Download High negative differential resistance in silicon quantum dot metal

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

Switched-mode power supply wikipedia , lookup

Memristor wikipedia , lookup

Electrical ballast wikipedia , lookup

Transistor wikipedia , lookup

Voltage optimisation wikipedia , lookup

Rectifier wikipedia , lookup

Stray voltage wikipedia , lookup

Shockley–Queisser limit wikipedia , lookup

Triode wikipedia , lookup

Opto-isolator wikipedia , lookup

Diode wikipedia , lookup

Mains electricity wikipedia , lookup

TRIAC wikipedia , lookup

Current source wikipedia , lookup

Buck converter wikipedia , lookup

Ohm's law wikipedia , lookup

Surge protector wikipedia , lookup

Microelectromechanical systems wikipedia , lookup

Resistive opto-isolator wikipedia , lookup

Metadyne wikipedia , lookup

Rectiverter wikipedia , lookup

Alternating current wikipedia , lookup

Current mirror wikipedia , lookup

Transcript
APPLIED PHYSICS LETTERS 89, 153117 共2006兲
High negative differential resistance in silicon quantum dot
metal-insulator-semiconductor structure
Nae-Man Park,a兲 Sang Hyeob Kim, and Sunglyul Maeng
Cambridge-ETRI Joint R&D Center, Electronics and Telecommunications Research Institute,
Daejeon 305-700, Korea
Seong-Ju Park
Department of Materials Science and Engineering, Gwangju Institute of Science and Technology,
Gwangju 500-712, Korea
共Received 18 May 2006; accepted 23 August 2006; published online 12 October 2006兲
Metal-insulator-semiconductor structures, comprised of silicon quantum dot films grown by
plasma-enhanced chemical vapor deposition, were fabricated on Si wafers. The devices showed a
negative differential resistance, as a result of the resonant tunneling and the very high peak-to-valley
current ratios of 2240 under illumination and 390 in the dark at room temperature, which are much
higher than the corresponding values of other Si tunneling devices. The peak voltage was reduced
down to 1.9 V by increasing the doping concentration of the wafer and reducing the device area.
The structure shows promise for use in solid-state switch applications. © 2006 American Institute
of Physics. 关DOI: 10.1063/1.2360888兴
Through innovations in chemical and material sciences,
Si quantum dots 共QDs兲 have led to a bright future in Si
photonics for optical interconnections. The Si QD is one of
the most interesting research areas in recent years because
of its efficient light emission and quantum confinement
effects.1–3 Therefore, studies in the chemistry, materials science, physics, and optical device development have been
reported.4,5 Nanoelectronics using Si QDs is also a promising
topic. A resonant tunneling diode 共RTD兲 is a representative
example. RTDs are excellent candidates for use as nanodevices in digital circuits because of high switching speed, low
power consumption, and reduced complexity in implementing a given function. A number of attempts to produce RTDs
in a Si system with good current-voltage characteristics at
room temperature have not, however, been encouraging. The
special operating condition needed, such as a cryogenic system, restricts commercial applications of Si RTDs. Several
reports have showed the tunneling properties in Si QD structures and the promising results of Si QD RTD.6–10
In this letter, metal-insulator-semiconductor 共MIS兲 structures comprised of Si QD films were fabricated and the negative differential resistance by carrier tunneling was characterized. A room-temperature peak-to-valley current ratio
共PVCR兲 as high as 2200 and the on/off large voltage difference 共i.e., the voltage difference between the maximum current and the minimum current兲 were observed, which make
the device control easy to switch on and off. Si QD films
were grown on 共100兲 Si wafers with various doping concentrations at 300 ° C in the presence of 5% silane 关10 SCCM兴
共SCCM denotes cubic centimeter per minute at STP兲 diluted
with nitrogen and additional nitrogen 共500 SCCM兲 gases by
conventional plasma-enhanced chemical vapor deposition.11
Si QDs are embedded in a silicon nitride film. The total
pressure and the plasma power were fixed at 0.5 torr and
6 W, respectively. The contact electrodes using nickel metal
were formed on the silicon nitride film and the backside of
a兲
Author to whom correspondence should be addressed; electronic mail:
[email protected]
the Si wafer. The size of the top metal contact was 5
⫻ 5 mm2. A Si QD in silicon nitride film serves as the insulating layer in the MIS structure. The thicknesses of the Si
QD film and the top nickel metal are about 50 and 60 nm,
respectively. A normal silicon nitride film was also fabricated, and its characteristics compared with those of Si QD
film. A normal silicon nitride film is defined as the film that
does not contain Si QDs. The current-voltage characteristics
of the device were measured in a dark condition using HP
4155A, a semiconductor parameter analyzer. A transmission
electron microscopic study clearly showed that the Si QDs
are well organized in the silicon nitride film and the average
size is about 1.9 nm.12 Dot size could be controlled with a
high reproducibility in excess of 95%.
A simple schematic band diagram of a Si QD MIS device is shown in Fig. 1. In reality, QDs are randomly distributed in a film and, as a result, many QDs are involved in one
current path and there are many paths in one device, which
result in the asymmetric barrier widths between quantum
dots and a broad current-voltage peak curve. However, a
multiple QD structure is desirable for a high PVCR 共Refs. 10
FIG. 1. Schematic band diagram of Si QD MIS structure. Electron tunneling
occurs through QDs in the case of n-type Si wafer and the hole tunneling in
the case of p-type Si wafer under a forward bias. The bound states of the Si
QDs are in the range of 2.1– 2.5 eV depending on their sizes.
0003-6951/2006/89共15兲/153117/3/$23.00
89, 153117-1
© 2006 American Institute of Physics
Downloaded 13 Oct 2006 to 203.237.48.62. Redistribution subject to AIP license or copyright, see http://apl.aip.org/apl/copyright.jsp
153117-2
Park et al.
Appl. Phys. Lett. 89, 153117 共2006兲
FIG. 3. 共Color online兲 Current-voltage characteristics of a Si QD MIS device grown on a p-type Si wafer 共1 – 10 ⍀ cm兲. The PVCR is about 22240 at
9.7 V under conditions of illumination with a 20 W lamp at room temperature. In a reverse bias, the negative resistance disappeared when illuminated
due to the field screening effect from electron charging in the Si QDs.
FIG. 2. Current-voltage characteristics of 共a兲 a normal silicon nitride film
and 共b兲 a Si QD MIS device. Both films were grown on a n-type Si wafer
共0.1– 1 ⍀ cm兲. Negative resistance by electron tunneling is observed only in
the Si QD MIS device, in which the PVCR is about 76 at 7.2 V at room
temperature.
and 13兲 and a large difference of on/off voltage. Even in
multiple QD structure, we can consider that QDs are connected in series in one current path. Figure 1 shows three
QDs with asymmetric barriers, which demonstrates their
complex structure. The barrier width appears to be thicker on
the wafer side of the Si QD than on the top nickel metal side
in the samples because a few nanometer thick silicon nitride
film without Si QDs was grown during the initial QD growth
process. The carrier tunneling from the Si wafer through Si
QDs occurs towards the top Ni metal. When the carrier state
of a Si wafer 共conduction band or valence band edge兲 is
resonant to the bound state in the quantum dot, the number
of tunneling carriers reaches a maximum, resulting in a
maximum current peak in the current-voltage curve. Only a
few carriers are able to tunnel into the quantum dot thereafter
and the current drastically decreases with increasing bias
voltage. Therefore, the current behavior shows a negative
resistance.
For a n-type Si wafer 共0.1– 1 ⍀ cm兲, the maximum peak
current occurs at 7.2 V, as shown in Fig. 2共b兲. However, a
small current flow through a normal silicon nitride film in
which a negative differential resistance is not observed can
be detected, as shown in Fig. 2共a兲. Therefore, it can be concluded that the negative resistance is associated with the Si
QD. In the case of a n-type Si wafer, the conduction band
edge of the Si wafer would rise up under a forward bias.
Therefore, electron tunneling occurs. Although a theoretical
investigation of resonant tunneling through multiple QDs
shows the sharper peak current than other RTD structures,13
our results show that the peak current is broad over a wide
range of voltage. This is due to fluctuations in the size of the
dots in the multiple dot states that cause a small shift in the
bound state level of the dots. Nevertheless, the peak-tovalley current ratio of 76 is much larger than the previously
reported value of 4 for a Si system.14 Hole resonant tunneling can also occur, which could be observed in a reverse bias
at −5.5 V, as shown in the inset of Fig. 2共b兲. Hole tunneling
current is observed to be much smaller than electron one due
to the small mobility of hole.
When a p-type Si wafer 共1 – 10 ⍀ cm兲 was used, the current peak was observed at 9.7 V in a forward bias, as shown
in Fig. 3 共black line兲. In this case, the valence band edge
goes down and hole tunneling occurs. Electron tunneling was
clearly observed at −7.7 V for a reverse bias. The negative
resistance could also be due to transient current from the
charging effect,15 in which charged QDs screen the applied
field, preventing a current flow, resulting in a current dip
共negative resistance兲. This is similar to the negative resistance from resonant tunneling. In the case of a charging effect, the same carriers as those charged in QDs, however, do
not flow through the QDs and, as a result, the current dips
more sharply when the carriers are charged in QDs. To further investigate the origin of the negative resistance in our
samples, we charged Si QDs by generating carriers with a
tungsten-halogen lamp 共20 W兲 and measured the photocurrent characteristics. This is shown as a red line in Fig. 3. Si
QDs in silicon nitride films are charged with electrons because of nitrogen dangling bonds at the dot surface.16 Therefore, if the negative resistance is related to the charging effect in our samples, the current dip must be sharper than the
dark case and the PVCR must increas in a reverse bias under
illumination because the electron charging prevents electron
tunneling. However, the result appears to be the opposite.
Therefore, the negative resistance in our samples is related to
resonant tunneling. The absence of a current peak in the
Downloaded 13 Oct 2006 to 203.237.48.62. Redistribution subject to AIP license or copyright, see http://apl.aip.org/apl/copyright.jsp
153117-3
Appl. Phys. Lett. 89, 153117 共2006兲
Park et al.
of magnitude, compared with the sample on a low doped
wafer shown in Fig. 3. A hole tunneling device has the advantage of fabricating a simpler logic for the given function
than electron one. The high PVCR and low peak voltage also
provide the potential for the system to act as a solid-state
switch, but there is room to improve performance even more.
In summary, the results effectively demonstrate the
preparation of a Si-based tunneling device using Si QD films
at room temperature, with a peak-to-valley current ratio of
2240 under the illumination condition. Such functional electronic devices satisfy the required factors needed for miniaturization, multiplicity, and low manufacturing cost in silicon
ultra-large-scale integration technology in the future.
This work was partially supported by the Ministry of
Information and Communication, Republic of Korea under
Project No. 2005-S-605.
FIG. 4. Current-voltage characteristics of a Si QD MIS device grown on a
highly doped p-type Si wafer 共0.01– 0.1 ⍀ cm兲. The peak voltage at which
the maximum current density peak occurred was shifted to a lower bias
voltage of 1.9 V compared to the sample shown in Fig. 3, accomplished by
using a highly doped wafer and reducing diode area. The device shows a
PVCR of 254.
photocurrent curve in a reverse bias is, on the contrary, the
result of the charging effect; the charging of photogenerated
electrons screens the applied bias and prevents resonant tunneling. In a forward bias, the PVCR is increased up to 2240
because of the production of photoinduced hole carriers,
which results in the best value for the whole RTD system.
However, the peak voltage at which the maximum current peak is observed is typically large. To overcome this
problem, we used a highly doped Si wafer and reduced the
device size by 1 / 25. The high doping in the wafer reduces
the depletion width in the wafer and the resulting voltage
drop in the depletion region is small.17 In addition, the small
device area causes the peak current to be shifted to a lower
voltage due to the reduction in series resistance.18 Therefore,
the peak voltage was shifted to the low bias voltage side as a
result of these two factors. Figure 4 shows the currentvoltage characteristics of a Si QD MIS device fabricated on a
highly doped p-type Si wafer 共0.01– 0.1 ⍀ cm兲. As expected,
the current peak shifted downwards 1.9 V, and the PVCR is
about 250. The peak current is also increased about one order
W. L. Wilson, P. F. Szajowski, and L. E. Brus, Science 262, 1242 共1993兲.
L. Pavesi, L. Dal Negro, C. Mazzoleni, G. Franzo, and F. Priolo, Nature
共London兲 408, 440 共2000兲.
3
P. M. Fauchet, Am. Heart J. 8, 26 共2005兲.
4
Light Emission in Silicon: From Physics to Devices, edited by D. J. Lockwood 共Academic, San Diego, 1998兲.
5
Silicon Photonics, edited by L. Pavesi and D. J. Lockwood 共Springer,
Berlin, 2004兲, pp. 1–238.
6
K. Nishiguchi and S. Oda, J. Appl. Phys. 88, 4186 共2000兲.
7
L. Wu, M. Dai, X. Huang, W. Li, and K. Chen, J. Vac. Sci. Technol. B 22,
678 共2004兲.
8
B. V. Kamenev, G. F. Grom, D. J. Lockwood, J. P. McCafrey, B. Laikhtman, and L. Tsybeskov, Phys. Rev. B 69, 235306 共2004兲.
9
X. Y. Chen and W. Z. Shen, Appl. Phys. Lett. 85, 287 共2004兲.
10
Z. Pei, Alex Y. K. Su, H. L. Hwang, and H. L. Hsiao, Appl. Phys. Lett.
86, 63503 共2005兲.
11
N.-M. Park, S. H. Kim, G. Y. Sung, and S.-J. Park, Chem. Vap. Deposition
8, 254 共2002兲.
12
N.-M. Park, C.-J. Choi, T.-Y. Seong, and S.-J. Park, Phys. Rev. Lett. 86,
1355 共2001兲.
13
W. B. Garnett, Phys. Rev. B 44, 3064 共1991兲.
14
V. M. Franks, K. F. Hulme, and J. R. Morgan, Solid-State Electron. 8, 343
共1965兲.
15
D. N. Kouvatsos, V. Ioannou-Sougleridis, and A. G. Nassiopoulou, Appl.
Phys. Lett. 82, 397 共2003兲.
16
N.-M. Park, S.-H. Choi, and S.-J. Park, Appl. Phys. Lett. 81, 1092 共2002兲.
17
Y. Hu and S. P. Stapleton, IEEE J. Quantum Electron. 29, 327 共1993兲.
18
S. L. Rommel, T. E. Dillon, M. W. Dashiell, H. Feng, J. Kolodzey, P. R.
Berger, P. E. Thompson, K. D. Hobart, R. Lake, A. C. Seabaugh, G.
Klimeck, and D. K. Blanks, Appl. Phys. Lett. 73, 2191 共1998兲.
1
2
Downloaded 13 Oct 2006 to 203.237.48.62. Redistribution subject to AIP license or copyright, see http://apl.aip.org/apl/copyright.jsp