cal simulation approach is proposed to explore the time
response of a diamond device to ultra-short and intense
high-energy electron pulses. Such a two-step simulation
(Sentaurus +LTspice) can interpret the measured data
and extrapolate the performance of diamond device in
similar radiation environments.
An experimental program is ongoing at the electron-
linac of the FERMI free-electron laser in Trieste [14], to
study the response of the diamond detectors to the irra-
diation by intense, sub-picosecond bunches of electrons
accelerated to about 1 GeV. We compare our simula-
tions with preliminary data from Ref. [15].
The experimental setup is briefed in Section 2. The
two-step simulation approach is elucidated in Section 3.
The two steps with regard to the response of the dia-
mond detector and the effect of the electronic circuit are
demonstrated in Section 4 and Section 5, respectively.
A validation of the approach has been devised in Sec-
tion 6. Final results of the numerical simulation overlaid
with the experimental data are shown in Section 7.
2. Experimental setup
We briefly describe here the diamond detectors de-
veloped for the Belle II experiment and the beam test
facility set up to characterize their behaviour under ex-
treme irradiation conditions.
2.1. Diamond detectors
The diamond sensors consist of (4.5×4.5×0.5) mm3
high-purity sCVD diamond crystals provided by Ele-
ment Six Ltd [16] and two (4.0×4.0) mm2electrodes
on opposite faces. The electrodes are made of Ti/Pt/Au
layers with (100+120+250) nm thickness, processed
by CIVIDEC [17]. More details can be found else-
where [6]. During the measurements, one electrode of
the diamond sensor is connected to a custom-made HV
power supply [18], the other is connected to a LeCroy
HDO9000 oscilloscope [19].
2.2. Beam facility
Collimated electron bunches of about 1 ps duration
with energy up to 1.5 GeV are available at the linac of
the FERMI free-electron laser in Trieste [14]. Bunch
charge can be tuned from tens to 1000 pC and trans-
verse size down to about 0.1 mm. The first data tak-
ing is carried out using 0.9 GeV electron bunches of
35 pC charge (<1% of bunch charge at SuperKEKB)
and 120 µm transverse size.
3. Simulation workflow
TCAD-Sentaurus is regarded as the industry standard
for the numerical simulation of the electrical character-
istics of silicon-based and compound semiconductor de-
vices [10]. The predictive power of its numerical calcu-
lation on the response of semiconductor devices to ex-
ternal electrical, thermal, and optical sources has been
verified via numerous applications in the semiconductor
industry. Using TCAD-Sentaurus, a series of sequential
processes of the diamond detector are simulated, includ-
ing a radiation beam interacting with the diamond crys-
tal, the creation of electron-hole pairs, the drift of charge
carriers, and the evolution of the induced voltage drop
on electrodes. In addition, the space-time evolution of
the concentration of charge carriers inside the diamond
crystal is also obtained. These processes are regarded
as the first step of the simulation workflow and are de-
scribed in Section 4.
Though a SPICE-like utility is embedded in the
TCAD-Sentaurus package, its functionalities and mod-
els are not adequate to fully take into account the ef-
fect of transmission cables, which plays a leading role
in dealing with fast pulses. To overcome these limita-
tions, LTspice, the most widely used SPICE software in
the industry [13] is employed to take over as the second
step. In our model described in Section 5, the diamond
detector is implemented by a combination of a voltage
source in series with a resistor and in parallel with a ca-
pacitor. The simulated result of the evolution of voltage
drop on electrodes from the first step serves as an input
to the LTspice simulation for the voltage source. The
resistance in the diamond detector model is obtained
from the time evolution of voltage and current in TCAD.
Coaxial cables, power supply, and oscilloscope all are
properly modeled to take into account the transmission
effects on the electrical signal such as reflection, atten-
uation, and distortion.
4. Response of the diamond detector
The First step of the simulation is described here,
including our choices for the physical parameters and
models in TCAD and the generation of excess charge
carrier by the ionizing electron beam.
4.1. Diamond physical parameters
The default values of parameters for the physi-
cal properties of diamond in the database of TCAD-
Sentaurus are set according to Ref. [20]. In addition,
we update several sets of parameters. The mobility
and saturation velocity of charge carriers are updated
2