Development of High-Quality FBAR Devicesfor Wireless Applications Employing Two-Step Annealing Treatments Eunju Lee, Linh Mai, and Giwan Yoon, Member, IEEE Abstract—In this letter, a new
Trang 1Development of High-Quality FBAR Devices
for Wireless Applications Employing Two-Step Annealing Treatments
Eunju Lee, Linh Mai, and Giwan Yoon, Member, IEEE
Abstract—In this letter, a new two-step annealing technique is
presented that can more effectively improve the resonance
charac-teristics of the film bulk acoustic wave resonator (FBAR) devices
in terms of return loss, -factor, and effective electromechanical
coupling coefficient ( 2 ) In the case of the SMR-type FBAR
de-vices, the use of this approach has considerably improved the
res-onance performance ( 8000 of Q-factor value, 2% of 2 )
at the operating frequency of 1 8 GHz, as compared to the
con-ventional annealing techniques.
Index Terms—Film bulk acoustic wave resonator (FBAR)
de-vices, high quality factor, resonance characteristics, two-step
an-nealing, ZnO films.
I INTRODUCTION
R ECENTLY, a considerable technology progress in
micro-electronics has enabled most of the RF components to be
integrated into a single-chip or transceiver of the wireless
mo-bile systems Unfortunately, the portion alone of the RF filter
has been still used as an off-chip type This is largely because the
conventional RF filters can hardly be integrated compatibly with
the current Si-based CMOS technologies [1] From this
stand-point, the film bulk acoustic wave resonator (FBAR) devices and
their process technologies are most likely to be a very promising
candidate to resolve the above issue, mainly due to the high
compatibility of the materials, device structures, and fabrication
processing with the current CMOS technology Thus, the FBAR
device technology has a strong potential for more extensive
ap-plication of the RF/IF filters, duplexers and voltage-controlled
oscillators [2]–[4]
In this work, a new FBAR device fabrication method has been
studied that can more effectively enhance the resonance
charac-teristics The proposed technique appears to be very useful and
promising especially for the solidly mounted resonator
(SMR)-type FBAR devices [5]
Manuscript received February 18, 2011; accepted September 02, 2011 Date
of publication October 06, 2011; date of current version November 09, 2011.
This work was supported in part by the Basic Science Research Program through
National Research Foundation of Korea (NRF) funded by the Ministry of
Edu-cation, Science and Technology (20090063076).
E Lee and G Yoon are with the Electrical Engineering Department, the
Korea Advanced Institute of Science and Technology (KAIST), Daejeon, Korea
(e-mail: celkaist@gmail.com; gwyoon@kaist.ac.kr).
L Mai is with the School of Engineering, International University,
Vietnam National University-HCMC, Ho Chi Minh City, Vietnam (e-mail:
mlinh@hcmiu.edu.vn).
Color versions of one or more of the figures in this letter are available online
at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/LMWC.2011.2168200
Fig 1 (a) 3-D and 2-D schematic diagrams of the FBAR devices, (b) Cross-sectional SEM image of the fabricated FBAR device, and (c) Schematic descrip-tion of the two-step annealing concept proposed in this work.
II DEVICEDESIGN ANDFABRICATION
A two-step thermal annealing treatment was performed on the SMR-type FBAR devices particularly with Co or Al electrodes Fig 1 shows the schematic diagrams (a) and the cross-sectional SEM image of the FBAR device (b), accompanied by the con-cept of the two-step annealing process employed in this work (c) The two-step annealing process includes both the so-called Bragg reflector-annealing step [6] and the post-annealing step [7] The former step is the thermal annealing treatment per-formed just on the acoustic Bragg reflectors prior to the deposi-tion of the bottom electrode, and the latter step is the annealing treatment performed immediately after the top electrode deposi-tion was completed Overall, the most significant improvement has been obtained by the proposed two-step annealing technique rather than by either Bragg reflector-annealing or the post-an-nealing alone
The FBAR devices were fabricated using a thin-film deposi-tion technique, as shown in Fig 1 The fabricadeposi-tion of the SMR-type FBAR devices is classified largely into four stages The first stage, namely Bragg reflector deposition, was carried out
as follows The acoustic Bragg reflector of /W multi-layers was formed by the thin-film deposition on the 4-inch p-type (100) silicon (Si) wafer with the 6000 Å-thick thermally ox-idized layer Here, four more layers were additionally formed by the alternate-deposition of 0.6 -thick tungsten
1531-1309/$26.00 © 2011 IEEE
Trang 2(W) and 0.6 -thick silicon dioxide ( ) in a magnetron
sputtering system The second stage was named the Bragg
re-flector annealing process Immediately after the Bragg rere-flector
formation, the wafer with the five-layer Bragg reflector was
di-vided into five samples (A~E) Then, only the three samples
(B, D, and E) were thermally annealed in a dehydrate furnace
at 400 /30 min, while keeping the other two samples (A, C)
non-annealed In the third stage of the so-called resonator part
fabrication process, the 0.2 -thick Co bottom electrodes were
formed on the 4 samples (A~D) while the 0.2 -thick Al
bottom electrode was formed on the sample E These bottom
electrodes can act as a floating ground, which has the
semi-in-finite ground effect Then, the 1.2 -thick zinc oxide (ZnO)
was deposited on all five samples at the same time Three
dif-ferent top electrode patterns (patterns 1, 2, and 3 in the insets of
Fig 2) were employed in this work in order to confirm whether
the proposed process could work well, regardless of the
dimen-sions, configurations and shapes of the top electrodes The top
Co electrode (0.2 -thick) was formed on the four samples
(A~D) while the top Al electrode on the sample E As a
re-sult, the five FBAR device samples (A~E) were divided into
two groups where the first group is the Co-FBAR devices
(sam-ples A~D) and the second is the Al-FBAR devices (sample E)
The last stage, called the post annealing process, was carried
out only on the three resonator samples (C~E) in argon ambient
at 200 /120 min The thermal treatment conditions of the five
samples are summarized in Table I Finally, the parameters for
return loss ( ) were extracted from the FBAR devices using
a measurement system along with a probe station and HP 8722D
network analyzer
III RESULTS ANDDISCUSSION
The values for return loss ( ) of the Co-FBAR devices
for the three patterns were plotted and summarized for the
com-parison of the annealing effects, as shown in Table II and Fig 2
The resonance characteristics of the three annealed samples, i.e.,
Bragg reflector annealing only (sample B), post-annealing only
(sample C) and two-step annealing (sample D) were compared
with the non-annealed sample A
On the other hand, the effective electromechanical coupling
coefficient (1) and series/parallel -factors (2) are generally
used as the figure of merits (FOMs) to estimate the device
per-formance [8]
(1) (2)
where the and are the series and parallel resonance
fre-quencies, respectively, and also the is the input impedance
phase The slopes of the are plotted as a function of the
frequency for the different annealing conditions on the FBAR
devices with the pattern 1, as shown in Fig 3 The calculated
effective electromechanical coupling coefficient and the
series/parallel quality factor values for the FBAR devices
with pattern 1 are tabulated in Table III
Fig 2 S measurement results for the three different top electrode patterns
of the FBAR device.
TABLE I
T HERMAL T REATMENT C ONDITIONS FOR FBAR D EVICE S AMPLES
TABLE II
R ETURN L OSS M EASUREMENT R ESULTS FOR T HREE D IFFERENT P ATTERNS
First, as shown in Table II, the return losses of the sample
B were around 3.18, 1.38, and 0.96 dB better than those of the non-annealed sample A for the three different patterns, respec-tively This indicates that the resonance characteristics of the FBAR devices (sample B) depend surely on the annealing ditions applied to the Bragg reflectors As a result, it was con-firmed that the devices with the Bragg reflectors annealed at
400 /30 min could show a good resonator performance [6], [9] Second, the return losses of the sample C were around 4.87, 4.24 and 8.99 dB better than those of the sample A Clearly,
Trang 3Fig 3 Slopes of Z plotted as a function of the frequency for the different
annealing conditions on the FBAR devices with pattern 1.
TABLE III
C ALCULATED Q-F ACTOR AND E LECTROMECHANICAL C OUPLING
C OEFFICIENTS FOR FBAR D EVICES W ITH P ATTERN 1
the post-annealing process alone seems to affect the sandwiched
structure and help to further enhance the resonator performance
of the FBAR devices Last, the return losses of the sample D
were around 10.37, 11.61, and 12.81 dB better than those of the
sample A, indicating a significantly more improvement by the
proposed two-step annealing method The combined use of both
the Bragg reflector-annealing and the post-annealing (namely,
two-step annealing) is expected to more effectively improve the
resonance characteristics of the FBAR devices These trends of
the resonance characteristics could be further supported by the
values of the , and , as summarized in Table III
Par-ticularly in the case of the SMR-type FBAR devices employing
this two-step annealing approach (sample D), a considerably
higher performance could be achieved at , as
com-pared to the conventional annealing techniques (samples B, C)
Overall, the Bragg reflector-annealing at 400 /30 min, the
first step, is considered to eliminate any possibly existing
imper-fect microstructures and incomplete adhesions in the Bragg
re-flectors The post-annealing at 200 /120 min, the second step,
is likely to get rid of any physical imperfections, such as the
incomplete adhesions and micro-defects of the FBAR device
it-self, further reducing any incomplete microstructures and
adhe-sions in the Bragg reflectors
Moreover, the effects of the two-step annealing on the
reso-nance characteristics of the FBAR devices particularly with the
Co electrodes were investigated in comparison with those with
the Al electrodes, respectively called the Co-FBAR (sample D)
and the Al-FBAR (sample E) The two critical factors that deter-mine the ZnO-based FBAR characteristics are believed to be the quality of both Bragg reflector and the piezoelectric property of the ZnO film as well Comparing the Al-FBAR and Co-FBAR devices with the Bragg reflectors annealed under the same con-dition, the difference in the device performance appears to be related to the degree of the preferred orientation of the crystal grains of the ZnO film The ZnO/Co structured film was reported
to be more strongly oriented toward the -axis perpendicular
to the surface of the substrate compared with the ZnO/Al film Thus, the resonance characteristics could be further improved
by using the Co electrodes, instead of the Al electrodes [6], [9]
As shown in Table II, the return losses of the sample D were around 1.61, 5.13 and 3.51 dB better than those of sample E for the 3 different patterns, respectively Also, the and -factor values of the sample D were observed to show better resonance characteristics than those of the sample E, as shown
in Table III Therefore, we believe that the combined use of the two-step thermal annealing and the Co electrodes can more ef-fectively improve the resonance characteristics of the FBAR de-vices as compared to the other cases [6], [9]
IV CONCLUSION
We for the first time presented a new two-step annealing method that could more effectively improve the resonance char-acteristics of the FBAR devices We have also demonstrated that the resonance characteristics of the FBAR devices could
be further enhanced in terms of return loss, -factor, and
by the optimization of the fabrication processing technique The proposed two-step annealing approach seems very useful and promising especially for the fabrication of the SMR-type Co-FBAR devices with the W/ multi-layer acoustic Bragg reflectors
REFERENCES [1] P R Gray and R G Meyer, “Future directions in silicon ICs for RF
personal communications,” in Proc IEEE Custom Integr Circuits Conf., 1995, pp 83–90.
[2] M Aissi, E Tournier, M A Dubois, G Parat, and R Plana 1, “A 5.4 GHz 0.35 m BiCMOS FBAR resonator oscillator in above-IC
tech-nology,” in Proc IEEE ISSCC, 2006, pp 1228–1235.
[3] C H Tai, T K Shing, Y D Lee, and C C Tien, “A novel thin film
bulk acoustic resonator duplexer for wireless applications,” Tamkang
J Sci Eng., vol 7, pp 67–71, 2004.
[4] P Bar, A Giry, P Triolet, G Parat, D Pache, P Ancey, and J F Car-pentier, “Full-duplex receiver and PA integration with BAW devices,”
in Proc SiRF, 2008, pp 9–12.
[5] K M Lakin, “Thin film resonator technology,” IEEE Trans Ultrason Ferroelectr Freq Control, vol 52, no 5, pp 707–716, May 2005.
[6] M H Yim, D H Kim, D K Chai, and G W Yoon, “Effects of thermal annealing of W/ SiO multilayer Bragg reflectors on resonance charac-teristics of film bulk acoustic resonator devices with cobalt electrodes,”
J Vac Sci Technol A, vol 22, pp 465–471, 2004.
[7] S Y Chu, W Water, and J T Liaw, “Influence of post deposition annealing on the properties of ZnO films prepared by RF magnetron
sputtering,” J Eur Ceram Soc., vol 23, pp 1593–1598, 2003.
[8] K M Lakin, G R Kline, and K T McCarron, “High- Q acoustic
mi-crowave resonators and filters,” IEEE Trans Microw Theory Tech., vol.
41, no 12, pp 2139–2146, Dec 1993.
[9] M H Yim, D H Kim, D K Chai, and G W Yoon, “Significant res-onance characteristic improvements by combined thermal annealing
and Co electrode in ZnO-based FBARs,” Electron Lett., vol 39, pp.
1638–1640, Nov 2003.