Electrically benign Ru wet etching method for fabricating Ru/TiO 2 /Ru capacitor

Article abstract of DOI:10.1149/1.3530789

Ru top electrode etching techniques for Ru/Ti O₂ /Ru (RTR) thin film capacitor fabrication were examined. A dry etching process using a plasma mixture of O₂, Cl₂, and Ar gases deteriorated the leakage current properties significantly, which were not recovered by postannealing processes. The surface roughness was not a critical factor in determining the leakage characteristics. The etching damage along the etched edges was not the main cause of the leakage degradation but it was observed over the entire area, which was confirmed according to a comparison of capacitors with different perimeter/area ratios. For the wet etching of Ru films, the etch rates were evaluated using a periodic acid solution at various concentrations at 60°C. The Ru films etched using a 14 wt % periodic acid solution showed a moderate etch rate and a reasonable etching selectivity on the Ti O₂ and Al₂ O₃ films. The as-wet-etched RTR capacitors showed a lower leakage current level than the dry-etched capacitors. Furthermore, the electrical properties of the wet-etched capacitor were improved significantly by a postannealing process.

Full text of DOI:10.1149/1.3530789

Journal of The Electrochemical Society, 158 ͑3͒ G47-G51 ͑2011͒
G47
0013-4651/2011/158͑3͒/G47/5/$28.00 © The Electrochemical Society
Electrically Benign Ru Wet Etching Method for Fabricating
RuÕTiO2ÕRu Capacitor
*
**
Sang Young Lee, Seong Keun Kim, Kyung Min Kim, Gyu-Jin Choi,
**,z
Jeong Hwan Han, and Cheol Seong Hwang
WCU Hybrid Materials Program, Department of Materials Science and Engineering and Inter-university
Semiconductor Research Center, Seoul National University, Seoul 151-744, Korea
Ru top electrode etching techniques for Ru/TiO2/Ru ͑RTR͒ thin film capacitor fabrication were examined. A dry etching process
using a plasma mixture of O2, Cl2, and Ar gases deteriorated the leakage current properties significantly, which were not recovered
by postannealing processes. The surface roughness was not a critical factor in determining the leakage characteristics. The etching
damage along the etched edges was not the main cause of the leakage degradation but it was observed over the entire area, which
was confirmed according to a comparison of capacitors with different perimeter/area ratios. For the wet etching of Ru films, the
etch rates were evaluated using a periodic acid solution at various concentrations at 60°C. The Ru films etched using a 14 wt %
periodic acid solution showed a moderate etch rate and a reasonable etching selectivity on the TiO2 and Al2O3 films. The
as-wet-etched RTR capacitors showed a lower leakage current level than the dry-etched capacitors. Furthermore, the electrical
properties of the wet-etched capacitor were improved significantly by a postannealing process.
© 2011 The Electrochemical Society. ͓DOI: 10.1149/1.3530789͔ All rights reserved.
Manuscript submitted September 20, 2010; revised manuscript received November 29, 2010. Published January 6, 2011.
High performance and high density metal-insulator-metal ͑MIM͒
capacitors have attracted considerable interest for applications in
dynamic random access memory, radio frequency integrated chips,
and passive multilayer ceramic capacitors.1 To increase the capaci-
tance and reduce the leakage current of these MIM capacitors, noble
metals, such as Pt, Ir, and Ru, have been studied as electrode mate-
rials on account of their high work function, low resistivity, and
chemical stability. Among the noble metals, Ru is a step ahead of its
contenders in electrode applications owing to its more facile etching
behavior than Pt and Ir.2-4 In addition, there are many Ru precursors
available, which have proven themselves suitable for atomic layer
deposition ͑ALD͒ and chemical vapor deposition.
The authors reported several studies on MIM capacitors with Ru
electrodes.5-10 In particular, rutile structured TiO2 films with a di-
electric constant as high as 80–100 could be deposited on a Ru
electrode by ALD using a metallorganic Ti precursor and an O3 or
N2O-plasma oxidant.1,5,8 This is because the in situ formed RuO2 at
the interface is structurally similar to the rutile TiO2 with small
lattice mismatch between them.5-8 In this study, MIM capacitors
consisting of a rutile TiO2 dielectric and Ru electrodes were fabri-
cated using dry and wet etching processes. The effect of top elec-
trode ͑TE͒ etching on the dielectric properties of the capacitors was
examined for both processes. Although the electrical properties of
the as-etched samples from both approaches were rather unsatisfac-
tory, the deterioration was much less severe in the samples fabri-
cated using the wet etching process. Furthermore, the damage
caused by the wet etching process was recovered fully after a post-
annealing process. Therefore, based on the superior characteristics
of wet-etched samples, a wet etching process of a Ru layer using a
periodic acid solution is suggested as an alternative for the pattern-
ing of Ru electrodes.
rication of the MIM capacitors, 35 nm thick bottom electrode ͑BE͒
Ru layers were deposited on TiO2/SiO2/Si substrates using the same
ALD process. A 100 nm thick thermally grown SiO2 layer was used
as the reaction barrier and a 5 nm thick TiO2 film was used as a
nucleation enhancing layer by another ALD process at 250°C using
Ti͑OC3H7͒4 and O3 as the Ti source and oxygen source,
respectively.10 The TiO2 dielectric films were grown on the Ru BE
using the same ALD process. The typical dielectric film thickness
was 19.5 nm. Top electrode Ru layers were formed on the TiO2
films by ALD under the same conditions used for Ru BE. An Al2O3
layer was deposited at 250°C on the Ru TE as a hard mask for the
wet etching process by a third ALD system using Al͑CH3͒3 and O3
as the Al precursor and oxygen source, respectively.
Both dry and wet etching processes were examined for Ru TE
patterning. In the dry etching process, the Ru TE was patterned by
photolithography and a dry etching process using a capacitively
coupled plasma type reactive ion etching system ͑Oxford RIE 80
plus model͒. A mixture of O2, Cl2, and Ar gases was used for the dry
etching process. The flow rate of O2, Cl2, and Ar was 96, 24, and
30 sccm, respectively. The plasma power and working pressure were
30 mTorr and 100 W, respectively. Under these optimized condi-
tions, the Ru etch rate was reported to be ϳ34 nm/min.11,12 In this
study, a relatively thick photoresist ͑PR͒ layer ͑1.4 ␮m͒ was used
for dry etching a 35 nm thick Ru TE because the PR is also etched
by O2 plasma. For the wet etching process, periodic acid ͑H5IO6͒
dissolved in deionized water was used as the Ru etchant. An Al2O3
hard mask was employed because the PR was dissolved easily in the
periodic acid solution. After patterning the Al2O3 layer by photoli-
thography and wet etching using a diluted hydrofluoric acid ͑HF͒
solution, the Ru TE was wet-etched in a 14 wt % periodic acid
solution at 60°C. In addition, a TE etch-free capacitor was fabri-
cated as a control sample by the electron-beam evaporation of Pt
through a shadow mask ͑hole diameter = 0.3 mm͒. After fabricating
the MIM capacitors, some samples were annealed in a furnace for
30 min under a N2 + O2 or N2 atmosphere to mitigate the damage
caused by the etching process.
Experimental
Ru thin films were deposited by ALD using 2,4-
͑dimethylpentadienyl͒͑ethylcyclopentadienyl͒Ru ͑DER, from Tosoh
Co.͒, as the Ru source, dissolved in ethylcyclohexane at a concen-
tration of 0.2 M and O2 as the reactant. The Ru solution was in-
jected into the vaporizer for 6 ms and flash-evaporated. The vapor
was then transferred to the reactor by Ar carrier gas at a flow rate of
50 sccm. O2 at a flow rate of 1000 sccm was used as the reactant.
The Ru films were grown at a substrate temperature of 250°C. More
detailed deposition conditions are reported elsewhere.9 For the fab-
The sheet resistance of the Ru films was measured using a four-
point probe. The layer density of the Ru and TiO2 films was evalu-
ated by x-ray fluorescence spectroscopy ͑XRF, Themoscientific,
ARL Quant’X͒. The thickness of the TiO2 and Al2O3 films was
measured by single wavelength ellipsometry. The surface morphol-
ogy of the Ru films was observed by atomic force microscopy
͑AFM, JEOL, JSPM-5200͒. The capacitance–voltage and leakage
current density–voltage ͑J-V͒ properties of the MIM capacitors were
examined using a Hewlett-Packard 4194 impedance analyzer and a
*
Electrochemical Society Student Member.
Electrochemical Society Active Member.
**
z E-mail: cheolsh@snu.ac.kr
Downloaded on 2014-06-26 to IP 137.99.31.134 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

G48
Journal of The Electrochemical Society, 158 ͑3͒ G47-G51 ͑2011͒
100
100
10-2
10-4
10-6
10-8
10-10
(a)
(c)
(b)
(a)
(b)
PR
PR
10-2
10-4
10-6
10-8
10-10
Al2O3
TE Ru
TiO2
Al2O3
TE Ru
TiO2
BE Ru
BE Ru
as-etched
400oC
Solid : as-dep.
Open : anneal.
Large
Small
450oC
TE Pt
(d)
-2
0
1
2
Ap-p1lied voltage [V]
-2
-1
0
1
2
Applied voltage [V]
PR
Al2O3
TE Ru
TiO2
Figure 2. ͑Color online͒ J-V curves of the Ru/TiO2/Ru capacitors fabricated
by the dry etching process, as-etched and after annealing under a N2/O2 ͑5%͒
atmosphere at 400°C and 450°C for 30 min. J-V curves of the etch-free Pt
TE capacitor is also included as a reference.
TE Ru
TiO2
BE Ru
BE Ru
Figure 1. ͑Color online͒ Schematic cross section of the Ru/TiO2/Ru struc-
ture fabricated by the wet etching process: ͑a͒ PR patterning by photolithog-
raphy, ͑b͒ Al2O3 etching by HF solution, ͑c͒ Ru TE etching in a periodic acid
solution, and ͑d͒ Al2O3 on Ru TE removal in a HF solution for the TE Ru
contact.
a N2/O2 ͑5%͒ atmosphere at 400 and 450°C for 30 min. The capaci-
tor annealed at 400°C showed better leakage properties than the
capacitor annealed at 450°C. Although the annealing process at
400°C improved the leakage properties of the capacitor, the J level
of the capacitor was still much higher than that of the etch-free
capacitor with a Pt TE.
The surface morphology of the electrodes was observed by AFM
with a scan area of 2 ␮m ϫ 2 ␮m to explain the high leakage cur-
rents in the dry-etched samples. These measurements were per-
formed under the assumption that a rough surface could result in
more defects at the interface or the electric field concentration effect
acts as leakage sources. However, the root-mean-square ͑rms͒
roughness of the as-etched sample was 3.83 nm, whereas the sample
annealed at 400°C showed a rather higher rms roughness of 4.95 nm
͑the measured AFM images are not shown͒. Therefore, the surface
roughness is not a critical factor determining the leakage character-
istics in this case.
To determine the origin of the high leakage current paths, two
possible leakage paths—along the periphery and the entire plane of
the TE—were considered. Accordingly, the electrical properties of
two RTR capacitors with different TE areas were compared. If the
leakage paths are in the entire plane, the leakage current density
should have no areal dependence. However, if the leakage paths are
distributed along the periphery, the capacitor with the smaller elec-
trode, which has a larger ratio of the periphery/whole electrode area,
should show inferior electrical properties. Two electrodes with dif-
ferent electrode areas ͑29,400 and 623,200 ␮m2 with a periphery
proportion of 40:1 and 110:1, respectively͒ were prepared. The areal
ratio was 21. Figure 2b shows the leakage current behavior of the
large ͑triangle symbol͒ and small ͑rectangular symbol͒ area sample
before ͑solid symbol͒ and after ͑open symbol͒ annealing. The
equivalent oxide thickness ͑toxequiv͒ was almost identical in all
samples ͑data not shown͒. Interestingly, the J-V results showed a
larger leakage current in the larger sample. This suggests that the
number of defects might increase with increasing TE area, even
though the sample was coated with a 1.4 ␮m thick PR layer. The
abnormal increase in leakage according to the electrode size can be
explained by assuming the presence of extremely large defects.
These large defects would cause considerable leakage. Moreover,
larger electrodes would show a higher leakage current density be-
cause electrodes with large areas are more likely to have such de-
fects. In all cases, the damage from dry etching could not be recov-
ered by postannealing because the damage had spread throughout
the entire capacitor area despite PR screening during the etching
process.
Hewlett-Packard 4140B picoammeter, respectively. During the elec-
trical measurements, the Ru TE was biased and the Ru BE was
grounded.
Results and Discussion
Figure 1 shows a schematic diagram of the wet etching process
used in this experiment. The sample structure used for this wet etch-
ing test is similar to that used for the dry etching test except for the
use of a hard mask ͑Al2O3 layer͒. Figure 1a shows a schematic
diagram of a cross section of the Ru/TiO2/Ru ͑RTR͒ structure after
PR patterning. The 50 nm thick Al2O3 layer was inserted between
the Ru TE and PR layers, which not only worked as a hard mask,
but also improved the adhesion between the PR and the Ru TE. The
PR was first patterned by photolithography. The exposed Al2O3
layer was removed by dipping the sample into a HF solution, as
indicated in Fig. 1b. During this HF dipping step, the Ru TE layer
was barely etched ͑data not shown͒. This process was followed by
wet etching in a periodic acid solution to pattern the Ru TE ͑Fig.
1c͒. The PR was also removed during these processes. Finally the
remaining Al2O3 hard mask was removed for TE contact by dipping
the sample into a HF solution ͑Fig. 1d͒. The TiO2 layer was barely
etched during this final HF-dipping step. Therefore, a RTR struc-
tured capacitor was fabricated successfully using this wet etching
process.
Although Ru thin films are also easily dry-etched by a plasma
mixture of O2, Cl2, and Ar gases because oxidized species of Ru,
such as RuO3 and RuO4, are volatile and removed readily,11 this dry
etching process has an adverse effect on the electrical properties of
MIM capacitors. Figure 2 shows the J-V curves of the capacitors
fabricated by the dry etching process before and after annealing. The
J-V curve of a capacitor with an etch-free Pt TE is also included in
Fig. 2a to show the effect of the dry etching process on the leakage
properties of the capacitors. The J-V data for the Pt/TiO2/Ru ͑PTR͒
capacitor shows only the positive bias region because the data in the
negative bias region is affected by the higher work function of the Pt
electrode. The data in this region is affected by the higher work
function of Pt compared to Ru, and was actually better than the data
from the positive bias region. The very low leakage current density
of the PTR capacitor suggests that the TiO2 film itself and the
TiO2/Ru BE interface remain intact. The as-dry-etched capacitor in
Fig. 2a exhibited a very high leakage current suggesting serious
damage during the etching process. This damage effect degrades the
Ru BE/TiO2 interface properties as well as the Ru TE/TiO2 inter-
face properties, as can be seen from the high J level in the positive
bias region. To relieve this damage, the capacitors were annealed in
The authors recently reported a non-negligible amount of Cl ions
remaining in the TiO2 and Ru films after the Ru layer was dry
etched with Cl2 and Ar plasma.12 Residual Cl from the dry etching
process penetrates the Ru TE layer and even the TiO2 film, resulting
Downloaded on 2014-06-26 to IP 137.99.31.134 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 158 ͑3͒ G47-G51 ͑2011͒
G49
107
106
105
104
103
102
101
100
107
106
105
104
103
102
101
100
107
106
105
104
103
102
101
100
50
40
30
20
10
0
50
40
30
20
10
0
50
40
30
20
10
0
(a)
(b)
(c)
17wt%
14wt%
10wt%
Figure 3. ͑Color online͒ Sheet resistance
and Ru layer density of Ru thin film on Si
substrate according to the concentration of
the periodic acid solution: ͑a͒ 17, ͑b͒ 14,
and ͑c͒ 10 wt % at 60°C, ͑d͒ sheet resis-
tance and Ru layer density of Ru film on
the TiO2 substrate, ͑e͒ thickness and layer
density of Ti atoms in the TiO2 film, and
͑f͒ thickness of the Al2O3 film as a func-
tion of the etching time in a periodic acid
solution of 14 wt %, respectively.
0
100
200
300
0
300
600
900
0
30
60
90
Etch Time [s]
Etch Time [s]
Etch Time [s]
104
103
102
101
100
50
9
35
30
25
20
15
10
5
7
6
5
4
3
(d) Ru
(e) TiO2
(f) Al2O3
40
30
20
10
0
8
7
6
5
0
0
30 60 90 120
Etching time [s]
0
60
120
180
0
60
120
180
Etching time [s]
Etching time [s]
in a large leakage current. These adverse effects caused by Cl con-
tamination cannot be recovered by annealing because the residual Cl
cannot be removed under the given conditions.12
and in many cases, it is quite difficult to avoid over-etching. In
contrast, the 10 wt % periodic acid solution took more than 6 min to
etch the Ru layer, which is too slow for practical use. The 14 wt %
periodic acid solution required approximately 2 min to etch the
35 nm thick Ru layer so that the periodic solution was chosen as the
etchant to fabricate the RTR capacitors. In the cases of 14 and
10 wt % etching solutions, the Ru layer density did not decrease to
ϳ0 even after a very long time, whereas the value for the 17 wt %
case reached ϳ0 after an etching time of ϳ20 s. This may be re-
lated to interfacial Ru-silicide formation, which must be more resis-
tant to wet etching. Therefore, the wet etching rate of a Ru film on
the TiO2 layer was evaluated again to further optimize the etching
conditions using a 14 wt % periodic acid solution at 60°C, which
was the optimal condition determined by previous etchant concen-
tration tests.
Figure 3d shows the changes in the sheet resistance and layer
density of the Ru films on a TiO2/SiO2/Si substrate as a function of
the etching time. The sheet resistance of the Ru films increased with
increasing etching time and reached a certain value after etching for
more than 90 s. The layer density of Ru atoms in the films also
decreased with the increasing etching time and approached the de-
tection limit after an etching time of ϳ90 s. This shows that the Ru
thin films were well etched in the periodic acid solution. In addition,
a 35 nm thick Ru TE was removed completely in this periodic acid
solution after an etching time greater than ϳ90 s.
To examine the etching selectivity of the periodic solution con-
cerning Ru and oxide films, the wet etching process was attempted
for a 23 nm thick ALD TiO2 film and an 8 nm thick ALD Al2O3
film using the same periodic acid solution. Figures 3e and 3f show
the change in the TiO2 and Al2O3 film thickness as a function of the
etching time. Figure 3e also shows the change in the layer density of
Ti atoms in the TiO2 film as a function of the etching time. Figure 3e
showed no change in the TiO2 film thickness and Ti layer density up
to an etching time of 180 s, which suggests that TiO2 does not react
with this etchant. However, the thickness of the Al2O3 film de-
creased slightly with increasing etching time. The etching rate of the
Al2O3 film in this periodic acid solution was ϳ1.1 nm/min, which
was much lower than that of the Ru films. Therefore, Al2O3 is a
suitable candidate for a hard mask in the wet etching process. TiO2
cannot be used as a hard mask because it is not removed even after
the wet etching of Ru is complete ͑see Fig. 1͒.
The addition of Cl2 in the dry etching process of Ru films was
reported to significantly enhance the etching rate of Ru and improve
the etching selectivity of Ru to oxides.13 In addition, in the absence
of Cl2, the by-products of the etching process were redeposited on
the chamber walls.13 Therefore, the addition of Cl2 is essential for
the dry etching of Ru films. However, this means that the dry etch-
ing process fatally deteriorates the leakage properties of the capaci-
tor due to severe Cl contamination.
As discussed above, Cl contamination ͑and other possible dam-
age by the energetic particle bombardment͒ was the source of the
high leakage in the dry-etched sample. This problem was not solved
even with a thick enough photoresist. Although a hard mask has not
been attempted in the dry etching test, it may not solve the Cl
contamination problem because Cl was already detected, even in the
Ru BE after dry etching.12 The wet etching process in this study
provides a much easier way of fabricating the MIM capacitors with-
out such adverse effects.
Therefore, the wet etching process as shown in Fig. 1 is sug-
gested for the patterning of the Ru TE. Although the wet etching
process shows isotropic etching characteristics, the effect of side
etching might be minimized when the size of the patterned cells is
much larger than the thickness of Ru TE. In this study, periodic acid
+
was used as the etchant for Ru films. The active ͓I͑OH͒ ͔ is pro-
6
duced in the water solution by the following reaction14
+
͓I͑OH͒ ͔ ↔ H+ + H5IO6
͓1͔
6
Ru can easily react with ͓I͑OH͒ ͔+, resulting in Ru etching ac-
6
cording to Eq. 2 ͑Ref. 15͒
+
+
Ru + 4͓I͑OH͒ ͔ → 4IO + RuO4 + 8H + 8H2O
͓2͔
6
3
The process should be performed with care because the wet etch-
ing process produces hazardous RuO4 gas.
The etching rate of Ru films is strongly dependent on the con-
centration of periodic acid. Therefore, the etching rate of Ru films in
a periodic acid solution was examined to optimize the etching pro-
cess. Figure 3 shows the etching rate of a 35 nm thick Ru film on a
Si substrate in a ͑a͒ 17, ͑b͒ 14, and ͑c͒ 10 wt % periodic acid solu-
tion at 60°C. The etching rate was evaluated by the changes in the
sheet resistance and areal density of the film as a function of the
etching time. The 17 wt % periodic acid solution etched the Ru film
too rapidly. The 35 nm thick Ru film was removed almost com-
pletely in the 17 wt % periodic solution within 15 s. It is difficult to
accurately tailor the etching conditions under these circumstances;
Wet etching uniformity is another important factor for a repro-
ducible process. The surface of TiO2 films before and after wet
etching the Ru layer was observed by AFM to examine the wet
etching uniformity. Figures 4a and 4b show AFM images of the
Downloaded on 2014-06-26 to IP 137.99.31.134 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

G50
Journal of The Electrochemical Society, 158 ͑3͒ G47-G51 ͑2011͒
under a N2/O2 ͑5%͒ atmosphere at the same temperature reduced
(a)
(b)
the leakage current significantly, resulting in superior leakage char-
acteristics to the etch-free capacitor with a Pt TE. This suggests that
some of the unidentified damage from the ALD step of Ru TE and
wet etching processes of the Ru TE had been removed by annealing.
Annealing at temperatures ranging from 400 to 450°C resulted in
similarly improved properties but temperatures Ͼ450°C degraded
them.
Figure 5c shows the leakage current density at 0.8 V versus
toxequiv projections of both the dry- and wet-etched RTR capacitors.
Here, the PTR capacitor reference data from the paper by Kim et al.7
was included. The data grouped by red circles mean that the same
TiO2 film thicknesses of 19.5 ͑for larger toxequiv͒ and 16.3 nm ͑for
smaller toxequiv͒ were used. The 19.5 nm thick TiO2 dielectrics were
evaluated for both dry- and wet-etched RTR capacitors and the PTR
capacitor. The 16.3 nm thick TiO2 layer was tested only in the wet-
etched case for better electrical properties in the lower toxequiv re-
gion. The star shaped symbols are the leakage current density at a
voltage of 0.8 V for the wet-etched capacitors. The wet-etched data
show a considerably lower leakage current density than the dry-
etched data. Furthermore, the leakage current density of the wet-
etched samples was even lower than that of the PTR capacitor,
which is significant because PTR capacitors have no etch damage.
This type of behavior can be seen in the case of the 16.3 nm thick
TiO2 layer. This means that wet etching has beneficial effects on the
electrical properties in the lower toxequiv region. This phenomenon
can also be understood from the good interface between the Ru TE
and the TiO2. Here, annealing under a N2/O2 atmosphere was per-
formed at 400°C.
Figure 4. ͑Color online͒ AFM images of the TiO2 films ͑a͒ before deposition
of the Ru TE and ͑b͒ after wet etching of Ru, respectively. The rms rough-
ness of the etched TiO2 film is ͑a͒ 1.34 and ͑b͒ 1.40 nm, respectively.
TiO2 films before deposition of the Ru TE and after the wet etching
of Ru, respectively. The rms roughness after wet etching of the TiO2
film was similar to that before wet etching: 1.34 and 1.40 nm, re-
spectively. There were no particles or residue remaining on the TiO2
surface after etching, suggesting that the wet etching process is
highly uniform and effective in etching Ru thin films.
The electrical properties of the RTR capacitors fabricated by the
wet etching process shown in Fig. 1 were evaluated. Figures 5a and
5b show the changes in the dielectric constant of the TiO2 films in
the RTR capacitors and the J-V curves of the capacitors with respect
to the annealing conditions, respectively. The dielectric constant of
the TiO2 films was in the 86–95 range, which is consistent with the
previous results on the formation of rutile structured TiO2 on a Ru
substrate.12 The leakage current density of the as-wet-etched sample
was lower than that of the as-dry-etched capacitor in Fig. 2. Anneal-
ing in a N2 atmosphere at 400°C for 30 min could not improve the
leakage properties of the capacitor. However, the annealing process
The RTR capacitor fabricated by the wet etching process showed
excellent properties comparable to those of PTR capacitors. This is
meaningful because a device that was fabricated only by ALD and
then wet etched showed better characteristics than a device without
any etch damage fabricated by other means. This all ALD-processed
capacitor is applicable to three dimensional structures, such as deep
holes or thin film type multilayer ceramic capacitors.
100
120
100
10-2
10-4
10-6
Conclusions
80
This study examined the wet etching processes of Ru TEs for
RTR thin film capacitor fabrication. The dry etching process of Ru
using a plasma mixture of O2, Cl2, and Ar deteriorated the leakage
properties of the capacitors due to severe Cl contamination. Cl ions
remaining in the Ru and TiO2 films could not be removed by post-
annealing. Therefore, a wet etching process for patterning the Ru TE
was attempted to fundamentally solve this Cl contamination prob-
lem. The wet etching rate of Ru in the periodic acid solutions with
various compositions were evaluated at 60°C. As a result, the
14 wt % periodic acid solution showed a moderate Ru etching rate
with reasonable etching selectivity on TiO2 and Al2O3 films. The
as-wet-etched RTR capacitors apparently showed comparable leak-
age characteristics compared to the dry-etched capacitors. After an-
nealing under a N2/O2 ͑5%͒ atmosphere at 400°C, the capacitors
fabricated by the wet etching process showed significantly improved
electrical properties. The electrical properties of the RTR capacitor
fabricated solely by ALD and then wet etched stacked in an ALD
Ru/ALD TiO2/ALD Ru sequence shows better electrical properties
than the etch-free sputtered Ru/TiO2/E-beam Pt stack.
60
40
as-etched
10-8
20
0
(b)
TE Pt
1
N2
(a)
N2/O2
10-10
-2
-1
2
Applied0voltage[V]
as-etched
N2
N2/O2
Annealing condition
100
PTR capacitor
Ref. 7
(c)
10-2
10-4
10-6
10-8
10-10
Pt
dry etch as dep.
dry etch N2+O2 anneal.
wet etch as dep.
wet etch N2+O2 anneal.
wet etch N2 anneal.
Acknowledgment
4
5
6
8
9
10
toxeq.7[Å]
This study was supported by the IT R&D program of MKE/IITA,
“Capacitor technology for next generation DRAMs having mass-
production compatibility” 2009-F-013-01, Converging Research
Center Program through National Research Foundation of Korea
͑NRF͒ funded by the Ministry of Education, Science and Technol-
ogy ͑2010K000977͒, and World Class University program through
the National Research Foundation of Korea funded by the Ministry
of Education, Science and Technology ͑R31-2008-000-10075-0͒.
Figure 5. ͑Color online͒ Variations in ͑a͒ the dielectric constant of TiO2
film, ͑b͒ J-V curves in Ru/TiO2/Ru capacitor fabricated by wet etching with
respect to the annealing conditions under a N2 and a N2/O2 ͑5%͒ atmosphere
at 400°C, ͑c͒ leakage current density at 0.8 V vs toxequiv of the dry- and
wet-etched RTR capacitors with reference to the etch-free PTR capacitor,
respectively.
Downloaded on 2014-06-26 to IP 137.99.31.134 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 158 ͑3͒ G47-G51 ͑2011͒
G51
Lett., 9, F5 ͑2006͒.
Seoul National University assisted in meeting the publication costs of this
article.
7. S. K. Kim, G.-J. Choi, S. Y. Lee, M. Seo, S. W. Lee, J. H. Han, H.-S. Ahn, S. Han,
and C. S. Hwang, Adv. Mater., 20, 1429 ͑2008͒.
8. G.-J. Choi, S. K. Kim, S.-J. Won, H. J. Kim, and C. S. Hwang, J. Electrochem.
Soc., 156, G138 ͑2009͒.
9. S. K. Kim, S. Y. Lee, S. W. Lee, G. W. Hwang, C. S. Hwang, J. W. Lee, and J.
Jeong, J. Electrochem. Soc., 154, D95 ͑2007͒.
10. S. K. Kim, J. H. Han, G. H. Kim, and C. S. Hwang, Chem. Mater., 22, 2850
͑2010͒.
11. H. W. Kim and C.-J. Kang, Microelectron. Eng., 69, 89 ͑2003͒.
12. K. M. Kim, S. Y. Lee, G. J. Choi, J. H. Han, and C. S. Hwang, Electrochem.
Solid-State Lett., 13, G1 ͑2010͒.
13. C. C. Hsu, J. W. Coburn, and D. B. Graves, J. Vac. Sci. Technol. A, 24, 1 ͑2006͒.
14. H. C. Mishra and M. C. R. Symons, J. Chem. Soc., 1962, 1194.
15. I. Asano, Y. Nakamura, M. Hiratani, T. Nabatame, S. Iijima, T. Saeki, T. Futase, S.
Yamomoto, T. Saito, and T. Sekiguchi, Electron. Commun. Jpn., 87, 26 ͑2004͒.
References
1. S. K. Kim, S. W. Lee, J. H. Han, B. Lee, S. Han, and C. S. Hwang, Adv. Funct.
Mater., 20, 2989 ͑2010͒.
2. S. Won, W. Kim, C. Yoo, S. Kim, Y. Park, J. Moon, and M. Lee, Tech. Dig. - Int.
Electron Devices Meet., 2000, 789.
3. S. Y. Kang, H. J. Lim, C. S. Hwang, and H. J. Kim, J. Electrochem. Soc., 149,
C317 ͑2002͒.
4. K. Ono, T. Horikawa, T. Shibano, N. Mikami, T. Kuroiwa, T. Kawahara, S. Mat-
suno, F. Uchikawa, S. Satoh, and H. Abe, Tech. Dig. - Int. Electron Devices Meet.,
1998, 802 ͑1998͒.
5. S. K. Kim, W.-D. Kim, K.-M. Kim, C. S. Hwang, and J. Jeong, Appl. Phys. Lett.,
85, 4112 ͑2004͒.
6. S. K. Kim, G. W. Hwang, W.-D. Kim, and C. S. Hwang, Electrochem. Solid-State
Downloaded on 2014-06-26 to IP 137.99.31.134 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).