Oxygen stoichiometry of adhesion layer dictates thermal boundary conductance: Paper published in APL – Congrats Hans Olson!

Congrats to Hans Olson for his recent publication demonstrating that the oxygen stoichiometry of a Ti adhesion layer between an Au films and a non-metal substrate will impact the thermal boundary conductance across Au/Ti/substrate interfaces.  To maximize TBC, the Ti layer should be a pure metal (i.e., no oxygen defects), which is rarely achieved in high vacuum conditions, but can be achieved by depositing Ti layers in ultra-high vacuum.

Olson, D.H., Freedy, K.M., McDonnell, S., Hopkins, P.E., “The influence of titanium adhesion layer layer oxygen stoichiometry on thermal boundary conductance at gold contacts,” Applied Physics Letters 112, 171602 (2018). PDF (Supporting Information).


We experimentally demonstrate the role of oxygen stoichiometry on the thermal boundary conductance across Au/TiOx/substrate interfaces. By evaporating two different sets of Au/TiOx/ substrate samples under both high vacuum and ultrahigh vacuum conditions, we vary the oxygen composition in the TiOx layer from 0 x 2.85. We measure the thermal boundary conductance across the Au/TiOx/substrate interfaces with time-domain thermoreflectance and characterize the interfacial chemistry with x-ray photoemission spectroscopy. Under high vacuum conditions, we speculate that the environment provides a sufficient flux of oxidizing species to the sample surface such that one essentially co-deposits Ti and these oxidizing species. We show that slower deposi- tion rates correspond to a higher oxygen content in the TiOx layer, which results in a lower thermal boundary conductance across the Au/TiOx/substrate interfacial region. Under the ultrahigh vacuum evaporation conditions, pure metallic Ti is deposited on the substrate surface. In the case of quartz substrates, the metallic Ti reacts with the substrate and getters oxygen, leading to a TiOx layer. Our results suggest that Ti layers with relatively low oxygen compositions are best suited to maximize the thermal boundary conductance.

D. H. Olson would like to thank the Virginia Space Grant Consortium (VSGC) for their continued funding and support. We also appreciate the support from the Army Research Office, Grant Nos. W911NF-16-1-0320 and W911NF-16-1-0406.

Low frequency phonon transport in C60 fullerite: Paper published in Phys. Rev. B – Congrats Ash Giri!

Congrats to Ash Giri for his recent publication in Phys. Rev. B on the pronounced low-frequency vibrational thermal transport in C60 fullerite realized through pressure-dependent molecular dynamics simulations.

Giri, A., Hopkins, P.E., “Pronounced low-frequency vibrational thermal transport in C60 fullerite realized through pressure-dependent molecular dynamics simulations,” Physical Review B 96, 220303(R) (2017). PDF.

Abstract: Fullerene condensed-matter solids can possess thermal conductivities below their minimum glassy limit while theorized to be stiffer than diamond when crystallized under pressure. These seemingly disparate extremes in thermal and mechanical properties raise questions into the pressure dependence on the thermal conductivity of C60 fullerite crystals, and how the spectral contributions to vibrational thermal conductivity changes under applied pressure. To answer these questions, we investigate the effect of strain on the thermal conductivity of C60 fullerite crystals via pressure-dependent molecular dynamics simulations under the Green-Kubo formalism. We show that the thermal conductivity increases rapidly with compressive strain, which demonstrates a power-law relationship similar to their stress-strain relationship for the C60 crystals. Calculations of the density of states for the crystals under compressive strains reveal that the librational modes characteristic in the unstrained case are diminished due to densification of the molecular crystal. Over a large compression range (0–20 GPa), the Leibfried-Schlömann equation is shown to adequately describe the pressure dependence of thermal conductivity, suggesting that low-frequency intermolecular vibrations dictate heat flow in the C60 crystals. A spectral decomposition of the thermal conductivity supports this hypothesis.

We would like to thank the Army Research Office for support (Grant No. W911NF-16-1-0320).

High temperature TDTR up to 1000 K using HfN transducers: Paper published in Appl. Phys. Lett – Congrats Tina!

Dr. Tina Rost’s work has recently appeared in Applied Physics Letters.

Rost, C.M., Braun, J.L., Ferri, K., Backman, L., Giri, A., Opila, E., Maria, J.-P., Hopkins, P.E., “Hafnium nitride films for thermoreflectance transducers at high temperatures: Potential based on heating from laser absorption,” Applied Physics Letters 111, 151902 (2017). PDF.

In a typical TDTR experiment, a thin metal transducer is deposited on top of a sample to measure the sample’s thermal properties.  Ultimately, TDTR can be limited by the stability of this transducer.  In this work, we have demonstrated the ability to extend TDTR measurements up to 1000 K using HfN as a metal transducer.  Note only does HfN demonstrate one of the highest thermoreflectance coefficients at 800 nm measured to date, but it’s high temperature phase stability make it attractive for use as a metal transducer at high temperatures.


Time domain thermoreflectance (TDTR) and frequency domain thermoreflectance (FDTR) are common pump-probe techniques that are used to measure the thermal properties of materials. At elevated temperatures, transducers used in these techniques can become limited by melting or other phase transitions. In this work, time domain thermoreflectance is used to determine the viability of HfN thin film transducers grown on SiO2 through measurements of the SiO2 thermal conductivity up to approximately 1000 K. Further, the reliability of HfN as a transducer is determined by measuring the thermal conductivities of MgO, Al2O3, and diamond at room temperature. The thermoreflectance coefficient of HfN was found to be 1.4 10^-4 K^-1 at 800 nm, one of the highest thermoreflectance coefficients measured at this standard TDTR probe wavelength. Additionally, the high absorption of HfN at 400 nm is shown to enable reliable laser heating to elevate the sample temperature during a measurement, relative to other transducers.

We acknowledge the financial support from the Office of Naval Research MURI program (Grant No. N00014-15-1- 2863).

Thermal boundary resistance limits the ablation thresholds of thin films: Paper published in Physical Review B – Congrats John Tomko!

John Tomko’s works has recently appeared in Physical Review B.

Tomko, J.A., Giri, A., Donovan, B.F., Bubb, D.M., O’Malley, S.M., Hopkins, P.E., “Energy confinement and thermal boundary conductance effects on short-pulsed thermal ablation thresholds in thin films,” Physical Review B 96, 014108 (2017). PDF.

In this work,we demonstrated a direct correlation between thermal ablation of thin gold films and the thermal boundary conductance across the film/substrate interface.  We used high energy pulsed lasers to induce the film ablation and showed a dependence on substrate.  However, the ablation threshold did not trend with the substrate thermal properties, but instead the thermal boundary conductance across the film/substrate interface.


For this paper, single-pulse ablation mechanisms of ultrafast laser pulses (25 ps) were studied for thin gold films (65 nm) on an array of substrates with varying physical properties. Using time-domain thermoreflectance, the interfacial properties of the thin-film systems are measured: in particular, the thermal boundary conductance. We find that an often used, and widely accepted relation describing threshold fluences of homogeneous bulk targets breaks down at the nanoscale. Rather than relying solely on the properties of the ablated Au film, the ablation threshold of these Au/substrate systems is found to be dependent on the measured thermal boundary conductance; we additionally find no discernible trend between the damage threshold and properties of the underlying substrate. These results are discussed in terms of diffusive thermal transport and the interfacial bond strength.

This material is based upon work supported by the Air Force Office of Scientific Research under Award No. FA9550-15-1-0079. D.M.B. and S.M.O. acknowledge Awards CMMI- 1531783 and CMMI-0922946 from the National Science Foundation.

Functional chemistry impacts interfacial heat transport from fullerene derivatives to liquids: Paper published in ACS Nano – Congrats Chet!

Chet Szwejkowski’s work on the thermal boundary conductance across fullerene derivative/liquid interfaces has recently appeared in ACS Nano.

Szwejkowski, C.J., Giri, A., Kaehr, B., Warzoha, R.J., Donovan, B.F., Hopkins, P.E., “Molecular tuning of the vibrational thermal transport mechanisms in fullerene derivative solutions,” ACS Nano 11, 1389-1396 (2017).

In this work, we showed that the functional group on fullerene derivatives can be used to control the thermal boundary conductance across the molecule/liquid interfaces.  Based on the geometry of the functional group, the density of states of low frequency modes will change and impact the transmission of thermal energy across the molecule/liquid interface.  We demonstrated this by measuring the thermal boundary conductance across these interfaces in samples of dilute fullerene derivative suspensions.  The measurements were conducted using time domain thermotransmission.  Our experimental results are supported with molecular dynamic simulations.


Control over the thermal conductance from excited molecules into an external environment is essential for the development of customized photothermal therapies and chemical processes. This control could be achieved through molecule tuning of the chemical moieties in fullerene derivatives. For example, the thermal transport properties in the fullerene derivatives indene-C60 monoadduct (ICMA), indene-C60 bisadduct (ICBA), [6,6]-phenyl C61 butyric acid methyl ester (PCBM), [6,6]-phenyl C61 butyric acid butyl ester (PCBB), and [6,6]-phenyl C61 butyric acid octyl ester (PCBO) could be tuned by choosing a functional group such that its intrinsic vibrational density of states bridge that of the parent molecule and a liquid. However, this effect has never been experimentally realized for molecular interfaces in liquid suspensions. Using the pump−probe technique time domain thermotransmittance, we measure the vibrational relaxation times of photoexcited fullerene derivatives in solutions and calculate an effective thermal boundary conductance from the opto-thermally excited molecule into the liquid. We relate the thermal boundary conductance to the vibrational modes of the functional groups using density of states calculations from molecular dynamics. Our findings indicate that the attachment of an ester group to a C60 molecule, such as in PCBM, PCBB, and PCBO, provides low-frequency modes which facilitate thermal coupling with the liquid. This offers a channel for heat flow in addition to direct coupling between the buckyball and the liquid. In contrast, the attachment of indene rings to C60 does not supply the same low-frequency modes and, thus, does not generate the same enhancement in thermal boundary conductance. Understanding how chemical functionalization of C60 affects the vibrational thermal transport in molecule/liquid systems allows the thermal boundary conductance to be manipulated and adapted for medical and chemical applications.

Patrick Wins ASME Bergles-Rohsenow Young Investigator Award in Heat Transfer

Patrick has been awarded the 2016 ASME Bergles-Rohsenow Young Investigator Award in Heat Transfer for significant heat transfer research that has produced experimental and analytical advancements in areas including thermal transport across interfaces, reduced thermal conductivity materials, electron-phonon coupling, and transport of electrons and phonons.

The Bergles-Rohsenow Young Investigator Award in Heat Transfer is given to a young engineer that is under 36 years of age and has received a Ph.D. or equivalent degree in Engineering. The individual must be committed to pursuing research in heat transfer, and must have demonstrated the potential to make significant contributions to the field of heat transfer. Such contributions may take the form of, but are not limited to, analytical/numerical methods, equipment/instrumentation, or experimentation – any of which should lead to peer-reviewed publications.

Established by the Heat Transfer Division in 2003, the award was funded through the efforts of Arthur Bergles and Warren Rohsenow who are well known for their accomplishments in heat transfer research and for their mentoring of young researchers.

The official announcement can be found here




Ordering effects on the thermal transport mechanisms in metallic alloys: Paper published in Scientific Reports – Congrats Ash!

Ash Giri’s work on the electron and phonon thermal transport mechanisms in ordered and disordered metallic alloys has recently appeared in Scientific Reports.  In this work, we show, via both experimental measurements and molecular dynamics simulations, that the thermal conductivity of an ordered metallic alloy (FePt) can have the same thermal conductivity as a disordered alloy at high temperatures.  This is due to a decreasing phonon thermal conductivity in the ordered alloy driven by three phonon scattering events.  This work has great implications for the thermal management involved in heat assisted magnetic recording applications.



We report on the out-of-plane thermal conductivities of tetragonal L10FePt (001) easy-axis and cubic A1 FePt thin films via time-domain thermoreflectance over a temperature range from 133 K to 500 K. The out-of-plane thermal conductivity of the chemically ordered L10 phase with alternating Fe and Pt layers is ~23% greater than the thermal conductivity of the disordered A1 phase at room temperature and below. However, as temperature is increased above room temperature, the thermal conductivities of the two phases begin to converge. Molecular dynamics simulations on model FePt structures support our experimental findings and help shed more light into the relative vibrational thermal transport properties of the L10 and A1 phases. Furthermore, unlike the varying temperature trends in the thermal conductivities of the two phases, the electronic scattering rates in the out-of-plane direction of the two phases are similar for the temperature range studied in this work.

P.E.H. and A.G. appreciate support from the Air Force Office of Scientific Research, Grant No. FA9550-15-1-0079.