Begona Abad Mayor1 Joshua Knobloch1 Travis Frazer1 Jorge Nicolas Hernandez Charpak1 Hiu Yan Cheng2 3 Alex Grede2 5 Andrew Glaid2 5 Noel Giebink2 5 Tom Mallouk2 3 4 Pratibha Mahale4 Weinan Chen2 6 Yihuang Xiong2 6 Ismaila Dabo2 6 Vincent Crespi2 3 6 Disha Talreja2 Venkat Gopalan2 6 John Badding2 3 6 Henry Kapteyn1 Margaret Murnane1

1, JILA - CU Boulder, Boulder, Colorado, United States
2, Materials Research Institute, The Pennsylvania State University, State College, Pennsylvania, United States
3, Department of Chemistry, The Pennsylvania State University, State College, Pennsylvania, United States
5, Department of Electrical Engineering, The Pennsylvania State University, State College, Pennsylvania, United States
4, Department of Biochemistry and Molecular Biology, The Pennsylvania State University, State College, Pennsylvania, United States
6, Department of Materials Science and Engineering, The Pennsylvania State University, State College, Pennsylvania, United States

Nanoscale phononic metamaterials make it possible to engineer the thermal, magnetic, and electronic properties of materials, which is essential for nanoelectronics, thermoelectric and data storage devices, or nanoparticle-mediated thermal therapies [1]. Specifically, nanoscale metalattices are a powerful bottom-up approach to tune the propagation of high frequency phonons - highly ordered nanoscale opal templates allow for precise control over nanoscale structure in which different materials can be infiltrated. Moreover, these metamaterials can be organized into hierarchical structures on length scales from nanometers to micrometers which enable unique properties that cannot be accessed using bulk/layered materials [2,3]. In the case of nanoscale thermal transport, metalattices are a powerful approach to further advance our understanding, which is critical since macroscopic diffusive models completely break down at dimensions that are comparable to the mean free path of the heat carriers [4,5]. However, characterizing energy flow in these metalattices is extremely challenging. Most current techniques rely on visible light, which is limited in wavelength to probing heat flow away from nanostructures >100s of nanometers. We can overcome this limitation by utilizing coherent extreme ultraviolet (EUV) light generated by tabletop high harmonic generation (HHG) [6]. The short pulse duration (≈10fs) and wavelength (≈30nm) of tabletop HHG sources are an excellent match for probing the intrinsic length- and time-scales relevant to nanoscale dynamics. In this unique EUV nanometrology technique, we use an ultrafast 800nm femtosecond laser to impulsively heat periodic arrays of nickel nanolines deposited on top of a silicon metalattice. We then probe the cooling of the nanolines, and the resulting thermal transport properties of the metalattice, by monitoring the thermally-induced surface deformation using coherent EUV light. In this work, we show how the thermal transport in silicon metalattices is modified due to the metalattice structure. First, we observe a slow thermal decay of the nanolines through the metalattice, suggesting very low thermal conductivity of the metalattice. Further analysis indicates that, not only do these metalattices have lower thermal conductivity than expected from macroscopic predictions, but that the heat flow is a function of the geometry of the heat sources. In addition, these findings are supported by equilibrium Green–Kubo atomistic simulations which show that silicon metalattices are capable of significantly reducing the thermal conductivity below the prediction of continuum models. This suggests that nanostructured metalattices may be able impede heat flow even more than had been initially thought. The ability to impede the flow of phonons, while allowing electrical current to flow, can dramatically impact applications such as optimized thermoelectric materials, as well as providing routes for enhanced functionality of other nanodevices.

[1] Maldovan, M., Nature 503, 209–217 (2013).
[2] Liu, Y. et al., Nano Lett. 18(1), 546-552 (2018).
[3] Han, J.E. et al. Phys. Rev. Lett. 86(4), 696-699 (2001).
[4] M. E. Siemens et al., Nat. Mater., 9(1), 26–30 (2010).
[5] K. M. Hoogeboom-Pot et al., Proc. Natl. Acad. Sci., 112(16), 4846–4851 (2015).
[6] A. Rundquist et al., Science 280 (5368), 1412–1415 (1998).