Formation of ohmic contacts to n(p)-gan and measurement of their contact resistivity
We propose multilayer ohmic contacts to n- and p-GaN layers, with titanium boride as diffusion barrier. It is shown that the optimal method of contact resistivity measurement is the transmission line method (TLM) with circular contact geometry. The Ti−Al−TiBx−Au contact metallization to n-GaN ret...
Збережено в:
| Опубліковано в: : | Semiconductor Physics Quantum Electronics & Optoelectronics |
|---|---|
| Дата: | 2010 |
| Автори: | , , , , , , , , |
| Формат: | Стаття |
| Мова: | English |
| Опубліковано: |
Інститут фізики напівпровідників імені В.Є. Лашкарьова НАН України
2010
|
| Онлайн доступ: | https://nasplib.isofts.kiev.ua/handle/123456789/118559 |
| Теги: |
Додати тег
Немає тегів, Будьте першим, хто поставить тег для цього запису!
|
| Назва журналу: | Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| Цитувати: | Formation of ohmic contacts to n(p)-gan and measurement of their contact resistivity / M.S. Boltovets, V.M. Ivanov, R.V. Konakova, Ya.Ya. Kudryk, V.V. Milenin, V.V. Shynkarenko, V.M. Sheremet, Yu.N. Sveshnikov, B.S. Yavich // Semiconductor Physics Quantum Electronics & Optoelectronics. — 2010. — Т. 13, № 4. — С. 337-342. — Бібліогр.: 37 назв. — англ. |
Репозитарії
Digital Library of Periodicals of National Academy of Sciences of Ukraine| id |
nasplib_isofts_kiev_ua-123456789-118559 |
|---|---|
| record_format |
dspace |
| spelling |
Boltovets, M.S. Ivanov, V.M. Konakova, R.V. Kudryk, Ya.Ya. Milenin, V.V. Shynkarenko, V.V. Sheremet, V.M. Sveshnikov, Yu.N. Yavich, B.S. 2017-05-30T16:13:17Z 2017-05-30T16:13:17Z 2010 Formation of ohmic contacts to n(p)-gan and measurement of their contact resistivity / M.S. Boltovets, V.M. Ivanov, R.V. Konakova, Ya.Ya. Kudryk, V.V. Milenin, V.V. Shynkarenko, V.M. Sheremet, Yu.N. Sveshnikov, B.S. Yavich // Semiconductor Physics Quantum Electronics & Optoelectronics. — 2010. — Т. 13, № 4. — С. 337-342. — Бібліогр.: 37 назв. — англ. 1560-8034 PACS 73.40.Cg, 73.40.Ns, 85.30.-z https://nasplib.isofts.kiev.ua/handle/123456789/118559 We propose multilayer ohmic contacts to n- and p-GaN layers, with titanium boride as diffusion barrier. It is shown that the optimal method of contact resistivity measurement is the transmission line method (TLM) with circular contact geometry. The Ti−Al−TiBx−Au contact metallization to n-GaN retains its layer structure after thermal annealing at temperatures up to 900 °C. The contact resistivity ρс is (6.69±1.67)×10⁻⁵ Ω⋅cm². For the Au−TiBx−Ni−p-GaN contact structure, the contact resistivity is (1±0.15)×10⁻³ Ω⋅cm² . This work was supported by the Project No 31/4.2.3.1/33 of the Governmental task scientific and technical program “Development and implementation of energy-saving light sources and illumination systems based on them” (Regulation of the Cabinet of Ministers of Ukraine No 632 from July 9, 2008). The development of varactor diodes was carried out under the INCO−COPERNICUS Program (Project No 977131 “MEMSWAVE”). en Інститут фізики напівпровідників імені В.Є. Лашкарьова НАН України Semiconductor Physics Quantum Electronics & Optoelectronics Formation of ohmic contacts to n(p)-gan and measurement of their contact resistivity Article published earlier |
| institution |
Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| collection |
DSpace DC |
| title |
Formation of ohmic contacts to n(p)-gan and measurement of their contact resistivity |
| spellingShingle |
Formation of ohmic contacts to n(p)-gan and measurement of their contact resistivity Boltovets, M.S. Ivanov, V.M. Konakova, R.V. Kudryk, Ya.Ya. Milenin, V.V. Shynkarenko, V.V. Sheremet, V.M. Sveshnikov, Yu.N. Yavich, B.S. |
| title_short |
Formation of ohmic contacts to n(p)-gan and measurement of their contact resistivity |
| title_full |
Formation of ohmic contacts to n(p)-gan and measurement of their contact resistivity |
| title_fullStr |
Formation of ohmic contacts to n(p)-gan and measurement of their contact resistivity |
| title_full_unstemmed |
Formation of ohmic contacts to n(p)-gan and measurement of their contact resistivity |
| title_sort |
formation of ohmic contacts to n(p)-gan and measurement of their contact resistivity |
| author |
Boltovets, M.S. Ivanov, V.M. Konakova, R.V. Kudryk, Ya.Ya. Milenin, V.V. Shynkarenko, V.V. Sheremet, V.M. Sveshnikov, Yu.N. Yavich, B.S. |
| author_facet |
Boltovets, M.S. Ivanov, V.M. Konakova, R.V. Kudryk, Ya.Ya. Milenin, V.V. Shynkarenko, V.V. Sheremet, V.M. Sveshnikov, Yu.N. Yavich, B.S. |
| publishDate |
2010 |
| language |
English |
| container_title |
Semiconductor Physics Quantum Electronics & Optoelectronics |
| publisher |
Інститут фізики напівпровідників імені В.Є. Лашкарьова НАН України |
| format |
Article |
| description |
We propose multilayer ohmic contacts to n- and p-GaN layers, with titanium
boride as diffusion barrier. It is shown that the optimal method of contact resistivity
measurement is the transmission line method (TLM) with circular contact geometry. The
Ti−Al−TiBx−Au contact metallization to n-GaN retains its layer structure after thermal
annealing at temperatures up to 900 °C. The contact resistivity ρс is (6.69±1.67)×10⁻⁵
Ω⋅cm². For the Au−TiBx−Ni−p-GaN contact structure, the contact resistivity is
(1±0.15)×10⁻³ Ω⋅cm²
.
|
| issn |
1560-8034 |
| url |
https://nasplib.isofts.kiev.ua/handle/123456789/118559 |
| citation_txt |
Formation of ohmic contacts to n(p)-gan and measurement of their contact resistivity / M.S. Boltovets, V.M. Ivanov, R.V. Konakova, Ya.Ya. Kudryk, V.V. Milenin, V.V. Shynkarenko, V.M. Sheremet, Yu.N. Sveshnikov, B.S. Yavich // Semiconductor Physics Quantum Electronics & Optoelectronics. — 2010. — Т. 13, № 4. — С. 337-342. — Бібліогр.: 37 назв. — англ. |
| work_keys_str_mv |
AT boltovetsms formationofohmiccontactstonpganandmeasurementoftheircontactresistivity AT ivanovvm formationofohmiccontactstonpganandmeasurementoftheircontactresistivity AT konakovarv formationofohmiccontactstonpganandmeasurementoftheircontactresistivity AT kudrykyaya formationofohmiccontactstonpganandmeasurementoftheircontactresistivity AT mileninvv formationofohmiccontactstonpganandmeasurementoftheircontactresistivity AT shynkarenkovv formationofohmiccontactstonpganandmeasurementoftheircontactresistivity AT sheremetvm formationofohmiccontactstonpganandmeasurementoftheircontactresistivity AT sveshnikovyun formationofohmiccontactstonpganandmeasurementoftheircontactresistivity AT yavichbs formationofohmiccontactstonpganandmeasurementoftheircontactresistivity |
| first_indexed |
2025-11-26T01:42:45Z |
| last_indexed |
2025-11-26T01:42:45Z |
| _version_ |
1850605164322357248 |
| fulltext |
Semiconductor Physics, Quantum Electronics & Optoelectronics, 2010. V. 13, N 4. P. 337-342.
PACS 73.40.Cg, 73.40.Ns, 85.30.-z
Formation of ohmic contacts to n(p)-gan and measurement
of their contact resistivity
M.S. Boltovets1, V.M. Ivanov1, R.V. Konakova2, Ya.Ya. Kudryk2, V.V. Milenin2, V.V. Shynkarenko2,
V.M. Sheremet2, Yu.N. Sveshnikov3, B.S. Yavich4
1State Enterprise Research Institute “Orion”, 8a Eugene Pottier St., Kyiv 03057, Ukraine,
2V. Lashkaryov Institute of Semiconductor Physics, NAS Ukraine
41 Prospect Nauky, Kyiv 03028, Ukraine
Tel.: (380-44) 525-61-82; Fax: (380-44) 525-83-42; e-mail: konakova@isp.kiev.ua,
3Close Corporation “Elma-Malakhit”, Zelenograd, Russia; e-mail: info@emal.zelcom.ru
4Close Corporation “Svetlana-Optoelektronika”, Sankt-Peterburg, Russia
Abstract. We propose multilayer ohmic contacts to n- and p-GaN layers, with titanium
boride as diffusion barrier. It is shown that the optimal method of contact resistivity
measurement is the transmission line method (TLM) with circular contact geometry. The
Ti−Al−TiBx−Au contact metallization to n-GaN retains its layer structure after thermal
annealing at temperatures up to 900 °C. The contact resistivity ρс is (6.69±1.67)×10-5
Ω⋅cm2. For the Au−TiBx−Ni−p-GaN contact structure, the contact resistivity is
(1±0.15)×10-3 Ω⋅cm2.
Keywords: ohmic contact, gallium nitride, contact resistivity, transmission line method.
Manuscript received 22.07.10; accepted for publication 02.12.10; published online 30.12.10.
1. Introduction
In recent years, the lighting engineering dealing with
production of light-emitting diodes (LEDs) for various
purposes has been developing intensely. Having higher
light efficiency and longer service life, LEDs rather
actively displace other illuminators such as incandescent
and luminescent lamps [1, 2]. The process has become
more evident after appearance of highly efficient LEDs
based on the III group nitrides that made it possible to
develop blue LEDs and, closing the RGB range, obtain
white light using concurrently red, green and blue LEDs.
Such three-color cells are of importance for developing
of LED monitors.
It is known that one cannot realize such properties
of gallium nitride (that is the most advanced material of
the III group nitrides) as high value of avalanche
breakdown field and thermal stability without highly
reliable ohmic contacts. The reason is that application of
LEDs for illumination requires increase of power of the
present-day diodes. In this case, LED heating in both the
base region and contacts will lead to device degradation.
Therefore, search for contact metallization systems to
gallium nitride LEDS that would be high-tech, thermally
stable and compatible with technological processes for
microwave devices makes a topical problem.
Another (no less important) problem in development
of investigations of contact metallization to GaN is
making of low-resistance thermally stable contacts to p-
GaN. The existing experimental data on the resistivity ρс
for contacts to p-GaN indicate wide spread of ρс values
depending on the level of doping with acceptor impurity
and crystallographic orientation of GaN surface as well as
on the work function ϕm of the material of contact-
forming layer. The latter can be formed with
multicomponent compounds with high ϕm value [3-8].
Reduction of resistivity ρс of contact to p-GaN can
be achieved (along with choosing the appropriate value
of φm) by doping the near-surface layer with an acceptor
impurity that provides tunnel current transport in the
ohmic contact. The theoretical studies of ρс for contact to
p-GaN showed [8] that, in the case of a tunnel contact
and with allowance made for tunneling of light and
heavy holes, ρс value depends on the crystallographic
orientation of GaN. The calculated ρс value of about
2×10-4 Ω⋅cm2 corresponds to concentration of
noncompensated acceptors ~1019 cm-3 for GaN [0001]
and 5×1018 cm-3 for GaN [1010], respectively.
© 2010, V. Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine
337
mailto:konakova@isp.kiev.ua
mailto:bms@i.kiev.ua
Semiconductor Physics, Quantum Electronics & Optoelectronics, 2010. V. 13, N 4. P. 337-342.
At the same time, it turned out that, along with the
well-known mechanisms of current transport in ohmic
contacts to n-GaN grown on sapphire, metallic
conductance is observed in such contacts over a wide
temperature range. The reason is high density of
structural defects in GaN. Increase of ρс as temperature
grows (and contributes to device structure heating) is
undesirable not only for LEDs but for other types of
GaN-based electronic devices as well. Thus, formation
of metal−GaN interface with preset parameters as well
as making of buffer layers (diffusion barriers) between
the upper metallization layer and contact-forming layer
that impede mass transfer of contact components are
topical from the viewpoint of production of highly
reliable thermally stable ohmic contacts to n(p)-GaN
epitaxial layers grown on foreign substrates.
Therefore, one should use multilayer contact
metallizations (made of refractory metals and their
compounds) serving as diffusion barriers. In this
connection, borides and nitrides of refractory metals
seem most promising [9-11]. Choice of procedures and
estimation of ρс measurement errors are of great
importance too. In this work, we studied efficiency of
TiBx as diffusion barrier in ohmic contacts to GaN using
the transmission line method (TLM) to measure ρс.
2. Methods of contact resistance measurement
Contact resistivity is one of the main parameters that
characterizes both an ohmic contact and processes of
heat release in it. The principal difficulty in
measurement of contact resistance is its determination
from the measured resistance R. The latter can be
represented as a sum of resistances of the contacts (Rc),
semiconductor (Rs), probes and lead-ins (Rp) and the
resistance provoked by current spreading (RT) [12]:
pTsc RRRRR +++= . (1)
It was shown in [13] that the biggest error in
determination of contact resistivity is characteristic of
the method with vertical geometry of test structure (the
Cox−Strack method) [14-16]. In this method, the
minimal relative error of ρс measurement (at ρс/ρs = 0.01
cm, where ρs is the semiconductor resistivity) is as high
as 90% if one uses lithography with accuracy of 2.5 μm
[13]. Therefore, it is reasonable to apply the Cox−Strack
method [14] only if one has to measure the parameters of
contacts to bulk material and in the case ρс ≥ ρs. The
minimal value of error is ensured by the method of
interface probing [17] and Kelvin method [16, 18, 19].
At ρс/ρs = 0.01 cm, the relative error of ρс measurement
is 25% and 15%, respectively, if one uses lithography
with accuracy of 2.5 μm [13]. When using the probing
method, the lithography should enable one to deposit
sufficiently small external terminals, while application
of the Kelvin method for measurement of low
resistances requires contacts smaller than 1 μm×1 μm
[19].
The most convenient and sufficiently accurate
methods are those TLM versions [20-27] that make it
possible to measure low contact resistances, and the
lithography used meets the requirements to making test
structures for measurements. TLM has many
modifications that differ in geometry and ρс calculation
technique. One should set off the versions of TLM with
linear and circular contact geometry.
The necessary condition for resistance
measurement by TLM with linear geometry of contact
pads is absence of edge surface spreading of current. To
this end, each group of contacts has to be separated from
the rest part of semiconductor plate. This requires
application of additional technological process. Contrary
to this, use of structures with circular geometry of
contact pads makes it possible to avoid edge spreading
of current without additional treatment [25, 26].
In the TLM with circular geometry, the resistance
between contact pads is determined with the following
expression [25]:
)(
)(1
2
ln
2 11
10
11
2
rI
rI
r
R
r
rR
R SHSH
α
α
απ
+
π
= (2)
Here ( ) is the inner (outer) radius of the
corresponding contact pad, and
1r 2r
)( 10 rI α )( 11 rI α the
modified Bessel functions of the zeroth and first order,
respectively, TL1=α the damping coefficient,
scT tL ρρ= transfer length, t thickness of the
semiconductor epitaxial layer.
If and 1rLT >> ( ) Crr =12ln , then the
approximation 11110 2)()( rrIrI α→αα is valid, and
the resistance measured is determined as [25]:
2
12 r
CR
R cSH
π
ρ
+
π
= . (3)
Thus, by plotting the dependence ( )2
1/1 rfR π= ,
one can calculate cρ from the slope of the straight line.
If 1rLT ≤ , then the value obtained in the above way will
give the upper estimate for contact resistivity [25].
One should note that, to determine contact
resistivity accurately, the difference between the outer
and inner radii of contact pad has to be minimal. In that
case, the contribution from semiconductor resistance to
the total resistance becomes smaller [21]. The relative
error of contact resistivity measurement can be
determined from the following expression [13]:
⎟
⎠
⎞
⎜
⎝
⎛ Δ
+
Δ
ρ
ρ
+
Δ
+
Δ
=
ρ
ρΔ
t
t
R
R
t
Cr
R
R
r
l
c
s
c
c
2
2
2
, (4)
where ∆t is inaccuracy of the grown epitaxial layer
thickness, ∆l inaccuracy of lithography when depositing
contacts. The results of calculation of the dependence
f(r) = Δρс/ρс are shown in Fig. 1. At best, if ρc/ρs = 0.01
cm-1, the relative inaccuracy of ρc measurement is 28%.
© 2010, V. Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine
338
Semiconductor Physics, Quantum Electronics & Optoelectronics, 2010. V. 13, N 4. P. 337-342.
0 20 40 60 80 100 120 140 160 180 200
0,0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
0,8
0,9
1,0
4
3
2
Δρ
c/ρ
c
r, μm
1- ρs/ρc=100 cm-1
2- ρs/ρc=10 cm-1
3- ρs/ρc=1 cm-1
4- ρs/ρc=0,1 cm-1
1
Fig. 1. Dependence of relative inaccuracy of contact resistivity
measurement on contact radius when using TLM with circular
contact geometry [13].
If one uses TLM with linear geometry of contact
pads and ρс/ρs = 0.01 cm and ρs = 10-4 Ω⋅cm, the relative
inaccuracy of ρс measurement is 16% [13]. This value is
below that obtained in the case of TLM with circular
geometry. However, as ρs increases to 10-2 Ω⋅cm, the
relative inaccuracy in the case of TLM with linear
geometry grows to 50%, contrary to the case of TLM
with circular geometry in which the error does not
depend on the value of semiconductor resistivity.
3. Formation of contact systems
To measure resistivity of ohmic contacts to n- and p-
GaN, we chose TLM with circular geometry of contact
pads. We made a test pattern with circular geometry of
contacts (Fig. 2) and test TLM structures to measure ρс.
Fig. 2. Contact structure pattern for ρс measurement (a) and
design of a contact structure formed on GaN to measure ρс (b).
(The sizes are given in μm.)
The GaN epitaxial structures were MOCVD-grown
on the sapphire (0001) substrates 400 μm thick. The
LED structures were made at the Close Corporation
“Svetlana-Optoelektronika”, Sankt-Peterburg, Russia. A
low-temperature nucleating layer, n-GaN layer (3.5 μm),
active area with five InGaN/GaN quantum wells, p-
AlGaN current-limiting layer (20 nm) and p-GaN layer
(0.1 μm) with dopant concentration > 2×1017 cm-3 were
grown successively on the sapphire substrate. The n+-n-
n+-GaN−Al2O3 epitaxial structures were grown at the
Close Corporation “Elma-Malakhit”, Zelenograd,
Russia. The layer parameters were as follows: n+ ≈ 1018
cm-3, dn+ ≈ 0.8 μm; n ≈ 1017 cm-3, dn ≈ 3 μm; buffer
layer: n+ ≈ 1018 cm-3, dn+ ≈ 3 μm.
The dislocation density in structures of both types
was over 108 cm-2. The successive Ti−Al−TiBx−Au
(Ni−TiBx−Au) metallization layers to form ohmic
contact to n+-GaN (p-GaN) were deposited with
magnetron sputtering in the argon atmosphere onto the
gallium nitride surface subjected to photon cleaning.
After sputtering the Ti−Al (Ni) layer to n-GaN (p-GaN),
the samples were subjected to rapid thermal annealing
(RTA) at Т = 900 °С in the nitrogen atmosphere for 30 s.
Then the TiBx and Au layers were deposited. As a result,
the ohmic contacts Au−TiBx−Al−Ti-n-GaN and
Au−TiBx−Ni−p-GaN were formed.
The contact resistivity was studied for the
Au−TiBx−Al−Ti−n-GaN contact structures before and
after RTA at Т = 700°С, 870°С and 900°С for 60 s and
Au−TiBx−Ni−p-GaN contact structures before and after
RTA at 700°С for 60 s. We used the developed plant for
measurement of parameters of ohmic contacts to
semiconductors (Fig. 3), as well as Auger electron
spectroscopy with ion etching by an argon beam (energy
of 1 keV) to determine concentration depth profiles of
metallization components.
Fig. 3. The plant for measurement of parameters of ohmic
contact to semiconductor.
© 2010, V. Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine
339
Semiconductor Physics, Quantum Electronics & Optoelectronics, 2010. V. 13, N 4. P. 337-342.
4. Structure of contacts before and after RTA
© 2010, V. Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine
0 2 4 6 8 10
0
20
40
60
80
100
120
140 ρc=6.69·10-5 Ω•cm2
1/πr2, ·104 cm-2
R
, Ω
Fig. 4. The total resistance of Au−TiBx−Al−Ti−n-GaN
contacts as function of 2
1
rπ
.
It was shown in [29] that the contacts with TiBx
antidiffusion layer (barrier) demonstrate better thermal
stability than those with more widespread metallization
with Ti and Al as diffusion barrier. As in [29], our
analysis of the concentration depth profiles for
components of the Ti−Al−TiBBx−Au contacts to n-GaN
taken before and after RTA at Т = 700 °С in the nitrogen
atmosphere for 60 s showed that no interactions between
phases in the contact metallization. Some changes in the
layer structure of such contact were observed after RTA
at a temperature of 900 °С only. One should also note
considerable oxygen concentration in the TiBx layer.
Therefore, one may assume presence of titanium
oxyboride instead of single-phase of TiB2B in it. This may
be the reason for contact metallization structural failure
in the course of RTA at a temperature of 900 °С.
It was noted in a number of works [3-5, 30-32] that
formation of ohmic contact to n-GaN with contact-
forming layers of Al−Ti occurs owing to appearance at
RTA of TiN composition whose work function is 3.87 eV,
i.e., below that of pure Ti (3.95 eV). In this case, nitrogen
vacancies are produced in the GaN near-surface layer
because of N atoms coming to the titanium film. These
nitrogen vacancies are shallow donors in n-GaN. They
produce a thin n+-layer in the near-contact region, thus
ensuring, along with TiN, barrier lowering and decrease
of ρс. In our case, however, the x-ray diffraction pattern
had no indications at presence of TiN phase. This may be
owing to either small amount of titanium nitride or quasi-
amorphous nature of the compound obtained.
Formation of ohmic contact to p-GaN at RTA of
nickel film occurs due to production of intermetallic
compounds of nickel and gallium, with appearance in
the GaN near-contact region of gallium vacancies
serving as acceptors [6, 7].
5. Electrophysical properties of ohmic contacts
5.1. Au−TiBx−Al−Ti−n+-n-n+-GaN ohmic contacts.
The I−V curves of Ti−Al−TiBBx−Au contacts to the n -n-
n -GaN structure taken in the 90−380 K temperature
range turned out to be linear over the whole this range.
The contact resistivity at room temperature was
(6.69±1.67)·10 Ω⋅cm (Fig. 4). The semiconductor
resistivity estimated from the dependence R(1/πr ) was
~10 Ω⋅cm.
+
+
-5 2
2
-2
The current through the Au−TiBx−Al−Ti−n+-n-n+-
GaN contact structure is determined by the
semiconductor resistance because structure conductance
increases as temperature grows up to 150 K. Further
temperature increase led to decrease of conductance.
This fact correlates with the temperature dependence of
electron mobility in GaN considered in [33].
The contact resistivity measurements in the
225−380 К temperature range showed that there are two
regions of ρс variation. In the 225−335 К temperature
range, ρс practically does not depend on temperature.
This is characteristic of the tunnel mechanism of current
transport in the contact. In the 335−380 К temperature
range, ρс decreases exponentially with temperature and
is linear in the coordinates ln(ρcT) = f(1/T). This is
characteristic of the thermionic mechanism of current
transport. The potential barrier height ϕВ determined
from the dependence ln(ρсT) = f(q/kT) in the 335−380 К
temperature range was 0.16 eV.
In the 225−335 К temperature range, the tunnel
mechanism via dislocations of current transport is
characteristic of the contact structure. The dislocation
density in some samples was ~108 cm-2. The current
transport mechanism via dislocations was observed in
[34] for gallium nitride Schottky barrier diodes. In that
case, the dislocation density in GaN was ~108–1010 cm-2.
Such mechanism was advanced earlier to describe the
temperature dependence of saturation current in the
GaP-based surface-barrier diodes [35].
For the best gallium nitride epitaxial structures with
Au−TiBx−Al−Ti−n-GaN ohmic contacts, the contact
resistivity was ~10-6 Ω⋅cm2. After repeated RTA of the
Au−TiBx−Al−Ti−n-GaN contacts at Т = 700 °С and T =
870 °С, the current transport mechanism and ohmic
contact parameters did not change considerably. This
indicates their thermal resistance.
5.2. Au−TiBx−Ni−p-GaN ohmic contacts.
The I−V curves of the Au−TiBx−Ni−p-GaN contact
structure subjected to RTA at Т = 700 °С for 60 s were
linear. The contact resistivity was (1±0.15)×10-3 Ω⋅cm2
(Fig. 5). This is much higher than that of the ohmic
contacts to n-GaN; the ρs value estimated from the cutoff
in the R(1/πr2) curve is ~10-2 Ω⋅cm. This value is in
agreement with the data of other authors who studied
ohmic contacts formed in the course of thermal
annealing (at Т = 700 °С) of a gilded nickel film
(Au−Ni) deposited onto the surface of p-GaN with p ≈
340
Semiconductor Physics, Quantum Electronics & Optoelectronics, 2010. V. 13, N 4. P. 337-342.
© 2010, V. Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine
1 2 3 4
0
20
40
60
80
100
120
140
ρc=10-3 Ω·cm2
R
, Ω
1/πr2, cm-2
Fig. 5. The total resistance of Au−TiBx−Ni−p-GaN contacts as
function of 2
1
rπ
.
1017 cm-3 [36]. In that case, the contact was formed by
the Ga4Ni3, Ga3Ni2, GaAu and GaAu2 intermetallic
phases that reduce the barrier at the interface between
the above phases and p-GaN. Production of Ga
vacancies in the near-contact region that serve as
acceptors [7] also favors reduction of ρс to p-GaN.
When comparing our data on ρс value with those of
other authors [6, 7], one may assume that in our case,
with TiBx diffusion barrier, the contact-forming layer
(after RTA at Т = 700 °С) is formed by the GaNi3 and
Ga3Ni2 intermetallic phases, since the TiBx layer
prevents from penetration of Au atoms to the interface
between Ni and GaP. However, this should be confirmed
with x-ray diffraction studies. In this case, one should
not exclude also additional (positive) influence of Ga
vacancies in the p-GaN near-contact region on decrease
of ρс. The repeated RTA (700 °С, 60 s) did not change ρс
value for the Au−TiBx−Ni−p-GaN ohmic contact. This
fact indicates completeness of metallurgical reactions in
the near-contact region and at the metal−GaN interface
in the course of the first RTA.
The fact that ρс values are higher than those in the
ohmic contact to n-GaN is determined by the features of
doping of p-GaN, e.g., owing to impurities passivation
by hydrogen atoms in the course of MOCVD-growth
and doping of GaN with uncontrolled donor impurities
[37]. The complication results also from absence of
metals with work function over 7.5 eV. Therefore, to
make low-barrier ohmic contacts to p-GaN, one should
either use contact-forming compounds with high work
function values or form a heavily doped narrow-gap
near-surface layer in GaN [3-5, 30].
6. Conclusion
A comparison between the methods of measurement of
contact resistivity of ohmic contacts to n(p)-GaN showed
that TLM with circular contact geometry is optimal. Using
the developed plant for measurement of ohmic contact
parameters, we applied this method for study of
electrophysical properties of ohmic contacts to n- and p-
GaN. The contacts were multilayer Au−TiBx−Al−Ti−n-
GaN structures (that remained thermally stable after RTA
at temperatures up to 900 °С [29]) and Au−TiBx−Ni−p-
GaN structures (that remained thermally stable after RTA
at temperatures up to 700 °С). The I−V curves of the
Ti−Al−TiBBx−Au contacts to the n -n-n -GaN structure
were linear in the 90−380 К temperature range; the
contact resistivity ρ
+ +
с at room temperature was
(6.69±1.67)×10 Ω⋅cm . For the Au−TiB-5 2
x−Ni−p-GaN
contact structure, ρс = (1±0.15)×10 Ω⋅cm , and the
contact I−V curves were.
-3 2
Thus, the experimental results obtained by us
indicate high thermal stability of ohmic contacts, with
titanium boride as diffusion barrier, to n- and p-GaN
layers.
Acknowledgements
This work was supported by the Project No 31/4.2.3.1/33
of the Governmental task scientific and technical program
“Development and implementation of energy-saving light
sources and illumination systems based on them”
(Regulation of the Cabinet of Ministers of Ukraine
No 632 from July 9, 2008). The development of varactor
diodes was carried out under the INCO−COPERNICUS
Program (Project No 977131 “MEMSWAVE”).
References
1. F.E. Shubert, Light-Emitting Diodes, Cambridge
University Press (2006).
2. Yu. Davidenko, High-efficiency modern LEDs //
Sovremennaya Elektronika no 8, p. 36-43 (2004)
(in Russian).
3. H. Morkoç, Handbook of Nitride Semiconductors
and Devices, Willey-VCH (2008).
4. V.N. Danilin, Yu.P. Dokuchaev, T.A. Zhukova,
M.A. Komarov, High-Power High-Temperature
Capable and Radiation-Resistant Microwave New-
Generation Devices with AlGaN/GaN Wide-Gap
Heterojunction Structures. Reviews of Electronic
Equipment, GUP NPP “Pulsar”, Moscow (2001) (in
Russian).
5. Yu.G. Shretter, Yu.T. Rebane, V.A. Zykov,
V.G. Sidorov, Wide-Gap Semiconductors, Nauka,
Sankt-Peterburg (2001) (in Russian).
6. T.V. Blank, Yu.A. Gol’dberg, Semiconductor
photoconverters for ultraviolet spectral region. A
review // Fiz. Tekh. Poluprov. 37(9), p. 1025-1055
(2003) (in Russian).
7. T.V. Blank, Yu.A. Gol’dberg, The mechanism of
current transport in ohmic metal−semiconductor
contacts. A review // Fiz. Tekh. Poluprov. 41(11),
p. 1281-1308 (2007) (in Russian).
8. N. Mochida, T. Honda, T. Shirasawa, A. Inoue,
T. Sakaguchi, F. Koyama, K. Iga // J. Cryst.
Growth 189-190, p. 716(1998).
341
Semiconductor Physics, Quantum Electronics & Optoelectronics, 2010. V. 13, N 4. P. 337-342.
9. M.A. Nicolet, Diffusion barriers in thin films //
Thin Solid Films 52(3), p. 415-443 (1978).
10. O.A. Ageev, A.E. Belyaev, N.S. Boltovets,
R.V. Konakova, V.V. Milenin, V.A. Pilipenko,
Interstitial Phases in Technology for
Semiconductor Devices and VLSI, NTK “Institute
of Single Crystals”, Kharkov (2008) (in Russian).
11. V.N. Sheremet, The making features and electro-
physical properties of ohmic contacts to gallium
nitride (a review) // Optoelektronika i Poluprovod-
nikovaya Tekhnika 44, p. 41-59 (2010) (in Russian).
12. Chung-Yu Ting, Charles Y. Chen, A study of the
contacts of a diffused resistor // Solid St. Electr.
14(6), p. 433-438 (1971).
13. V.N. Sheremet, Metrological aspects of measure-
ment of ohmic contacts resistance // Izvestiya Vuzov.
Radioelektronika 53(3), p. 3-12 (2010) (in Russian).
14. R.H. Cox, H. Strack, Ohmic contacts for GaAs
devices // Solid St. Electr. 10(12), p. 1213-1218
(1967).
15. R.D. Brooks, H.G. Mattes, Spreading resistance
between constant potential surfaces // Bell Sys.
Tech. J. 50(3), p. 775-784 (1971).
16. S.S. Cohen, Contact resistance and methods for its
determination // Thin Solid Films 104(3-4), p. 361-
379 (1983).
17. E.B. Kaganovich, S.V. Svechnikov, The methods
of measurement of contact resistance of
semiconductor planar structures (a review) //
Optoelektronika i Poluprovodnikovaya Tekhnika
21, p. 1-11 (1991) (in Russian).
18. L.G. Russel, J.H. Michael, Y.R. Gary, The effect of
lateral current spreading on the specific contact
resistivity in D-Resistor Kelvin devices // IEEE
Trans. Electr. Dev. 34(3), p. 537-543 (1987).
19. N. Stavitski, M.J.H. van Dal, J.H. Klootwijk,
R.A.M. Wolters, A.Y. Kovalgin, J. Schmitz, Cross-
Bridge Kelvin Resistor (CBKR) structures for
silicide-semiconductor junctions characterization,
in Proc. 9th Annual Workshop on Semiconductor
Advances for Future Electronics and Sensors, 23-
24 Nov 2006, Veldhoven, The Netherlands, p. 436-
438 (2006).
20. H.H. Berger, Contact resistance and contact
resistivity // J. Electrochem. Soc. 119(4), p. 507-
514 (1972).
21. H.H. Berger, Models for contacts to planar devices
// Solid St. Electr. 15(2), p. 145-158 (1972).
22. H. Murrmann, D. Wiedmann, Current crowding on
metal contacts to planar devices // IEEE Trans.
Electr. Dev. 16(12), p. 1022-1024 (1969).
23. H. Murrmann, D. Wiedmann, Messung des
Übergangswiderstandes zwischen Metall und
Diffusionsschicht in Si-Planarelementen // Solid St.
Electr. 12(11), p. 879-886 (1969).
24. G.K. Reeves, H.B. Harrison, Obtaining the specific
contact resistance from transmission line model
measurement // IEEE Electron Device Lett. 3(5), p.
111-113 (1982).
25. A.N. Andreev, M.G. Rastegaeva, V.P. Rastegaev,
S.A. Reshanov, To the problem of accounting for
current spreading in a semiconductor when
determining the transient resistance of ohmic
contacts // Fiz. Tekh. Poluprov. 32(7), p. 832-838
(1998) (in Russian).
26. G.K. Reeves, Specific contact resistance using a
circular transmission line model // Solid St. Electr.
23(5), p. 487-490 (1980).
27. D. Sawdai, Enhanced transmission line model
structures for accurate resistance evaluation of
small-size contacts and for more reliable
fabrication // IEEE Trans. Electron. Dev. 46(7), p.
1302-1311 (1999).
28. M. Lijadi, F. Pardo, N. Bardou, J. Pelouard,
Floating contact transmission line modelling: An
improved method for ohmic contact resistance
measurement // Solid St. Electr. 49(10) p. 1655-
1661 (2005).
29. A.E. Belyaev, N.S. Boltovets, V.N. Ivanov,
L.M. Kapitanchuk, V.P. Kladko, R.V. Konakova,
Ya.Ya. Kudryk, A.V. Kuchuk, O.S. Lytvyn,
V.V. Milenin, V.N. Sheremet, Yu.N. Sveshnikov,
Development of high-stable contact systems to
gallium nitride microwave diodes // SQO 10(4), p. 1-8
(2007).
30. S. Noor Mohammad, Contact mechanisms and
design principles for alloyed ohmic contacts to n-
GaN // J. Appl. Phys. 95(12), p. 7940-7953 (2004).
31. M. Pidun, P. Karduck, J. Mayer, K. Heime,
B. Shineller, T. Walther, Auger depth profile
analysis and EFTEM analysis of annealed Ti/Al-
contacts on Si-doped GaN // Appl. Surf. Sci. 179(1-
4), p. 213-221 (2001).
32. S. Ruvimov, Z. Liliental-Weber, J. Washburn,
D. Qiao, S.S. Lau, Paul K. Chu, Microstructure of
Ti/Al ohmic contacts for n-AlGaN // Appl. Phys.
Lett. 73(18), p. 2582-2584 (1998).
33. W. Gotz, N.M. Johnson, C. Chen, H. Liu, C. Kuo,
W. Imler, Activation energies of Si donors in GaN
// Appl. Phys. Lett. 68(22), p. 3144-3146 (1996).
34. A.E. Belyaev, N.S. Boltovets, V.N. Ivanov,
V.P. Klad’ko, R.V. Konakova, Ya.Ya. Kudryk,
A.V. Kuchuk, V.V. Milenin, Yu.N. Sveshnikov,
V.N. Sheremet, On the tunnel mechanism of
current flow in Au−TiBx−n-GaN-i-Al2O3 Schottky
barrier diodes // SQO 10(3), p. 1-5 (2007).
35. V.V. Evstropov, Yu.V. Zhilyaev, M. Dzhumaeva,
N. Nazarov, Tunnel-excess current in III−V nonde-
generate barrier p-n and m-s structures on Si // Fiz.
Tekh. Poluprov. 31(2), p .152-158 (1997) (in Russian).
36. J.K. Shen, Y.K. Su, G.C. Chi, W.C. Chen, C.Y. Chen,
C.N. Huang, J.M. Hong, Y.C. Yu, C.W. Wang,
E.K. Lin, The effect of thermal annealing on thermal
annealing on the Au/Ni contact of p-GaN // J. Appl.
Phys. 83(6), p. 3172-3175 (1998).
© 2010, V. Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine
37. S.J. Pearton, F. Ren, A.P. Zhang, K.P. Lee,
Fabrication and performance of GaN electronic
devices // Mat. Sci. Eng. R 30(3-6), p. 55-212 (2000).
342
Acknowledgements
|