Anisotropic thermal conductivity

measurement of organic thin film with

bidirectional 3ω method
Cite as: Rev. Sci. Instrum. 92, 034902 (2021); https://doi.org/10.1063/5.0030982
Submitted: 26 September 2020 . Accepted: 24 February 2021 . Published Online: 15 March 2021

Shingi Yamaguchi, Takuma Shiga, Shun Ishioka, Tsuguyuki Saito, Takashi Kodama, and Junichiro
Shiomi

ARTICLES YOU MAY BE INTERESTED IN

Thermal conduction through individual cellulose nanofibers

Applied Physics Letters 118, 053701 (2021); https://doi.org/10.1063/5.0042463

Thermal conductivity measurement from 30 to 750 K: the 3ω method

Review of Scientific Instruments 61, 802 (1990); https://doi.org/10.1063/1.1141498

One-directional thermal transport in densely aligned single-wall carbon nanotube films

Applied Physics Letters 115, 223104 (2019); https://doi.org/10.1063/1.5127209

Rev. Sci. Instrum. 92, 034902 (2021); https://doi.org/10.1063/5.0030982 92, 034902

© 2021 Author(s).
 Review of ARTICLE scitation.org/journal/rsi
Scientific Instruments

Anisotropic thermal conductivity measurement

of organic thin film with bidirectional 3ω method
Cite as: Rev. Sci. Instrum. 92, 034902 (2021); doi: 10.1063/5.0030982
Submitted: 26 September 2020 • Accepted: 24 February 2021 •
Published Online: 15 March 2021

Shingi Yamaguchi,1 Takuma Shiga,1 Shun Ishioka,2 Tsuguyuki Saito,2 Takashi Kodama,1
1,a)
and Junichiro Shiomi

AFFILIATIONS
1
Department of Mechanical Engineering, The University of Tokyo, 7-3-1, Hongo, Bunkyo-ku, Tokyo 113-8656, Japan
2
Department of Biomaterial Sciences, Graduate School of Agricultural and Life Sciences, The University of Tokyo,
1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-8657, Japan

a)
Author to whom correspondence should be addressed: shiomi@photon.t.u-tokyo.ac.jp

ABSTRACT
Organic thin film materials with molecular ordering are gaining attention as they exhibit semiconductor characteristics. When using them
for electronics, the thermal management becomes important, where heat dissipation is directional owing to the anisotropic thermal con-
ductivity arising from the molecular ordering. However, it is difficult to evaluate the anisotropy by simultaneously measuring in-plane
and cross-plane thermal conductivities of the film on a substrate because the film is typically as thin as tens to hundreds of nanome-
ters and its in-plane thermal conductivity is low. Here, we develop a novel bidirectional 3ω system that measures the anisotropic ther-
mal conductivity of thin films by patterning two metal wires with different widths and preparing the films on top and extracting the
in-plane and cross-plane thermal conductivities using the difference in their sensitivities to the metal-wire width. Using the developed
system, the thermal conductivity of spin-coated poly(3,4-ethylenedioxythiophene) polystyrene sulfonate (PEDOT:PSS) with thickness of
70 nm was successfully measured. The measured in-plane thermal conductivity of PEDOT:PSS film was as high as 2.9 W m−1 K−1 pre-
sumably due to the high structural ordering, giving an anisotropy of 10. The calculations of measurement sensitivity to the film thick-
ness and thermal conductivities suggest that the device can be applied to much thinner films by utilizing metal wires with a smaller
width.
Published under license by AIP Publishing. https://doi.org/10.1063/5.0030982

INTRODUCTION oscillation signal with various widths of metal wires deposited on

the sample. The measurement is made sensitive to η by making the
In the field of electronics, organic thin films are gaining atten- width of a narrowest wire as small as the thickness of the thin film.
tion because of their semiconductor features.1,2 Thermal manage- Such metal wires are usually prepared on the target sample by lithog-
ment of these materials is an important issue as heat is generated due raphy techniques, which can damage organic materials with harsh
to self-Joule heating, and therefore, the evaluation of their thermal chemicals in the developing or etching process. Some alternative
conductivity (κ) is essential. Many of these materials take a highly methods have been proposed for the κ∥ measurement of organic thin
ordered structure in the form of thin film with a thickness less than films by fabricating metal wires with shadow mask17,18 or applying
100 nm,3,4 and this structural anisotropy and small thickness have to 3ω measurement to a film on a suspended membrane with metal
be considered in the κ measurements. Among the developed mea- wires.19,20 However, these methods also have limits on the appre-
surement techniques for in-plane thermal conductivity (κ∥ ) such ciable sample range; the shadow mask cannot prepare nanometers-
as the offset time domain thermoreflectance (TDTR) method,5–7 wide wire required for the measurement of the submicron-thin film,
suspended microdevice,8,9 and offset laser flash method,10 the 3ω and the suspended membrane can be destructed by the common
method11–16 has the advantage of measuring such thin samples. The spin coating process used for film preparation. Therefore, a general
3ω method separately measures cross-plane thermal conductivity method with wider applicability is required to measure η of organic
(κ– ) and thermal anisotropy (η = κ∥ /κ– ) by detecting the voltage thin films.

Rev. Sci. Instrum. 92, 034902 (2021); doi: 10.1063/5.0030982 92, 034902-1
Published under license by AIP Publishing
 Review of ARTICLE scitation.org/journal/rsi
Scientific Instruments

There also exists the bidirectional 3ω method where a sample TABLE I. Length between the current and voltage probes for each wire width.
is placed on metal wires, which is prepared beforehand on a sub-
Wire width (μm) 50 20 10 2 1 0.5 0.3
strate.21–24 Here, “bidirectional” is named after the directions of heat
conduction, in comparison to the conventional 3ω method where Length A (μm) 4500 1800 1800 1800 1800 900 540
the heat conducts unidirectionally [Figs. 1(a) and 1(b)]. While this Length B (μm) 1500 600 600 600 600 300 180
method can avoid damaging the sample by chemicals, only a sin-
gle narrow wire has been used to extract κ– and η, since the mea-
surement with a wider wire would be insensitive to both κ– and the device as an insulation layer by atomic layer deposition (ALD),
η.25 In this work, we develop a bidirectional 3ω system with wide while Al2 O3 on contact pads was removed by the wet-etching pro-
and narrow wires to measure the anisotropic thermal conductivi- cess using the photoresist and mask aligner. The wafer was cut
ties of organic thin films. In analogy to the conventional 3ω system into square-shaped devices with the size of 1 cm2 . As the wire with
[Fig. 1(a)], κ– was obtained first from the measurement with the 100-nm width did not survive the dry etching process, the 300-nm-
wide wire, where thermally conductive Al layer was deposited to wide wire was used as the narrowest one in this study. Polymethyl
selectively suppress the sensitivity to η [Fig. 1(b)], and then, η was methacrylate (PMMA) and PEDOT:PSS films were prepared by spin
obtained from the measurement with the narrow wire without the coating, and part of the film on the contact pads was wiped off before
Al layer. By preparing the wires with different widths on one sub- the measurement. The details of sample preparation are provided in
strate, the two parameters (κ– and η) can be obtained from the same the supplementary material.
sample. The method can measure the thermal anisotropy of the thin
organic film with a thickness of less than 100 nm, and the results Solution to the bidirectional heat diffusion equation
clarify that κ∥ of highly ordered organic thin film can be more than
three times larger than that ever reported. The two-dimensional equation of Feldman’s algorithm26 was
chosen to take the thermal anisotropy into account. The calculated
EXPERIMENTAL METHOD temperature rise (ΔT) at the wire can be written as

Device and sample preparation P ∞ (A+ + A− )(B+ + B− ) 1 sin2 (kb)

ΔT = ∫ dk, (1)
The measurement device was designed to have eight metal 2πl 0 A+ B− − A− B+ γj (kb)2
wires of Ti/Pt with different widths [Fig. 1(c)]. The wire width and
the lengths between current probes and between voltage probes √
[lengths A and B in Fig. 1(d), respectively] are summarized in 2iω
γj = κj,z ηj k2 + , (2)
Table I. Thirty devices were fabricated on the 4-in. quartz wafer at Dj,z
once, and all the processes were conducted in the Takeda cleanroom
at the University of Tokyo. Quartz was chosen as the substrate mate- where P, l, and b are the heating power, length, and half width
rial to suppress heat conduction to the substrate. Ti/Pt (5-nm and of the wire, respectively. κj,z and Dj,z are the thermal conductivity
20-nm-thick, respectively) was sputtered on the substrate first, and and thermal diffusivity of the j-th layer in the z direction, and ηj is
negative resist for electron beam lithography was spin-coated. The anisotropy (κj,x /κj,z ). k is a variable of integration, and A+ , A− , B+ ,
patterns were prepared by direct exposure, development, and subse- and B− are dimensionless parameters obtained by solving the equa-
quent dry etching. A 100 nm alumina (Al2 O3 ) layer was deposited on tion by a recursive matrix method. The details of the mathematical
solution and material parameters required for calculation (specific
heat capacity and mass density) are included in the supplementary
material.
The absolute value of measurement sensitivities (∣Sβ ∣) to the
parameters of interest (β) can be calculated using Eq. (1) as
RRR −1 R
Δβ ΔT(β + Δβ) − ΔT(β) RRRR
∣Sβ ∣ = RRRR( ) RRR, (3)
RRR β ΔT(β) RRR
R
where ∆β/β was set to be 0.1 referring to the previous work.21

Measurement setup and analysis procedure

For the measurement, the metal wire on the 3ω device was con-
nected in series with a potentiometer. A schematic of the setup is
shown in Fig. 1(e). The potentiometer is adjusted to have the same
electrical resistance with that of the wire (R). The temperature coef-
ficient of the resistance (α) of the wire was obtained in advance by
FIG. 1. Schematic image of (a) the conventional 3ω method and (b) bidirectional measuring the R from 20 ○ C to 40 ○ C with the increment of 5 ○ C.
3ω method in this work. (c) Picture of the measurement device. (d) Enlarged image
All the measurements were conducted at room temperature (25 ○ C).
of the metal wire. (e) Schematic image of the measurement setup.
Wave current with a frequency ω (10–1500 Hz) was applied, and

Rev. Sci. Instrum. 92, 034902 (2021); doi: 10.1063/5.0030982 92, 034902-2
Published under license by AIP Publishing
 Review of ARTICLE scitation.org/journal/rsi
Scientific Instruments

the third harmonic voltage across the wire (V 3ω ) was obtained RESULTS AND DISCUSSION
by subtracting the first harmonic voltage across the potentiometer
Before measuring the organic thin films, the thermal conduc-
(V 1ω ) from that across the device (V 1ω+3ω ). The voltage was mea-
tivity of the 100-nm-thick insulation layer κAl2 O3 was determined.
sured by using DAQ NI-9239 (National Instruments). The temper-
Figure 2(a) shows the measurement sensitivities to κAl2 O3 with wide
ature rise (∆T) at the wire was calculated by
(50 μm) and narrow (300 nm) wires, which were calculated assum-
2 V3ω ing that κAl2 O3 is the reported value27 of 2 W m−1 K−1 and thermal
ΔT = , (4)
α V1ω anisotropy ηAl2 O3 is one. Sensitivities to κSiO2 were also shown as a
reference. While the 50-μm-wide wire has sensitivity only to κSiO2
and the current amplitude (I 1ω ) was controlled so that the highest
and not to κAl2 O3 in the whole frequency range, the 300-nm-wide
∆T is around 2 K. The target parameters were obtained by fitting the
wire gives the sensitivity to κAl2 O3 as high as 0.2. Therefore, the
theoretical ∆T [Eq. (1)] with the experimental ones [Eq. (4)] with a
300-nm-wide wire was used for the measurement of κAl2 O3 . The
least squares method.
reason for this sensitivity difference with wide- and narrow-wires
The measurement was first conducted with bare device, and the
is discussed in the supplementary material. As input parameters,
κ of insulation layer (Al2 O3 ) was determined for differential mea-
measured values were used for thickness and specific heat capacity,
surement. Next, measurements with (i) the film and (ii) the film with
while reference values were used for density. The actual values of the
Al deposited on the top were conducted to get ∆T. Finally, κ– and η
parameters are summarized in Table S1. The theoretical curve cal-
of the film were determined in order by fitting the data of (ii) and (i),
culated by Eq. (1) was fitted with the experimental values with a least
respectively.
squares method by altering κAl2 O3 as a fitting parameter. Figure 2(b)
shows the experimental temperature rise and best fitted curve for
Sources of error
the 300-nm-wide wire. The theoretical curve shows good agree-
In this study, measurements on a target material were con- ment with the experimental data, and the fitting gives κAl2 O3 of 1.50
ducted three times with different substrates, and a standard devia- ± 0.72 W m−1 K−1 (N = 6), which is consistent with the reported
tion of the obtained values is denoted as the error. Possible sources value.27 The relatively large error may be due to the uncertainty in
of the deviation are (1) the fluctuation of the measured signal and the device configuration, such as the wire width and Al2 O3 thickness,
(2) the geometric uncertainty of the measurement setup. As for (1), which deviate depending on the device position on the wafer. This
in our system, DAQ functioning as a lock in amplifier can detect uncertainty is eliminated in the measurement of the organic films by
the V 3ω signal with a fluctuation as low as 0.2%, which is negligi- the differential technique as described later.
ble. Regarding (2), the uncertainties of the wire width and the layer Next, anisotropic κ of the cellulose nanofiber (CNF) oriented
thickness influence the fitting results. As will be discussed later, dif- film was measured to validate the system. The oriented CNF film was
ferential measurements were conducted when obtaining η to elimi- selected as a model sample because its anisotropic κ has been already
nate the uncertainty of Al2 O3 layer thickness, but those of the wire reported,28 and in addition, it can be formed either into a film on
width and the sample thickness remain. Their impact on the tar- a substrate or into a self-standing film, which makes it a suitable
get properties can be estimated by inputting the geometries altered sample for validation that can be measured by both bidirectional 3ω
by 5% into the physical model to simulate the signal and then fit- method and other existing methods. The CNF film thickness mea-
ting it with the physical model, assuming unaltered geometries. For sured by using the micrometer was 26 μm. To calculate the measure-
example, with the PMMA film case, the deviation was estimated to ment sensitivities, reported values of CNF cross-plane thermal con-
be 8% in the cross-plane thermal conductivity and 29% in the ther- ductivity (κCNF,– ) and its thermal anisotropy (ηCNF ) were used.28 For
mal anisotropy, which reasonably match with the above standard thermal boundary resistance (TBR) values at all organic/inorganic
deviation obtained from multiple experiments. interfaces, a reported value of 3 × 10−8 m2 K W−1 for the interface

FIG. 2. (a) Sensitivities to κAl2 O3 and κSiO2 when 300-nm- and 50-μm-wide wires are used. (b) Temperature rise and the fitting curve for the measurement of κAl2 O3 with the
300-nm-wide wire.

Rev. Sci. Instrum. 92, 034902 (2021); doi: 10.1063/5.0030982 92, 034902-3
Published under license by AIP Publishing
 Review of ARTICLE scitation.org/journal/rsi
Scientific Instruments

between silicon (Si) and spin-coated organic polymer measured by Al deposition. To get those parameters from an identical sample, the
the TDTR method29 was employed. The absolute value of TBR is not measurements of the temperature rises were conducted in reverse
likely to affect the results, as the measurement sensitivity to TBR is order: first without Al case and second with Al case.
less than 0.003 (ee Fig. S1). Sensitivities to κCNF,– and ηCNF with the κCNF,– and ηCNF were obtained by the process described above.
300-nm-wide wires are shown in Fig. 3(a). The sensitivities to κCNF,– In addition to the thickness, specific heat capacity, and density
and ηCNF are as high as 0.1–0.2, and another independent measure- (actual values summarized in Table S1), the reference values28 of
ment is required to suppress uncertainty by determining either one κCNF,– and ηCNF were used as initial input parameters, while the mea-
of the parameters. sured value of κCNF,– was used when obtaining ηCNF . A least squares
In conventional 3ω measurement, the films are sandwiched by method was used for fitting by altering κCNF,– or ηCNF as a fitting
a metal wire and heat conducting substrate such as Si and the gen- parameter in each case. Figures 3(c) and 3(d) show the measured
erated heat passes through the film to the substrate [Fig. 1(a)]. In temperature rise and its fitting curve of each case for CNF mea-
analogy to the conventional system, 500 nm of heat conducting alu- surement. κCNF,– was determined to be 0.28 ± 0.02 W m−1 K−1 (N
minum (Al) was deposited on the CNF film with resistive thermal = 3) from the measurement with the 50-μm-wide wire [Fig. 3(d)].
evaporation so that the heat can conduct in the vertical direction A differential technique was used for the measurement of ηCNF ;
[Fig. 1(b)]. The temperature rise of the sample during deposition κAl2 O3 was measured beforehand and the value was used in the
is less than several tens of degrees as the melting point of Al is measurement of ηCNF . This method can eliminate the uncertainty
low30 and the heating effect to the sample is negligible. Al was also resulting from the device configuration as it is included in κAl2 O3 .
deposited on a bare device simultaneously to measure κ of Al (κAl ) in Figure 3(c) shows the temperature rise of the wire with and with-
advance (the representative temperature rise and its fitting curve to out CNF, where the difference results from the increased heat con-
extract κAl is shown in Fig. S2). Figure 3(b) shows sensitivity to each duction to CNF. The measured ηCNF was 3.46 ± 0.15 (N = 3), and
parameter when the 50-μm-wide wire is used to measure CNF with correspondingly, the κ∥ of CNF (κCNF,∥ ) was 0.95 ± 0.04 W m−1 K−1 .
the Al deposition. In contrast to the previous 300-nm case, the sensi- To validate these values, κCNF,– and κCNF,∥ of the self-standing CNF
tivity to ηCNF is less than 25% of that to κCNF,– , which means that the film were also measured by laser flash analysis (LFA)31,32 and T-type
measurement is far sensitive to κCNF,– . Therefore, κCNF,– was first method,33–35 respectively. The details of each method are described
determined with the 50-μm-wide wire with the Al deposition, and in the supplementary material. κCNF,– obtained from LFA was 0.30
then, ηCNF was determined with the 300-nm-wide wire without the ± 0.03 W m−1 K−1 , and κCNF,∥ obtained from T-type method was

FIG. 3. Sensitivities to κCNF,– and ηCNF when (a) the 300-nm-wide wire is used without Al deposition and (b) the 50-μm-wide wire is used with Al deposition. Representative
temperature rise and its fitting curve for the measurement of κCNF,– and ηCNF with (c) the 300-nm-wide wire and (d) the 50-μm-wide wire. Control measurement without
CNF is also shown in (c).

Rev. Sci. Instrum. 92, 034902 (2021); doi: 10.1063/5.0030982 92, 034902-4
Published under license by AIP Publishing
 Review of ARTICLE scitation.org/journal/rsi
Scientific Instruments

1.09 ± 0.07 W m−1 K−1 . These values agree well with the ones and therefore, the 70-nm-thin PEDOT:PSS film is suggested to have
obtained from the bidirectional 3ω method, which proves the reli- high κ∥ because of its highly ordered structure.
ability of this novel method. To examine the range of the sample properties and geometries
Thermal conductivities of spin-casted polymer films were also that the current method is applicable to, sensitivities to κ– and κ∥
measured to reveal the anisotropic thermal properties of submicron- of model sample with various thicknesses (d) were calculated. The
thin films. PMMA and PEDOT:PSS were selected as representa- density and specific heat capacity of the model sample were set to
tive materials, and their films were spin-coated on the measure- 1000 J/kg⋅K and 1000 kg/m3 , respectively, as representative values of
ment device. The thicknesses of PMMA and PEDOT:PSS films organic materials. The sensitivity of 0.02 was taken as lower bound,
were 308 nm and 70 nm (measured by interferometry and AFM, considering that κPEDOT:PSS,∥ can be determined with 30% uncer-
respectively). Figure S3 shows measurement sensitivities calculated tainty when the sensitivity is 0.02 (Fig. S4). First, the sensitivity to κ–
with the reported cross-plane thermal conductivity and thermal of model sample was calculated when the 50-μm-wide wire was used
anisotropy of PMMA17 and PEDOT:PSS.36 Same as in the CNF with 1000 Hz alternating current and 500-nm-thick Al is deposited
case, the sensitivities to both κ– and η are large when the narrow on the sample. An isotropic model sample was assumed in this anal-
wire was used, whereas sensitivities to η are negligible when a wide ysis to show the lowest bound, while the actual sample is expected to
wire is used with Al deposition. Therefore, the same process was have thermal anisotropy. The contour of 0.02 in Fig. 4(a) indicates
conducted to obtain κ– and η of each material (input parameters the lower bound of the measurement, which corresponds to a ther-
summarized in Table S1). The measured temperature rise and fitting mal resistance (=d/κ– ) of ∼10−7 m2 K W−1 . The measurable smallest
curves are shown in Fig. S4, and obtained κ– , η, and κ∥ are shown in film thickness increases from 20 nm to 800 nm as κ– increases from
Table II together with the reference values. While κ– of PMMA and 0.1 W m−1 K−1 to 10 W m−1 K−1 . This lower bound of thickness can
PEDOT:PSS are close to the reported ones, their κ∥ are much differ- become high if the sample has a thermal anisotropy. Next, the sensi-
ent from the reference values. The lower value of the measured κ∥ of tivity to κ∥ of a model sample was estimated when the 300-nm-wide
PMMA could be due to the difference in the degree of polymer align- wire was used with 1000 Hz alternating current without Al deposi-
ment. The PMMA film in this study has a larger thickness (308 nm) tion. Here, κ– of the sample is assumed to be 0.1 W m−1 K−1 , and
than the reported one (170 nm), presumably because of the lower the thermal anisotropy was changed from 1 to 100. The contour
rotation speed in the spin-coat process, which might have led to the of 0.02 in Fig. 4(b) indicates the lower bound of the measurement,
lower degree of alignment and hence the lower thermal conductiv- which corresponds to a thermal conductance (=d × κ∥ ) of ∼10−7 W
ity.37,38 On the other hand, κ∥ of PEDOT:PSS is three to ten times m−2 K−1 . The measurable lowest κ∥ increases from 0.2 W m−1 K−1
higher than the reported values obtained from thicker films. This to 6 W m−1 K−1 as the film thickness decreases from 1 μm to 10 nm.
can also be explained by the difference of molecular ordering as the These bounds can be broadened when a wire with narrower width
thicker films were formed by drop-casting or repeated spin-coatings, such as 50 nm is used.

TABLE II. Anisotropic thermal properties of spin-casted polymer films (N = 3).

κ– (W m−1 K−1 ) κ∥ (W m−1 K−1 ) η(= κ∥ /κ– )

This study 0.22 ± 0.01 0.72 ± 0.18 3.25 ± 0.83
PMMA
Reference 0.20–0.25 ( 17,29,39–41 ) 2.33 ± 1.55 ( 17 ) ...
This study 0.29 ± 0.11 2.94 ± 0.81 10.1 ± 2.8
PEDOT:PSS
Reference 0.2–0.3 ( 36,42,43 ) 0.3–0.8 ( 20,36,42,43 ) ...

FIG. 4. Contour of the measurement sensitivity to (a) κ– and (b) κ∥ of the model film sample with various thicknesses and κ.

Rev. Sci. Instrum. 92, 034902 (2021); doi: 10.1063/5.0030982 92, 034902-5
Published under license by AIP Publishing
 Review of ARTICLE scitation.org/journal/rsi
Scientific Instruments

9
CONCLUSION A. Weathers, Z. U. Khan, R. Brooke, D. Evans, M. T. Pettes, J. W. Andreasen, X.
Crispin, and L. Shi, Adv. Mater. 27, 2101 (2015).
A novel bidirectional 3ω system with two different width wires 10
P. Goli, H. Ning, X. Li, C. Y. Lu, K. S. Novoselov, and A. A. Balandin, Nano Lett.
to measure the anisotropic thermal conductivity of an organic thin 14, 1497 (2014).
11
film was developed. The metal wires of Ti/Pt with widths of 300 nm T. Borca-Tasciuc, A. R. Kumar, and G. Chen, Rev. Sci. Instrum. 72, 2139 (2001).
12
and 50 μm are patterned on a quartz substrate beforehand, and CNF, T. Tong and A. Majumdar, Rev. Sci. Instrum. 77, 104902 (2006).
13
PMMA, and PEDOT:PSS thin films were prepared on the substrate. Z. Li, S. Tan, E. Bozorg-grayeli, T. Kodama, M. Asheghi, G. Delgado, M. Panzer,
Cross-plane and in-plane thermal conductivity were measured in A. Pokrovsky, D. Wack, and K. E. Goodson, Nano Lett. 12, 3121 (2012).
14
the sample with the 50-μm- and the 300-nm-wide wires, and this J. Lee, Z. Li, J. P. Reifenberg, S. Lee, R. Sinclair, M. Asheghi, and K. E. Goodson,
J. Appl. Phys. 109, 084902 (2011).
selective measurement was enabled by placing an aluminum layer 15
D. G. Cahill, Rev. Sci. Instrum. 61, 802 (1990).
on the samples in the measurement of cross-plane thermal conduc- 16
K. Makoto, L. Yuxuan, J. Shenghong, M. Asuka, K. Shota, S. Takuma, K.
tivity and by suppressing the sensitivity to in-plane thermal conduc- Takashi, and S. Junichiro, ACS Appl. Energy Mater. 2, 7083 (2019).
tivity. The in-plane thermal conductivity of 70-nm-thick film was 17
J. S. Katz, W. Park, M. T. Barako, A. Sood, M. Asheghi, and K. E. Goodson, in
successfully measured, revealing that high molecular ordering in the Solid-State, Actuators, Microsystems Workshop (Transducer Research Foundation,
film can increase the in-plane thermal conductivity of the thin film. 2018), pp. 280–283.
18
Sensitivity calculations for various samples with different thermal J. S. Katz, M. T. Barako, W. Park, A. Sood, M. Asheghi, and K. E. Goodson,
conductivities and thicknesses clarified the applicable ranges, which in 2018 17th IEEE Intersociety Conference on Thermal and Thermomechanical
suggest the possibility for further improvement by narrowing the Phenomena in Electronic Systems (IEEE, 2018), pp. 477–481.
19
wire width. H. Ushirokita and H. Tada, Chem. Lett. 45, 735 (2016).
20
V. Linseis, F. Völklein, H. Reith, K. Nielsch, and P. Woias, Rev. Sci. Instrum. 89,
015110 (2018).
21
SUPPLEMENTARY MATERIAL M. L. Bauer and P. M. Norris, Rev. Sci. Instrum. 85, 064903 (2014).
22
S. Roy-Panzer, T. Kodama, S. Lingamneni, M. A. Panzer, M. Asheghi, and K. E.
See the supplementary material for additional information Goodson, Rev. Sci. Instrum. 86, 024901 (2015).
23
regarding the film preparation, schematic image and solution to the S. D. Lubner, J. Choi, G. Wehmeyer, B. Waag, V. Mishra, H. Natesan, J. C.
bidirectional heat transfer equation, material parameters, and signal Bischof, and C. Dames, Rev. Sci. Instrum. 86, 014905 (2015).
24
with a fitting curve for 3ω measurement. M. T. Barako, A. Sood, C. Zhang, J. Wang, T. Kodama, M. Asheghi, X. Zheng,
P. V. Braun, and K. E. Goodson, Nano Lett. 16, 2754 (2016).
25
M. T. Barako, S. Roy-Panzer, T. S. English, T. Kodama, M. Asheghi, T. W.
Kenny, and K. E. Goodson, ACS Appl. Mater. Interfaces 7, 19251 (2015).
ACKNOWLEDGMENTS 26
A. Feldman, High Temp. - High Pressures 31, 293 (1999).
27
This work was supported, in part, by JSPS KAKENHI (Grant M. E. DeCostera, K. E. Meyerb, B. D. Piercyc, J. T. Gaskinsa, B. F. Donovand,
Nos. JP18J21726 and 19H00744) and JST CREST (Grant Nos. A. Giria, F. Nicholas, A. Strnade, D. M. Potrepkag, A. A. Wilsonh, M. D. Losegoc,
and P. E. Hopkinsa, Thin Solid Films 650, 71 (2018).
JPMJCR20Q3 and JPMJCR19I2). This work was conducted at the 28
K. Uetani, T. Okada, and H. T. Oyama, Biomacromolecules 16, 2220 (2015).
Takeda Sentanchi Supercleanroom, the University of Tokyo, sup- 29
M. D. Losego, L. Moh, K. A. Arpin, D. G. Cahill, and P. V. Braun, Appl. Phys.
ported by the “Nanotechnology Platform Program” of the Ministry Lett. 97, 011908 (2010).
of Education, Culture, Sports, Science and Technology (MEXT), 30
M. Kosaka, J. Met. Finish. Soc. Jpn. 29, 344 (1978).
Japan (Grant No. JPMXP09F20UT0043). 31
W. J. Parker, R. J. Jenkins, C. P. Butler, and G. L. Abbott, J. Appl. Phys. 32, 1679
(1961).
32
D. Zhao, X. Qian, X. Gu, S. A. Jajja, and R. Yang, J. Electron. Packag. 138, 040802
DATA AVAILABILITY (2016).
33
The data that support the findings of this study are available C. Dames, S. Chen, C. T. Harris, J. Y. Huang, Z. F. Ren, M. S. Dresselhaus, and
G. Chen, Rev. Sci. Instrum. 78, 104903 (2007).
from the corresponding author upon reasonable request. 34
M. Fujii, X. Zhang, H. Xie, H. Ago, K. Takahashi, T. Ikuta, H. Abe, and T.
Shimizu, Phys. Rev. Lett. 95, 065502 (2005).
35
S. Yamaguchi, I. Tsunekawa, N. Komatsu, W. Gao, T. Shiga, T. Kodama, J.
REFERENCES
Kono, and J. Shiomi, Appl. Phys. Lett. 115, 223104 (2019).
1 36
C. Wang, H. Dong, L. Jiang, and W. Hu, Chem. Soc. Rev. 47, 422 (2018). G.-H. Kim, L. Shao, K. Zhang, and K. P. Pipe, Nat. Mater. 12, 719 (2013).
2 37
H. Jiang and W. Hu, Angew. Chem., Int. Ed. 59, 1408 (2020). B. D. Washo and D. Hansen, J. Appl. Phys. 40, 2423 (1969).
3 38
Y. Zhang, J. Qiao, S. Gao, F. Hu, D. He, B. Wu, Z. Yang, B. Xu, Y. Li, Y. Shi, W. X. T. Hao, T. Hosokai, N. Mitsuo, S. Kera, K. K. Okudaira, K. Mase, and N.
Ji, P. Wang, X. Wang, M. Xiao, H. Xu, J. B. Xu, and X. Wang, Phys. Rev. Lett. 116, Ueno, J. Phys. Chem. B 111, 10365 (2007).
016602 (2016). 39
X. Xie, D. Li, T.-H. Tsai, J. Liu, P. V. Braun, and D. G. Cahill, Macromolecules
4
A. Yamamura, S. Watanabe, M. Uno, M. Mitani, C. Mitsui, J. Tsurumi, 49, 972 (2016).
N. Isahaya, Y. Kanaoka, T. Okamoto, and J. Takeya, Sci. Adv. 4, eaao5758 40
X. Xie, K. Yang, D. Li, T. H. Tsai, J. Shin, P. V. Braun, and D. G. Cahill, Phys.
(2018). Rev. B 95, 035406 (2017).
5 41
P. Jiang, X. Qian, and R. Yang, Rev. Sci. Instrum. 88, 074901 (2017). X. Wang, C. D. Liman, N. D. Treat, M. L. Chabinyc, and D. G. Cahill, Phys. Rev.
6
A. J. Schmidt, X. Chen, and G. Chen, Rev. Sci. Instrum. 79, 114902 B: Condens. Matter Mater. Phys. 88, 075310 (2013).
(2008). 42
J. Liu, X. Wang, D. Li, N. E. Coates, R. A. Segalman, and D. G. Cahill,
7
J. P. Feser and D. G. Cahill, Rev. Sci. Instrum. 83, 104901 (2012). Macromolecules 48, 585 (2015).
8 43
I. Jo, M. T. Pettes, J. Kim, K. Watanabe, T. Taniguchi, Z. Yao, and L. Shi, Nano Q. Wei, C. Uehara, M. Mukaida, K. Kirihara, and T. Ishida, AIP Adv. 6, 045315
Lett. 13, 550 (2013). (2016).

Rev. Sci. Instrum. 92, 034902 (2021); doi: 10.1063/5.0030982 92, 034902-6
Published under license by AIP Publishing