm2 V−1 s−1 at room temperature, and 1.8 × 106 cm2 V−1 s−1 at 9 K. We also assemble heterostructures using graphene intentionally exposed to polymers and solvents. After cleaning, these samples reach similar mobilities. This demonstrates that exposure of graphene to process-related contaminants is compatible with the realization of high mobility samples, paving the way to the development of wafer-scale processes for the integration of layered materials in (opto)electronic devices.

Optical harmonic generation occurs when high intensity light (>10 W m) interacts with a nonlinear material. Electrical control of the nonlinear optical response enables applications such as gate-tunable switches and frequency converters. Graphene displays exceptionally strong light-matter interaction and electrically and broadband tunable third-order nonlinear susceptibility. Here, we show that the third-harmonic generation efficiency in graphene can be increased by almost two orders of magnitude by controlling the Fermi energy and the incident photon energy. This enhancement is due to logarithmic resonances in the imaginary part of the nonlinear conductivity arising from resonant multiphoton transitions. Thanks to the linear dispersion of the massless Dirac fermions, gate controllable third-harmonic enhancement can be achieved over an ultrabroad bandwidth, paving the way for electrically tunable broadband frequency converters for applications in optical communications and signal processing.

Van der Waals heterostructures have emerged as promising building blocks that offer access to new physics, novel device functionalities, and superior electrical and optoelectronic properties [1][2][3][4][5][6][7]. Applications such as thermal management, photodetection, light emission, data communication, high-speed electronics and light harvesting [8][9][10][11][12][13][14][15][16] require a thorough understanding of (nanoscale) heat flow. Here, using time-resolved photocurrent measurements we identify an efficient out-of-plane energy transfer channel, where charge carriers in graphene couple to hyperbolic phonon polaritons [17][18][19] in the encapsulating layered material. This hyperbolic cooling is particularly efficient, giving picosecond cooling times, for hexagonal BN, where the high-momentum hyperbolic phonon polaritons enable efficient near-field energy transfer. We study this heat transfer mechanism through distinct control knobs to vary carrier density and lattice temperature, and find excellent agreement with theory without any adjustable parameters. These insights may lead to the ability to control heat flow in van der Waals heterostructures.Owing to its large in-plane thermal conductivity, graphene has been suggested as material for the thermal management of nanoscale devices [8]. At the same time, graphene is well-known for its ability to convert incident light into electrical heat, i.e. hot electrons that can be used to generate photocurrent, with applications in photodetection, data communication and light harvesting [10,20,21]. Understanding, and ultimately controlling, heat flow in graphene-van der Waals heterostructures is therefore of paramount importance. For example, a short cooling time of graphene hot carriers is advantageous for thermal management and for high switching rates of photodetectors (PDs) for data communication, whereas a long cooling time is favorable for photodetection sensitivity [10,20,21]. Of particular relevance are heterostructure devices that contain high-quality graphene encapsulated by layered materials, such as hexagonal BN (hBN) and MoS 2 , which have the potential to crucially improve the performance of electronic and optoelectronic devices [1,2]. * Electronic address: Correspondence: klaas-jan.tielrooij@icfo.eu, frank.koppens@icfo.eu ‡ Equal contribution.A number of cooling pathways for graphene carriers have been proposed, involving among others strongly coupled optical phonons [22][23][24], acoustic phonons [25][26][27][28], substrate phonons [29] and plasmons [30] (see also Appendix 1). Here, using several experimental approaches, we show that cooling in graphene encapsulated by hBN is governed by out-of-plane coupling of graphene electrons to special polar phonon modes that occur in layered materials (LMs): hyperbolic phonon polaritons, where xx zz < 0, with xx and zz the permittivity parallel and perpendicular to the LM plane. Owing to this property, these materials carry deep sub-wavelength, raylike optical phonon polaritons. For hBN, within the two Rest...

‡) equal contributionWe report high room-temperature mobility in single layer graphene grown by Chemical Vapor Deposition (CVD) after wet transfer on SiO2 and hexagonal boron nitride (hBN) encapsulation. By removing contaminations trapped at the interfaces between single-crystal graphene and hBN, we achieve mobilities up to∼ 70000cm 2 V −1 s −1 at room temperature and∼ 120000cm 2 V −1 s −1 at 9K. These are over twice those of previous wet transferred graphene and comparable to samples prepared by dry transfer. We also investigate the combined approach of thermal annealing and encapsulation in polycrystalline graphene, achieving room temperature mobilities∼ 30000cm 2 V −1 s −1 . These results show that, with appropriate encapsulation and cleaning, room temperature mobilities well above 10000cm 2 V −1 s −1 can be obtained in samples grown by CVD and transferred using a conventional, easily scalable PMMA-based wet approach. * al515@cam.ac.uk[1] S. M. Sze and K. K. Ng, Physics of semiconductor devices, (John Wiley & Sons, 2006).

Uncooled terahertz photodetectors (PDs) showing fast (ps) response and high sensitivity (noise equivalent power (NEP) < nW/Hz1/2) over a broad (0.5–10 THz) frequency range are needed for applications in high-resolution spectroscopy (relative accuracy ∼10–11), metrology, quantum information, security, imaging, optical communications. However, present terahertz receivers cannot provide the required balance between sensitivity, speed, operation temperature, and frequency range. Here, we demonstrate uncooled terahertz PDs combining the low (∼2000 k B μm–2) electronic specific heat of high mobility (>50 000 cm2 V–1 s–1) hexagonal boron nitride-encapsulated graphene, with asymmetric field enhancement produced by a bow-tie antenna, resonating at 3 THz. This produces a strong photo-thermoelectric conversion, which simultaneously leads to a combination of high sensitivity (NEP ≤ 160 pW Hz–1/2), fast response time (≤3.3 ns), and a 4 orders of magnitude dynamic range, making our devices the fastest, broad-band, low-noise, room-temperature terahertz PD, to date.

Hexagonal boron nitride (h-BN) is the only known material aside from graphite with a structure composed of simple, stable, non-corrugated atomically thin layers. While historically used as lubricant in powder form, h-BN layers have become particularly attractive as an ultimately thin insulator. Practically all emerging electronic and photonic device concepts rely on h-BN exfoliated from small bulk crystallites, which limits device dimensions and process scalability.Here, we address this integration challenge for mono-layer h-BN via a chemical vapour deposition process that enables crystal sizes exceeding 0.5 mm starting from commercial, reusable platinum foils, and in unison allows a delamination process for easy and clean layer transfer. We demonstrate sequential pick-up for the assembly of graphene/h-BN heterostructures with atomic layer precision, while minimizing interfacial contamination. Our process development builds on a systematic understanding of the underlying mechanisms. The approach can be readily combined with other layered materials and opens a scalable route to h-BN layer integration and reliable 2D material device layer stacks.

Optical harmonic generation occurs when high intensity light (> 10 10 W/m 2 ) interacts with a nonlinear material. Electrical control of the nonlinear optical response enables applications such as gate-tunable switches and frequency converters. Graphene displays exceptionally strong-light matter interaction and electrically and broadband tunable third order nonlinear susceptibility. Here we show that the third harmonic generation efficiency in graphene can be tuned by over two orders of magnitude by controlling the Fermi energy and the incident photon energy. This is due to logarithmic resonances in the imaginary part of the nonlinear conductivity arising from multi-photon transitions. Thanks to the linear dispersion of the massless Dirac fermions, ultrabroadband electrical tunability can be achieved, paving the way to electrically-tuneable broadband frequency converters for applications in optical communications and signal processing.The response of a material to interaction with an optical field can be described by its polarization[1]:where E is the incident electric field and ǫ 0 is the permittivity of free space. χ (1) (dimensionless) is the linear susceptibility, while the tensorsare the second-and third-order nonlinear susceptibilities [2]. Thanks to the nonlinear terms of P , new frequencies can be generated inside a material due to harmonic generation[3] and frequency mixing[4]. E.g., in Second Harmonic Generation (SHG) an incident electromagnetic wave with angular frequency ω 0 = 2πν, with ν the photon frequency, generates via χ (2) a new electromagnetic wave with frequency 2ω 0 [3]. The SHG efficiency (SHGE) is defined as the ratio between the SH intensity and the intensity of the incoming light. Analogously, Third Harmonic Generation (THG) is the emission of a photon with energy triple that of the incident one. The THG efficiency (THGE) is defined as the ratio between the TH intensity and the intensity of the incoming light. Second-order nonlinear processes are also known as three-wave-mixing, as they mix two optical fields to produce a third one [5]. Third-order nonlinear processes are known as four-wave-mixing (FWM)[5], as they mix three fields to produce a fourth one. Nonlinear optical effects are exploited in a variety of applications, including laser technology[6], material processing[7] and telecommunications [8]. E.g., to generate new photon frequencies (532nm from SHG in a Nd:YAG laser at 1.06µm) [9] or broadly tuneable ultrashort pulses (fs-ps) by optical parametric amplifiers (OPAs)[10] and optical parametric oscillators (OPOs) [11]. High harmonic generation is also used for extreme UV light [12] and attosecond pulse generation [13], while difference frequency generation is used to create photons in the THz range [14].Second order nonlinear effects can only occur in materials without inversion symmetry, while third order ones occur in any system independent of symmetry[15], and they thus represent the main intrinsic nonlinear response for most materials. THG intensity enhancement was achieved...

We combine a graphene mode-locked oscillator with an external compressor and achieve∼29fs pulses with∼52mW average power. This is a simple, low-cost, and robust setup, entirely fiber based, with no free-space optics, for applications requiring high temporal resolution.Ultrafast light pulses in the femtosecond range are needed for advanced photonics applications. E.g. in pump-probe spectroscopy, photophysical and photochemical relaxation processes are monitored by exciting a sample with an ultrashort light pulse. The maximum temporal resolution is determined by the duration, ∆τ , of the pulse. This is usually defined as the full width at half maximum (FWHM) of its intensity profile in the time domain, I(t) [1]. Alternatively ∆τ may be defined by the number of oscillation periods of the electric field carrier wave (optical cycles) within the pulse[2] N = ∆τ T0 = ν 0 ∆τ , where T 0 is the optical cycle of frequency ν 0 . The ultimate pulse duration is set by a single cycle of light, i.e. T 0 , given by[2] λ c , where λ is the wavelength and c is the speed of light. Finally, the uncertainty relation ∆ν∆τ ≃ 1 π provides a measure of the minimum frequency bandwidth ∆ν required for an ultrashort pulse formation[2], i.e. the broader the bandwidth, the shorter the supported pulse. In the visible and near infrared (NIR), T 0 lies, e.g, between 2fs at λ ∼600nm and 5fs at λ ∼1.5µm, which set the ultimate speed limit for devices operating in this wavelength range. Achieving shorter pulses therefore requires moving to shorter wavelengths.Pulses as short as 2-cycles can be generated directly from laser cavities using passive mode-locking [1][2][3]. Ti:Saphire lasers have become established tools for few-cycle generation [2], with the shortest pulses produced to date having ∆τ ∼5fs[4] at a centre wavelength, λ 0 ∼800nm, corresponding to less than 2-cycles, with spectral width ∆λ ∼600nm [4]. Ti:Saphire lasers able to generate few-cycle durations are typically optimized to make use of the maximum ∆λ gain available[2], consequently they have no wavelength tunability[2]. Tunable Ti:Saphire operate with a much longer pulse duration, e.g. ∆τ ∼150fs in a typical∼680-1080nm commercially available spectral range [5]. Tunable few-cycle pulses can be achieved by exploiting nonlinear optical effects in optical parametric amplifiers (OPAs). These can be described by expressing the polarization (P ) as a power series in the applied optical field (E) [6,7]:, where ǫ 0 is the free space permittivity, χ (1) is the linear and χ (2) and χ (3) are the second-and third-order nonlinear susceptibilities. OPAs are optical amplifiers based on the χ (2) nonlinearity of a crystal [6][7][8], in a process, called parametric [6,7], where there is no net transfer of energy and momentum between E and the crystal [6,7]. This can be visualized, by considering energy transfer from a pump pulse of frequency ω p to two pulses of lower frequencies ω s and ω i , called signal and idler [6,7], with the requirement ω p = ω s + ω i [6,7]. Under this condition, OPAs can...