electronics

Article
Development of an Advanced TDDB Analysis Model
for Temperature Dependency
Kiron Park, Keonho Park, Sujin Im, SeungEui Hong, Kwonjoo Son and Jongwook Jeon *
Department of Electrical and Electronics Engineering, Konkuk University, Seoul 143-701, Korea
* Correspondence: jwjeon@konkuk.ac.kr; Tel.: +82-2-450-3494




Received: 5 August 2019; Accepted: 23 August 2019; Published: 27 August 2019


Abstract: This paper proposes a hybrid model to describe the temperature dependence of the
time-dependent dielectric breakdown (TDDB) phenomenon. TDDB can be expressed in terms of two
well-known representative degradation mechanisms: The thermo-chemical (TC) mechanism and the
anode hole injection (AHI) mechanism. A single model does not account for the measured lifetime,
due to TDDB under different temperature conditions. Hence, in the proposed model, two different
degradation mechanisms are considered simultaneously in an appropriate manner to describe the
trap generation in the dielectric layer. The proposed model can be used to simulate the generation of
the percolation path in a dielectric layer, and it is in agreement with the measured lifetime because of
TDDB at different temperatures. Therefore, the proposed model can be used to predict guarantee
time or initial failure detection, using the accelerated life test for industrial purposes.

Keywords: reliability; time-dependent dielectric breakdown (TDDB); degradation; oxide dielectric;

dielectric breakdown

1. Introduction
The time-dependent dielectric breakdown (TDDB) phenomenon is a major issue affecting the
reliability of semiconductor devices, often limiting their high voltage applications. In particular,
the reduction of lifetime in high voltage devices such as power semiconductors, as a result of
TDDB, is crucial to device development [1]. Accordingly, many theories have been proposed for
interpreting TDDB. Previously published research confirms that the effect of TDDB varies with respect
to semiconductor device structure, process variation and physical degradation mechanism [2]. The most
widely used and prominent concepts among the physical theories of TDDB are the thermo-chemical
(TC) model, anode hole injection (AHI) model, and hydrogen release (HR) model. Each TDDB model
has the following characteristics: (i) TC model for analyzing the trap with bond breakage; (ii) an HR
model, where the released hydrogen from the Si–H bond manifests itself at the interface of Si/SiO2
and acts as a trap in the oxide; and (iii) the AHI model, where hot holes are generated by impact
ionization, creating defects in the oxide. These three theories of TDDB have been designed differently,
depending on the structure of the device, the electrical condition, and the oxide material. In other
words, the TDDB models are highly influenced by the device environment elements, which differ with
the semiconductor material. Based on this theoretical background, there have been many attempts
to explain the measured device lifetime in presence of the TDDB phenomenon and to predict their
lifetime in different operating conditions, such as temperature, bias, and stress time.
A single TDDB model cannot adequately describe the temperature dependence; there is a
discrepancy between the predicted lifetime and the actual measured data at high temperatures. Hence,
the temperature-dependence, in the high temperature range, is described using a mixture of several
TDDB models. This is because, in the high temperature region, more than one TDDB model becomes
dominant and trap generations, by each TDDB mechanism becomes active: (i) in TC model, the trap

Electronics 2019, 8, 942; doi:10.3390/electronics8090942 www.mdpi.com/journal/electronics

 Electronics 2019, 8, x FOR PEER REVIEW 2 of 8

becomes
Electronics dominant
2019, 8, 942 and trap
generations, by each TDDB mechanism becomes active: (i) in TC model,2 of 8
the trap generation increases because the bond breakage rate is increased by the thermal energy; (ii)
in AHI mode, hot holes, generated by impact ionization, drift into the oxide increasing trap
generation increases because the bond breakage rate is increased by the thermal energy; (ii) in AHI
generation rate, due to thermal energy. Thus, there is a limitation in estimating the lifetime using a
mode, hot holes, generated by impact ionization, drift into the oxide increasing trap generation rate,
single traditional model of TDDB. Therefore, in this paper, we propose a hybrid TDDB model, in
due to thermal energy. Thus, there is a limitation in estimating the lifetime using a single traditional
which two mechanisms operate simultaneously.
model of TDDB. Therefore, in this paper, we propose a hybrid TDDB model, in which two mechanisms
Additionally, the hybrid TDDB model, proposed in this paper, is used to simulate the
operate simultaneously.
percolation path of oxide, using a cell-based analytical method, and to analyze the lifetime by using
Additionally, the hybrid TDDB model, proposed in this paper, is used to simulate the percolation
a lifetime prediction model incorporating temperature variables. The results of this model were
path of oxide, using a cell-based analytical method, and to analyze the lifetime by using a lifetime
compared with measured TDDB data, both under low and high temperatures, and were found to
prediction model incorporating temperature variables. The results of this model were compared with
closely agree. Thus, the developed model is useful in performing early failure detection by using an
measured TDDB data, both under low and high temperatures, and were found to closely agree. Thus,
accelerated lifetime test and in providing long-term quality assurance [3].
the developed model is useful in performing early failure detection by using an accelerated lifetime
test and in providing
2. TDDB long-term quality assurance [3].
Hybrid Model
2. TDDB Hybrid
In this Model
paper, we propose a TDDB hybrid model in which the TC and AHI models operate
simultaneously. Figure 1 shows the behavior model that is changed by oxide thickness (𝑇 ) and oxide
In this paper, we propose a TDDB hybrid model in which the TC and AHI models operate
electric field (𝐸 ), when predicting lifetime, using a single TDDB model. Among the three TDDB
simultaneously. Figure 1 shows the behavior model that is changed by oxide thickness (Tox ) and oxide
models (TC/AHI/HR), the HR model is the proposed model for hyper-thin oxide with 𝑇 less than
electric field (Eox ), when predicting lifetime, using a single TDDB model. Among the three TDDB
5nm as the TDDB effect on thick oxide with 𝑇 > 5 nm can be negligible. In addition, the HR model
models (TC/AHI/HR), the HR model is the proposed model for hyper-thin oxide with Tox less than
operates when the energy of the injected electron from the substrate (𝐸 ) is 3 eV or more, the
5nm as the TDDB effect on thick oxide with Tox > 5 nm can be negligible. In addition, the HR model
threshold energy (𝐸 ) of the HR model. The threshold energy of the AHI model, which is affected
operates when the energy of the injected electron from the substrate (EIN ) is 3 eV or more, the threshold
by hot holes generated from impact ionization, is 6 eV [4]. This is the highest impact ionization among
energy (ETH ) of the HR model. The threshold energy of the AHI model, which is affected by hot
the three TDDB models (i.e., it operates mainly in highest electric fields). In the TC model, the TDDB
holes generated from impact ionization, is 6 eV [4]. This is the highest impact ionization among the
phenomenon is interpreted as a trap, due to breakage of oxide bonding; therefore, the threshold
three TDDB models (i.e., it operates mainly in highest electric fields). In the TC model, the TDDB
energy is extremely low (i.e., it operates mainly in lowest electric fields). The red line in Figure 1
phenomenon is interpreted as a trap, due to breakage of oxide bonding; therefore, the threshold energy
means the threshold energy of the HR model (𝐸 ), according to the 𝑇 , and the black line means
is extremely low (i.e., it operates mainly in lowest, electric fields). The red line in Figure 1 means
the threshold energy of the AHI model (𝐸 , ).
the threshold energy of the HR model (ETH,HR ), according to the Tox , and the black line means the
threshold energy of the AHI model (ETH,AHI ).

Figure
Figure Time-dependent
1. 1. Time-dependentdielectric breakdown
dielectric (TDDB)
breakdown model
(TDDB) which
model changes
which with oxide
changes with electric field
oxide electric
(Efield
ox ) and oxide thickness (Tox
(𝐸 ) and oxide thickness ). (𝑇 ).

In summary, for thin-oxides with Tox < 5 nm, the behavior single TDDB model, which operates as
In summary, for thin-oxides with 𝑇 < 5 nm, the behavior single TDDB model, which operates
the electric field increases, predicts the device lifetime in the order of TC–HR–AHI model. For thick
as the electric field increases, predicts the device lifetime in the order of TC–HR–AHI model. For thick
oxides
oxideswith 𝑇 >>55nm,
withTox nm,the thebehavior
behaviorsingle
singleTDDB
TDDBmodel
modelpredicts
predictsthe thedevice
devicelifetime
lifetimeininorder
orderofofthe
the
TC–AHI model, since the influence of the HR model is negligible. In this paper, the target
TC–AHI model, since the influence of the HR model is negligible. In this paper, the target oxide was oxide was
a athick
thickoxide witha T
oxidewith Toxofof9.9
a ox 9.9nm.
nm.For
Forpower
powerdevices
devices(e.g.,
(e.g.,lateral
lateraldouble
doublediffused
diffusedMOSMOS(LDMOS),
(LDMOS),
etc.), the channel length is still longer, and thick oxides are often used [5]. This paper analyzed the
Electronics 2019, 8, 942 3 of 8

oxide reliability of these power devices. TC model operates under low field condition and AHI model
operates under high field condition. Thus, the modeling process in this study was performed with TC
modeling and AHI modeling. In this process, Nhybrid is obtained by summing the number of traps per
unit volume, due to the TC mechanism, denoted as NTC , and the number of traps caused by the AHI
mechanism, denoted as NAHI .
NTC is calculated from the TC model. Polarization, denoted as P, occurs when an external electric
field, Eox , is applied to a dielectric with polarity. The local electric field, Eloc , applied to the bonding of
the oxide because of this polarization, is calculated as follows [6],

1+χ
!
P/0 2+k
 
Eloc = Eox + = Eox = Eox (1)
3 3 3

where 0 is the permittivity of free space, χ is the electric susceptibility, and k is a dielectric constant.
With this local electric field, the activation energy, ∆H, of the bond breakage is defined by the
following equation,
∆H = ∆H0 − pe f f ·Eloc (2)

where ∆H0 is the field free activation energy and Pe f f is the effective dipole moment. According to the
TC model, the field-enhanced thermal bond-breakage (dNTC /dt) can be expressed by the first-order
reaction rate equation as,
dNTC /dt = −kbreak ·NTC (t) (3)

where NTC is a trap number per 1 cm3 under given conditions (time, temperature, and electric field)
and kbreak is the bond breakage rate. kbreak can be expressed using the Boltzmann probability of a
molecule in an insulator, due to its interaction with the surrounding lattice, by,

kbreak = v0 ·exp(−∆H/rkb T ) (4)

where v0 is a lattice vibration frequency (~1013 /sec), ∆H is the activation energy due to the local electric
field, kb is the Boltzmann constant, r is the fitting parameter, and T is the temperature. Therefore,
NTC can be obtained using (1)–(4).
In addition, NAHI is calculated as follows [4]: first, we calculate the injected current from the
substrate due to the Fowler–Nordheim (FN) tunneling current density equation by,

JFN = AEox 2 exp(−B/Eox ) (5)

where A and B are the FN tunneling coefficients. Then, quantum yield can be expressed as Υ [7]. Because
the current density of the generated holes is JH = ΥJFN , the charge amount of the generated holes is
JH × Aox × t (where Aox is the insulator area and t is the observation time). This value, when divided by
elementary charge, q, gives the number of generated holes. As the generated holes drift to the insulator,
they can generate defects. The defect generation probability, denoted as Pdefect , is a newly defined
fitting parameter. If the number of defects generated in the insulator is divided by the insulator volume
Vox , the expected value of defects per unit volume in the insulator can be defined as:

JH × Aox × t
NAHI = ( ·Pde f ect )/Vox . (6)
q

Then, the summation of the two values yields the number of traps per unit volume calculated
using the two models simultaneously, i.e., Nhybrid = NTC + NAHI .
Figure 2 describes the process of calculating oxide survival probability from the trap density
obtained in the above step, and extracting lifetime from it. First, the percolation path of the oxide is
calculated from the trap density and the survival probability of the oxide is obtained from it. Then,
the device lifetime can be obtained by interpreting the Weibull distribution, popularly used in plotting
Electronics 2019, 8, 942 4 of 8

semiconductor failure mechanisms, such as TDDB from the obtained probability. The detailed process
Electronics 2019, 8, x FOR PEER REVIEW
is as follows. 4 of 8

Figure
Figure 2. Description
2. Description of the
of the proposedTDDB
proposed TDDBhybrid
hybrid model
modelwith
withone
oneinput, oneone
input, output, and and
output, threethree
core core
modules for TDDB analysis.
modules for TDDB analysis.
Figure 2 describes the process of calculating oxide survival probability from the trap density
We used the cell-based analytical percolation model to analyze the lifetime of dielectric breakdown,
obtained in the above step, and extracting lifetime from it. First, the percolation path of the oxide is
which is caused by traps in the insulator, as described by the TDDB model. In the percolation model,
calculated from the trap density and the survival probability of the oxide is obtained from it. Then,
trapstheinside
devicethelifetime
insulator,can between
be obtained two byelectrodes,
interpretingare therandomly generated popularly
Weibull distribution, by stress used and formin a
conductive path between the electrodes, causing electrical breakdown
plotting semiconductor failure mechanisms, such as TDDB from the obtained probability. The [8]. In this paper, the oxide
survival probability
detailed process iswas calculated using, not only a simple single percolation path, but also a
as follows.
percolation Wepath usedconsidering
the cell-based the analytical
nine nearest cells. Theoretically,
percolation model to analyze this approach
the lifetime is more accurate in
of dielectric
breakdown,
computing which isprobability
the survival caused by [9]. traps in the insulator, as described by the TDDB model. In the
percolation
First, model, trapslayer
the insulating insidewas
the insulator,
divided into betweenunittwo electrodes,
cells. At thisare point,
randomly thegenerated
volume by of the
stress and form a conductive path between the electrodes,
3
unit cell is the cube of the lattice constant (a0 ). That is, the total oxide is divided causing electrical breakdown [8]. In thisinto N
paper, the oxide survival probability was calculated using, not only a simple single percolation path,
columns (N = (oxide area)/a0 2 ), and each column is divided into n unit cells(n = (oxide thickness)/a0 ).
but also a percolation path considering the nine nearest cells. Theoretically, this approach is more
If the Poisson distribution parameter, that was obtained in the previous step, is b = Nhybrid ×
accurate in computing
  the survival probability [9].
volume o f First,
unit cell a0 3 , the probability
the insulating of generating k-traps in the unit cell can be obtained by using the
 layer was divided into unit cells. At this point, the volume of the unit cell is
(k = 0, 1,( 𝑎2, 3,).. That ( 𝑁 = is a

PMF,the = k)of=thebk lattice
P(Xcube e−b /k!constant . .) of is,
thethe total oxide
Poisson is divided
distribution [10].into N columns
Poisson distribution
(𝑜𝑥𝑖𝑑𝑒 𝑎𝑟𝑒𝑎)/𝑎 ), and each column is divided into n unit cells(𝑛
discrete probability distribution that expresses the probability of a given number of events occurring= (𝑜𝑥𝑖𝑑𝑒 𝑡ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠)/𝑎 ). If the in
Poisson distribution parameter, that was obtained in the previous step, is 𝑏 = 𝑁 ×
a fixed interval of space. If there is at least one trap in the unit cell, the cell is determined as defective-
𝑣𝑜𝑙𝑢𝑚𝑒 𝑜𝑓 𝑢𝑛𝑖𝑡 𝑐𝑒𝑙𝑙(𝑎 ), the probability of generating k-traps in the unit cell can be obtained by
and P(de f ective cell) = 1 − P(no trap in the unit cell) = 1 − e−b . The failure rate of a cell is denoted as
using the PMF, 𝑃(𝑋 = 𝑘) = (𝑏 𝑒 )/𝑘! (𝑘 = 0,1,2,3, … ) of the Poisson distribution [10]. Poisson
λ. The probability
Fcell =distribution that a column becomes the percolation path is Fperc =ofλangiven
is a discrete probability distribution that expresses the probability
, and number
the probability
of
N
that the percolation
events occurringpath is notinterval
in a fixed generatedof space.in allIfNthere
columns is 1 −
is at least oneFperc
trap= in − λunit
(1the n ) .cell,
Then,
the the
cell Weibit
is
 
(WBDdetermined as defective-
) can be calculated as WBD and
= ln𝑃(𝑑𝑒𝑓𝑒𝑐𝑡𝑖𝑣𝑒
[−ln(1 − F𝑐𝑒𝑙𝑙) BD )] ==1 −
ln 𝑃(𝑛𝑜
( N ) 𝑡𝑟𝑎𝑝
+ ln 𝑖𝑛1 𝑡ℎ𝑒 𝑢𝑛𝑖𝑡 𝑐𝑒𝑙𝑙) = 1 − 𝑒
9 + nln ( 9λ ) [8,9]. . The
failure
The rate of adistribution,
Weibull cell is denoted as 𝐹 in =
used this𝜆. The
paper,probability
is a very thatwidely
a column becomes
used the percolation
reliability distribution,
path is 𝐹 = 𝜆
especially for plotting semiconductor failure mechanisms, such as TDDB [11]. AsNTDDB
, and the probability that the percolation path is not generated in all columns is at
occurs
1−𝐹 = (1 − 𝜆 ) . Then, the Weibit ( 𝑊 ) can be calculated as 𝑊 =𝑙𝑛 [− 𝑙𝑛 (1 − 𝐹 ) ] =
small localized regions (usually called a percolation region/path) degrading rapidly in the entire oxide
region,𝑙𝑛 (𝑁)
this +phenomenon
𝑙𝑛 + 𝑛𝑙𝑛(9𝜆) [8][9].
is well-fitted by the Weibull distribution. The Weibull probability density
The Weibull
function (PDF) is defineddistribution,
by [12] as,used in this paper, is a very widely used reliability distribution,
especially for plotting semiconductor failure mechanisms, such as TDDB [11]. As TDDB occurs at
β  1 β−1 region/path)
small localized regions (usually called a percolation
! "  β # degrading rapidly in the entire
t
(t) =
oxide region, this phenomenon is fwell-fitted by the exp − distribution. The Weibull probability
Weibull (7)
α α α
density function (PDF) is defined by [12] as,
where α is the characteristic time-to-failure (usually,
𝛽 1 𝑡 as α = t63% ), β is the shape (dispersion
expressed
𝑓(𝑡) =
or Weibull slope) parameter. In addition, the𝛼Weibull 𝑒𝑥𝑝 −
cumulative (7)
𝛼 𝛼 density function (CDF) is defined by:
"  β # as 𝛼 = 𝑡
where α is the characteristic time-to-failure (usually, expressed % ), 𝛽 is the shape
Z t
t
F(t) =
(dispersion or Weibull slope) parameter. Inf (addition,
t)dt = 1the
− exp − cumulative
Weibull . density function (CDF) (8)
0 α
is defined by:
 𝑡
𝐹(𝑡) = 𝑓(𝑡)𝑑𝑡 = 1 −𝑒𝑥𝑝 − . (8)
𝛼
Rearranging this CDF ( 𝐹(𝑡) ) and taking the appropriate logarithms, results in a quantity,
Electronics 2019, 8, x FOR PEER REVIEW 5 of 8
commonly known as the Weibit (𝑊 ):
Electronics 2019, 8, 942 5 of 8
𝑡
𝑊 = 𝑙𝑛[− 𝑙𝑛 (1 − 𝐹) ] = 𝛽[𝑙𝑛( 𝑡 )] . (9)
𝐹(𝑡) = 𝑓(𝑡)𝑑𝑡 = 1 −𝑒𝑥𝑝 − 𝛼 . (8)
Rearranging this CDF (F(t)) and taking the appropriate logarithms, 𝛼 results in a quantity, commonly
When F = 50% in the above equation, 𝑡 = (𝑡 % = 50% 𝑏𝑟𝑒𝑎𝑘𝑑𝑜𝑤𝑛 𝑡𝑖𝑚𝑒 𝑓𝑜𝑟 𝑜𝑥𝑖𝑑𝑒) can be
known
obtained
as the
Rearranging Weibit
from thethis
plot
(W
CDFofBD( 𝐹(𝑡)vs
):
Weibit ) and taking 3).
𝑡 (Figure theInappropriate
this process,logarithms, results in a quantity,
tthe Weibit plot should be linear. For
commonly known as the Weibit (𝑊 W
this, we used the ‘polyfit’ function in BD): = ln [ −ln ( 1 − F ) ]
Python’s NumPy library to α= β [ ln ( )] .
linearly approximate the bent Weibit (9)
plot. 𝑡
When F = 50% in the above𝑊equation, 𝑙𝑛 (1
= 𝑙𝑛[−tBD =− 𝐹) ]==50%
(t50% 𝛽[𝑙𝑛( )] .
breakdown time f or oxide) can be obtained
(9)
𝛼
from the plot of Weibit vs tBD (Figure 3). In this process, the Weibit plot should be linear. For this,
3. Results Fand Discussion 𝑡 = (𝑡
weWhen
used the=‘polyfit’
50% infunction
the above equation,
in Python’s NumPy % =to
library 50% 𝑏𝑟𝑒𝑎𝑘𝑑𝑜𝑤𝑛
linearly 𝑡𝑖𝑚𝑒 𝑓𝑜𝑟
approximate the𝑜𝑥𝑖𝑑𝑒) can be
bent Weibit plot.
obtained from the plot of Weibit vs 𝑡 (Figure 3). In this process, the Weibit plot should be linear. For
this, we used the ‘polyfit’ function in Python’s NumPy library to linearly approximate the bent Weibit
plot.

3. Results and Discussion

(a) (b)
Figure
Figure3.3.Weibit
Weibit(𝑊 ) )vsvs𝑡 tBDplot
(WBD plotobtained
obtainedfrom
fromthe
theproposed
proposedTDDB
TDDBhybrid
hybridmodel:
model:(a)
(a)Weibit
Weibitatatlow
low
temperature
temperature(298.15
(298.15K);
K);(b)(b)Weibit
Weibit at
at high
high temperature
temperature (398.15
(398.15 K).

3. Results
Figureand Discussion
3 shows the Weibit slope obtained from the proposed TDDB hybrid model. Figure 3a
showsFigure 3 shows (a)
the Weibit slope,
theatWeibit
a low slope
temperature
obtainedof 298.15
from theK, proposed
while FigureTDDB (b)shows
3b hybridthe WeibitFigure
model. slope at3a
ashows
high temperature
the Weibit of 398.15
slope, at a K. In
low the encircled
temperature of region,K,the
298.15 influence
while Figure of3bthe AHIthe
shows model, affected
Weibit slope bya
at
Figure 3. Weibit (𝑊 ) vs 𝑡 plot obtained from the proposed TDDB hybrid model: (a) Weibit at low
the
highFNtemperature
tunneling, increases
of 398.15 as
K.the electric
In the field increases.
encircled theThis
region,(398.15 indicates
influence thatAHI
of the the model,
Weibit tends
affected to bend.
by the
temperature (298.15 K); (b) Weibit at high temperature K).
The lifetime vs 𝐸 plots obtained from the proposed TDDB hybrid model
FN tunneling, increases as the electric field increases. This indicates that the Weibit tends to bend.are shown in Figure
4. The target
The
Figure structure
lifetime
3 shows Eoxis
vsthe a thick
plots
Weibit oxide
obtained
slope from 𝑇the
with
obtained of 9.9 nm
proposed
from the and
TDDB
proposed𝐴 hybrid
of 0.6 model
TDDB mm 2. The range of 𝐸
hybrid are shownFigure
model. is
3a9–4.
in Figure
12 MV/cm
The the
shows target at low
structure
Weibit temperature,
slope, is
at aa thick and 7–11 MV/cm,
oxide with of
low temperature at
Tox298.15 a high
of 9.9K,nm temperature.
Aox of3b
andFigure
while Under
0.6shows 2
mm .the both
The low
range
Weibit and high
of Eat
slope ox is
9–12temperature
a high MV/cm at low of temperature,
398.15 K. In theandencircled
7–11 MV/cm, at the
region, a high temperature.
influence Under
of the AHI both affected
model, low andby high
thetemperatures,
FN tunneling,there is a proper
increases fit withfield
as the electric the actual measured
increases. data [13].
This indicates Figure
that 4 also tends
the Weibit showstothat there
bend.
is The
a discrepancy 𝐸 plotsthe
lifetime vsbetween actual from
obtained measured data andTDDB
the proposed the estimated lifetime
hybrid model areifshown
the existing single
in Figure
AHItarget
4. The modelstructure
is used. isThis discrepancy
a thick oxide with can𝑇 beofsuccessfully 𝐴 of 0.6 by
9.9 nm and overcome mmusing
2. Theour of 𝐸 isTDDB
proposed
range 9–
12 hybrid
MV/cmmodel.
at low temperature, and 7–11 MV/cm, at a high temperature. Under both low and high

(a) (b)
Figure 4. Lifetime versus 𝐸 plot obtained from the proposed TDDB hybrid model and actual measured
data: (a) Lifetime at low temperature (298.15 K); (b) Lifetime at high temperature (398.15 K).

(a) (b)
Figure 4. Lifetime versus 𝐸 plot obtained from the proposed TDDB hybrid model and actual measured
Figure 4. Lifetime versus Eox plot obtained from the proposed TDDB hybrid model and actual measured
data: (a) Lifetime at low temperature (298.15 K); (b) Lifetime at high temperature (398.15 K).
data: (a) Lifetime at low temperature (298.15 K); (b) Lifetime at high temperature (398.15 K).
Electronics 2019, 8, 942 6 of 8

In addition, we analyzed the factors affecting the lifetime, as time passes in both the TC and AHI
mechanisms. Figure 5 shows the trap density of each TDDB mechanism under a stress condition that
guarantees a lifetime of 10 years, assuming 8 hours of operation per day. The Eox value guaranteeing
a 10-year lifetime is 7.81 MV/cm at low temperature and 5.54 MV/cm at high temperature. N0 is the
number of initial weak bonds in the TC model. Upon solving the differential equation in (3):

Electronics 2019, 8, x FOR PEER REVIEW NTC (t) = N0 (1 − exp(−kbreak t) ). 6(10)

of 8

(a) (b)
Figure 5. Trap
Trap density (NTC, ,𝑁NAHI
density (𝑁 ) vs) vs
time plot
time obtained
plot from
obtained thethe
from proposed TDDB
proposed TDDB hybrid model:
hybrid (a)
model:
Trap
(a) density
Trap at low
density temperature
at low temperature(298.15
(298.15K);K);
(b)(b)
Trap density
Trap at at
density high temperature
high (398.15
temperature K).K).
(398.15

N0 refers there
temperatures, to theismaximum
a proper fit value
withofthethe traps,
actual which can
measured databe[13].generated
Figure by theshows
4 also TC mechanism,
that there
i.e.,
is a discrepancy between the actual measured data and the estimated lifetime if theyears,
breakage of the bonds by polarization. To guarantee the same lifetime of 10 it can
existing be
single
operated
AHI model in aisrelatively
used. This higher Eox region,
discrepancy can under low temperature
be successfully overcome conditions.
by using As ourthe AHI model
proposed TDDB is
more dependent
hybrid model. on the field than the TC model, the point at which the saturation value of the TC
model In is exceeded
addition, weby NAHI atthe
analyzed ‘low temperature
factors affectingunder high Eas
the lifetime, ox ’time
condition
passes isin lower
both thethan
TCthe
and‘high
AHI
temperature
mechanisms.under Figurelow Eox ’ condition.
5 shows In other
the trap density ofwords, Figure
each TDDB 5 shows that
mechanism the atrap
under generation
stress conditionofthat
the
guarantees a lifetime of 10 years, assuming 8 hours of operation per day. The 𝐸 value guaranteeing
TC model has a higher temperature dependence than that of the AHI model. The trap generation of
the AHI model
a 10-year lifetimealsois has
7.81aMV/cm
relativelyat greater field dependence
low temperature and 5.54than MV/cm that of
at the
highTC model.
temperature. 𝑁 is the
number of initial weak bonds in the TC model. Upon solving the differential equation in (3): model.
Figure 6 is the lifetime versus temperature plot, obtained by the proposed TDDB hybrid
The temperature conditions for simulation were in the range of 25–175 ◦ C. The oxide electric field was
𝑁 (𝑡) = 𝑁 (1 −𝑒𝑥𝑝 (−𝑘 𝑡) ). (10)
varied in the range of 7.38–10 MV/cm. The lifetime predicted from the proposed hybrid model matches
𝑁 refers
well with the actual
to themeasured
maximumTDDB valuedata [13–15].
of the The dotted-line
traps, which of Figureby
can be generated 6 is
thetheTClifetime predicted
mechanism, i.e.,
by a single
breakage of TDDB
the bonds model. The single TDDB
by polarization. model the
To guarantee doessame
not lifetime
adequately of 10describe
years, itthecantemperature
be operated
dependence
in a relatively higherin𝐸 theregion,
of TDDB high temperature
under low range.
temperature conditions. As the AHI model is more
dependent on the field than the TC model, the point at which the saturation value of the TC model is
 𝑁 refers to the maximum value of the traps, which can be generated by the TC mechanism, i.e.,
breakage of the bonds by polarization. To guarantee the same lifetime of 10 years, it can be operated
in a relatively higher 𝐸 region, under low temperature conditions. As the AHI model is more
dependent on the field than the TC model, the point at which the saturation value of the TC model is
Electronics 2019, 8, 942 7 of 8

Figure Lifetime
6. 6.
Figure versus
Lifetime temperature
versus plot plot
temperature obtained from from
obtained the proposed TDDB TDDB
the proposed hybrid hybrid
model and actual
model and
measured data.
actual measured data.
4. Conclusions
A single TDDB model cannot correctly predict the lifetime at various temperature levels. To solve
this problem, we propose a hybrid model that allows two TDDB mechanisms (TC and AHI) to operate
simultaneously. We compared the proposed hybrid model with the actual measured data and confirm
that the lifetime is accurately predicted at both low and high temperatures.

Author Contributions: Methodology, Validation, Paper Writing, K.P. (Kiron Park), K.P. (Keonho Park), S.I., S.H.,
and K.S.; Conceptualization, Funding Acquisition, Investigation, Supervision, Writing-Review & Editing, J.J.
Funding: This research was supported by the MOTIE (Ministry of Trade, Industry & Energy (10085645) and KSRC
(Korea Semiconductor Research Consortium) support program for the development of future semiconductor
devices, and in part, by the Korea Electric Power Corporation. (Grant number: R18XA06-78).
Acknowledgments: The EDA tool was supported by the IC Design Education Center (IDEC), Korea.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Gurfinkel, M.; Horst, J.C.; Suehle, J.S.; Bernstein, J.B.; Shapira, Y.; Matocha, K.S.; Dunne, G.; Beaupre, R.A.
Time-Dependent Dielectric Breakdown of 4H-SiC/SiO2 MOS Capacitors. IEEE Trans. Device Mater. Reliab.
2008, 8, 635–641. [CrossRef]
2. Chen, F.; Bravo, O.; Chanda, K.; McLaughlin, P.; Sullivan, T.; Gill, J.; Lloyd, J.; Kontra, R.; Aitken, J.
A comprehensive study of low-k SiCOH TDDB phenomena and its reliability lifetime model development.
In Proceedings of the 2006 IEEE International Reliability Physics Symposium Proceedings, San Jose, CA,
USA, 26–30 March 2006.
3. McPherson, J.W. Time dependent dielectric breakdown physics–Models revisited. Microelectron. Reliab. 2012,
52, 1753–1760. [CrossRef]
4. Alam, M.A.; Bude, J.; Ghettiet, A. Field acceleration for oxide breakdown-can an accurate anode hole injection
model resolve the E vs. 1/E controversy? In Proceedings of the 2000 IEEE International Reliability Physics
Symposium Proceedings. 38th Annual (Cat. No.00CH37059), San Jose, CA, USA, USA, 10–13 April 2000.
5. Samanta, P.; Mandal, K.C. Hole injection and dielectric breakdown in 6H–SiC and 4H–SiC metal–oxide–
semiconductor structures during substrate electron injection via Fowler–Nordheim tunneling. Solid-State
Electron. 2015, 114, 60–68. [CrossRef]
6. McPherson, J.W.; Mogul, H.C. Underlying physics of the thermochemical E model in describing low-field
time-dependent dielectric breakdown in SiO2 thin films. J. Appl. Phys. 1998, 84, 1513–1523. [CrossRef]
7. Chang, C.; Hu, C.; Brodersen, R.W. Quantum yield of electron impact ionization in silicon. J. Appl. Phys.
1985, 57, 302–309. [CrossRef]
Electronics 2019, 8, 942 8 of 8

8. Suñe, J.; Jimenez, D.; Miranda, E. Breakdown modes and breakdown statistics of ultrathin SiO2 gate oxides.
In Oxide Reliability: A Summary of Silicon Oxide Wearout, Breakdown, and Reliability; Dumin, D.J., Ed.; World
Scientific: Singapore, 2002; pp. 173–232. [CrossRef]
9. Krishnan, A.T.; Nicollian, P.E. Analytic extension of thecell-based oxide breakdown model to full percolation
and its implications. In Proceedings of the IEEE 45th Annual International Reliability Physics Symposium
Proceedings, Phoenix, AZ, USA, 15–19 April 2007.
10. Stigler, S.M. Poisson on the Poisson distribution. Stat. Probab. Lett. 1982, 1, 33–35. [CrossRef]
11. Mcpherson, J.W. Reliability Physics and Engineering; Springer: New York, NY, USA, 2010.
12. Rinne, H. The Weibull distribution: A handbook, 1st ed.; CRS Press: New York, NY, USA, 2008.
13. Kimura, M. Field and temperature acceleration model for time-dependent dielectric breakdown. IEEE Trans.
Electron Devices 1999, 46, 220–229. [CrossRef]
14. Cheung, K.P. SiC power MOSFET gate oxide breakdown reliability—Current status. In Proceedings of the
2018 IEEE International Reliability Physics Symposium (IRPS), Burlingame, CA, USA, 11–15 March 2018.
15. Teramoto, A.; Umeda, H.; Azamawari, K.; Kobayashi, K.; Shiga, K.; Komori, J.; Ohno, Y.; Miyoshi, H. Study
of oxide breakdown under very low electric field. In Proceedings of the 1999 IEEE International Reliability
Physics Symposium Proceedings, San Diego, CA, USA, 23–25 March 1999.

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).