Z. E. Nataj, et al.

, Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

Cryogenic Characteristics of Graphene Composites –

Evolution from Thermal Conductors to Thermal Insulators

Zahra Ebrahim Nataj,1,* Youming Xu,2,* Jonas Brown,1 Jivtesh Garg,3 Xi Chen, 2 Fariborz
Kargar,1,† and Alexander A. Balandin1

1
Phonon Optimized Engineered Materials Center, Department of Electrical and Computer
Engineering, University of California, Riverside, California 92521 USA

2
Department of Electrical and Computer Engineering, University of California, Riverside,
California 92521 USA

3
Department of Aerospace and Mechanical Engineering, University of Oklahoma, Norman,
Oklahoma 73019 USA

*
These authors contributed equally to the manuscript.
†
Corresponding authors: fkargar@ece.ucr.edu ; balandin@ece.ucr.edu

1|Page
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

ABSTRACT

The development of cryogenic semiconductor electronics and superconducting quantum

computing requires composite materials that can provide both thermal conduction and thermal
insulation. We demonstrated that at cryogenic temperatures, the thermal conductivity of graphene
composites can be both higher and lower than that of the reference pristine epoxy, depending on
the graphene filler loading and temperature. There exists a well-defined cross-over temperature –
above it, the thermal conductivity of composites increases with the addition of graphene; below it,
the thermal conductivity decreases with the addition of graphene. The counter-intuitive trend was
explained by the specificity of heat conduction at low temperatures: graphene fillers can serve as,
both, the scattering centers for phonons in the matrix material and as the conduits of heat. We offer
a physical model that explains the experimental trends by the increasing effect of the thermal
boundary resistance at cryogenic temperatures and the anomalous thermal percolation threshold,
which becomes temperature dependent. The obtained results suggest the possibility of using
graphene composites for, both, removing the heat and thermally insulating components at
cryogenic temperatures – a capability important for quantum computing and cryogenically cooled
conventional electronics.

KEYWORDS: cryogenic electronics; thermal management; thermal conductivity; graphene

composites; thermal percolation; quantum technologies

2|Page
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

There is a rapidly emerging need for thermal management at cryogenic temperatures. It is driven
by several trends and developments. There is a strong motivation to run conventional
semiconductor electronics at low temperatures to implement “cold computing,” which allows one
to increase computational and energy efficiency while reducing power consumption.1,2 The main
progress with quantum computing technologies is associated with superconducting qubits, which
require cryogenic temperatures.3 Space exploration needs electronics operating in harsh low-
temperature environments. Cryogenic thermal management usually relies on both thermal
conductors and thermal insulators.4–6 The former is the polymer-based thermal interface materials
(TIMs) with fillers that conduct heat well, facilitating heat removal, and the latter are polymer
materials, which have low thermal conductivity and can act as thermal insulators between
electronic components operating at different temperatures. Examples of low-temperature thermal
management applications include protective coatings of superconductive power cables7, adhesives
in cryogenic low-noise amplifiers for radio-astronomy and space communication systems,8 optical
mounts of cryogenic refractive optics and cryo-sorption pumps.9,10

Polymers are poor heat conductors with thermal conductivity in the range from ~0.2 to
0.5 Wm−1 K −1 at room temperature (RT).11,12 Polymers are used in TIMs as the base, i.e. matrix
material, that fills the air gaps between two adjacent solid interfaces and provides adhesive
functionality when used in curing composites. The common strategy to increase the thermal
conductivity of polymers is to add micrometer- and nanometer-scale fillers with a higher intrinsic
thermal conductivity that can couple well with the base polymer. A mixture of the single-layer and
few-layer graphene flakes, termed “graphene” in the thermal context, has proven to be an efficient
filler material for a variety of TIMs, including non-curing mineral oil-based thermal pastes,13 and
curing epoxies.12,14,15 Graphene for thermal management applications can be mass-produced via
liquid-phase exfoliation, graphene oxide reduction, or other techniques.16,17 Graphene TIMs with
thermal conductivity above ~12 Wm−1 K −1 near RT, which exceeds the metric of conventional
commercial TIMs, have been reported by several research groups.12,14,15 The excellent
performance of graphene TIMs near RT originates in the extraordinarily high intrinsic thermal
conductivity of graphene and few-layer graphene,18,19 strong coupling to the matrix, good
dispersion, and appropriate viscosity range of the resulting composites.20 The enhancement of

3|Page
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

thermal conductivity is achieved both below and above the thermal percolation threshold – a
loading fraction at which the graphene fillers start to form continuous thermally conductive
paths.21,22 The thermal percolation threshold can be identified when the dependence of the thermal
conductivity on filler loading becomes super-linear.21,23 One should note that the thermal
properties of graphene composites have only been studied at RT and above – the temperature range
of interest for conventional electronics. We are not aware of any report on the cryogenic thermal
characteristics of graphene composites. In general, the data for the thermal properties of any
polymer composites at cryogenic temperatures are scares. The understanding of heat propagation
in amorphous polymers at low temperatures is far from complete, even if one does not consider
the solid inclusions, i.e. fillers.24,25

Here, we investigated the thermal properties of epoxy–graphene composites at temperatures from

2 K to RT. Epoxy is a practically important material and it is often used as a reference material to
compare the effect of different types of fillers on its thermal conductivity. We found that at
cryogenic temperatures, the thermal conductivity of graphene composites can be both higher and
lower than that of the reference pristine epoxy, depending on the graphene filler loading and
temperature. This is drastically different from what is observed near RT. Moreover, there exists a
well-defined cross-over temperature that above it, the thermal conductivity increases with the
addition of graphene, whereas, below it, the thermal conductivity decreases with the addition of
graphene. Graphene composites are unique in a way that they can provide both the strongest
enhancement in thermal conductivity and the strongest suppression. We offer a physical model
explanation of the counter-intuitive trends and provide numerical simulation data which agree with
the measurements. The obtained results suggest the possibility of using composites with the same
constituent materials for, both, removing the heat and thermally insulating electronic components
at cryogenic temperatures. The latter constitutes a conceptual change for thermal management,
which typically rely on different materials for heat conduction and isolation.

4|Page
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

EXPERIMENTAL SECTION

Materials: The polymer matrix is a thermoset epoxy set consisting of a base resin (Bisphenol-A,
Allied High-Tech Products, Inc.) and a hardening agent (Triethylenetetramine; Allied High-Tech
Products, Inc.). We used few-layer graphene with the specified average lateral dimension of 25 µm
and an average surface area of 50 to 80 m2 g −1 (xGnP, Grade H, XGSciences, the US) as the fillers
for the preparation of the composites. The graphene fillers were subjected to further exfoliation
during the high-speed mixing process with the base polymer matrix. The resulting fillers constitute
a mixture of FLG and single-layer graphene. The lateral dimensions of graphene fillers are an
important parameter for tuning the thermal conductivity of the composites.26 To maximize the
thermal conductivity of the composite, one normally wants to keep the lateral dimensions above
the gray phonon mean free path (MFP) in graphene, which is about ~1 m near RT.27 In contrast,
to improve the insulating properties of composites, fillers with smaller lateral dimensions are
preferred.28

Composite Preparation and Characterization: Several composite samples were prepared by

mixing precalculated quantities of the resin, the hardener, and the FLG fillers to hit a targeted filler
loading level. First, a certain amount of FLG fillers was distributed into the epoxy resin using a
high-speed shear mixer (Flacktek, Inc.). The hardener was then added to the mixture in the mass
ratio of 12:100 with respect to the base epoxy resin’s weight. The blend was vacuumed several
times to eliminate any air bubbles that might have been trapped during the composite preparation
process. The mixture was poured into special molds to cure and solidify. The results of the mass
density measurements confirms that the porosity of the samples is negligible (Figure S1).The full
recipe of the preparation is detailed in the Supplemental Information. An optical image of
representative samples with various graphene loadings is presented in Figure 1 (a). Figure 1 (b) is
an SEM image of a composite sample with the filler concentration of 𝑓 = 21.7 vol%. As the
volume fraction of fillers increases beyond a certain loading, referred to as the percolation
threshold, the fillers start to overlap (see SEM image in Figure 1 (b)). At and beyond the
percolation regime, fillers create a network of electrically and thermally conductive pathways

5|Page
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

within the base polymer matrix.21,22,29 The percolation results in significant enhancements in both
the electrical and thermal characteristics of the composites.21,22,29 The composites were further
characterized using Raman spectroscopy (Renishaw InVia) with the laser excitation wavelength
of 𝜆 = 633 nm in the conventional backscattering configuration. In all experiments, the laser
power was kept at ~3 mW. Raman spectra of several samples at random spots are presented in
Figure 1(c). In all composites, the characteristic G-peak and 2D-peak of graphene were observed
at ~1580 cm-1 and ~2700 cm-1 [Ref. 30] confirming an even distribution of graphene throughout the
samples. The intensity of the 2D peak is much lower than the G-peak indicating the presence of a
mixture of single-layer and few-layer graphene in the composite samples.

6|Page
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

[Figure 1: Sample preparation and characterization. (a) From left to right, optical images of
pristine epoxy and composites with 2.6 vol% and 5.4 vol% loading of few-layer graphene. (b)
SEM image of an epoxy sample with 21.7 vol% graphene loading. Note the regions where the
fillers overlap. (c) Raman spectra of pristine epoxy, few-layer graphene, and composites with
various graphene concentrations.]

RESULTS AND DISCUSSIONS

Specific heat of graphene composites at cryogenic temperatures: We start with the

investigation of the specific heat and illustrate its typical temperature trends in disordered

7|Page
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

amorphous materials. Figure 2 (a) and (b) show the results of the specific heat, 𝑐𝑝 , measurements
of pristine epoxy and selected composites as a function of temperature for different filler
concentrations. The data are presented in two different temperature ranges to make the trend more
explicit (also, see Figure S2). The shaded area around the experimental data points indicates the
standard error in the measurements conducted with Physical Property Measurement System
(PPMS, DynaCool, Quantum Design). The results for the neat epoxy agree well with the
literature.31 The variation of the specific heat with temperature resembles that in other amorphous
polymers.31–33 In all samples, 𝑐𝑝 increases more rapidly in the low-temperature region; the rate of
increase slows down as the temperature rises. To better visualize the temperature characteristics
of 𝑐𝑝 , we also plotted the data in a log-log scale, indicating the polynomial functional dependencies
(Figure 2 (c)). One can distinguish three regions: quasi-cubic, 𝑐𝑝 ~𝑇 3+𝛿 , in the temperature range
of 2 K ≤ 𝑇 ≤ 6 K; parabolic, 𝑐𝑝 ~𝑇 2, in the interval of 6 K ≤ 𝑇 ≤ 35 K; and linear, 𝑐𝑝 ~𝑇, in the
range of 𝑇 ≥ 35 K. Among these regions, the 𝑐𝑝 behavior in the low-temperature limits is of
particular interest since it deviates from the classical Debye model for crystalline materials. We
plotted the “Debye-reduced” heat capacity, 𝑐𝑝 𝑇 −3, as a function of temperature in a log-log scale
in Figure 2 (d). In the insulating crystalline materials, 𝑐𝑝 ~𝑇 3, and therefore one expects to see a
flat curve in the low-temperature limits of the 𝑐𝑝 𝑇 −3 vs. 𝑇 plot. For the pristine epoxy and
composites with graphene, there is no 𝑇 3 dependency of 𝑐𝑝 . The latter stems from the amorphous,
disordered natures of these materials.

One can notice from Figure 2 (d) that as the graphene loading increases, the curves flatten out,
getting closer to the 𝑐𝑝 ~𝑇 3 characteristic of crystalline materials. It is well-established that the
specific heat of amorphous materials exhibits a “universal” characteristic in the low-temperature
limits.34,35 Below 𝑇~1 K, the heat capacity considerably exceeds the Debye model predictions,
dominated by a quasi-linear temperature dependence.34,35 This behavior is explained by the
tunneling model of two-level systems for amorphous materials at low temperatures.36 This
anomalous quasi-linear trend is followed by a hump in the 𝑐𝑝 𝑇 −3 data, referred to as the “boson
peak” in the 3 K ≤ 𝑇 ≤ 10 K region.36–38 The “boson peak” shown in Figure 2 (d) is attributed to
the low-frequency vibrational modes present in amorphous materials.38 The temperature at which

8|Page
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

the boson peak occurs decreases slightly with increasing the filler loading. The “boson peak”
eventually fades away with increasing the filler content as the material acquires more “crystalline”
characteristics owing to the graphene content. The overall value of 𝑐𝑝 decreases with increasing
graphene loading since 𝑐𝑝 of FLG is much lower than that of the epoxy matrix. We established the
general trends of the specific heat of graphene composites that follow those of amorphous
materials but become more crystalline-like with increasing graphene content.

[Figure 2: Specific heat of graphene composites. (a) Temperature-dependent specific heat of

composites in the temperature range of 2 K ≤ 𝑇 ≤ 300 K. (b) The same data as in panel (a)
shown in the low-temperature limits. (c) Specific heat of the composites plotted in the log-log
scale, revealing the quasi-cubic, parabolic, and linear temperature dependence in the low,
intermediate, and high-temperature ranges. (d) The “Debye-reduced” specific heat of graphene
composites as a function of temperature in a the log-log scale. The dashed line is a guide to the
eye, showing the behavior of the so-called “boson peak” as a function of filler loading and
temperature.]

9|Page
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

Thermal conductivity of graphene composites at cryogenic temperatures: We now turn to the

main topic of this study – the thermal conductivity of graphene composites at cryogenic
temperatures. Figure 3 (a) shows the thermal conductivity, 𝑘, measured using PPMS in the
temperature range from 2 K to RT in a log-log scale. The details of the measurements are described
in the Supplemental Materials. The thermal conductivity of the pristine epoxy initially increases
sup-linearly in the temperature range of 2 K ≤ 𝑇 ≤ 6 K, followed by a “plateau region” in the
interval of 6 K ≤ 𝑇 ≤ 17 K, where 𝑘 remains nearly constant. After the plateau, 𝑘 increases
linearly again until 𝑇~80 K at which point a second plateau extending up to 𝑇~175 K occurs. The
existence of the first plateau in 𝑘 vs. T dependence is universal for amorphous materials, and it
occurs almost in the same temperature range. It starts at the temperature where the “boson peak”
appears in the specific heat (see Figure 2 (d)).39 The first plateau region is explained by the two-
level systems and the tunneling model.36,39 According to this model, the dependence is due to the
cross-over at which phonons with shorter MFP become dominant heat carriers instead of phonons
with longer MFP.36,39 The thermal conductivity is given by 𝑘 = (1⁄3) 𝐶𝑣Λ in which 𝐶, 𝑣, and Λ
are the specific heat, average group velocity and the mean free path of phonons, respectively. In
the plateau region, while 𝐶 increases with the temperature, Λ decreases almost at the same rate. In
other words, the plateau forms at the temperature range where the product of Λ and 𝐶 becomes
independent of temperature. The second plateau region was explained by similar considerations
but its origin is still the subject of debate.40

Before proceeding further with the analysis of the experimental thermal data we should clarify the
use of the “phonon” concept and terminology. Naturally, the introduction of phonons – quanta of
crystal lattice vibrations – requires translational symmetry encountered in crystalline solids.41
Amorphous materials lack translation symmetry. Complex models involving different descriptions
of atomic vibrations referred to as propagons, diffusons, and locons, corresponding to propagating,
diffusion, and localized modes, have been introduced to describe thermal transport in amorphous
materials.42,43 However, in the context of the present study of heat conduction in epoxy-based
composites, we still can use the concept of acoustic phonons as the elastic vibrations in the
continuous medium of the base material. The acoustic phonons are the dominant heat carriers in
dielectric materials, including amorphous epoxy. We can talk about the acoustic phonons, i.e.

10 | P a g e
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

vibrations, propagating through the matrix that carry heat and interact with graphene fillers. For
analysis of heat conduction in graphene composites, it is also useful to keep in mind the concept
of the two-phase semi-crystalline medium where crystalline fillers interface with the amorphous
polymer host.34

Let us now consider the temperature-dependent characteristics of the thermal conductivity of the
graphene–epoxy composites in more detail. In Figure 3 (b), we replotted the thermal conductivity
data for the same graphene loadings on a linear scale. There is no plateau region in graphene
composites similar to the one observed for pristine epoxy. For composites with the filler loading
𝑓 ≤ 8.3 vol%, there is a well-defined cross-over temperature, 𝑇𝑐 ~11 K, at which 𝜅 is the same for
the pristine epoxy and graphene composites with variable graphene loading. The existence of a
cross-over temperature and the absence of a plateau have been reported previously in the thermal
conductivity data of semi-crystalline materials44 and polymer-based composites with filler
inclusions, which can be considered semi-crystalline media.34 For example, while the plateau in
𝑘 emerges in the amorphous polyethylene terephthalate (PET), it vanishes in semicrystalline PET
containing crystalline zones with volume fraction as low as 𝑓 = 9 vol%.44 The 𝑇𝑐 ~15 K is
reported for PET with crystalline regions of 0 ≤ 𝑓 ≤ 51 vol% dispersed inside amorphous PET.44
Several studies have reported a cross-over temperature for different composite systems in a similar
temperature range. In all these studies, the plateau region vanished as soon as crystalline fillers
were added to the amorphous polymer.28,45–48

The peculiar features in the temperature-dependent thermal transport characteristics can be

explained by the thermal boundary resistance (TBR), 𝑟𝑏 , at the interface of the amorphous polymer
and crystalline FLG fillers.28,44 According to the acoustic mismatch theory, at sufficiently low
temperatures, 𝑟𝑏 ~𝑇 −3, and thereby 𝑟𝑏 shows strong effects in the low-temperature limits.49,50 The
effect becomes small at high temperatures. Note that, below 𝑇𝑐 , 𝜅 of the pristine epoxy is larger
than that of the epoxy composites with graphene fillers. This means that the inclusion of FLG
fillers into the amorphous epoxy reduces its 𝑘 when the temperature is below 𝑇𝑐 but improves it
when the temperature is above the cross-over temperature. This is a direct consequence of the

11 | P a g e
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

strong temperature dependence of TBR at the polymer—filler interfaces. This will be further
discussed in the theory section.

Figure 3 (c) shows the thermal conductivity of graphene composites as a function of the filler
loading, 𝑓, at constant temperatures of 2 K, 𝑇𝑐 = 11 K, and RT. The systematic errors for the data
points at low temperatures were smaller than the size of the symbols and not shown for clarity. At
𝑇 = 2 K, the thermal conductivity of the epoxy falls approximately two times with the addition
of only 2.6 vol% of graphene fillers. The sharp decrease in thermal conductivity suggests that FLG
fillers contribute significantly to the phonon scattering processes. At this temperature, as the
concentration of the filler increases, the thermal conductivity declines linearly. At RT, however,
the trend is the opposite. The addition of graphene fillers improves the thermal conductivity
monotonically in line with many prior reports.12,21,23 The variation of 𝑘 as a function of 𝑇 is more
intriguing in the vicinity of the cross-over temperature, 𝑇𝑐 ~11 K. With increasing the graphene
filler loading, thermal conductivity remains the same up to 𝑓 = 8.3 vol%; then, linearly increases
in the interval of 8.3 vol% ≤ 𝑓 ≤ 14.6 vol%; and after that, remains unchanged with adding more
fillers. Typically, a nonlinear behavior such as the one observed at 𝑇𝑐 ~11 K might indicate that
the composite system at 𝑓 = 14.6 vol% enters the thermal percolation regime. The absence of
such nonlinear behavior at 𝑇 = 2 K and RT suggests an intriguing possibility of the temperature-
dependent percolation threshold.

Typically, one thinks about the percolation threshold in terms of the filler loading, 𝑓𝐻 , as the point
where the fillers start to mechanically touch each other, forming a continuous conductive network.
Thermal percolation is less abrupt than electrical percolation because the heat can be conducted
by the matrix materials, contrary to the electrical current that cannot be conducted by the dielectric
matrix.21,51,52 The ratio of the thermal conductivity of the fillers to the matrix, 𝑘𝑓 /𝑘𝑚 , is several
orders of magnitude smaller than the ratio of the electrical conductivity of the fillers to the matrix,
𝜎𝑓 ⁄𝜎𝑚 .21 The wavelength of the thermal phonons, which make the dominant contribution to heat
conduction, is 𝜆 𝑇 ~ℎ𝑉𝑠 /𝑘𝑏 𝑇, where ℎ is the Plank’s constant, 𝑘𝑏 is the Boltzmann’s constant and
𝑉𝑠 is the sound velocity, or more accurately, the phonon group velocity.34 For a typical

12 | P a g e
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

semiconductor material 𝜆 𝑇 is on the order of 1 nm – 2 nm at RT.53 This means that for the thermal
percolation to occur at RT, the fillers should be in physical contact or close to each other, i.e., on
the order of ~ 1 nm. At low temperatures, 𝜆 𝑇 increases by more than an order of magnitude. Thus,
phonons with long wavelengths might provide thermal “cross-talk” to the fillers over some
distance. The average distance between the fillers at ~10 vol % is still larger than 𝜆 𝑇 . However,
one should remember that more refined theories of thermal conductivity attribute a more
significant contribution to heat conduction to the phonons with wavelength above 𝜆 𝑇 .54 If we
accept this picture of the process, then the dependence of the thermal conductivity at 2 K and 300
K are those below and above thermal percolation, correspondingly (see Figure 3 (c)). It is
illustrative to analyze them further with the specific heat dependence on the loading fraction shown
in Figure 3 (d). The specific heat decreases with increasing 𝑓 for both temperatures since 𝑐𝑝 of
FLG is lower than that of the matrix. The thermal conductivity decreases at 2 K with 𝑓 either
following the 𝑐𝑝 trend or because graphene fillers are acting more like scattering centers for the
relevant low-wavelength phonons. The thermal conductivity at 300 K increases with 𝑓 despite the
decrease owing to the addition of more percolated FLG conducting channels. The total cross-
section of the percolated channels increases faster than the decrease in 𝑐𝑝 .

13 | P a g e
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

[Figure 3: Temperature-dependent thermal transport characteristics of graphene

composites. (a) Thermal conductivity of graphene composites in the temperature range of 2 K ≤
𝑇 ≤ 300 K in the log-log scale. The symbols are the experimental data points. The shaded area
indicates the experimental uncertainty. (b) Thermal conductivity of the graphene composites in
the low-temperature region, showing the cross-over temperature, 𝑇𝑐 . (c) Thermal conductivity
of the graphene composites as a function of the filler loading at 2 K, 𝑇𝑐 ~11 K, and RT. (d)
Specific heat of the graphene composites as a function of the filler loading at constant
temperatures of 2 K and 300 K.]

Effective medium model for cryogenic heat conduction for the low-loading composites: We
now develop a model of cryogenic heat conduction in graphene composites using Nan’s effective
medium model as the starting point.55 According to this model, the effective thermal conductivity
of a composite with randomly oriented low-loading fillers is given as55

14 | P a g e
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

𝑘 = 𝑘𝑚
3+𝑓[2𝛽11 (1−𝐿11 )+ 𝛽33 (1−𝐿33 )]
. (1)
3−𝑓[2𝛽11 𝐿11 + 𝛽33 𝐿33 ]

Here, 𝑘𝑚 is the thermal conductivity of the pristine epoxy, 𝑘 is the effective thermal conductivity
of epoxy–graphene nanocomposite with the filler volume fraction 𝑓, 𝐿𝑖𝑖 are the geometrical
parameters that depend upon the aspect ratio, 𝑝 = 𝑡/𝐿, of graphene fillers with 𝑡 and 𝐿 being the
thickness and lateral dimensions of the fillers. The details of 𝐿𝑖𝑖 parameters and their definition
can be found in the Supplementary Information. The parameters 𝛽𝑖𝑖 contains information about
the thermal boundary resistance at the filler-epoxy interface and is defined as:

𝑐
𝐾𝑖𝑖 −𝑘𝑚
𝛽𝑖𝑖 = 𝑐 −𝑘 ) , (2)
𝑘𝑚 +𝐿𝑖𝑖 (𝐾𝑖𝑖 𝑚

where, 𝐾𝑖𝑖𝑐 are the effective values of FLG thermal conductivity along different cartesian directions,
that take into account the effect of interface thermal resistance. The effective thermal
𝑐 𝑐 𝑐
conductivities of the FLG fillers along the in-plane (𝐾11 ~𝐾22 ) and through-plane (𝐾33 ) are,
respectively,

𝑐 𝑐 𝑘𝑖𝑛
𝐾11 = 𝐾22 = 1+𝛾𝐿 , (3)
11 𝑘𝑖𝑛 /𝑘𝑚

𝑐 𝑘𝑜𝑢𝑡
𝐾33 = 1+𝛾𝐿 , (4)
33 𝑘𝑜𝑢𝑡 /𝑘𝑚

where, 𝛾 = (1 + 2𝑝)𝛼 in which 𝛼 = 𝑟𝑏 𝑘𝑚 /𝑡 is a dimensionless parameter related to the interface

thermal resistance, 𝑟𝑏 , between the epoxy and filler, 𝑘𝑖𝑛 and 𝑘𝑜𝑢𝑡 represent the in-plane and
through-plane thermal conductivity of pristine graphene fillers, respectively. To obtain the
temperature dependence of the effective thermal conductivity, all involved parameters in Eq. (1)
including 𝑘𝑖𝑛 , 𝑘𝑜𝑢𝑡 , 𝑘𝑚 , and 𝑟𝑏 were taken to be temperature-dependent (see the Supplementary
Information and Figure S3 for details). The combined effect of the high interfacial thermal
resistance and low through-plane thermal conductivity of FLG at low temperatures (see Figure S4
(b) and Figure S3 (b)) results in a significantly small in-plane and through-plane “effective thermal

15 | P a g e
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

𝑐 𝑐
conductivity” of FLG, 𝐾11 and 𝐾33 , computed using Eq. (3) and (4). The results are shown in
Figure 4 (a). The effective through-plane thermal conductivity of FLG becomes lower than that of
the neat epoxy through the whole temperature range of 2 K up to 300 K. The graphene fillers
oriented perpendicular to the heat flux serve as the extra thermal boundary resistance, a scattering
𝑐
center, rather than the conduit of heat. Note that 𝐾33 is more than two orders of magnitude lower
than the thermal conductivity of the pristine epoxy at 2 K. The in-plane thermal conductivity of
FLG is only three times higher than that of the pristine epoxy at 𝑇 = 2 K, whereas, at RT, its
effective in-plane thermal conductivity is ~160 times higher than that of the neat epoxy.

The results of the calculations based on this effective medium model for composites with 𝑓 ≤
8.3 vol% are presented in Figure 4 (b) in the temperature range of 2 K up to 50 K. The model
successfully reflects the experimental thermal conductivity characteristics for the low-
concentration composites and correctly estimates the experimental cross-over temperature. Based
on this model, we infer that the low effective through-plane thermal conductivity of FLG
outweighs its highly effective in-plane thermal conductivity, causing the composite thermal
conductivity to become lower than that of pure epoxy at low-temperature limits. As the loading of
FLG increases, the effect of the low through-plane thermal conductivity of fillers dominates,
resulting in decreasing the composite’s thermal conductivity even more. In contrast, in high-
temperature limits, both the in-plane and through-plane thermal conductivities increase, while
simultaneously the interfacial TBR decreases as 𝑟𝑏 ~𝑇 −3 . This causes a rise in the effective thermal
conductivity of the filler in both directions. At ~15 K, the effective thermal conductivities of FLG
fillers recover enough, to result in a thermal transport enhancement. Above this temperature, the
composite thermal conductivity becomes higher than that of the neat epoxy and it grows with
increasing FLG loading. Thus, there is a transition in the thermal conductivity trend, with 𝑘
decreasing with increasing FLG concentration at low temperatures, and the opposite trend at higher
temperatures. The latter leads to a cross-over temperature effect at 15 K for composite’s 𝑘.

Effective medium model for cryogenic heat conduction for the high-loading composites: The
considered effective medium model is unable to predict the thermal conductivity of the high-
concentration composites since it does not include the effect of filler-filler contact. As the filler

16 | P a g e
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

loading increases, the probability that fillers can physically contact each other within the host
polymer grows. Increasing the filler content can result in entering into a strong thermal percolation
regime.21–23 In this regime, heat can conduct along a network of highly conductive connected FLG
fillers within the polymer matrix. The disappearance of the cross-over temperature in high-
concentration samples implies that the favorable effect of heat conduction along such networks
might have overcome the negative effect of high interfacial thermal resistance at low temperatures.
To predict the thermal conductivity characteristics at higher FLG concentrations, we use a recently
introduced percolation-based effective medium model.56 Through this model the composite
thermal conductivity 𝑘 is determined by solving the following equation:

𝑘0 − 𝑘 𝑓 2(𝑘11 − 𝑘) (𝑘33 − 𝑘)
(1 − 𝑓) + [ + ] = 0, (5)
𝑘0 − 𝑘 3 𝑘 + 𝑆11 (𝑘0 − 𝑘) 𝑘 + 𝑆33 (𝑘𝑜𝑢𝑡 − 𝑘)
𝑘+ 3

where 𝑘11 and 𝑘33 are the effective in-plane and through-plane thermal conductivity of FLG fillers,
which account for the interfacial thermal resistance. These parameters are calculated using the
actual in-plane and through-plane thermal conductivities of FLG, 𝑘𝑖𝑛 and 𝑘𝑜𝑢𝑡 . The shape
parameters 𝑆11 and 𝑆33 are related to the aspect ratio of the FLG fillers, and 𝑘0 is the thermal
conductivity of an interlayer surrounding FLG fillers. This interlayer represents the interface
thermal resistance surrounding the graphene fillers. Its role is to include the combined effect of
interfacial thermal resistance at graphene-polymer and graphene-graphene contacts. The
definitions of these parameters are explained in more detail in the Supplementary Information. The
values of different parameters used in this model are listed in Table S1. Figure 4 (c) shows the
results of the calculations for the high-concentration composites with 𝑓 ≥ 11.4 vol% along with
the results of the low-loading effective medium model for the composites with 𝑓 ≤ 8.3 vol%. The
predicted thermal conductivity in both cases is in good agreement with the experimental results
(see Figure 3). One can conclude that the superior heat conduction along the percolated graphene
fillers channel overcomes the negative effect of large interfacial thermal resistance at low
temperatures causing the cross-over effect to disappear at compositions greater than 𝑓 ≥
11.4 vol%.

17 | P a g e
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

[Figure 4: Thermal conductivity and thermal performance benchmarking. (a) Calculated

𝑐 𝑐
effective in-plane, 𝐾33 , and through-plane, 𝐾11 , thermal conductivity of graphene composites
with FLG loading 𝑓 ≤ 8.3 vol%. (b) Calculated thermal conductivity of graphene composites
with FLG loading of 𝑓 ≤ 8.3 vol%. The model successfully predicts the cross-over temperature,
𝑇𝑐,𝑡ℎ ~15 K, which is in agreement with the experimental value. (c) Prediction of the percolative-
based effective medium theory used to describe the thermal conductivity of composites with the
high filler loading, 𝑓 ≥ 11.4 vol%. The data is presentred in the log-log scale together with the
results of the low-loading model prediction presented in (b). (d) Comparison of the measured
heat conduction properties of graphene composites with other materials. Note that graphene
composites demonstrate better thermal insulation at cryogenic temperatures and superior
thermal conductivity at RT. The experimental data for composites with other fillers are from
Refs. 28,46.]

As follows from the above discussions, the thermal characteristics of semi-crystalline systems can
drastically change in the low-temperature limits due to the increase in the dominant phonon
wavelength and the changes in the phonon scattering mechanisms.28,34,44 In this regime, fillers can
act as phonon scattering centers rather than conductive inclusions, which suppresses the thermal
transport of the composites even below the limit of its pristine amorphous polymer matrix.28,34,44

18 | P a g e
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

The temperature-dependent thermal transport data of composites shows a cross-over temperature,

𝑇𝑐 , usually in the interval of 5 K to 20 K, at which the thermal conductivity of the composite is
lower than the pristine polymer host at temperatures below 𝑇𝑐 and vice versa. The inclusion of
fillers causes the plateau region observed in amorphous polymers to disappear. These two features
are attributed to TBR at the interface of polymer-filler, which becomes dominant at low
temperatures.28 Cryogenic characteristics of graphene composites are unique in the sense that they
offer the strongest suppression of the thermal conductivity below the cross-over temperature and
the highest enhancement of the thermal conductivity above the cross-over temperature (see Figure
4 (d)). This is due to the atomic thickness of graphene and FLG, its geometry and exceptionally
high intrinsic in-plane thermal conductivity. For practical applications, graphene composites offer
dual functionality for the circuits and systems where both cooling and thermal insulation are
required. At RT, the thermal conductivity of the graphene composite with 𝑓 = 21.7 vol% reaches
6 Wm−1 K −1, which is suitable for cooling semiconductor electronics, whereas a dilute composite
with only 8.3 vol% filler loading reveals 𝑘~0.02 Wm−1 K −1 at 𝑇 = 2 K, providing excellent
thermal insulating properties for superconducting electronics.

CONCLUSIONS

We demonstrated that at cryogenic temperatures, the thermal conductivity of graphene composites

can be both higher and lower than that of the reference pristine epoxy, depending on the graphene
filler loading and specific temperature. There exists a well-defined cross-over temperature – above
it, the thermal conductivity increases with the addition of graphene; below it, the thermal
conductivity decreases with the addition of graphene. The counter-intuitive trend was explained
by the specificity of heat conduction at low temperatures. The randomly distributed graphene
fillers can serve, simultaneously, as the scattering centers for acoustic phonons in the matrix
material and as the conduits of heat. We also argued that the onset of the thermal percolation
threshold can undergo modification owing to the dominance of the low-wavelength phonons that
facilitate the filler-to-filler heat conduction even before the fillers are physically connected. The
obtained results suggest the possibility of using composites with the same constituent materials
for, both, removing the heat and thermally insulating electronic components at cryogenic

19 | P a g e
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

temperatures. The latter is an important capability for the development of quantum computing
technologies and cryogenically cooled conventional semiconductor electronics.

CONTRIBUTIONS

F.K. and A.A.B. conceived the idea of the low-temperature study of graphene composites for
thermal management applications in cryogenic electronics and quantum computing technologies.
F.K. and A.A.B. coordinated the project and contributed to the experimental data analysis. Z.E.N.
exfoliated graphene fillers, prepared the graphene composites, performed Raman measurements,
and contributed to the thermal data analysis. Y.X. conducted specific heat and thermal conductivity
measurements using PPMS. J.B. assisted with the thermal data analysis. J.G. performed numerical
modeling of thermal conductivity and contributed to data analysis. X.C. supervised thermal
measurements and contributed to the experimental data analysis. F.K. and A.A.B. led the
manuscript preparation. All authors contributed to the writing and editing of the manuscript.

Acknowledgements

A.A.B. acknowledges the support of the Vannevar Bush Faculty Fellowship from the Office of
Secretary of Defense (OSD), under the office of Naval Research (ONR) contract N00014-21-1-
2947.

20 | P a g e
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

REFERENCES

(1) Tan, K. Y.; Partanen, M.; Lake, R. E.; Govenius, J.; Masuda, S.; Möttönen, M. Quantum-
Circuit Refrigerator. Nat. Commun. 2017, 8 (1), 1–8.
https://doi.org/10.1038/ncomms15189.

(2) DARPA LTLT program developing Cryogenic microelectronics for future

Supercomputers and data centres – International Defense Security & Technology
https://idstch.com/technology/electronics/darpa-ltlt-program-developing-cryogenic-
microelectronics-for-future-supercomputers-and-data-centres/ (accessed Jan 18, 2023).

(3) Arute, F.; Arya, K.; Babbush, R.; Bacon, D.; Bardin, J. C.; Barends, R.; Biswas, R.;
Boixo, S.; Brandao, F. G. S. L.; Buell, D. A.; Burkett, B.; Chen, Y.; Chen, Z.; Chiaro, B.;
Collins, R.; Courtney, W.; Dunsworth, A.; Farhi, E.; Foxen, B.; Fowler, A.; Gidney, C.;
Giustina, M.; Graff, R.; Guerin, K.; Habegger, S.; Harrigan, M. P.; Hartmann, M. J.; Ho,
A.; Hoffmann, M.; Huang, T.; Humble, T. S.; Isakov, S. V; Jeffrey, E.; Jiang, Z.; Kafri,
D.; Kechedzhi, K.; Kelly, J.; Klimov, P. V; Knysh, S.; Korotkov, A.; Kostritsa, F.;
Landhuis, D.; Lindmark, M.; Lucero, E.; Lyakh, D.; Mandrà, S.; McClean, J. R.;
McEwen, M.; Megrant, A.; Mi, X.; Michielsen, K.; Mohseni, M.; Mutus, J.; Naaman, O.;
Neeley, M.; Neill, C.; Niu, M. Y.; Ostby, E.; Petukhov, A.; Platt, J. C.; Quintana, C.;
Rieffel, E. G.; Roushan, P.; Rubin, N. C.; Sank, D.; Satzinger, K. J.; Smelyanskiy, V.;
Sung, K. J.; Trevithick, M. D.; Vainsencher, A.; Villalonga, B.; White, T.; Yao, Z. J.;
Yeh, P.; Zalcman, A.; Neven, H.; Martinis, J. M. Quantum Supremacy Using a
Programmable Superconducting Processor. Nature 2019, 574 (7779), 505–510.
https://doi.org/10.1038/s41586-019-1666-5.

(4) Demko, J. A.; Fesmire, J. E.; Shu, Q.-S. Cryogenic Heat Management: Technology and
Applications for Science and Industry; CRC Press, 2022.

(5) Chen, D.; Li, J.; Yuan, Y.; Gao, C.; Cui, Y.; Li, S.; Liu, X.; Wang, H.; Peng, C.; Wu, Z. A
Review of the Polymer for Cryogenic Application: Methods, Mechanisms and
Perspectives. Polymers. 2021, pp 1–29. https://doi.org/10.3390/polym13030320.

21 | P a g e
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

(6) Sápi, Z.; Butler, R. Properties of Cryogenic and Low Temperature Composite Materials –
A Review. Cryogenics (Guildf). 2020, 111, 103190.
https://doi.org/10.1016/J.CRYOGENICS.2020.103190.

(7) Choi, Y. S.; Kim, D. L. Thermal Property Measurement of Insulating Material Used in
HTS Power Device. J. Mech. Sci. Technol. 2012, 26 (7), 2125–2128.
https://doi.org/10.1007/s12206-012-0528-y.

(8) Amils, R. I.; Gallego, J. D.; Sebastián, J. L.; Muñoz, S.; Martín, A.; Leuther, A. Thermal
Conductivity of Silver Loaded Conductive Epoxy from Cryogenic to Ambient
Temperature and Its Application for Precision Cryogenic Noise Measurements.
Cryogenics (Guildf). 2016, 76, 23–28.
https://doi.org/10.1016/J.CRYOGENICS.2016.03.001.

(9) Verma, R.; Behera, U.; Kasthurirengan, S.; Shivaprakash, N. C.; Udgata, S. S.;
Gangradey, R. Measurement of Thermal Conductivity of Materials down to 4.5 K for
Development of Cryosorption Pumps. In IOP Conference Series: Materials Science and
Engineering; IOP Publishing, 2017; Vol. 171, p 012098. https://doi.org/10.1088/1757-
899X/171/1/012098.

(10) Rondeaux, F.; Bredy, P.; Rey, J. M. Thermal Conductivity Measurements of Epoxy
Systems at Low Temperature. AIP Conf. Proc. 2002, 614 (1), 197.
https://doi.org/10.1063/1.1472543.

(11) Xie, X.; Li, D.; Tsai, T.-H. H.; Liu, J.; Braun, P. V.; Cahill, D. G. Thermal Conductivity,
Heat Capacity, and Elastic Constants of Water-Soluble Polymers and Polymer Blends.
Macromolecules 2016, 49 (3), 972–978. https://doi.org/10.1021/acs.macromol.5b02477.

(12) Shahil, K. M. F.; Balandin, A. A. Graphene-Multilayer Graphene Nanocomposites as

Highly Efficient Thermal Interface Materials. Nano Lett. 2012, 12 (2), 861–867.
https://doi.org/10.1021/nl203906r.

(13) Naghibi, S.; Kargar, F.; Wright, D.; Huang, C. Y. T.; Mohammadzadeh, A.; Barani, Z.;
Salgado, R.; Balandin, A. A. Noncuring Graphene Thermal Interface Materials for
Advanced Electronics. Adv. Electron. Mater. 2020, 6 (4), 1901303.

22 | P a g e
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

https://doi.org/10.1002/aelm.201901303.

(14) Barani, Z.; Mohammadzadeh, A.; Geremew, A.; Huang, C. Y.; Coleman, D.; Mangolini,
L.; Kargar, F.; Balandin, A. A. Thermal Properties of the Binary-Filler Hybrid Composites
with Graphene and Copper Nanoparticles. Adv. Funct. Mater. 2020, 30 (8), 1904008.
https://doi.org/10.1002/adfm.201904008.

(15) Ohayon-Lavi, A.; Buzaglo, M.; Ligati, S.; Peretz-Damari, S.; Shachar, G.; Pinsk, N.;
Riskin, M.; Schatzberg, Y.; Genish, I.; Regev, O. Compression-Enhanced Thermal
Conductivity of Carbon Loaded Polymer Composites. Carbon 2020, 163, 333–340.
https://doi.org/10.1016/j.carbon.2020.03.026.

(16) Hernandez, Y.; Nicolosi, V.; Lotya, M.; Blighe, F. M.; Sun, Z.; De, S.; McGovern, I. T.;
Holland, B.; Byrne, M.; Gun’Ko, Y. K.; Boland, J. J.; Niraj, P.; Duesberg, G.;
Krishnamurthy, S.; Goodhue, R.; Hutchison, J.; Scardaci, V.; Ferrari, A. C.; Coleman, J.
N. High-Yield Production of Graphene by Liquid-Phase Exfoliation of Graphite. Nat.
Nanotechnol. 2008, 3 (9), 563–568. https://doi.org/10.1038/nnano.2008.215.

(17) Lotya, M.; Hernandez, Y.; King, P. J.; Smith, R. J.; Nicolosi, V.; Karlsson, L. S.; Blighe,
F. M.; De, S.; Wang, Z.; McGovern, I. T.; Duesberg, G. S.; Coleman, J. N. Liquid Phase
Production of Graphene by Exfoliation of Graphite in Surfactant/Water Solutions. J. Am.
Chem. Soc. 2009, 131 (10), 3611–3620. https://doi.org/10.1021/ja807449u.

(18) Balandin, A. A. Thermal Properties of Graphene and Nanostructured Carbon Materials.

Nat. Mater. 2011, 10 (8), 569–581. https://doi.org/10.1038/nmat3064.

(19) Balandin, A. A. Phononics of Graphene and Related Materials. ACS Nano 2020, 14 (5),
5170–5178. https://doi.org/10.1021/acsnano.0c02718.

(20) Sun, X.; Huang, C.; Wang, L.; Liang, L.; Cheng, Y.; Fei, W.; Li, Y. Recent Progress in
Graphene/Polymer Nanocomposites. Adv. Mater. 2020, 2001105.
https://doi.org/10.1002/adma.202001105.

(21) Kargar, F.; Barani, Z.; Salgado, R.; Debnath, B.; Lewis, J. S.; Aytan, E.; Lake, R. K.;
Balandin, A. A. Thermal Percolation Threshold and Thermal Properties of Composites
with High Loading of Graphene and Boron Nitride Fillers. ACS Appl. Mater. Interfaces

23 | P a g e
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

2018, 10 (43), 37555–37565. https://doi.org/10.1021/acsami.8b16616.

(22) Sarikhani, N.; Arabshahi, Z. S.; Saberi, A. A.; Moshfegh, A. Z. Unified Modeling and
Experimental Realization of Electrical and Thermal Percolation in Polymer Composites.
Appl. Phys. Rev. 2022, 9 (4), 041403. https://doi.org/10.1063/5.0089445.

(23) Shtein, M.; Nadiv, R.; Buzaglo, M.; Kahil, K.; Regev, O. Thermally Conductive
Graphene-Polymer Composites: Size, Percolation, and Synergy Effects. Chem. Mater.
2015, 27 (6), 2100–2106. https://doi.org/10.1021/cm504550e.

(24) Minnich, A. J. Advances in the Measurement and Computation of Thermal Phonon

Transport Properties. J. Phys. Condens. Matter 2015, 27 (5), 053202.
https://doi.org/10.1088/0953-8984/27/5/053202.

(25) Kim, T.; Drakopoulos, S. X.; Ronca, S.; Minnich, A. J. Origin of High Thermal
Conductivity in Disentangled Ultra-High Molecular Weight Polyethylene Films: Ballistic
Phonons within Enlarged Crystals. Nat. Commun. 2022, 13 (1), 1–9.
https://doi.org/10.1038/s41467-022-29904-2.

(26) Sudhindra, S.; Rashvand, F.; Wright, D.; Barani, Z.; Drozdov, A. D.; Baraghani, S.;
Backes, C.; Kargar, F.; Balandin, A. A. Specifics of Thermal Transport in Graphene
Composites: Effect of Lateral Dimensions of Graphene Fillers. ACS Appl. Mater.
Interfaces 2021, 13 (44), 53073–53082.

(27) Ghosh, S.; Calizo, I.; Teweldebrhan, D.; Pokatilov, E. P.; Nika, D. L.; Balandin, A. A.;
Bao, W.; Miao, F.; Lau, C. N. Extremely High Thermal Conductivity of Graphene:
Prospects for Thermal Management Applications in Nanoelectronic Circuits. Appl. Phys.
Lett. 2008, 92 (15), 151911. https://doi.org/10.1063/1.2907977.

(28) Garrett, K. W.; Rosenberg, H. M. The Thermal Conductivity of Epoxy-Resin / Powder

Composite Materials. J. Phys. D. Appl. Phys. 1974, 7 (9), 1247–1258.
https://doi.org/10.1088/0022-3727/7/9/311.

(29) Stankovich, S.; Dikin, D. A.; Dommett, G. H. B.; Kohlhaas, K. M.; Zimney, E. J.; Stach,
E. A.; Piner, R. D.; Nguyen, S. B. T.; Ruoff, R. S. Graphene-Based Composite Materials.
Nature 2006, 442 (7100), 282–286. https://doi.org/10.1038/nature04969.

24 | P a g e
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

(30) Ferrari, A. C.; Meyer, J. C.; Scardaci, V.; Casiraghi, C.; Lazzeri, M.; Mauri, F.; Piscanec,
S.; Jiang, D. Da; Novoselov, K. S.; Roth, S.; others; Geim, a. K. Raman Spectrum of
Graphene and Graphene Layers. Phys. Rev. Lett. 2006, 97 (18), 1–4.
https://doi.org/10.1103/PhysRevLett.97.187401.

(31) Kelham, S.; Rosenberg, H. M. The Thermal Conductivity and Specific Heat of Epoxy-
Resin from 0.1-80K. J. Phys. C Solid State Phys. 1981, 14 (12), 1737.
https://doi.org/10.1088/0022-3719/14/12/007.

(32) Nakamura, S.; Fujii, T.; Matsukawa, S.; Katagiri, M.; Fukuyama, H. Specific Heat,
Thermal Conductivity, and Magnetic Susceptibility of Cyanate Ester Resins – An
Alternative to Commonly Used Epoxy Resins. Cryogenics (Guildf). 2018, 95, 76–81.
https://doi.org/10.1016/j.cryogenics.2018.09.001.

(33) Evseeva, I. E.; Tanaeva, S. A. Thermophysical Properties of Epoxy Composite Materials

at Low Temperatures. Cryogenics (Guildf). 1995, 35 (4), 277–279.
https://doi.org/10.1016/0011-2275(95)90833-2.

(34) Choy, C. L. Thermal Conductivity of Polymers. Polymer. Elsevier October 1, 1977, pp

984–1004. https://doi.org/10.1016/0032-3861(77)90002-7.

(35) Ramos, M. A. Introduction: About 50 Years of Two-Level Systems and Boson Peak. Low-
Temperature Therm. Vib. Prop. Disord. Solids 2022, 1–20.
https://doi.org/10.1142/9781800612587_0001.

(36) Zeller, R. C.; Pohl, R. O. Thermal Conductivity and Specific Heat of Noncrystalline
Solids. Phys. Rev. B 1971, 4 (6), 2029. https://doi.org/10.1103/PhysRevB.4.2029.

(37) Pohl, R. O.; Liu, X.; Thompson, E. J. Low-Temperature Thermal Conductivity and
Acoustic Attenuation in Amorphous Solids. Rev. Mod. Phys. 2002, 74 (4), 991–1013.
https://doi.org/10.1103/RevModPhys.74.991.

(38) Talón, C.; Zou, Q. W.; Ramos, M. A.; Villar, R.; Vieira, S. Low-Temperature Specific
Heat and Thermal Conductivity of Glycerol. Phys. Rev. B - Condens. Matter Mater. Phys.
2002, 65 (1), 122031–122034. https://doi.org/10.1103/physrevb.65.012203.

(39) Ramos, M. A. Low-Temperature Thermal and Vibrational Properties of Disordered

25 | P a g e
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

Solids. Low-Temperature Therm. Vib. Prop. Disord. Solids 2022.

https://doi.org/10.1142/Q0371/SUPPL_FILE/Q0371_PREFACE.PDF.

(40) Scheibner, W.; Jäckel, M. Thermal Conductivity and Specific Heat of an Epoxy
Resin/Epoxy Resin Composite Material at Low Temperatures. Phys. status solidi 1985, 87
(2), 543–547. https://doi.org/10.1002/pssa.2210870216.

(41) Srivastava, G. P. The Physics of Phonons, 1st ed.; CRC Press, 1990.

(42) DeAngelis, F.; Muraleedharan, M. G.; Moon, J.; Seyf, H. R.; Minnich, A. J.; McGaughey,
A. J. H.; Henry, A. Thermal Transport in Disordered Materials.
https://doi.org/10.1080/15567265.2018.1519004 2018, 23 (2), 81–116.
https://doi.org/10.1080/15567265.2018.1519004.

(43) Wingert, M. C.; Zheng, J.; Kwon, S.; Chen, R. Thermal Transport in Amorphous
Materials: A Review. Semicond. Sci. Technol. 2016, 31 (11), 113003.
https://doi.org/10.1088/0268-1242/31/11/113003.

(44) Choy, C. L.; Greig, D. The Low-Temperature Thermal Conductivity of a Semi-Crystalline

Polymer, Polyethylene Terephthalate. J. Phys. C Solid State Phys. 1975, 8 (19), 3121–
3130. https://doi.org/10.1088/0022-3719/8/19/012.

(45) Anderson, A. C.; Rauch, R. B. Low-Temperature Thermal Conductivity of a Suspension

of Copper Particles. J. Appl. Phys. 1970, 41 (9), 3648–3651.
https://doi.org/10.1063/1.1659486.

(46) Araujo, F. F. T.; Rosenberg, H. M.; De Araujo, F. F. T.; Rosenberg, H. M. The Thermal
Conductivity of Epoxy-Resin/Metal-Powder Composite Materials from 1.7 to 300K. J.
Phys. D. Appl. Phys. 1976, 9 (4), 665–675. https://doi.org/10.1088/0022-3727/9/4/017.

(47) Hartwig, G.; Knaak, S. Fibre-Epoxy Composites at Low Temperatures. Cryogenics

(Guildf). 1984, 24 (11), 639–647. https://doi.org/10.1016/0011-2275(84)90083-3.

(48) Jäckel, M. Thermal Properties of Polymer/Particle Composites at Low Temperatures.

Cryogenics (Guildf). 1995, 35 (11), 713–716. https://doi.org/10.1016/0011-
2275(95)90896-N.

26 | P a g e
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

(49) Little, W. A. The Transport of Heat between Dissimilar Solids at Low Temperatures. Can.
J. Phys. 1959, 37 (3), 334–349. https://doi.org/10.1139/p59-037.

(50) Swartz, E. T.; Pohl, R. O. Thermal Boundary Resistance. Rev. Mod. Phys. 1989, 61 (3),
605–668. https://doi.org/10.1103/RevModPhys.61.605.

(51) Li, A.; Zhang, C.; Zhang, Y.-F. Thermal Conductivity of Graphene-Polymer Composites:
Mechanisms, Properties, and Applications. Polymers. 2017, 9 (9), 437.
https://doi.org/10.3390/polym9090437.

(52) Shenogina, N.; Shenogin, S.; Xue, L.; Keblinski, P. On the Lack of Thermal Percolation in
Carbon Nanotube Composites. Appl. Phys. Lett. 2005, 87 (13), 1–3.
https://doi.org/10.1063/1.2056591.

(53) Kargar, F.; Debnath, B.; Kakko, J.-P.; Saÿnätjoki, A.; Lipsanen, H.; Nika, D. L.; Lake, R.
K.; Balandin, A. A. Direct Observation of Confined Acoustic Phonon Polarization
Branches in Free-Standing Semiconductor Nanowires. Nat. Commun. 2016, 7, 13400.
https://doi.org/10.1038/ncomms13400.

(54) Yang, F.; Dames, C. Mean Free Path Spectra as a Tool to Understand Thermal
Conductivity in Bulk and Nanostructures. Phys. Rev. B 2013, 87 (3), 35437.
https://doi.org/10.1103/PhysRevB.87.035437.

(55) Nan, C. W.; Birringer, R.; Clarke, D. R.; Gleiter, H. Effective Thermal Conductivity of
Particulate Composites with Interfacial Thermal Resistance. J. Appl. Phys. 1998, 81 (10),
6692. https://doi.org/10.1063/1.365209.

(56) Su, Y.; Li, J. J.; Weng, G. J. Theory of Thermal Conductivity of Graphene-Polymer
Nanocomposites with Interfacial Kapitza Resistance and Graphene-Graphene Contact
Resistance. Carbon 2018, 137, 222–233.
https://doi.org/10.1016/J.CARBON.2018.05.033.

27 | P a g e
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

Supporting information

Low-Temperature Thermal Characteristics of Graphene Composites –

Applications in Thermal Management for Quantum Technologies

Zahra Ebrahim Nataj,1 Youming Xu,2 Jonas Brown,1 Jivtesh Garg,3 Xi Chen, 2 Fariborz
Kargar,1,‡ and Alexander A. Balandin1,*

1
Nano-Device Laboratory and Phonon Optimized Engineered Materials Center, Department of
Electrical and Computer Engineering, University of California, Riverside, California 92521 USA

2
Department of Electrical and Computer Engineering, University of California, Riverside,
California 92521 USA

3
Department of Aerospace and Mechanical Engineering, University of Oklahoma, Norman,
Oklahoma 73019 USA

‡
Corresponding authors: fkargar@ece.ucr.edu ; balandin@ece.ucr.edu ; web-site: http://balandingroup.ucr.edu/

28 | P a g e
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

Contents
1 Sample Preparation .......................................................................................................................... 29
2 Mass Density Measurement ............................................................................................................. 30
3 Thermal Conductivity Measurement and uncertainty analysis ................................................... 31
4 Heat capacity measurement and uncertainty analysis .................................................................. 32
5 Effective medium model for cryogenic heat conduction for low-loading composites ................. 34
6 Effective medium model for cryogenic heat conduction for high-loading composites (including
percolation effect)...................................................................................................................................... 36

1 Sample Preparation

For each composite sample with desired Few-layer graphene (FLG) loading, commonly produced
FLG (xGnP H-25, XG Sciences, US) with an average surface area of 65 𝑚2 𝑔−1 and average lateral
dimensions of 25 𝜇𝑚 were mixed with the base epoxy resin (bisphenol A (epichlorhydrin);
molecular weight, 700; Allied HighTech Products, Inc.) at predetermined proportions. In order to
have a uniform compound, FLG was added in several steps and mixed for 3 minutes at 800 rpm in
a high-shear speed mixer in each step. The hardener (triethylenetetramine, Allied HighTech
Products, Inc.) was then added at a mass ratio of 12:100 to the epoxy resin. The final compound
was mixed and vacuumed for 10 minutes to remove trapped air bubbles. This procedure was
performed three times to achieve a void-free combination. Then, the samples set aside at room
temperature for about 8 hours to harden. At higher graphene concentrations (from 20 wt% and
above), it was hard to reduce its porosity with just mixing and vacuuming the compound. So, one
step was added to the previous procedures. Before leaving the samples at room temperature to be
rigid, the completed mixture was poured into spherical molds, gently squeezed. Finally, all the
samples were put in a 130℃ furnace for 3 hours to cure and solidify. The final composite samples
were disks with a diameter of 25.4 mm and a thickness of 5 mm. the optical images of samples are
shown in Figure 1.

29 | P a g e
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

2 Mass Density Measurement

The mass density of the samples was determined using the Archimedes principle and an electronic
scale (Mettler Toledo). The density is calculated using the formula,

𝜌𝑐 = (𝑤𝑎 ⁄(𝑤𝑎 − 𝑤𝑤 )) × (𝜌𝑤 − 𝜌𝑎 ) + 𝜌𝑎 (S1)

where 𝑤𝑎 , 𝑤𝑤 are the sample's weight in air and water, respectively, and 𝜌𝑤 and 𝜌𝑎 are the density
of ionized water and air (0.0012 𝑔𝑐𝑚−3) at room temperature .The results of the mass density
measurements for different filler concentrations presented in Figure S1 indicate that addition of
graphene fillers to the composites would increase the density of the samples.

Figure S1: Mass density of composites as a function of, a) volume loading and b) mass loading
fraction of graphene.

30 | P a g e
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

3 Thermal Conductivity Measurement and uncertainty analysis

The samples were cut into a typical dimension of 1×1×10 mm for the thermal conductivity
measurement. A Quantum Design Physical Property Measurement System (PPMS) was employed
to measure the thermal conductivity from 2 to 300 K with a steady-state 4-probe continuous mode.
In this mode, the temperature sweeps between 2 and 300 K at low rate (0.3 K/min) and
measurements are taken continually, while the measurement parameters like heating power and
period are adjusted by software automatically.1 When a square-wave heat pulse is applied, the
temperature difference between the hot probe and cold probe is measured as a function of time and
a steady-state temperature difference is fitted using the following equation:

𝑡 𝑡
𝜏1 × exp (− 𝜏 ) − 𝜏2 × exp (− 𝜏 )
1 2
∆𝑇 = ∆𝑇∞ × (1 − ) (S2)
𝜏1 − 𝜏2

where ∆𝑇 is the measured temperature difference, ∆𝑇∞ is the steady-state temperature difference,
𝜏1 and 𝜏2 are time constants and can be obtained from the fitting.

When ∆𝑇∞ is calculated, the thermal conductance K is calculated as

𝐾 = 𝑃/∆𝑇∞ (S3)

where P is the heat flowing through the sample. And P is calculated using Joule heat minus
radiation heat:

𝑃 = 𝐼 2 𝑅 − 𝑃𝑟𝑎𝑑 (S4)

where I is the current flowing into the sample and R is the resistance of the heater, the radiation
loss is calculated as

𝑆 4 4 (S5)
𝑃𝑟𝑎𝑑 = 𝜎𝑇 × ( ) × 𝜀 × (𝑇ℎ𝑜𝑡 − 𝑇𝑐𝑜𝑙𝑑 )
2

31 | P a g e
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

where 𝜎𝑇 = 5.67 × 10-8 Wm-2K-4 is the Stefan-Boltzmann constant, S is the surface area of the
sample, 𝜀 is the emissivity, 𝑇ℎ𝑜𝑡 and 𝑇𝑐𝑜𝑙𝑑 are the temperature of hot probe and cold probe.

Then the shoe assembly’s thermal conductance 𝐾𝑠ℎ𝑜𝑒 is deduced from K to get sample thermal
conductance 𝐾𝑠𝑎𝑚𝑝𝑙𝑒 as:

𝐾𝑠𝑎𝑚𝑝𝑙𝑒 = 𝐾 − 𝐾𝑠ℎ𝑜𝑒 (S6)

𝐾𝑠ℎ𝑜𝑒 = 𝑎𝑇 + 𝑏𝑇 2 + 𝑐𝑇 3 (S7)

where a, b, c are constants. The thermal conductivity of the sample is calculated as:

𝜅 = 𝐾𝑠𝑎𝑚𝑝𝑙𝑒 × 𝑙/(𝑤 × 𝑡) (S8)

where l, w, t are the distance between the hot and cold probe, width and thickness of the sample,
respectively.

The error of the thermal conductivity measurement includes the fitting of ∆𝑇, the error of the
heating power, the error in the estimation of the sample radiation due to the error of sample surface
area and emissivity, the error in 𝐾𝑠ℎ𝑜𝑒 ,the error from the measurements of the size of sample and
the error from the measurement of the distance between hot and cold probes d. So the error is
calculated as

𝑅 𝐼𝑅𝜕𝐼 2 0.2×𝑃𝑟𝑎𝑑 2 0.1×𝑇∞ ×𝐾𝑠ℎ𝑜𝑒 2 ∆𝑙 ∆𝑤 ∆𝑡

𝜎(𝜅) = 𝜅 × √(∆𝑇∆𝑇 )2 + (2 ) +( ) +( ) + ( 𝑙 )2 + ( 𝑤 )2 +( 𝑡 )2 (S9)
∞ 𝑃 𝑃 𝑃

where 𝑅∆𝑇 is the residual term from the ∆𝑇 versus t fitting.

4 Heat capacity measurement and uncertainty analysis

The sample of a typical mass of 5 mg was used for the heat capacity measurement from 2 to 300
K with the PPMS. The measurement of heat capacity assumes that the temperature of the sample
𝑇𝑠 is different from the temperature of the platform 𝑇𝑝 because the thermal contact between the

32 | P a g e
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

sample and the platform is not good enough.2 The heat transfer between the sample and the
platform is described by the following equations:

𝑑𝑇𝑝 (S10)
𝐶𝑝𝑙𝑎𝑡𝑓𝑜𝑟𝑚 = 𝑃(𝑡) − 𝐾𝑤 (𝑇𝑝 (𝑡) − 𝑇𝑏 ) + 𝐾𝑔 (𝑇𝑠 (𝑡) − 𝑇𝑝 (𝑡))
𝑑𝑥
𝑑𝑇𝑠 (S11)
𝐶𝑠𝑎𝑚𝑝𝑙𝑒 = −𝐾𝑔 (𝑇𝑠 (𝑡) − 𝑇𝑝 (𝑡))
𝑑𝑥

where 𝑃(𝑡) is the heater power. 𝐶𝑝𝑙𝑎𝑡𝑓𝑜𝑟𝑚 and 𝐶𝑠𝑎𝑚𝑝𝑙𝑒 are the heat capacity of the platform and
the sample, respectively. 𝐾𝑤 is the thermal conductance of the supporting wires, 𝐾𝑔 is the thermal
conductance between the platform and the sample due to the grease. 𝑇𝑏 is the temperature of the
thermal bath. 𝑇𝑠 and 𝑇𝑝 are function of time.

The error of the heat capacity measurements includes the total heat capacity error from the fitting
of the heat capacity parameters, the addenda heat capacity error from the addenda measurement,
sample mass error and the fitting deviation error from the modeling. Therefore, the error of heat
capacity 𝐶𝑝 is calculated as

𝑅𝐶𝑝𝑙𝑎𝑡𝑓𝑜𝑟𝑚 (S12)
𝑅𝐶 ∆𝑀 𝑅∆𝑇 2
𝜎(𝐶𝑝 ) = 𝐶𝑝 × √( 𝑡𝑜𝑡𝑎𝑙 )2 + ( )2 + ( )2 + ( )
𝐶𝑡𝑜𝑡𝑎𝑙 𝐶𝑝𝑙𝑎𝑡𝑓𝑜𝑟𝑚 𝑀 ∆𝑇

where 𝑅𝐶𝑡𝑜𝑡𝑎𝑙 and 𝑅𝐶𝑝𝑙𝑎𝑡𝑓𝑜𝑟𝑚 are the mean-square deviation of the fit to variations in the total heat

capacity and platform heat capacity, M is the mass of sample and 𝑅∆𝑇 is the fitting deviation error.

Figure S2 shows the heat capacity of all samples as a function of temperature in the range of
100 K ≤ 𝑇 ≤ 300 K. Based on the figure, one can see that heat capacity of each filler
concentration increases as temperatures increase with a linear relation. Also, by increasing the
filler amount the heat capacity decreased. However, there is an irregularity in this behavior for
filler concentration of 5.4 vol% which might be because of systematic errors.

33 | P a g e
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

Figure S2: Heat capacity of composites with different graphene loading as a function of
temperature. The shaded area around the experimental data points displays the errors involved
in the measurements.

5 Effective medium model for cryogenic heat conduction for low-loading

composites

In the equation (1) of the main text, 𝐿𝑖𝑖 are geometrical parameters and depend upon the aspect
ratio, 𝑝 = 𝑡/𝐿, of graphene fillers with 𝑡 and 𝐿 being the thickness and lateral dimensions of fillers.
For oblate inclusions such as nanoplatelets, where 𝑝 < 1, these geometrical parameters, 𝐿𝑖𝑖 , are
computed using the following equations,

𝑝2 𝑝
𝐿11 = 𝐿22 = 2(𝑝2 −1) + 2(1−𝑝2 )3/2 cos−1 𝑝 (S13)

𝐿33 = 1 − 2𝐿11 (S14)

Temperature Dependence of thermal conductivity of graphite:

Temperature dependent in-plane thermal conductivity of natural graphite was obtained from 3.
Temperature dependent through-plane thermal conductivity was obtained from 4. These data are

34 | P a g e
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

presented in Figures S3a and S3b. It is visible that thermal conductivity of graphite can reach very
low values at low temperatures. At 4 K, the in-plane thermal conductivity of graphite decreases to
~ 1 W/mK, while the through-plane thermal conductivity reaches 0.23 W/mK.

Figure S3: Temperature dependence of, a) in-plane and b) through-plane thermal

conductivities of natural graphite.

The explicit temperature dependence of in-plane and through-plane thermal conductivities (fitted
to the data in Figs. S3a and b) is given by following equations:

𝑘𝑖𝑛 = exp(0.00116[ln(𝑇)]6 − 0.02386[ln(𝑇)]5 + 0.17360[ln(𝑇)]4 − 0.66809[ln(𝑇)]3 +

1.67342[ln(𝑇)]2 − 0.19473[ln(𝑇)]1 − 1.65861)
(S15)

𝑘𝑜𝑢𝑡 = exp(0.0975538[ln(𝑇)]4 − 1.6222796[ln(𝑇)]3 + 9.1860571[ln(𝑇)]2 −

19.9450584[ln(𝑇)]1 + 14.4337754)
(S16)

The thermal conductivity of epoxy was taken to be temperature dependent from measurements.
Following curve fits can provide good fit to the experimental data.

For 2 K< T <14 K

35 | P a g e
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

𝑘𝑒𝑝𝑜𝑥𝑦 = 1.5044 × 10−6 𝑇 5 − 7.2195 × 10−5 𝑇 4 + 1.3521 × 10−3 𝑇 3 − 1.2278 × 10−2 𝑇 2 +

5.5773 × 10−2 𝑇 − 2.6347 × 10−2
(S17)

For 14 K < T <300 K

𝑘𝑒𝑝𝑜𝑥𝑦 = 1.3105 × 10−14 𝑇 6 − 1.1288 × 10−11 𝑇 5 + 3.3133 × 10−9 𝑇 4 − 3.08119 ×

10−7 𝑇 3 − 2.01387 × 10−5 𝑇 2 + 4.9861 × 10−3 𝑇 + 1.13744 × 10−2
(S18)

Explicit temperature dependence of interface thermal resistance is provided below.

𝑅𝑖𝑛𝑡𝑒𝑟𝑓𝑎𝑐𝑒 = exp(0.0111[ln(𝑇)]4 − 0.1147[ln(𝑇)]3 + 0.3702[ln(𝑇)]2 −

1.5114[ln(𝑇)]1 − 8.7497) (S19)
The interface thermal conductance and thermal resistance (inverse of conductance) is shown in
Figure S4.

Figure S4: Temperature dependence of interfacial, a) thermal conductance and b) thermal resistance,
between graphene and pure epoxy.

6 Effective medium model for cryogenic heat conduction for high-loading

composites (including percolation effect)

In the equation (5) of main text, 𝑆11 and 𝑆33 are the shape parameters related to the aspect ratio of
graphitic nanosheets, given by following equations.

36 | P a g e
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

𝑝 (S20)
𝑆11 = 𝑆22 = [cos −1 𝑝 − 𝑝(1 − 𝑝2 )1/2 ], 𝑝<1
2(1 − 𝑝2 )3/2
𝑆33 = 1 − 𝑆11 (S21)

In the above equations, 𝑝 is the aspect ratio of the fillers. The effective in-plane and through-plane
thermal conductivities, 𝑘11 and 𝑘33 (in Eq. 5 in main text) are computed using

(1 − 𝑐𝑖𝑛𝑡 )(𝑘𝑖𝑛 − 𝑘0 ) (S22)

𝑘11 = 𝑘0 [1 + ]
𝑐𝑖𝑛𝑡 𝑆11 (𝑘𝑖𝑛 − 𝑘0 )+𝑘0

(1 − 𝑐𝑖𝑛𝑡 )(𝑘𝑜𝑢𝑡 − 𝑘0 ) (S23)

𝑘33 = 𝑘0 [1 + ]
𝑐𝑖𝑛𝑡 𝑆11 (𝑘𝑜𝑢𝑡 − 𝑘0 )+𝑘0

In above equations, 𝑘0 is thermal conductivity of an interlayer surrounding graphene sheets. This

interlayer represents the interface thermal resistance surrounding the graphene particles. The
interlayer is used to represent the combined effect of graphene-epoxy and graphene-graphene
contact resistance as described in the article. The values of different parameters used in this
percolation based effective medium model are described below. Once again, wherever appropriate,
we used temperature dependent properties. Table S1 below shows the values of different
parameters used in the above calculations.

Table S1: Thermal conductivity calculations parameters

Material Parameters Values
Average graphene lateral length, l, 15 m
Average graphene thickness 15 nm
Aspect ratio of the graphene filler 0.001
Thermal conductivity of epoxy Temperature dependent
Thermal conductivity of graphene filler, 𝑘1 and 𝑘3 Temperature dependent
(W/mK)

37 | P a g e
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

Thermal conductivity of interlayer with Kapitza Temperature dependent

resistance
Thermal conductivity of the interlayer with a firmly Temperature dependent
developed graphene-graphene contact state,

The in-plane (𝑘𝑖𝑛 ) and through-plane thermal conductivities (𝑘𝑜𝑢𝑡 ) of graphene were taken to
betemperature dependent as shown in Figs. S3a and b. . The combined interfacial resistance of
graphene/epoxy and graphene/graphene contactis modeled as an interlayer in above theory with
an effective thermal conductivity of 𝑘0 which is taken to a be weighted sum of thermal conductivity
of graphene/epoxy and graphene/graphene contact. The thickness of this interlayer was nominally
taken to be 1.0 nm. Thermal conductivity of the interlayer based on graphene-epoxy contact was
computed from the thickness and interfacial resistance of graphene/epoxy contact which is taken
to the same as given by equation (S7). The thermal conductivity of interlayer based on
graphene/graphene contact was computed from the interfacial resistance at graphene-graphene
contact which was was taken to be lower than graphene-epoxy interfacial thermal resistance by a
factor ranging from 4.5 for 11.4 vol% composition to 6.0 for 21.7 vol% composition.

REFERENCES

(1) Maldonado, O. Pulse Method for Simultaneous Measurement of Electric Thermopower

and Heat Conductivity at Low Temperatures. Cryogenics (Guildf). 1992, 32 (10), 908–
912. https://doi.org/10.1016/0011-2275(92)90358-H.
(2) Lashley, J. C.; Hundley, M. F.; Migliori, A.; Sarrao, J. L.; Pagliuso, P. G.; Darling, T. W.;
Jaime, M.; Cooley, J. C.; Hults, W. L.; Morales, L.; Thoma, D. J.; Smith, J. L.; Boerio-
Goates, J.; Woodfield, B. F.; Stewart, G. R.; Fisher, R. A.; Phillips, N. E. Critical
Examination of Heat Capacity Measurements Made on a Quantum Design Physical
Property Measurement System. Cryogenics (Guildf). 2003, 43 (6), 369–378.
https://doi.org/10.1016/S0011-2275(03)00092-4.
(3) Nakamura, S.; Miyafuji, D.; Fujii, T.; Matsui, T.; Fukuyama, H. Low Temperature
Transport Properties of Pyrolytic Graphite Sheet. Cryogenics (Guildf). 2017, 86, 118–122.

38 | P a g e
 Z. E. Nataj, et al., Cryogenic Characteristics of Graphene Composites (University of California, Riverside – 2023)

https://doi.org/10.1016/J.CRYOGENICS.2017.08.004.
(4) Ho, C. Y.; Powell, R. W.; Liley, P. E. Thermal Conductivity of the Elements. J. Phys.
Chem. Ref. Data 1972, 1 (2), 279–421. https://doi.org/10.1063/1.3253100.

39 | P a g e