Significantly Enhanced Thermoelectric Performance of Graphene through Atomic-Scale Defect Engineering via Mobile Hot-Wire Chemical Vapor Deposition Systems

Abstract

Over the years, numerous studies have attempted to develop two-dimensional (2D) materials for improving both the applicability and performance of thermoelectric devices. Among the 2D materials, graphene is one of the promising candidates for thermoelectric materials owing to its extraordinary electrical properties, flexibility, and nontoxicity. However, graphene synthesized through traditional methods suffers from a low Seebeck coefficient and high thermal conductivity, resulting in an extremely low thermoelectric figure of merit (ZT). Here, we present an atomic-scale defect engineering strategy to improve the thermoelectric properties of graphene using embedded high-angle tilt boundary (HATB) domains in graphene films. These HATB domains serve as both energy filtering sites to filter out lower-energy charge carriers and scattering sites for phonons. Compared to the conventionally grown chemical vapor deposited graphene, the graphene with HATB domains shows an improved Seebeck coefficient (50.1 vs 21.1 μV K −1 ) and reduced thermal conductivity (382 vs 952 W m −1 K −1 ), resulting in a ZT value that is ∼7 times greater at 350 K. This defect engineering strategy is promising not only for graphene-based materials but also for 2D materials, in general, where further research and optimization could overcome the limitations of conventional bulk thermoelectric materials in energy-harvesting systems.

Improved Crystallinity of Graphene Grown on Cu/Ni (111) through Sequential Mobile Hot-Wire Heat Treatment

“…In our previous work, we reported that HATB domains can induce effective phonon scattering; consequently, the k of graphene grown on polycrystalline Cu showed a significantly low value (∼381 W m −1 K −1 at 350 K), which is lower than that of conventional polycrystalline graphene (∼952 W m −1 K −1 at 350 K). 21,35 In contrast, the measured k of graphene grown on Cu/Ni (111) was ∼1348 W m −1 K −1 at 350 K, which is a remarkable improvement compared to that on polycrystalline Cu (∼381 W m −1 K −1 ) and Cu (111) (∼646 W m −1 K −1 ). This result provides evidence of the crystallinity enhancement for graphene grown on Cu/Ni (111) in terms of structural defects which can cause phonon scattering.…”

“…A promising strategy to overcome the obstacles of the aforementioned approaches is to locally recrystallize the pregrown graphene nanograins unidirectionally from one edge to the other edge during growth before all the nanograins globally coalesce into a single film, as in the Czochralski method and floating-zone recrystallization, which have been known to form single crystals of three-dimensional materials. − However, this method is challenging for a conventional CVD system because it cannot separate nucleation from subsequent growth at the fixed temperature of the whole volume in the growth chamber. Our invention of a new CVD system by providing additional local heat energy with a mobile hot wire (MHW) over the growth substrate whose temperature is only high enough for nucleation and not for growth can enable this strategy. − In our previous work, we reported MHW-CVD growth of defect-controlled single-layer graphene on polycrystalline Cu , and Cu–Ni alloy substrates . During graphene growth in the MHW-CVD system, the pregrown graphene nanograins with random orientation on a catalytic metal substrate rotate and merge with each other along a specific orientation because the grains can align to form energetically stable bonds with neighboring grains depending on the growth parameters, such as the substrate temperature ( T sub ), hot-wire scan speed ( V w ) and temperature ( T w ), and flow rate of the carbon feedstock.…”

“…After hot-wire scanning, HATB domains of ∼30° were embedded in a large graphene domain (Figure a and Figures S3–S5). We previously reported that HATB domains were generated on polycrystalline Cu due to the lack of supplied thermal energy required to rotate the graphene grains located at a high angle with respect to neighboring grains. , The presence of the HATB domains was characterized by a false-colored DF-TEM image (Figure b). To remove the remaining HATB domains, we focused on modification of the catalytic metal substrate enabling graphene growth in terms of grain recrystallization, such as grain rotation, reconstruction by carbon adsorption/desorption, and merging.…”

“…To evaluate the thermal and electrical performance of the graphene grown on each substrate, we measured the thermal conductivity and electrical carrier mobility of graphene. At first, the thermal conductivity ( k ) of single-layer graphene was measured using a microfabricated thermoelectric measurement platform (MTMP) setup. ,− Figure a shows digital and SEM images of the graphene transferred onto the MTMP. A magnified SEM image (Figure b) indicates that the temperature gradient between the hot and cold zones of the transferred graphene is finely controlled by a micro-Joule heater.…”

“…system by providing additional local heat energy with a mobile hot wire (MHW) over the growth substrate whose temperature is only high enough for nucleation and not for growth can enable this strategy. 17−21 In our previous work, we reported MHW-CVD growth of defect-controlled single-layer graphene on polycrystalline Cu 17,21 and Cu−Ni alloy substrates. 19 During graphene growth in the MHW-CVD system, the pregrown graphene nanograins with random orientation on a catalytic metal substrate rotate and merge with each other along a specific orientation because the grains can align to form energetically stable bonds with neighboring grains depending on the growth parameters, such as the substrate temperature (T sub ), hot-wire scan speed (V w ) and temperature (T w ), and flow rate of the carbon feedstock.…”