The Effect of Silicon Nitride Stoichiometry on Charging Mechanisms ...

3518 IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES, VOL. 57, NO. 12, DECEMBER 2009
<strong>The</strong> <strong>Effect</strong> <strong>of</strong> <strong>Silicon</strong> <strong>Nitride</strong> <strong>Stoichiometry</strong>
on Charging Mechanisms in RF-MEMS
Capacitive Switches
Negar Tavassolian, Student Member, IEEE, M. Koutsoureli, E. Papandreou, Giorgos Papaioannou,
Benjamin Lacroix, Member, IEEE, Z. Liu, and John Papapolymerou, Senior Member, IEEE
Abstract—This paper discusses the mechanisms responsible for
charging <strong>of</strong> plasma enhanced chemical vapor deposition (PECVD)
silicon nitride films used in the fabrication <strong>of</strong> RF microelectromechanical
(MEMS) switches. <strong>Nitride</strong> films deposited at different
temperatures are characterized in order to better understand
the effect <strong>of</strong> deposition conditions on material stoichiometry and
stress. Both RF MEMS switches and metal–semiconductor–metal
capacitors with PECVD silicon nitride as the dielectric layer
were fabricated and their charging mechanisms were examined.
Measurements indicate that charging arises from the formation
<strong>of</strong> a defect band where charge transport occurs through
a Poole–Frenkel-like effect. <strong>The</strong> calculated activation energy
exhibits direct relation to material stoichiometry, and therefore to
the nitride bandgap. Finally, it is viewed that lower temperature
nitride is less prone to dielectric charging.
Index Terms—Dielectric materials, metal–semiconductor–metal
(MIM) capacitors, reliability RF microelectromechanical (MEMS)
switches, stoichiometry, stress.
I. INTRODUCTION
CAPACITIVE RF microelectromechanical (MEMS)
switches are among the most promising applications in
MEMS systems, being used especially at higher frequencies
(above 10 GHz). However, their commercialization is currently
hindered by reliability problems. <strong>The</strong> most important problem
is charging <strong>of</strong> the dielectric, causing unpredictable device
behavior [1]–[7]. <strong>The</strong> deposited dielectric films, such as silicon
nitride, contain a large density <strong>of</strong> traps associated with dangling
bonds [3], [4]. <strong>The</strong>se traps are amphoteric in nature, so they
can be negatively or positively charged. Under high electric
fields, charges can get injected into the dielectric and become
Manuscript received April 15, 2009; revised April 21, 2009. First published
November 13, 2009; current version published December 09, 2009. This work
was supported by the IMPACT Center for Advancement <strong>of</strong> MEMS/NEMS
VLSI, funded by DARPA, under Grant HR0011-06-1-0046, under the DARPA
N/MEMS Science and Technology Fundamentals Research Program.
N. Tavassolian, B. Lacroix, Z. Liu, and J. Papapolymerou are with
the School <strong>of</strong> Electrical and Computer Engineering, Georgia Institute <strong>of</strong>
Technology, Atlanta, GA 30332-0250 USA (e-mail: negar1@gatech.edu;
blacroix3@mail.gatech.edu; john.papapolymerou@ece.gatech.edu).
G. Papaioannou is with the School <strong>of</strong> Electrical and Computer Engineering,
Georgia Institute <strong>of</strong> Technology, Atlanta, GA 30332-0250 USA, and also with
the Physics Department, Solid State Physics Section, University <strong>of</strong> Athens,
Athens 15784, Greece (e-mail: gpapaioan@phys.uoa.gr).
M. Koutsoureli and E. Papandreou are with the Physics Department, Solid
State Physics Section, University <strong>of</strong> Athens, Athens 15784, Greece.
Color versions <strong>of</strong> one or more <strong>of</strong> the figures in this paper are available online
at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/TMTT.2009.2033865
0018-9480/$26.00 © 2009 IEEE
trapped in the bonds. Further, due to the insulating nature <strong>of</strong>
the silicon oxide or silicon nitride films, the recovery time can
take anywhere between a few seconds to several days or even
months.
Developing a good analytical model that would describe accumulating
<strong>of</strong> charges in the dielectric and their influence on
the device behavior would be the main step to achieving more
reliable switches. Presently, available models assume that dielectric
charging arises from charges distributed throughout the
dielectric material [5], the presence <strong>of</strong> charges at the dielectric
interface [8] and areas under compressive or tensile stress [9],
as well as injection <strong>of</strong> charges from the suspended bridge during
the on-state [10].
So far the effects <strong>of</strong> dielectric charging have been measured
by recording the transient current in permanently ON
switches [11], [12], the transient response <strong>of</strong> the ON capacitance
[13], [14], and the correlation <strong>of</strong> Poole–Frenkel (PF)
current intensity to the shift <strong>of</strong> the pull-out voltage [12]. <strong>The</strong>
simultaneous study <strong>of</strong> dielectric charging by recording the
capacitance-voltage characteristic in capacitive switches and
the thermally stimulated depolarization current (TSDC) in
metal–semiconductor–metal (MIM) structures with the same
dielectric has been recently proposed [15]. This method allows
for determination <strong>of</strong> both the main mechanisms contributing
to charging <strong>of</strong> the dielectric, which are charge injection and
dipole orientation. In spite <strong>of</strong> these research efforts, the exact
mechanism <strong>of</strong> charge injection that is responsible for dielectric
charging during the switch pull-down state has not yet been
entirely understood.
In thick dielectrics, the main electronic processes that lead to
dielectric charging are: 1) the trap-assisted-tunneling (TAT) <strong>of</strong>
electrons from the electrode to unoccupied defect or trap states
in the dielectric close to the electrode-dielectric interface and 2)
the time-dependent component <strong>of</strong> PF movement <strong>of</strong> the trapped
electrons from the defect levels to the conduction band <strong>of</strong> the
dielectric [16].
This paper elaborates more on the charging mechanisms <strong>of</strong>
silicon nitride films used in RF MEMS switches in order to identify
the most dominant one. In [17], the authors presented preliminary
results based on the the study <strong>of</strong> RF MEMS capacitive
switches with silicon nitride deposited at two different temperatures.
<strong>The</strong> charging investigation <strong>of</strong> the switches was based
on observing the dependence <strong>of</strong> capacitance-voltage characteristic
on both the bias sweep width and temperature. <strong>The</strong> latter
is used to determine the TAT activation energy <strong>of</strong> PF potential

TAVASSOLIAN et al.: EFFECT OF SILICON NITRIDE STOICHIOMETRY ON CHARGING MECHANISMS 3519
barrier in order to understand its relation to the dielectric material
stoichiometry. This paper expands the previously obtained
results by adding more data on both RF MEMS switches and
MIM capacitors, and by comparing their charging mechanisms.
Charging study on the MIM capacitors was done by applying
the TSDC method. In addition, more detailed information on the
stoichiometry <strong>of</strong> silicon nitride films deposited at various temperatures
was obtained using X-ray photoelectron spectroscopy
(XPS) and stress measurements. A relationship between the deposition
temperature and the silicon nitride stoichiometry was
derived for the first time. Consistent measurement results provide
a new insight into the charging mechanisms <strong>of</strong> silicon nitride
and its dependence on the material stoichiometry.
II. THEORETICAL BACKGROUND
A. Charging Mechanisms
In principle, TAT is determined by the spatial and energetic
distribution <strong>of</strong> trap states, the <strong>of</strong>fset between the Fermi energy
<strong>of</strong> the electrode and the conduction band <strong>of</strong> the dielectric, and
the lower limit <strong>of</strong> the tunneling relaxation times. <strong>The</strong> temperature
dependence <strong>of</strong> TAT is controlled primarily by the energy
distribution <strong>of</strong> the traps and the frequency dependence is controlled
by their spatial distribution. Traps distributed uniformly
in both energy and space result in an essentially temperature-independent
mechanism. <strong>The</strong> step-response currents for the tunnel
mechanism exhibit an inverse dependence on sample thickness
which becomes nonlinear at high applied electric fields, and the
detailed behavior is very sensitive to the trap distribution assumed.
For a particular trap distribution, it is possible to have a
polarity-dependent irreversible step response [13]–[16].
An expression was derived in [13] and [16] for the timedependent
current, assuming that TAT is the only operating
mechanism
where the current flows in the direction, is the sample thickness,
is the field-free trap distribution, is the energy,
is the electric field intensity, , and are the Fermi functions
defined as
and where , , , , and are the electronic
mass, Planck’s constant, Boltzmann’s constant, temperature,
and Fermi energy <strong>of</strong> the metal electrode, respectively.
<strong>The</strong> PF mechanism is driven by the electric field, which reduces
the barrier height on one side <strong>of</strong> the trap, thereby increasing
the probability <strong>of</strong> the electron escaping from the trap.
As the electric field increases, the potential barrier decreases on
(1)
Fig. 1. Typical capacitance-voltage characteristic <strong>of</strong> a capacitive RF MEMS
switch.
the right side <strong>of</strong> the trap, making it easier for the electron to
vacate the trap by thermal emission and enter the quasi-conduction
band <strong>of</strong> the host material. A first-order model for the current
due to the PF effect assumes that each trapping center is independent
<strong>of</strong> the other centers, i.e., the potentials do not overlap.
This assumption is valid only if the impurity density is low. Although
in amorphous materials the impurities density is usually
not very low, this model is generally used for the purpose <strong>of</strong>
simplicity. This procedure implicitly assumes just an effective
potential barrier .
<strong>The</strong> transient component <strong>of</strong> PF conduction was derived in
[13] and [16] assuming that a certain fraction <strong>of</strong> the trapped
electrons are lost to PF emission (thermal emission <strong>of</strong> electrons
from the trap states to the conduction band <strong>of</strong> the dielectric [15]).
<strong>The</strong> PF current density is given by a simple drift equation ,
where , and are the electronic mobility in the dielectric and
carrier density (per unit area <strong>of</strong> the dielectric) that participate
in the PF process respectively. Since there is a distribution <strong>of</strong>
trap states, the transient component <strong>of</strong> PF current density was
defined as
where is the high-frequency dielectric constant, and is
the density <strong>of</strong> trapped electrons at a depth from conduction
band. Here, it must be pointed out that at any given time the
PF process alters the trapped distribution <strong>of</strong> electrons, which in
turn alters the TAT process at subsequent times. Furthermore,
materials with higher dielectric constants will be less sensitive to
the field-induced trap barrier lowering inherent in the PF effect.
B. RF MEMs
<strong>The</strong> minimum <strong>of</strong> the capacitance-voltage characteristic
(Fig. 1) occurs when the electrostatic force is at its lowest
value. <strong>The</strong> bias value for the minimum capacitance
is proportional to the charge stored in dielectric and/or the
dielectric surface
(2)
(3)

3520 IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES, VOL. 57, NO. 12, DECEMBER 2009
Fig. 2. Top view <strong>of</strong> the capacitive RF MEMS switch used in this study.
where is the dielectric film thickness. Assuming the presence
<strong>of</strong> a background charge <strong>of</strong> introduced during fabrication
and/or handling, the total stored charge (at the top dielectric
surface) will be determined by the charge injected from the
contacting electrode and transported from the dielectric
surface. <strong>The</strong> charge polarity will be determined by the
polarity <strong>of</strong> the contacting electrode and can be written as
where is the time the switch is placed in the pull-down state.
Taking these into account, it is evident that the charging mechanism
can be determined by the shift <strong>of</strong> with either temperature
or the bias voltage in the pull-down state.
III. EXPERIMENTAL ANALYSIS
<strong>The</strong> switches utilized in the present experiment were fabricated
with a standard photolithographic process on high resistivity
silicon wafers k cm . <strong>The</strong> dielectric film is
silicon nitride deposited using plasma-enhanced chemical vapor
deposition (PECVD) method at 100 C–300 C in 100 C steps.
<strong>The</strong> stoichiometry <strong>of</strong> the dielectric film was determined using
XPS and its internal stress using both pr<strong>of</strong>ilometry and laser
methods. <strong>The</strong> dielectric is 200 nm thick in all cases. This thickness
is generally chosen as a regular dielectric thickness for RF
MEMS switches. <strong>The</strong> beam is an evaporated titanium–gold seed
layer electroplated to a thickness <strong>of</strong> m. Under no applied
force, the beam is suspended about m above the dielectric.
<strong>The</strong> sacrificial layer was removed with resist stripper and the
switches were dried using a CO critical point dryer. Fig. 2
shows the top view <strong>of</strong> the capacitive RF MEMS switch used
in this work.
Capacitance–voltage (C–V) characteristics were performed
on the RF MEMS switches. <strong>The</strong> capacitance was monitored
with a Boonton 72B capacitance meter while sweeping the
voltage in 0.5-V steps in the temperature range <strong>of</strong> 300 to 450 K.
MIM capacitors were fabricated in a similar way to the
MEMS switches. <strong>The</strong>ir charging mechanisms were then
compared with the MEMS switches in order to gain a better
understanding <strong>of</strong> the dielectric charging in silicon nitride. <strong>The</strong>
(4)
Fig. 3. Ni/S ratio for 200-nm silicon nitride on high-resistivity silicon. <strong>The</strong>
straight line was drawn to show the stoichiometry trend.
Fig. 4. Film stress for 200-nm silicon nitride deposited on silicon. <strong>The</strong> linear
fit was performed to show the average stress trend.
silicon nitride dielectric film is deposited using PECVD method
at 100 C–300 Cin50 C steps. <strong>The</strong> dielectric is 200 nm thick
in all cases. <strong>The</strong> charging process was investigated by applying
the TSDC method in the same manner as in [15]. This also
allows the calculation <strong>of</strong> stored charge.
<strong>The</strong> TSDC current was measured with a Keithley 6487
voltage source-picoampere meter in the temperature range <strong>of</strong>
200–500 K. <strong>The</strong> polarization bias was adjusted for each MIM
capacitor to ensure a constant electric field <strong>of</strong> 1 MV/cm for all
samples. Finally, the heating rate was K min.
IV. RESULTS AND DISCUSSIONS
A. <strong>Silicon</strong> <strong>Nitride</strong> Characterization and MIM Capacitors
Figs. 3 and 4 show the dependence <strong>of</strong> the nitride film stoichiometry
and stress on the deposition temperature. <strong>The</strong> increase
<strong>of</strong> deposition temperature clearly shows that the nitrogen
content increases. Taking into account that low temperature silicon
nitride is non homogeneous consisting <strong>of</strong> amorphous silicon
clusters [18] and various composition silicon nitride areas,
we conclude that by increasing the temperature the film average
bandgap also increases.
It is worth mentioning here that the results shown in Fig. 3
are consistent with the reported values in [17]. Less than 10%
variation is seen, which is within the tolerance range <strong>of</strong> the measurement
and fabrication process.

TAVASSOLIAN et al.: EFFECT OF SILICON NITRIDE STOICHIOMETRY ON CHARGING MECHANISMS 3521
Fig. 5. Dependence <strong>of</strong> TSDC spectra on film deposition temperature.
On the other hand, the increase <strong>of</strong> temperature leads to an increase
in the stress starting from compressive stress for films deposited
at 100 C to tensile when the deposition was performed
at 300 C [19]. <strong>The</strong> change <strong>of</strong> stress is expected to affect the
material electrical properties. This can be observed in the calculated
stored charge. <strong>The</strong> stored charge is calculated by integrating
the TSDC spectra obtained from MIM capacitors over
the whole temperature range. <strong>The</strong> TSDC spectra presented in
Fig. 5 show that in the low-temperature region, below 300 K,
the depolarization current is larger due to a larger amount <strong>of</strong>
stored charge when the nitride film deposition temperature increases.
In the high-temperature range, above 300 K, the TSDC
current increases as the temperature is increased with a practically
constant activation energy <strong>of</strong> about 0.22 eV, independent
<strong>of</strong> the film deposition temperature. <strong>The</strong> increase <strong>of</strong> stored charge
can be related to the increase <strong>of</strong> material nitrogen content hence
the bandgap. Here, we should point out that presently there is
no report on the dependence <strong>of</strong> silicon nitride charge trap properties
on the material stoichiometry and bandgap.
<strong>The</strong> TSDC spectrum allows the calculation <strong>of</strong> stored charge
as follows:
where is the capacitor area. This integration can be performed
since the temperature is varying linearly with time. <strong>The</strong> calculated
stored charge for the films deposited at 100 C, 150 C,
200 C, and 250 C showed an increasing trend with temperature.
This increasing trend is in good agreement with our previously
reported results [20] and also with [21].
As already mentioned, there is little known on the dependence
<strong>of</strong> the electrical properties <strong>of</strong> silicon nitride on stoichiometry
and stress and research work is presently in progress in this area.
B. RF MEMS Switches
Fig. 6 shows the dependence <strong>of</strong> the bias voltage at which the
minimum capacitance occurs on the maximum amplitude
<strong>of</strong> the applied bias for the capacitance-voltage
characteristic measurements. <strong>The</strong> data was obtained at 300 K
[Fig. 6(a)] and 340 K [Fig. 6(b)] from a switch with 200-nm
nitride film deposited at 250 C. <strong>The</strong> pull-in voltage was about
(5)
Fig. 6. Dependence <strong>of</strong> the bias voltage at which the minimum capacitance
occurs @† A on the maximum amplitude <strong>of</strong> the applied bias @† A for
(a) 300 K and (b) 340 K.
10 V. <strong>The</strong> switches were unpackaged and the measurement was
performed at 25% humidity conditions.
In order to determine the charging mechanism and simplify
the analysis, we assumed that the PF effect is dominant. <strong>The</strong>n
(3) and (4) result in
where , , , are fitting parameters and is the voltage
corresponding to . <strong>The</strong> pre-exponential parameter corresponds
to a maximum variation <strong>of</strong> bias shift due to charge
trapping in thermally activated traps and it is expected to relate
to the trap concentration through an equation analogous to (3).
<strong>The</strong> apparent excellent fit in Fig. 6 requires further investigation
since it is well known that the conductivity <strong>of</strong>
thin metal–semiconductor–metal structures containing hydrogenated
amorphous silicon alloys can be affected by the current
stressing conditions [22]. This phenomenon is attributed to the
formation <strong>of</strong> a metastable defect band through which carriers
can move by hopping between charged states in a manner similar
to the PF effect, the driving force for defect creation being
the energy released during hole-electron recombination in the
bulk <strong>of</strong> the device. For a small number <strong>of</strong> defects, a classical PF
behavior is observed over many orders <strong>of</strong> current. In the case
<strong>of</strong> disordered materials including those with carriers hopping
between neutral defect states in band tails, the charging process
(6)

3522 IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES, VOL. 57, NO. 12, DECEMBER 2009
Fig. 7. Temperature dependence <strong>of</strong> † in switches. <strong>The</strong> silicon nitride film
was deposited at (a) 150 C and (b) 250 C.
is expected to deviate from PF effect and is better represented
by (4), while macroscopically being better described by (6).
<strong>The</strong> main reason for observing a bigger change at 300 K compared
to 340 K in Fig. 6. is that trapping/detrapping kinetics increase
with temperature. <strong>The</strong> characteristic time is given by
which clearly indicates that during the OFF time, which is determined
by the C–V curve measurement process and not the
temperature, the detrapped charge is larger when experiments
take place at higher temperatures. This will lead to a smaller
change in excursion at higher temperatures.
<strong>The</strong> temperature dependence <strong>of</strong> , for V,
is presented in Fig. 7 for switches where the 200-nm nitride
film has been deposited at 150 C and 250 C. In both figures,
the fitting using (6) indicates thermally activated mechanisms
(the derived energy values represent the numerator <strong>of</strong>
the exponent). <strong>The</strong> activation energy is different in the 150 C
nitride which shows a smaller change over temperature.
This observed difference may be attributed to a change <strong>of</strong>
due to the different material electrical properties. Taking into
account that both the nitride stoichiometry and its hydrogen
content are affected by the deposition conditions, the former
was determined from XPS assessment which showed that in the
150 C nitride the stoichiometry is , while the stoichiometry
is N/Si in the 250 C one. This clearly shows
that the silicon content is larger in the low-temperature material
(7)
which leads to the conclusion that the bandgap in the 250 C
nitride is larger than in the 150 C one. Taking into account
the expected formation <strong>of</strong> metastable defect bands [22]–[24]
during high electric field stresses and the dependence <strong>of</strong> material
bandgap on the deposition method, it becomes clear that the
activation energy is related to the dielectric film bandgap. This
conclusion, although seeming to support the results obtained
from the PF effect and the low charging reported in low-temperature
silicon nitride MIM capacitors [25], requires further and
in depth investigation for better understanding <strong>of</strong> the charging
process and the predictability <strong>of</strong> MEMS capacitive switches
failure mechanism.
Finally, it is worth pointing out that the differences between
the 150 C and 250 C materials have to be attributed to different
trap concentrations, which increase for the higher temperature
materials as obviously obtained from the integral <strong>of</strong> TSDC over
temperature. <strong>The</strong> different activation energies (Fig. 7) may also
be attributed to different materials bandgap and band potential
fluctuation.
Since to the best <strong>of</strong> our knowledge there is currently no clear
information on the dependence <strong>of</strong> these parameters on silicon
nitride deposition conditions, this is the first work on understanding
these properties <strong>of</strong> silicon nitride.
C. <strong>Effect</strong> <strong>of</strong> Humidity
This paper presents results obtained in atmosphere with 25%
humidity conditions. Although recent experiments have demonstrated
that dielectric charging can be aggravated by humidity
[26], the way it affects dielectric charging has not yet been fully
clarified. Recently, the effect <strong>of</strong> humidity on dielectric charging
was investigated with the Kelvin Force microscopy method [27].
It was demonstrated that injected charge distribution is narrower
in dry nitrogen ambient and wider in normal atmosphere. It was
also found that the decay time constant (which depends on the
insulator bulk properties since the charges are collected by the
bottom electrode) increases by almost four orders <strong>of</strong> magnitude
in nitrogen ambient. Finally, it must be pointed out that no lateral
diffusion was observed in any <strong>of</strong> these experiments.
Regarding the present work, it is clear that the experimental
results in MIM capacitors are not affected by humidity. In the
case <strong>of</strong> MEMS switches, although humidity is expected to affect
dielectric charging, it is still not possible to correlate the effect <strong>of</strong>
humidity with the results that arise from different stoichiometry
and deposition conditions.
V. CONCLUSION
<strong>The</strong> mechanisms responsible for dielectric charging <strong>of</strong> RF
MEMS capacitive switches have been investigated. <strong>The</strong> investigation
includes films <strong>of</strong> silicon nitride deposited under different
substrate temperature conditions. <strong>The</strong> stoichiometry <strong>of</strong>
each silicon nitride film was carefully determined by XPS measurements.
Stress values were also measured for all films using
both pr<strong>of</strong>ilometry and laser methods. <strong>The</strong>se measurements allowed
for obtaining a direct relationship between the deposition
temperature and material properties (i.e., stress and stoichiometry).
It is seen that lower temperature silicon nitride has less
stress and is richer in the silicon component, which results in an
increase <strong>of</strong> the average bandgap with a temperature increase.

TAVASSOLIAN et al.: EFFECT OF SILICON NITRIDE STOICHIOMETRY ON CHARGING MECHANISMS 3523
Both RF MEMS switches and MIM capacitors were fabricated
using the characterized silicon nitride films. <strong>The</strong>ir
charging process was monitored and compared in order to
gain a better understanding <strong>of</strong> the mechanisms responsible for
charging. <strong>The</strong> comparison <strong>of</strong> the calculated activation energies
obtained from RF MEMS measurements clearly indicates that
the lower temperature PECVD nitride is less prone to dielectric
charging than the higher temperature ones. TSDC spectra <strong>of</strong> the
MIM capacitors at different silicon nitride temperatures show
that the differences between these materials arise from different
trap concentrations, which increase for the higher temperature
materials. Finally it is seen that charging in silicon nitride is
caused by the formation <strong>of</strong> a metastable defect band where
carriers can move by hopping between charged states in a way
similar to the PF effect. <strong>The</strong> formation <strong>of</strong> this band seems to
contribute to the dielectric charging <strong>of</strong> silicon nitride.
REFERENCES
[1] C. L. Goldsmith, J. Ehmke, A. Malczewski, B. Pillans, S. Eshelman,
Z. Yao, J. Brank, and M. Eberly, “Lifetime characterization <strong>of</strong> capacitive
RF MEMS switches,” in IEEE MTT-S Int. Microw. Symp. Dig.,
Phoenix, AZ, Jun. 2001, pp. 227–230.
[2] J. Wibbeler, G. Pfeifer, and M. Hietschold, “Parasitic charging <strong>of</strong>
dielectric surfaces in capacitive microelectromechanical systems
(MEMS),” Sens. Actuators A, vol. 71, no. 1–2, pp. 74–80, Nov. 1998.
[3] X. Yuan, S. Cherepko, J. Hwang, C. L. Goldsmith, C. Nordquist, and
C. Dyck, “Initial observation and analysis <strong>of</strong> dielectric-charging effects
on RF MEMS capacitive switches,” in IEEE MTT-S Int. Microw. Symp,
Dig., Fort Worth, TX, Jun. 2004, pp. 1943–1946.
[4] D. T. Krick, P. M. Lenahan, and J. Kanicki, “Electrically active point
defects in amorphous silicon nitride: An illumination and charge injection
study,” J. Appl. Phys., vol. 64, no. 7, pp. 3558–3563, Oct. 1988.
[5] G. M. Rebeiz, RF MEMS <strong>The</strong>ory, Design and Technology. Hoboken,
NJ: Wiley , 2003.
[6] J. DeNatale, R. Mihailovich, and J. Waldrop, “Techniques for reliability
analysis <strong>of</strong> RF MEMS switch,” in Proc. IEEE Int. Rel. Phys. Symp.,
Dallas, TX, 2002, pp. 116–117.
[7] W. M. van Spengen, R. Puers, R. Mertens, and I. de Wolf, “A comprehensive
model to predict the charging and reliability <strong>of</strong> capacitive
RF MEMS switches,” J. Micromech. Microeng., vol. 14, no. 4, pp.
514–521, Apr. 2004.
[8] P. Czarnecki, X. Rottenberg, P. Soussan, P. Ekkels, P. Muller, P. Nolmans,
W. De Raedt, H. A. C. Tilmans, R. Puers, L. Marchand, and I.
De Wolf, “Influence <strong>of</strong> the substrate on the lifetime <strong>of</strong> capacitive RF
MEMS switches,” in Proc. 2008 MEMS Conf., Tucson, AZ, Jan. 2008,
pp. 172–175.
[9] E. Papandreou, M. Lamhamdi, C. Skoulikidou, P. Pons, G. Papaioannou,
and R. Plana, “Structure dependent charging process in RF
MEMS capacitive switches,” Microelectron. Rel., vol. 47, no. 9–11,
pp. 1822–1827, Oct. 2007.
[10] X. Rottenberg, B. Nauwelaers, W. De Raedt, and H. A. C. Tilmans,
“Distributed dielectric charging and its impact on RF MEMS devices,”
in Proc. 12th GAAS Symp., Amsterdam, <strong>The</strong> Netherlands, Oct. 2004,
pp. 475–478.
[11] X. Yuan, J. C. M. Hwang, D. Forehand, and C. L. Goldsmith, “Modeling
and characterization <strong>of</strong> dielectric-charging effects in RF MEMS
capacitive switches,” in IEEE MTT-S Int. Microw. Symp. Dig., Long
Beach, CA, Jun. 2005, pp. 753–756.
[12] S. Melle, D. De Conto, L. Mazenq, D. Dubuc, B. Poussard, C. Bordas,
K. Grenier, L. Bary, O. Vendier, J. L. Muraro, J. L. Cazaux, and R.
Plana, “Failure predictive model <strong>of</strong> capacitive RF-MEMS,” Microelectron.
Rel., vol. 45, no. 9–11, pp. 1770–1775, Sep.–Nov. 2005.
[13] G. J. Papaioannou, M. Exarchos, V. <strong>The</strong>onas, G. Wang, and J. Papapolymerou,
“Temperature study <strong>of</strong> the dielectric polarization effects <strong>of</strong>
capacitive RF MEMS switches,” IEEE Trans. Microw. <strong>The</strong>ory Tech.,
vol. 53, no. 11, pp. 3467–3473, Nov. 2005.
[14] G. J. Papaioannou, M. Exarchos, V. <strong>The</strong>onas, J. Psychias, G. Konstantinidis,
D. Vasilache, A. Muller, and D. Neculoiu, “<strong>Effect</strong> <strong>of</strong> space
charge polarization in radio frequency microelectromechanical system
capacitive switch dielectric charging,” Appi. Phys. Lett., vol. 89, no. 10,
pp. 103512–4, Sep. 2006.
[15] R. Daigler, G. Papaioannou, E. Papandreou, and J. Papapolymerou,
“<strong>Effect</strong> <strong>of</strong> dielectric film thickness on dielectric charging <strong>of</strong> RF MEMS
capacitive switches,” in IEEE MTT-S Int. Microw. Symp. Dig, Jun.
2008, pp. 1275–1278.
[16] R. Ramprasad, “Phenomenological theory to model leakage currents
in metal-insulator–metal capacitor systems,” Physica Status Solidi (b),
vol. 239, no. 1, pp. 59–70, Sep. 2003.
[17] G. Papaioannou, N. Tavassolian, M. Koutsoureli, E. Papandreou, and J.
Papapolymerou, “Investigation <strong>of</strong> charging mechanisms in RF-MEMS
capacitive switches with silicon nitride: <strong>The</strong> effect <strong>of</strong> material stoichiometry,”
in IEEE MTT-S Int. Microw. Symp. Dig., Boston, MA, Jun.
2009, pp. 1653–1656.
[18] A. Zerga, M. Carrada, M. Amann, and A. Slaoui, “Si-nanostructures
formation in amorphous silicon nitride sinx:H deposited by remote
PECVD,” Physica E, vol. 38, no. 1–2, pp. 21–26, Apr. 2007.
[19] M. Martyniuk, J. Antoszewski, C. A. Musca, J. M. Dell, and L. Faraone,
“Environmental stability and cryogenic thermal cycling <strong>of</strong> low-temperature
plasma-deposited silicon nitride thin films,” J. Appl. Phys., vol.
99, no. 5, p. 053519, Mar. 2006.
[20] R. Daigler, E. Papandreou, N. Tavassolian, M. Koutsoureli, G. Papaioannou,
and I. Papapolymerou, “Dependence <strong>of</strong> dielectric charging
on film thickness and deposition conditions,” in Proc. Asia–Pacific
Microw. Conf., Dec. 2008, pp. 1–4.
[21] Z. Zhen Peng, Personal Communication, Semiconductor Technol.
Lab., Lehigh Univ., Bethlehem, PA.
[22] S. P. Lau, J. M. Shannon, and B. J. Sealy, “Changes in the
Poole–Frenkel coefficient with current induced defect band conductivity
<strong>of</strong> hydrogenated amorphous silicon nitride,” J. Non-Crystalline
Solids, vol. 221–230, pt. 1, pp. 533–537, May 1998.
[23] J. M. Shannon and B. A. Morgan, “Hole transport via dangling-bond
states in amorphous hydrogenated silicon nitride,” J. Appl. Phys., vol.
86, no. 3, pp. 1548–1550, Aug. 1999.
[24] J. M. Shannon, S. C. Deane, B. McGarvey, and J. N. Sandoe, “Current
induced drift mechanism in amorphous SiN :H thin film diodes,” Appl.
Phys. Lett., vol. 65, no. 23, pp. 2978–2980, Dec. 1994.
[25] R. Daigler, G. Papaioannou, E. Papandreou, and J. Papapolymerou,
“Dielectric charging in low temperature silicon nitride for RF-MEMS
capacitive switches,” in Proc. MRS Symp. J: Passive and Electromech.
Mater. Integration, San Francisco, CA, Apr. 2008.
[26] Z. Peng, C. Palego, J. C. M. Hwang, D. I. Forehand, C. L. Goldsmith,
C. Moody, A. Malczewski, B. W. Pillans, R. Daigler, and J. Papapolymerou,
“Impact <strong>of</strong> humidity on dielectric charging in RF MEMS capacitive
switches,” IEEE Microw. Wireless Compon. Lett., vol. 19, no.
5, pp. 299–301, May 2009.
[27] U. Z. Heiba, G. Papaioannou, A. Belarni, H. Wang, F. Coccetti, P. Pons,
and R. Plana, “<strong>Effect</strong> <strong>of</strong> humidity on dielectric charging process in capacitive
RF MEMS switches based on Kelvin force microscopy surface
potential measurements,” in Proc. 2009 MRS Fall Meeting, DD: Microelectromech.
Syst.—Mater. Devices III..
Negar Tavassolian (S’04) received the B.Sc. degree
in electrical engineering from Sharif University <strong>of</strong>
Technology, Tehran, Iran, in 2003 and the M.S.
degree (with honors) in electrical engineering from
McGill University, Montreal, QC, Canada, in 2006.
She is currently working towards the Ph.D. degree at
the Georgia Institute <strong>of</strong> Technology, Atlanta.
Her current research is focused on dielectric
charging <strong>of</strong> capacitive RF MEMS switches and
reliability <strong>of</strong> RF MEMS.
Ms. Tavassolian is a member <strong>of</strong> the IEEE
Microwave <strong>The</strong>ory and Techniques Society and a reviewer for the IEEE
TRANSACTIONS ON ANTENNAS AND PROPAGATION, the IEEE MICROWAVE AND
WIRELESS COMPONENTS LETTERS, PIER, and JEMWA journals.
M. Koutsoureli, photograph and biography not available at the time <strong>of</strong>
publication.
E. Papandreou, photograph and biography not available at the time <strong>of</strong>
publication.

3524 IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES, VOL. 57, NO. 12, DECEMBER 2009
Giorgos Papaioannou received the B.Sc. degree
in physics from the University <strong>of</strong> Athens, Athens,
Greece, the M.Sc. degree from University College
London, London, U.K., and the Ph.D. degree in
solid-state physics from the University <strong>of</strong> Athens.
In 1975, he joined the Physics Laboratory, Physics
Department, University <strong>of</strong> Athens, as a Research Assistant,
and in 1978 as a Lecturer conducting research
in III–V compound semiconductors, mainly in semiinsulating
GaAs. In 1982, he joined the Solid State
Physics Section, University <strong>of</strong> Athens, where, beginning
in 1982, he was a Senior Lecturer. He is currently an Associate Pr<strong>of</strong>essor
with the Solid State Physics Section, Athens University, where he leads a team
on the transport properties and the radiation effects in compound semiconductors
and III–V semiconductor devices and the interaction <strong>of</strong> light in III–V compound
semiconductor and silicon-on-insulator (SOI) structures and devices. He
is recently involved with the investigation <strong>of</strong> polarization effects in insulating
materials in (RF) microelectromechanical (MEMS) switches. He has authored
or coauthored 60 publications and over 90 conference presentations.
Benjamin Lacroix (M’08) received the Ph.D. degree
in electrical and telecommunications engineering
from the University <strong>of</strong> Limoges, Limoges, France,
in 2008.
As a Ph.D. student with the XLIM Research
Institute, he developed fast miniature (RF) microelectromechanical
(MEMS) switched capacitors
and high-speed reconfigurable low-loss distributed
MEMS transmission lines (DMTL) phase shifters.
He joined the MiRCTech Research Group, the
Georgia Institute <strong>of</strong> Technology, Atlanta, in December
2008, as a Postdoctoral Fellow. His current research is focused on
reconfigurable microwave filters using ferroelectric BST capacitors. He also
works on RF MEMS tunable devices and reliability <strong>of</strong> RF MEMS switches.
Z. Liu, photograph and biography not available at the time <strong>of</strong> publication.
John Papapolymerou (S’90–M99–SM’04) received
the B.S.E.E. degree from the National Technical
University <strong>of</strong> Athens, Athens, Greece, in 1993 and
the M.S.E.E. and Ph.D. degrees from <strong>The</strong> University
<strong>of</strong> Michigan at Ann Arbor, in 1994 and 1999,
respectively.
From 1999 to 2001, he was an Assistant Pr<strong>of</strong>essor
at the Department <strong>of</strong> Electrical and Computer Engineering,
University <strong>of</strong> Arizona, Tucson, and during
the summers <strong>of</strong> 2000 and 2003 he was a Visiting
Pr<strong>of</strong>essor at the University <strong>of</strong> Limoges, Limoges,
France. From 2001 to 2005 and 2005 to 2009, he was an Assistant and Associate
Pr<strong>of</strong>essor, respectively, at the School <strong>of</strong> Electrical and Computer Engineering,
Georgia Institute <strong>of</strong> Technology, Atlanta, where he is currently a Pr<strong>of</strong>essor. He
has authored or coauthored over 250 publications in peer-reviewed journals
and conferences. His research interests include the implementation <strong>of</strong> micromachining
techniques and microelectromechanical systems (MEMS) devices
in microwave, millimeter-wave and THz circuits and the development <strong>of</strong> both
passive and active planar circuits on semiconductor (Si/SiGe, GaAs) and organic
substrates (liquid crystal polymer-LCP, LTCC) for system-on-a-chip (SOC)/
system-on-a-package (SOP) radio frequency (RF) front ends.
Dr. Papapolymerou is the Chair for Commission D <strong>of</strong> the U.S. National Committee
<strong>of</strong> URSI. He is currently an Associate Editor for the IEEE TRANSACTIONS
ON ANTENNAS AND PROPAGATION. He served as an Associate Editor for the
IEEE MICROWAVE AND WIRELESS COMPONENTS LETTERS from 2006 to 2008.
During 2004, he was the Chair <strong>of</strong> the IEEE MTT/AP Atlanta Chapter. He was
the recipient <strong>of</strong> the 2009 IEEE MTT-S Outstanding Young Engineer Award,
the 2009 Georgia Tech School <strong>of</strong> ECE Outstanding Junior Faculty Award, the
2004 Army Research Office (ARO) Young Investigator Award, the 2002 National
Science Foundation (NSF) CAREER award, the Best Paper Award at
the Third IEEE International Conference on Microwave and Millimeter-Wave
Technology (ICMMT2002), Beijing, China, and the 1997 Outstanding Graduate
Student Instructional Assistant Award presented by the American Society
for Engineering Education (ASEE), <strong>The</strong> University <strong>of</strong> Michigan Chapter. His
students have also been recipients <strong>of</strong> several awards including the Best Student
Paper Award presented at the 2004 IEEE Topical Meeting on <strong>Silicon</strong> Monolithic
Integrated Circuits in RF Systems, the 2007 IEEE MTT-S Graduate Fellowship,
and the 2007/2008 and 2008/2009 IEEE MTT-S Undergraduate Scholarship/Fellowship.