high performance vertical tunneling diodes using graphene ...li.mit.edu/s/2d/paper/apl 2014 tuneling...
Post on 05-Apr-2018
227 Views
Preview:
TRANSCRIPT
High performance vertical tunneling diodes using graphene/hexagonal boronnitride/graphene hetero-structureSeung Hwan Lee, Min Sup Choi, Jia Lee, Chang Ho Ra, Xiaochi Liu, Euyheon Hwang, Jun Hee Choi, JianqiangZhong, Wei Chen, and Won Jong Yoo Citation: Applied Physics Letters 104, 053103 (2014); doi: 10.1063/1.4863840 View online: http://dx.doi.org/10.1063/1.4863840 View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/104/5?ver=pdfcov Published by the AIP Publishing Articles you may be interested in Spintronics with graphene-hexagonal boron nitride van der Waals heterostructures Appl. Phys. Lett. 105, 212405 (2014); 10.1063/1.4902814 A cohesive law for interfaces in graphene/hexagonal boron nitride heterostructure J. Appl. Phys. 115, 144308 (2014); 10.1063/1.4870825 Interlayer coupling enhancement in graphene/hexagonal boron nitride heterostructures by intercalated defects orvacancies J. Chem. Phys. 140, 134706 (2014); 10.1063/1.4870097 Tunneling characteristics in chemical vapor deposited graphene–hexagonal boron nitride–graphene junctions Appl. Phys. Lett. 104, 123506 (2014); 10.1063/1.4870073 In-plane and tunneling pressure sensors based on graphene/hexagonal boron nitride heterostructures Appl. Phys. Lett. 99, 133109 (2011); 10.1063/1.3643899
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 18.140.1.248
On: Wed, 14 Jan 2015 23:21:27
High performance vertical tunneling diodes using graphene/hexagonalboron nitride/graphene hetero-structure
Seung Hwan Lee,1,2,a) Min Sup Choi,2,3,a) Jia Lee,1,2 Chang Ho Ra,1,2 Xiaochi Liu,1,2
Euyheon Hwang,1,2 Jun Hee Choi,4 Jianqiang Zhong,5,6 Wei Chen,5,6
and Won Jong Yoo1,2,3,b)
1Samsung-SKKU Graphene Center (SSGC), Sungkyunkwan University, 2066, Seobu-ro, Jangan-gu, Suwon,Gyeonggi-do 440-746, South Korea2Department of Nano Science and Technology, SKKU Advanced Institute of Nano-Technology (SAINT),Sungkyunkwan University, 2066, Seobu-ro, Jangan-gu, Suwon, Gyeonggi-do 440-746, South Korea3Center for Human Interface Nano Technology (HINT), Sungkyunkwan University, 2066, Seobu-ro,Jangan-gu, Suwon, Gyeonggi-do 440-746, South Korea4Frontier Research Laboratory, Samsung Advanced Institute of Technology, Samsung Electronics Co., Ltd.,Yongin, Gyeonggi-do 446-711, South Korea5Department of Physics, National University of Singapore, 2 Science Drive 3, Singapore 1175426Department of Chemistry, National University of Singapore, 3 Science Drive 3, Singapore 117543
(Received 15 November 2013; accepted 18 January 2014; published online 3 February 2014)
A tunneling rectifier prepared from vertically stacked two-dimensional (2D) materials composed of
chemically doped graphene electrodes and hexagonal boron nitride (h-BN) tunneling barrier was
demonstrated. The asymmetric chemical doping to graphene with linear dispersion property
induces rectifying behavior effectively, by facilitating Fowler-Nordheim tunneling at high forward
biases. It results in excellent diode performances of a hetero-structured graphene/h-BN/graphene
tunneling diode, with an asymmetric factor exceeding 1000, a nonlinearity of �40, and a peak
sensitivity of �12 V�1, which are superior to contending metal-insulator-metal diodes, showing
great potential for future flexible and transparent electronic devices. VC 2014 AIP Publishing LLC.
[http://dx.doi.org/10.1063/1.4863840]
Since the initiation of studies of the electrical character-
istics of metal-insulator-metal (MIM) structured TDs, the
research community has recognized that these devices may
potentially be used as rectifiers with asymmetric tunneling
characteristics based on the difference between the work
functions of the two electrodes.1 These characteristics have
been utilized in a variety of applicable devices, including
antenna-coupled infrared detectors,2–4 rectennas for energy
harvesting,5,6 and high frequency mixers.7 The efficiency of
the TD characteristics may be improved by modifying the
layered structure, increasing the levels of asymmetry, nonli-
nearity, and sensitivity, and by reducing the diode resist-
ance.8,9 These characteristics are determined mainly by the
work function of the electrodes and the barrier height
between the insulator and the electrode materials. Hence, an
optimal TD can be fabricated by adjusting these parameters
and choosing proper materials. However, it is difficult to
design and fabricate optimal TDs, because the fabrication
process can be complicated.
The design of a 2D rectifier has been actively pursued
and is of great importance in such a device that provides a
building block for future electronic devices. Compared with
the MIM TDs, it does not show structural difference induc-
ing similar rectifying one-direction current flow. However,
the fabrication of a rectifier from 2D materials, such as gra-
phene (as the electrode) or h-BN (as the insulator) gives rise
to a variety of advantages: The work function of graphene
has been reported to be adjustable through the application of
an electrical field,10 chemical doping,11,12 metal deposi-
tion,13 or plasma treatment.14,15 In addition, as h-BN has a
crystalline structure similar to that of graphene and is ultra-
flat with no dangling bonds, h-BN reduces the resistance of
graphene. As the Fermi level (EF) of graphene is located
quite near the center of the h-BN band gap as high as 6 eV, it
may be used as an effective tunneling barrier.16–18 The
large-scale preparation of high-quality graphene and h-BN
with such advantages has become feasible through the
emerging chemical vapor deposition (CVD) processes; there-
fore, this technique is expected to be applicable to a variety
of flexible and transparent electronic devices and tunneling
diodes in the future.19,20
The hetero-structured graphene/h-BN/graphene tunnel-
ing diode (GBG-TD) was fabricated by transferring CVD
graphene and mechanically stacked thin h-BN layer on a Si
substrate covered by 90 nm SiO2 as reported previously.16
Rectangular graphene patterns were formed (W/L¼ 5/40
lm) using photolithography techniques, followed by oxygen
plasma etching. The work functions of graphene at either
side were adjusted using chemical doping methods with ben-
zyl viologen (BV) and AuCl3.11,12 The circuit and schematic
diagrams of the thus fabricated p-doped top graphene
(p-GrT)/h-BN/n-doped bottom graphene (n-GrB) TD are
shown in Fig. 1(a). Figs. 1(b) and 1(c) show the optical and
atomic force microscopic (AFM) images of a fabricated
GBG-TD. The white, yellow, and blue dotted lines indicate
the GrT, GrB, and h-BN, respectively. The thickness of h-BN
was measured by AFM (Innova Microscope in Veeco) and it
has 6 nm thickness. Electrical measurements were performed
using a semiconductor parameter analyzer (Agilent 4155 C)
a)S. H. Lee and M. S. Choi contributed equally to this work.b)yoowj@skku.edu
0003-6951/2014/104(5)/053103/4/$30.00 VC 2014 AIP Publishing LLC104, 053103-1
APPLIED PHYSICS LETTERS 104, 053103 (2014)
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 18.140.1.248
On: Wed, 14 Jan 2015 23:21:27
under vacuum (�10 mTorr) at room temperature. Raman
spectroscopy (Witec) and ultra-violet photoelectron spectros-
copy (UPS) (VG Esca-lab MK-II) were used to ensure the
quality and changes in the work function of graphene.
Raman spectroscopy was performed to identify the
property changes induced by the BV and AuCl3 doping of
the graphene on the SiO2 wafer. Fig. 2(a) shows the different
shifts in the G and G0 bands depending on the n- and p-type
doping processes. The Raman spectra indicate that the
observed G and G0 bands are up-shifted (7 and 2 cm�1) in
the case of p-Gr using AuCl3, whereas these bands are
down-shifted (4 and 3 cm�1) in the case of n-Gr using BV. It
is known that the position and intensity of the G band, which
indicate the phonon scattering at the C site, depend on the
doping states of the graphene.22 The G band may be shifted
due to phonon renormalization induced by electron and hole
transfers in the doped graphene. In AuCl3 doped graphene,
Au3þ ions collect electrons from graphene and are stabilized
by forming Auo. This reaction results in p-Gr, and the G
band shifts up due to enhanced electron–phonon coupling.
On the other hand, the neutralized BVo ions provide elec-
trons to graphene and are stabilized as BV2þ. This reaction
results in n-Gr, and the G band shifts down due to a reduc-
tion in the electron–phonon coupling. These tendencies were
apparent from the Raman spectra, which were consistent
with the spectra of p- and n-Gr, as reported previously.11,23
UPS measurement was conducted to identify the differ-
ences in the work functions of doped graphene using the
chemical methods as described above. Fig. 2(b) shows the
kinetic energy of the secondary electrons generated from
graphene by the introduced energy from a He light source.
Because the rapid increase in the secondary electrons corre-
sponds to the Fermi-level energy (EF) of graphene in the vac-
uum state, the different doping conditions of graphene could
be derived by calculating the work function (Ug) using the
equation, Ug¼ �hx� jEFE � Esecj, where �hx¼ 21.2 eV (He I
source), Esec is the onset of the secondary emission, and EFE
is the Fermi edge¼ 16.7 eV in this work (sample bias at
�10 V). Esec for pristine graphene was determined to be
10 eV and corresponds to a work function of 4.5 eV, similar
to the values obtained from previous reports.10,11 The
derived work functions from the 50 mM BV- and
AuCl3-doped graphene samples were 3.1 and 5.6 eV, respec-
tively, and a total difference of 2.5 eV was obtained. As men-
tioned previously, the large difference in the work function
of doped graphene improves the efficiency of operation and
the performance of MIM diodes.
Fig. 3(a) shows the changes in the tunneling characteris-
tics measured before (black square) and after (red circle)
applying p-type doping to GrT. The changes in the tunneling
characteristics depending on the temperature (100–300 K)
were verified [Fig. S1].26 Although the tunneling current
decreased slightly as the temperature decreased, the differ-
ence is insignificant in comparison with room temperature.
This alludes to the generation of Fowler-Nordheim (F-N)
tunneling at high bias regime among several field emission
tunneling. The n-type doping of GrB, even prior to applying
p-type doping of GrT, generates a difference between the
work functions of GrT and n-GrB, which gives rise to an
asymmetric tunneling characteristic. As shown in the inset
graph plotted on the logarithmic scale in Fig. 3(a), the asym-
metric characteristics were clearly observed with a current of
�1.2 pA at �7 V and 50 pA at þ 7 V. The above results are
obviously different from the symmetric tunneling character-
istics17,21 observed from the previous hetero-structured tun-
neling devices having the un-doped graphene layers. The
FIG. 1. (a) Circuit and schematic diagrams of a fabricated GBG-TD. The
AuCl3 solution was deposited on the GrT after device fabrication to compare
the doping effect. (b) and (c) show the optical microscopy (scale bar, 15 lm)
and AFM images (scale bar, 1 lm) of a fabricated GBG-TD, respectively.
FIG. 2. (a) Raman spectra of graphene
before and after doping (BV for
n-doping and AuCl3 for p-doping). (b)
UPS spectra around the secondary
electron threshold for the pristine,
n- and p-Gr. The extracted
work-functions of graphene are 3.1,
4.5, and 5.6 eV for BV-doped, pristine,
and AuCl3-doped.
053103-2 Lee et al. Appl. Phys. Lett. 104, 053103 (2014)
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 18.140.1.248
On: Wed, 14 Jan 2015 23:21:27
different values of the tunneling current under forward and
reverse bias conditions indicate the good applicability as a
diode that rectifies Vac to Vdc. The asymmetric rectifying
characteristics are significantly enhanced after p-doping of
GrT via spin-coating with the AuCl3 solution. Note that the
tunneling current changed to �5.2 pA at �7 V and to
11.5 nA at þ7 V after p-doping of GrT, about a 1000-fold
greater value at forward bias. For more precise analysis, the
asymmetry factor, which is a figure of merit for TDs and
expressed as the ratio of Iforward/Ireverse, is calculated as
shown in Fig. 3(b). After p-doping of the top graphene, the
rectifying behavior is improved significantly due to the dif-
ference in the work functions of the doped graphene on ei-
ther side. Thus, our observations agree well with the
previous reports1,2 of the asymmetric tunneling characteris-
tics measured using different types of electrode. This out-
standing asymmetry factor is more than 100 times the values
obtained from other types of electrodes (<7)24 or metal-insu-
lator-insulator-metal (MIIM) structure diodes (�10).9
Hence, the possibilities enabled by this hetero-structured
diode showing effective rectifying characteristics are
apparent.
After applying p-type doping to the GrT layer, the
tunneling current changed insignificantly under a reverse
bias, whereas the tunneling current changed significantly
under a forward bias. In addition, the point of the voltage
at where the F-N tunneling began decreased by as much
as 2 V. A previous study of the h-BN tunneling character-
istics has been reported that the dielectric strength of
h-BN is about 7.94 MV/cm.18 Hence, it can be predicted
that the tunneling current in a 6 nm thick h-BN layer rap-
idly increases at �5 V, and this value is similar to the
observed value in our experiment. In order to elucidate
the effects of chemical dopants on the h-BN, the electrical
characteristics before and after doping were confirmed by
fabricating a tunneling device without a top graphene
layer [Fig. S2].26 The difference between before and after
doping is insufficient to explain the great improvements of
our GBG-TD. Thus, it is thought that the improvements in
the TD arise from the asymmetric doping effects on only
graphene layers.
Figs. 3(c) and 3(d) reveal the nonlinearity and sensitivity
as the figures of merit for a TD which influence directly on
the rectified current and voltage. The fabricated GBG-TD
showed zero bias sensitivities of 2.75 V�1 and 0.32 V�1
before and after GrT doping, respectively, and peak sensitiv-
ities near 2.5 V of 12.5 V�1 and 11.8 V�1, respectively. These
figures are comparable with the zero bias sensitivities (0.00,
0.08, 0.45, and 0.74 V�1 measured for Al/AlOx/M (M¼Al,
Ti, Ni, and Pt))25 in conventional MIM-TD produced by
means of different electrodes and the peak sensitivity of a
MIIM structure diode (5.5 V�1 for Cr/Al2O3-HfO2/Cr).9
Moreover, the nonlinearity values are two or three times
higher than those obtained from conventional MIIM structure
diodes (�10) [Table S1].26
In order to understand the rectifying phenomenon, we
described the energy band diagrams under zero, reverse, and
forward bias conditions as shown in Fig. 4(a). A huge differ-
ence in the work functions of graphene on either side after
doping causes the significant band-bending of h-BN at zero
bias due to the EF alignment of graphene layers. The effec-
tive barrier thickness of this bended band structure of h-BN
can be tuned by applying forward and reverse biases. Under
reverse bias, the band-bending in h-BN is alleviated by an
increase in EF of GrB and the tunneling current remains rela-
tively low due to the thick effective barrier. By contrast, the
forward bias accelerates band-bending in h-BN to produce a
triangular band shape and induces a high tunneling current
by reducing the effective barrier thickness.
Based on the F-N tunneling equation, ln(I/V2)/-1/V, we
plotted the ln(I/V2) versus 1/V curve as shown in Figure 4(b)
to demonstrate linear behavior and extract barrier height. At
high bias regime, the linear characteristic is observed at
<0.24 V�1 which corresponds to >4.2 V. Thus, we can
FIG. 3. Electrical characteristics of the GBG-TD before and after p-doping
of GrT: (a) ID-VG transfer characteristics. (b) Asymmetry factor
(Iforward/Ireverse). (c) Nonlinearity dIdV=
IV
� �. (d) Sensitivity d2I
dV2=dIdV
� �.
FIG. 4. (a) Energy band diagrams under zero, reverse and forward bias con-
ditions. (b) ln(I/V2) vs. I/V curve for verifying F-N tunneling and evaluating
barrier height.
053103-3 Lee et al. Appl. Phys. Lett. 104, 053103 (2014)
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 18.140.1.248
On: Wed, 14 Jan 2015 23:21:27
estimate that the F-N tunneling starts at 4.2 V which is lower
than that without doping process of the previous report (F-N
tunneling starts at �6.6 V for similar thickness of the h-BN
of 6 nm).18 This lowered value is probably attributed to
the lowered tunneling barrier height. Using the slope of
the graph, we could extract the barrier height:
UB¼ �slope�3hq
8pffiffiffiffiffiffi2m�p
�d
� �23. The extracted value (�2.4 eV) is lower
than the previously reported value (�3 eV) and it seems that
this lowered value is caused by chemical doping on graphene
electrodes of both sides. This control of tunneling barrier
height using chemical doping can be applied for optimization
and development of TDs. Furthermore, we observed the in-
significant temperature dependency of tunneling current as
previously mentioned [Fig. S1]. The very little temperature
dependence of tunneling current is understood to support
F-N tunneling, as the other tunneling processes, such as
Schottky emission and Poole-Frenkel (P-F) emission are
strongly dependent on temperature.27
In summary, two typical 2D materials, graphene and h-
BN, were stacked to form a hetero-structured TD. The gra-
phene layers at either side of the h-BN were doped as p- or
n-type using chemical doping methods to control the work-
function and produce a rectifier which shows one-directional
current flow. After chemical doping, the figures of merit for
the TD, such as the asymmetry factor (�1000), nonlinearity
(�40), and sensitivity (2.75 V�1 zero bias sensitivity and
11.8 V�1 peak sensitivity) were enhanced significantly com-
pared with those obtained from conventional TD. These fig-
ures are superior to those of the conventional MIM-TD.
Therefore, this study could contribute to the design of TD
through controlling work-function of graphene and other 2D
materials in the future.
This work is supported by the Basic Science Research
Program through the National Research Foundation of Korea
(NRF): 2009-0083540, 2011-0010274, 2012H1A2A1004044,
and 2013-015516, and by the Global Frontier R&D Program
(2013-073298) on Center for Hybrid Interface Materials
(HIM) funded by the MOSIP, Korea.
1J. G. Simmons, J. Appl. Phys. 34, 2581–2590 (1963).2J. A. Bean, A. Weeks, and G. D. Boreman, IEEE J. Quantum Electron.
47(1), 126–135 (2011).
3J. A. Bean, B. B. Tiwari, G. H. Bernstein, P. Fay, and W. Porod, J. Vac.
Sci. Technol. B 27, 11 (2009).4M. R. Abdel-Rahman, F. J. Gonz�alez, and G. D. Boreman, Electron. Lett.
40, 116 (2004).5S. Grover and G. Moddel, IEEE J. Photovoltaics 1(1), 78–83 (2011).6M. Dagenais, K. Choi, F. Yesilkoy, A. N. Chryssis, and M. C. Peckerar,
Proc. SPIE 7605, 76050E (2010).7C. Fumeaux, W. Herrmann, F. K. Kneub€uhl, and H. Rothuizen, Infrared
Phys. Technol. 39, 123–183 (1998).8I. E. Hashem, N. H. Rafat, and E. A. Soliman, IEEE J. Quantum Electron.
49, 72–79 (2013).9P. Maraghechi, A. Foroughi-Abari, K. Cadien, and A. Y. Elezzabi, Appl.
Phys. Lett. 99, 253503 (2011).10Y.-J. Yu, Y. Zhao, S. Ryu, L. E. Brus, K. S. Kim, and P. Kim, Nano Lett.
9, 3430–3434 (2009).11H.-J. Shin, W. M. Choi, D. Choi, G. H. Han, S.-M. Yoon, H.-K. Park, S.-
W. Kim, Y. W. Jin, S. Y. Lee, J. M. Kim, J.-Y. Choi, and Y. H. Lee,
J. Am. Chem. Soc. 132, 15603–15609 (2010).12Y. Shi, K. K. Kim, A. Reina, M. Hofmann, L.-J. Li, and J. Kong, ACS
Nano 4, 2689–2694 (2010).13G. Giovannetti, P. A. Khomyakov, G. Brocks, V. M. Karpan, J. V. D.
Brink, and P. J. Kelly, Phys. Rev. Lett. 101, 026803 (2008).14J. Wu, L. Xie, Y. Li, H. Wang, Y. Ouyang, J. Guo, and H. Dai, J. Am.
Chem. Soc. 133, 19668–19671 (2011).15T. Kato, L. Jiao, X. Wang, H. Wang, X. Li, L. Zhang, R. Hatakeyama, and
H. Dai, Small 7, 574–577 (2011).16C. R. Dean, A. F. Young, I. Meric, C. Lee, L. Wang, S. Sorgenfrei, K.
Watanabe, T. Taniguchi, P. Kim, K. L. Shepard, and J. Hone, Nat.
Nanotechnol. 5, 722–726 (2010).17L. Britnell, R. V. Gorbachev, R. Jalil, B. D. Belle, F. Schedin, M. I.
Katsnelson, L. Eaves, S. V. Morozov, A. S. Mayorov, N. M. R. Peres, A.
H. Castro Neto, J. Leist, A. K. Geim, L. A. Ponomarenko, and K. S.
Novoselov, Nano Lett. 12, 1707–1710 (2012).18G.-H. Lee, Y.-J. Yu, C. Lee, C. Dean, K. L. Shepard, P. Kim, and J. Hone,
Appl. Phys. Lett. 99, 243114 (2011).19S. Bae, H. Kim, Y. Lee, X. Xu, J.-S. Park, Y. Zheng, J. Balakrishnan, T.
Lei, H. R. Kim, Y. I. Song, Y.-J. Kim, K. S. Kim, B. €Ozyilmaz, J.-H. Ahn,
B. H. Hong, and S. Iijima, Nat. Nanotechnol. 5, 574–578 (2010).20K. H. Lee, H.-J. Shin, J. Lee, I.-Y. Lee, G.-H. Kim, J.-Y. Choi, and S.-W.
Kim, Nano Lett. 12, 714–718 (2012).21L. Britnell, R. V. Gorbachev, R. Jalil, B. D. Belle, F. Schedin, A.
Mishchenko, T. Georgiou, M. I. Katsnelson, L. Eaves, S. V. Morozov, N.
M. R. Peres, J. Leist, A. K. Geim, K. S. Novoselov, and L. A.
Ponomarenko, Science 335, 947–950 (2012).22A. C. Ferrari and D. M. Basko, Nat. Nanotechnol. 8, 235–246 (2013).23K. K. Kim, A. Reina, Y. Shi, H. Park, L.-J. Li, Y. H. Lee, and J. Kong,
Nanotechnology 21, 285205 (2010).24S. Krishnan, E. Stefanakos, and S. Bhansali, Thin Solid Films 516,
2244–2250 (2008).25B. D. Kong, C. Zeng, D. K. Gaskill, K. L. Wang, and K. W. Kim, Appl.
Phys. Lett. 101, 263112 (2012).26See supplementary material at http://dx.doi.org/10.1063/1.4863840 for
electrical characteristics of GBG-TD depending on the low temperature,
chemical doping effects of h-BN layer, and comparison of rectifying char-
acteristics with conventional diodes.27A. Koszewski, F. Souchon, Ch. Dieppedale, D. Bloch, and T. Ouisse,
J. Micromech. Microeng. 23, 045019 (2013).
053103-4 Lee et al. Appl. Phys. Lett. 104, 053103 (2014)
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 18.140.1.248
On: Wed, 14 Jan 2015 23:21:27
top related