close

Вход

Забыли?

вход по аккаунту

?

1.4931062

код для вставкиСкачать
Hydrogenation effects on carrier transport in boron-doped ultrananocrystalline
diamond/amorphous carbon films prepared by coaxial arc plasma deposition
Yūki Katamune, Satoshi Takeichi, Shinya Ohmagari, and Tsuyoshi Yoshitake
Citation: Journal of Vacuum Science & Technology A: Vacuum, Surfaces, and Films 33, 061514 (2015);
View online: https://doi.org/10.1116/1.4931062
View Table of Contents: http://avs.scitation.org/toc/jva/33/6
Published by the American Vacuum Society
Articles you may be interested in
Bonding structure in nitrogen doped ultrananocrystalline diamond
Journal of Applied Physics 93, 5606 (2003); 10.1063/1.1564880
Hydrogenation effects on carrier transport in boron-doped
ultrananocrystalline diamond/amorphous carbon films
prepared by coaxial arc plasma deposition
ki Katamunea) and Satoshi Takeichi
Yu
Department of Applied Science for Electronics and Materials, Kyushu University, 6-1 Kasuga,
Fukuoka 816-8580, Japan
Shinya Ohmagari
Diamond Research Group, Research Institute for Ubiquitous Energy Devices (UBIQEN), National Institute of
Advanced Industrial Science and Technology (AIST), 1-8-31 Midorigaoka, Ikeda, Osaka 563-8577, Japan
Tsuyoshi Yoshitakeb)
Department of Applied Science for Electronics and Materials, Kyushu University, 6-1 Kasuga,
Fukuoka 816-8580, Japan; Research Center for Synchrotron Light Applications, Kyushu University,
6-1 Kasuga 816-8580, Japan; and Research and Education Center for Advanced Energy, Materials,
Devices, and Systems, Kyushu University, 6-1 Kasuga 816-8580, Japan
(Received 31 March 2015; accepted 3 September 2015; published 22 September 2015)
Boron-doped ultrananocrystalline diamond/hydrogenated amorphous carbon composite (UNCD/
a-C:H) films were deposited by coaxial arc plasma deposition with a boron-blended graphite target
at a base pressure of <103 Pa and at hydrogen pressures of 53.3 Pa. The hydrogenation effects
on the electrical properties of the films were investigated in terms of chemical bonding. Hydrogenscattering spectrometry showed that the maximum hydrogen content was 35 at. % for the film
produced at 53.3-Pa hydrogen pressure. The Fourier-transform infrared spectra showed strong
absorptions by sp3 C–H bonds, which were specific to the UNCD/a-C:H, and can be attributed to
hydrogen atoms terminating the dangling bonds at ultrananocrystalline diamond grain boundaries.
Temperature-dependence of the electrical conductivity showed that the films changed from semimetallic to semiconducting with increasing hydrogen pressure, i.e., with enhanced hydrogenation,
probably due to hydrogenation suppressing the formation of graphitic bonds, which are a source of
carriers. Carrier transport in semiconducting hydrogenated films can be explained by a variablerange hopping model. The rectifying action of heterojunctions comprising the hydrogenated films
C 2015 American Vacuum Society.
and n-type Si substrates implies carrier transport in tunneling. V
[http://dx.doi.org/10.1116/1.4931062]
I. INTRODUCTION
Nanodiamonds like ultrananocrystalline diamond/hydrogenated amorphous carbon composite (UNCD/a-C:H) films
and nanocrystalline diamond films are candidate materials
for use in microelectromechanical systems,1 field emission
devices,2 photodetectors,3 and biosensing.4,5 The physical
features of such nanodiamonds gradually deviate from those
of single-crystalline and polycrystalline diamonds with
decreasing diamond grain size, because purity of the sp3
bonds is gradually decreased. The hardness and Young’s
modulus values of the nanodiamonds decrease with increasing fraction of graphitic bonds. On the other hand, nanodiamond films have higher temperature stability than sp3-rich
a-C:H films, known as diamond-like carbon films. The
chemical inertness is high and the electrochemical potential
window is extremely wide for the nanodiamonds, similar to
single-crystalline and polycrystalline diamonds.6
UNCD/a-C:H, which comprises a large number of nanosized diamond grains and an a-C:H matrix, has the following
specific features arising from the large number of grain
boundaries (GBs): (1) large optical absorption coefficients in
a)
Electronic mail: yuki_katamune@kyudai.jp
Electronic mail: tsuyoshi_yoshitake@kyudai.jp
b)
061514-1 J. Vac. Sci. Technol. A 33(6), Nov/Dec 2015
the photon energy range of 3–6 eV (Refs. 7 and 8) and (2)
the potential for n-type and p-type conductions by doping
with nitrogen (N) and boron (B) atoms,9,10 respectively.
Here, GBs are interfaces among ultrananocrystalline diamond (UNCD) grains and between UNCD grains and the
a-C:H matrix.
The large optical absorption coefficients of nanodiamonds
are advantageous to their application as photovoltaic materials. Heterojunction diodes consisting of n-type Si substrates
and B-doped UNCD/a-C:H films, which are prepared by physical vapor depositions such as pulsed laser deposition (PLD)
and coaxial arc plasma deposition (CAPD), exhibit typical
rectifying actions, and photodetection of deep ultraviolet
light.3,11 In addition, it has been reported that carrier transport
might predominantly occur through GBs and a-C:H.12
For B-doped UNCD/a-C:H films prepared using PLD and
CAPD, doping with B produces p-type conduction, probably
due to the replacement of H atoms terminating the dangling
bonds at GBs by B atoms.9 It is clearly different from the
production of p-type conduction in B-doped crystalline diamonds, in which an acceptor energy level of 0.37 eV is
formed above the top of the valence band of diamond as a
result of B atoms being incorporated into the diamond lattices by substitution.13
0734-2101/2015/33(6)/061514/5/$30.00
C 2015 American Vacuum Society
V
061514-1
061514-2 Katamune et al.: Hydrogenation effects on carrier transport
B-doping effects have been studied for films prepared by
chemical vapor deposition (CVD), the main method used for
the preparation of nanodiamond films. For B-doped nanocrystalline diamond films, p-type conduction is produced by
the incorporation of B atoms into diamond lattices by substitution, and GBs do not have dominant roles in the carrier
transport. In cases of heavy doping, hopping conduction
between acceptor states occurs at low temperatures, similarly
to heavily doped single-crystalline, polycrystalline, and
microcrystalline diamonds.14,15 While GBs containing sp2bonding carbon have predominant roles in the carrier transport in undoped UNCD films, there are only few reports on
the semiconducting properties of B-doped UNCD films. For
a-C:H films, although B-doping produces p-type conduction,
it hardly enhances their electrical conductivities.
The electrical properties of UNCD/a-C:H films are
mainly due to GBs and a-C:H with disordered structures.
The electrical conductivities of disordered carbon are
strongly influenced by the ratio between the numbers of sp2
and sp3 bonds, and the H content.16 As yet, there has been
little research on the latter, because the incorporation of H
atoms into nanodiamond films from hydrocarbon source
gases occurs unintentionally during their preparation by
CVD. The hydrogenation effects on the electrical properties
of nanodiamond films have never been studied. The only
report in this regard states that the H content in UNCD films
increases with decreasing diamond grain size and the H
atoms are preferentially located at the grain surfaces and
GBs.17 In addition, it was reported that the H content of
B-doped UNCD films increases with decreasing deposition
temperature in the range <650 C due to the enhanced formation of C–H bonds at those temperatures.18
In contrast with CVD, CAPD does not necessarily require
a H2 atmosphere during deposition for the formation of
UNCD grains.19 The H content of UNCD/a-C:H films can be
tuned by adjusting the H2 pressure during deposition. H2
molecules, which are fed into the CAPD apparatus, are
effectively dissociated into atomic H in a plasma.20 In this
study, B-doped UNCD/a-C:H films were deposited in different H2 atmospheres, and the effects of hydrogenation on the
electrical properties of the UNCD/a-C:H films were studied.
Hydrogenation of the films was evaluated using hydrogen
forward-scattering (HFS) spectrometry and Fouriertransform infrared (FT-IR) spectroscopy. The electrical
properties, in particular, carrier transport, of the films were
explored in terms of temperature-dependence of electrical
conductivity and rectifying action of heterojunctions consisting of B-doped films and n-type Si substrates.
II. EXPERIMENT
B-doped UNCD/a-C:H films were prepared on quartz
(Daiko Manufacturing Co., Ltd.) and n-type Si (Sumco
Corp.) substrates at a substrate temperature of 550 C and
under H2 pressures up to 26.7 Pa by CAPD (Ulvac ARL-300)
with a 5-at. % B-blended graphite target (Toshima
Manufacturing Co., Ltd.). The electrical resistivity and thickness of the n-type Si substrates are 1–5 X cm and 260 lm,
J. Vac. Sci. Technol. A, Vol. 33, No. 6, Nov/Dec 2015
061514-2
respectively. A coaxial arc plasma gun (ARL-300; Ulvac)
with a capacity of 720 lF was operated at a voltage of 100 V.
The repetition rate of pulsed arc discharges was 5 Hz, and the
distance between the substrate and the target was 15 mm.
Heterojunctions consisting of B-doped films and n-type Si
substrates were fabricated by depositing Pd and Al ohmic
electrodes on the film and substrate surfaces, respectively,
using radio-frequency magnetron sputtering.
The B contents of the films were determined by x-ray
photoelectron spectroscopy using a monochromatic light
source from a Mg Ka line with a photon energy of
1253.6 eV. The H contents were quantitatively determined
by HFS spectrometry, and qualitatively evaluated using
FT-IR spectroscopy (FT/IR-4200; JASCO Corporation) at
room temperature. The HFS measurements were outsourced
to the Foundation for the Promotion of Material Science and
Technology of Japan, and the measurements were performed
using a Rutherford backscattering apparatus with a 2.275MeV 4He2þ ion beam under an incident angle of 30 . The
company does not officially announce a software for analyzing HFS spectra. The absolute H content values of the films
were calibrated using standard samples of Hþ-ion-planted Si
and muscovite. Surface treatments such as ion bombardment
etching were not used in the spectroscopic procedures, to
prevent the surface from being damaged. The temperaturedependence of the electrical conductivity was measured in
the temperature range of 120–500 K, using the van der Pauw
method (HL5500PC; Accent). The electrical properties of
the heterojunctions were investigated using current–voltage
(I–V) characteristic measurements in the dark at room
temperature. The positive voltage applied to the B-doped
UNCD/a-C:H film side is defined as the forward-biased
direction.
III. RESULTS AND DISCUSSION
The absolute H contents of the films were determined by
HFS spectroscopy. Measurements were performed on two
undoped UNCD/a-C:H films, prepared at a base pressure of
<103 Pa and under a H2 pressure of 53.3 Pa, respectively.
For the HFS spectra shown in Fig. 1, the H contents of the
nonhydrogenated and hydrogenated films were estimated to
be 2 at. % and 35 at. %, respectively, using simulation software. It is noteworthy that the value of 2 at. % indicates an
error in the estimation because hydrogenation is improbable
in deposition under a base pressure of <103 Pa. This is the
first known report of the absolute H content of a UNCD/
a-C:H film prepared by CAPD.
Chemical bonding structures involving H and C atoms in
the films were identified by FT-IR spectroscopy. The FT-IR
spectra of the films deposited at the base pressure, and at H2
pressures of 6.7 and 26.7 Pa are shown in Figs. 2(a)–2(c),
respectively. IR absorption by CHn bond stretching occurs in
the wavenumber range of 2800–3100 cm1. The spectrum of
the film prepared at the base pressure, i.e., the nonhydrogenated film, showed no peaks. In contrast, the FT-IR spectra of
the hydrogenated films showed absorption bands from CHn
bond stretching, and the total intensity of the absorptions
061514-3 Katamune et al.: Hydrogenation effects on carrier transport
FIG. 1. (Color online) Hydrogen forward-scattering spectra of undoped films
deposited (a) at base pressure of <103 Pa and (b) at H2 pressure of 53.3 Pa.
Squares and solid curve show the experimental and simulated data,
respectively.
increased with increasing H2 pressure. The CHn (n ¼ 1, 2, 3)
stretching band absorptions in the ranges of 2800–2950 and
2970–3100 cm1 are attributed to sp3 and sp2 C–H bonding,
respectively.21 The sp3-CHn/(sp2-CHn þ sp3-CHn) ratio
of the films is approximately estimated using the following
equation:
ð 3000
IðkÞdk
sp3 -CHn
2600
:
(1)
¼ ð 3000
ð 3200
sp2 -CHn þ sp3 -CHn
I ðkÞdk þ
IðkÞdk
2600
3000
061514-3
Here, I(k) is the intensity of the absorption at wavelength k.
Spectral areas due to sp2-CHn and sp3-CHn are values integrated in the ranges of 3000–3200 and 2600–3000 cm1,
respectively. The sp3-CHn/(sp2-CHn þ sp3-CHn) ratios of the
6.7- and 26.7-Pa films were estimated as 0.8 and 0.9. The
spectra indicate that H atoms predominantly form sp3 CHn
bonds, and the sp3-CHn fraction increases with the increasing
H2 pressure. The spectral profiles show that the sp3 C–H
stretching bands at 2905 cm1 are strong, which is typical of
UNCD/a-C:H films and may originate from H atoms that terminate the dangling bonds at GBs.8
The electrical conductivity r of the films is shown in Fig.
3 as a function of the inverse of the temperature T. The electrical conductivity clearly decreases with increasing H2 pressure. It was impossible to measure the electrical conductivity
of the 26.7-Pa film, because it was insulated. The 6.7-Pa film
shows typical semiconducting behavior, and the nonhydrogenated and 1.3-Pa films are semimetallic rather than semiconducting in nature.
The activation energy can be estimated from the
temperature-dependence of electrical conductivity using the
Arrhenius law
r ¼ r0 exp ðEa =kB TÞ:
(2)
Here, Ea, r0, and kB are the activation energy, a preexponential constant that can be extrapolated from experimental data, and the Boltzmann constant, respectively. The
gradient of the plot gradually changes with T1, which
implies that activation energy varies with temperature. The
gradient of the plot in Fig. 3, i.e., the actual activation energy
Eact, is calculated using the following equation, derived from
the transform of Eq. (2):22,23
Eact ¼ kB f½dðlnrÞ=½dð1=TÞg:
(3)
The calculated activation energies clearly vary with temperature (Fig. 4). In addition, the activation energy increases
with increasing extent of hydrogenation.
FIG. 2. (Color online) FT-IR spectra of B-doped films deposited (a) at base
pressure of <103 Pa, and at H2 pressures of (b) 6.7 and (c) 26.7 Pa.
JVST A - Vacuum, Surfaces, and Films
FIG. 3. (Color online) Temperature-dependences of electrical conductivities
of B-doped films deposited at base pressure of <103 Pa, and at H2 pressures of 1.3 and 6.7 Pa.
061514-4 Katamune et al.: Hydrogenation effects on carrier transport
061514-4
FIG. 4. (Color online) Actual activation energies of B-doped films deposited
at base pressure of <103 Pa, and at H2 pressures of 1.3 and 6.7 Pa.
The temperature-dependence of the electrical conductivity in hopping conduction is expressed by the following
equation:24
r ¼ r0 exp ðTm =TÞ1=m :
FIG. 5. Plots of electrical conductivities of films deposited (a) at base pressure of <103 Pa, and at H2 pressures of (b) 1.3 and (c) 6.7 Pa, as a function
of T1/m, where m ¼ 2, 3, or 4.
(4)
Here, r0 and Tm are the pre-exponential constant and a
material-dependent constant, respectively. Figure 5 shows
plots of the electrical conductivity as a function of T1/m. The
value of m depends on the carrier-transport mechanism as follows: m ¼ 1 indicates nearest-neighbor hopping,25 m ¼ 3 and
4 indicate two- and three-dimensional variable-range hopping
(VRH),26 and m ¼ 2 indicates Efros VRH.27,28 As shown in
Fig. 5, the plots for the 6.7-Pa film are linear, particularly
when m ¼ 4. Therefore, carrier transport in the hydrogenated
films probably follows three-dimensional VRH, assuming that
the density of states is constant near a Fermi level. In contrast,
the plots of the semimetallic films are not linear.
The IV characteristics of heterojunctions consisting of
B-doped films and n-type Si substrates are shown in Fig.
6(a). All the heterojunctions show rectifying actions. The
rectifying ratio gradually increases with increasing H2 pressure, mainly because the films change in nature from semimetallic to semiconducting with increasing H2 pressure,
which reduces the carrier concentration owing to hydrogenation of the films. Hydrogenation of B-doped UNCD/a-C:H
FIG. 6. (Color online) (a) IV characteristics of heterojunctions consisting of B-doped films deposited at base pressure of <103 Pa, and H2 pressures of 6.7,
26.7, and 53.3 Pa, and n-type Si substrates, and (b) magnification of I–V curve of heterojunction consisting of 53.3-Pa film and n-type Si substrate in low forward voltage range.
J. Vac. Sci. Technol. A, Vol. 33, No. 6, Nov/Dec 2015
061514-5 Katamune et al.: Hydrogenation effects on carrier transport
films suppresses the formation of graphitic bonds at GBs and
in an a-C:H.29 In the semimetallic films, graphitic bonds act
as a carrier source and the depletion layers in the heterojunctions hardly expand into the side of the UNCD/a-C:H film.
In addition, localized states caused by sp2 carbons might be
formed near the Fermi level and can act as leakage centers,
which may degrade the rectification ratio.
The ideality factor n was estimated for the forward
current, to clarify the carrier-transport mechanism at the
junction. Figure 6(b) shows a magnification of the I–V characteristics of the heterojunction comprising a 53.3-Pa film
and an n-type-Si substrate. The forward current in the low
carrier-injection range, in which the series resistance is negligible, is represented by the following equation:
061514-5
preferentially exit at GBs may play important roles in carrier
transport in UNCD/a-C:H films.
ACKNOWLEDGMENTS
This work was partly supported by the Grant-in-Aid for
Japan Society for the Promotion of Science (JSPS) Fellows
(13J07294), the Grant-in-Aid for Science Research
(24656389), the Advanced Low Carbon Technology
Research and Development Program (ALCA), Japan Science
and Technology (JST), a research grant from the Mazda
Foundation. The first author (Y.K.) was supported by a
research fellowship from the Japan Society for the
Promotion of Science for Young Scientists.
1
I ¼ I0 exp ðqV=nkB TÞ:
(5)
Here, I0 and q represent the saturation current at zero
applied voltage and the carrier electric charge, respectively.
When Eq. (5) is applied to the experimental I–V curves, the
symmetric background current for the bias voltage should be
subtracted from the experimental forward current. In practice,
as shown in Fig. 6(b), the I–V characteristic data in the
reverse range are subtracted from the experimental I–V characteristic data, i.e., the experimental forward current. Ideality
factors of 1 and 2 are derived from carrier transports in diffusion and generation–recombination processes, respectively.30
In heterojunctions consisting of nonhydrogenated and slightly
hydrogenated films, the depletion regions hardly expand into
the film side. Therefore, the ideality factor was estimated for
a heterojunction based on the 53.3-Pa film, and was 2.4,
implying that the carrier-transport mechanism involves a tunneling process in addition to the G–R process.25 This is consistent with the finding that carrier transport in UNCD/a-C:H
films follows the VRH model.
IV. SUMMARY AND CONCLUSIONS
The effect of hydrogenation on the electrical properties of
UNCD/a-C:H was studied in terms of chemical bonding
structures. Growth of UNCD/a-C:H films by CAPD does not
necessarily require H2 atmospheres; therefore, the films were
prepared using a wide range of H2 pressures. The H content
increased with increasing H2 pressure during film deposition,
and reached 35 at. % for films deposited at 53.3 Pa. The H2
atmospheres during the deposition process suppressed the
formation of graphitic bonds, resulting in enhanced electrical
resistivity. Carrier transport in the hydrogenated films followed the VRH model. Carrier transport in tunneling could
occur through GBs, from where H atoms preferentially
exit, and terminate the dangling bonds of UNCD grain surfaces. The electrical properties of UNCD/a-C:H films were
strongly affected by hydrogenation, and H atoms that
JVST A - Vacuum, Surfaces, and Films
A. R. Krauss et al., Diamond Relat. Mater. 10, 1952 (2001).
A. R. Krauss et al., J. Appl. Phys. 89, 2958 (2001).
3
S. Ohmagari and T. Yoshitake, Appl. Phys. Express 5, 065202 (2012).
4
C. Popov, W. Kulisch, M. Jelinek, A. Bock, and J. Strand, Thin Solid
Films 494, 92 (2006).
5
S. Raina, W. P. Kang, and J. L. Davidson, Diamond Relat. Mater. 18, 718
(2009).
6
T. Hara and T. Yoshitake (unpublished).
7
T. Yoshitake, A. Nagano, M. Itakura, N. Kuwano, T. Hara, and K.
Nagayama, Jpn. J. Appl. Phys. 46, L936 (2007).
8
T. Yoshitake, Y. Nakagawa, A. Nagano, R. Ohtani, H. Setoyama, E.
Kobayashi, K. Sumitani, Y. Agawa, and K. Nagayama, Jpn. J. Appl. Phys.
48, 020222 (2009).
9
S. Ohmagari, T. Yoshitake, A. Nagano, R. Ohtani, H. Setoyama, E.
Kobayashi, T. Hara, K. Nagayama, and T. Yoshitake, Jpn. J. Appl. Phys.
49, 031302 (2010).
10
S. Al-Riyami, S. Ohmagari, and T. Yoshitake, Appl. Phys. Express 3,
115102 (2010).
11
Y. Katamune, S. Ohmagari, S. Al-Riyami, S. Takagi, M. Shaban, and T.
Yoshitake, Jpn. J. Appl. Phys. 52, 065801 (2013).
12
S. Ohmagari, T. Hanada, Y. Katamune, S. Ali-Riyami, and T. Yoshitake,
Jpn. J. Appl. Phys. 53, 050307 (2014).
13
E. Rohrer, C. E. Nebel, M. Stutzmann, A. Floter, R. Zachai, X. Jiang, and
C.-P. Klages, Diamond Relat. Mater. 7, 879 (1998).
14
M. Werner et al., Appl. Phys. Lett. 64, 595 (1994).
15
W. Gajewski, P. Achatz, O. A. Williams, K. Haenen, E. Bustarret, M.
Stutzmann, and J. A. Garrido, Phys. Rev. B 79, 045206 (2009).
16
J. Robertson, Mater. Sci. Eng. R 37, 129 (2002).
17
Sh. Michaelson, O. Ternyak, R. Akhvlediani, A. Hoffman, A. Lafosse, R.
Azria, O. A. Williams, and D. M. Gruen, J. Appl. Phys. 102, 113516 (2007).
18
H. J. Zeng et al., Carbon 84, 103 (2015).
19
K. Hanada, T. Yoshitake, T. Nishiyama, and K. Nagayama, Jpn. J. Appl.
Phys. 49, 08JF09 (2010).
20
K. Hanada, T. Nishiyama, T. Yoshitake, and K. Nagayama, Diamond
Relat. Mater. 19, 899 (2010).
21
P. Couderc and Y. Catherine, Thin Solid Films 146, 93 (1987).
22
M. Ko
os, S. H. S. Moustafa, E. Szilagyi, and I. P
ocsik, Diamond Relat.
Mater. 8, 1919 (1999).
23
S. Bhattacharyya and S. R. P. Silva, Thin Solid Films 482, 94 (2005).
24
T. Sato, K. Ohashi, H. Sugai, T. Sumi, K. Haruna, H. Maeta, N.
Matsumoto, and H. Otsuka, Phys. Rev. B 61, 12970 (2000).
25
A. Miller and E. Abrahams, Phys. Rev. 120, 745 (1960).
26
N. F. Mott, J. Non-Cryst. Solids 1, 1 (1968).
27
M. Pollak, Discuss. Faraday Soc. 50, 13 (1970).
28
A. L. Efros and B. I. Shklovskki, J. Phys. C: Solid State Phys. 8, L49 (1975).
29
Y. Katamune, S. Takeichi, S. Ohmagari, H., Setoyama, and T. Voshitake,
Trans. Mater. Res. Soc. Jpn. 40, 243 (2015).
30
K. Mayes, A. Yasan, R. McClintock, D. Shiell, S. R. Darvish, P. Kung,
and M. Razeghi, Appl. Phys. Lett. 84, 1046 (2004).
2
Документ
Категория
Без категории
Просмотров
0
Размер файла
971 Кб
Теги
4931062
1/--страниц
Пожаловаться на содержимое документа