TCAD Analysis of a P-Type MOS Capacitor
* Digital Assignment 2
Chris Mathew
MTech VLSI Design
School of Electronics Engineering (SENSE)
25MVD0014
[email protected] Abstract—The electrical characteristics of a p-type trap density. The frequency of the AC excitation plays a
Metal–Oxide–Semiconductor Capacitor (MOSCAP) were crucial role:
investigated using Synopsys TCAD tools. A device with substrate
doping NA = 1016 cm−3 and a gate oxide thickness of 1 nm was • At low frequencies, minority carriers can respond to the
modeled. The threshold voltage was first calculated analytically gate modulation, and the inversion capacitance returns to
and then verified through simulation by examining the electron Cox .
density profiles under different gate biases. Capacitance–voltage • At high frequencies, minority carriers cannot respond,
(C–V) characteristics were obtained for both low-frequency
(1 Hz) and high-frequency (109 Hz) operation. The simulated and the capacitance saturates at the maximum depletion
results reveal a deviation from the expected low-frequency value.
response, which is explained in terms of minority carrier This frequency-dependent response makes C–V analysis an
dynamics and substrate doping. The study highlights the
influence of doping concentration on inversion-layer response
essential tool for device modeling, characterization, and
and frequency-dependent C–V behavior. technology optimization.
I. I NTRODUCTION II. T HEORETICAL BACKGROUND
The MOS capacitor (MOSCAP) is a fundamental test The MOS capacitor (MOSCAP) is a fundamental struc-
structure used to study the electrostatics of MOS devices. ture for understanding MOS device electrostatics. It consists
It consists of a gate electrode, a thin insulating oxide, and of a gate electrode, a thin insulating oxide, and a semi-
a semiconductor substrate. When a voltage is applied to the conductor substrate. The behavior of the device is governed
gate, the surface potential of the semiconductor is modulated, by Poisson’s equation and the continuity equations for
leading to three distinct modes of operation: electrons and holes, which describe how the electric potential
and carrier densities vary with applied gate voltage.
• Accumulation: Under negative gate bias for a p-type sub-
strate, majority carriers (holes) are attracted to the ox- For a p-type substrate, the three operating regions are:
ide–semiconductor interface. The measured capacitance accumulation, depletion, and inversion. The threshold voltage
approaches the oxide capacitance (Cox ). (Vth ) marks the onset of strong inversion, when the surface
• Depletion: As the gate voltage becomes positive, holes potential reaches twice the bulk Fermi potential and minority
are repelled, forming a depletion region with ionized carriers form a conductive channel.
acceptors. The depletion width increases with voltage, The threshold voltage can be estimated analytically using:
causing the capacitance to decrease. √
4ϵsi qNA ϕF
• Inversion: Beyond a critical voltage, known as the VT H = VF B + 2ϕF + (1)
Cox
threshold voltage (Vth ), the surface potential reaches
twice the bulk Fermi potential, and minority carriers where VF B is the flat-band voltage, ϕF is the bulk Fermi
(electrons) form an inversion layer. This strongly influ- potential, NA is the substrate doping, and Cox is the oxide
ences the capacitance depending on the measurement capacitance per unit area.
frequency. MOSCAP operation is also influenced by minority carrier
The resulting capacitance–voltage (C–V) characteristics dynamics, surface mobility, and quantum effects in ultra-
provide valuable information about substrate doping concen- thin oxides. These factors determine the frequency-dependent
tration, oxide thickness, flat-band voltage, and interface C–V response, which is key to extracting device parameters
such as doping concentration, oxide thickness, and interface
Vellore Institute of Technology, Vellore Campus. state density.
III. M ETHODOLOGY IV. R ESULTS AND D ISCUSSION
The MOSCAP was simulated using Synopsys Sentaurus
TCAD. The methodology involved three main steps: device
definition, physics modeling, and simulation.
A. Threshold Voltage Analysis
A. Device Definition
The two-dimensional MOSCAP structure was designed in 1) Analytical Calculation: Using Eq. 1 with NA =
Sentaurus Structure Editor (SDE). The device consists of a 1016 cm−3 , tox = 1 nm, and VF B = 0, the threshold voltage
TiN gate electrode, a 1 nm thick silicon dioxide (SiO2 ) gate was Vth ≈ 0.72 V (Calculations in Appendix).
oxide, and a uniformly boron-doped p-type silicon substrate
with a concentration of NA = 1016 cm−3 . Ohmic contacts 2) Simulated Threshold Voltage: The simulated threshold
were defined at both the gate and substrate to enable biasing. voltage was extracted from electron density profiles when
The cross-sectional schematic of the simulated device is surface electron density equaled bulk hole density (NA ) at
shown in Fig. 1. VG ≈ 0.65 V. Representative electron density profiles are
shown in Figs. 2 and 3.
Fig. 1: Cross-sectional schematic of the simulated MOS ca-
pacitor.
Fig. 2: Electron density profiles at gate voltages below, near,
and above threshold.
B. Physics Modeling
The simulation included the following physics models:
• Carrier transport: Poisson and continuity equations for
electrons and holes.
• Mobility: High-field saturation and Lombardi surface
mobility.
• Recombination: Shockley–Read–Hall (SRH) and Auger
processes.
• Quantum effects: eBarrier tunneling and bandgap nar-
rowing in thin oxides.
C. Simulation Procedure
1) A DC sweep of the gate voltage was performed to extract
electron density profiles and determine the threshold
voltage.
2) AC small-signal simulations were conducted at low Fig. 3: Electron density profile at gate voltage = 0.65 V.
(1 Hz) and high (109 Hz) frequencies to extract the
capacitance–voltage characteristics.
3) Plots of electron/hole density, potential, and capaci-
B. Capacitance–Voltage (C–V) Characteristics
tance were generated for analysis.
This approach allows the study of threshold voltage,
inversion-layer formation, and frequency-dependent C–V The simulated C–V curves for low and high frequencies are
behavior in a moderately doped p-type substrate. in Fig. 4.
quency response on substrate doping concentration and carrier
lifetime.
A PPENDIX
Bulk Fermi potential:
kT NA
ϕF = ln ≈ 0.358 V. (2)
q ni
Oxide capacitance:
ϵox
Cox = ≈ 3.45 × 10−2 F/m2 . (3)
tox
Depletion charge term:
√
4ϵsi qNA ϕF
≈ 0.0045 V. (4)
(a) Low-frequency (1 Hz) C–V curve. Cox
Threshold voltage:
Qdep
Vth = 2ϕF + ≈ 0.72 V. (5)
Cox
(b) High-frequency (106 Hz) C–V curve.
Fig. 4: Simulated C–V characteristics of the MOS capacitor.
1) Analysis of Frequency Dependence: The low-frequency
C–V curve was expected to return to Cox in strong in-
version; however, the simulated response closely resembled
the high-frequency curve. This behavior is attributed to the
long minority carrier response time in the lightly doped
substrate (NA = 1016 cm−3 ). Even at 1 Hz, the AC signal
period is insufficient for generating enough minority carriers
to establish a responsive inversion layer. Consequently, the
inversion capacitance behaves as if measured under high-
frequency conditions. By contrast, devices with much higher
doping concentrations exhibit faster minority carrier dynamics
and can recover Cox in the low-frequency regime.
V. C ONCLUSION
A comprehensive TCAD analysis of a p-type MOS capacitor
with NA = 1016 cm−3 and 1 nm oxide thickness was per-
formed. The threshold voltage obtained analytically (0.72 V)
was in good agreement with the simulated result (0.65 V). The
C–V analysis demonstrated that both low- and high-frequency
characteristics exhibited high-frequency behavior due to slow
minority carrier response in the lightly doped substrate. These
results emphasize the strong dependence of MOSCAP fre-
