PHOTODETECTION, SELF AMPLIFICATION AND DEMUX OPERATION IN TANDEM AMORPHOUS SI-C DEVICES


PHOTODETECTION, SELF AMPLIFICATION AND DEMUX 
OPERATION IN TANDEM AMORPHOUS SI-C DEVICES  

 
M. Vieira1,2,3, P. Louro1,2, M. A.Vieira1,2, A. Fantoni1,2, M. Fernandes1,2  

 
1 Electronics Telecommunications and Computer Dept, ISEL, Lisbon, Portugal 

2 CTS-UNINOVA, Lisbon, Portugal 
3 DEE-FCT-UNL, Quinta da Torre, Monte da Caparica, 2829-516, Caparica, Portugal 

 

Keywords: Optical devices, a-SiC heterostructures, optical communication, multiplexing and demultiplexing applications over 
POF. 

Abstract: In this paper we report the use of a monolithic system that combines the demultiplexing operation with the 
simultaneous photodetection and self amplification of the signal. The device is a double pi’n/pin a-SiC:H 
heterostructure with optical gate connections for light triggering in different spectral regions. Results show that when a 
polychromatic combination of different pulsed channels impinges on the device the output signal has a strong 
nonlinear dependence on the light absorption profile, (wavelength, bit rate and intensity). This effect is due to the self 
biasing of the junctions under unbalanced light generation of carriers. Self optical bias amplification under uniform 
irradiation and transient conditions is achieved. An optoelectronic model based on four essential elements: a voltage 
supply, a monolithic double pin photodiode, optical connections for light triggering, and optical power sources for 
light bias explains the operation of the optical system. 

 
1 INTRODUCTION 

There has been much research on semiconductor optical 
amplifiers as elements for optical signal processing, 
wavelength conversion, clock recovery, signal 
demultiplexing and pattern recognition [1]. Here, a 
specific band or frequency need to be filtered from a 
wider range of mixed signals. Active filter circuits can be 
designed to accomplish this task by combining the 
properties of high-pass and low-pass into a band-pass 
filter. Amorphous silicon carbon tandem structures, 
through an adequate engineering design of the multiple 
layers’ thickness, absorption coefficient and dark 
conductivities [2] can accomplish this function.  

Wavelength division multiplexing (WDM) devices 
are used when different optical signals are encoded in the 
same optical transmission path, in order to enhance the 
transmission capacity and the application flexibility of 
optical communication and sensor systems. Various types 
of available wavelength-division multiplexers and 
demultiplexers include prisms, interference filters, and 
diffraction gratings. Currently modern optical networks  
use Arrayed Waveguide Grating (AWG) as optical 
wavelength (de)multiplexers [3] based on multiple 
waveguides to carry the optical signals 

 

 
 
This paper reports results on the use of a double 

pi’n/pin a-SiC:H WDM heterostructure as an active 
band-pass filter transfer function whose operation 
depends on the wavelength of the trigger light and 
applied voltage and optical bias. The dynamic response 
can range from positive feedback (regeneration) under 
positive bias, to two different behaviours under negative 
bias: as an active multiple-feedback filter with internal 
gain or in a mode that preserves the amplitude of the 
signal, depending on the triggering light. An 
optoeletronic model gives insight on the physics of the 
system. 
 
2 EXPERIMENTAL DETAILS 

 

GLASS

pin 1
(a-SiC:H)
200 nm

pin 2
(a-Si H)
1000 nm

TCO

TCO
TCO

Applied voltage

Light
λ1
λ2
λ3

 
Figure 1.  Device configuration. 

The sensor element is a multilayered heterostructure 
based on a-Si:H and a-SiC:H produced by PE-CVD at 

i-ETC: ISEL Academic Journal of Electronics, 
Telecommunications and Computers 
CETC2011 Issue, Vol. 2, n. 1 (2013) ID-12

http://journals.isel.pt/index.php/IAJETC



13.56 MHz radio frequency. The configuration of the 
device, shown in Figure 1, includes two stacked p-i-n 
structures (p(a-SiC:H)- í'(a-SiC:H)-n(a-SiC:H)-p(a-
SiC:H)-i(a-Si:H)-n(a-Si:H)) sandwiched between two 
transparent contacts. The thicknesses and optical gap of 
the front í'- (200nm; 2.1 eV) and thick i- (1000nm; 
1.8eV) layers are optimized for light absorption in the 
blue and red ranges, respectively [4]. Experimental 
details on the preparations, characterizations and 
optoelectronic properties of the amorphous silicon 
carbide films and junctions were described elsewhere [5]. 
As a result, both front and back structures act as optical 
filters confining, respectively, the blue and the red optical 
carriers. The device operates within the visible range 
using as optical signals the modulated light (external 
regulation of frequency and intensity) supplied by a red 
(R: 626 nm; 51μW/cm2) a green (G: 524 nm; 73μW/cm2) 
and a blue (B: 470nm; 115μW/cm2) LED. Additionally, 
steady state red, green and blue illumination 
(background) was superimposed using similar LEDs. 

 
3  LIGHT FILTERING 

The characterization of the devices was performed 
through the analysis of the photocurrent dependence on 
the applied voltage and spectral response under different 
optical and electrical bias conditions. The responsivity 
was obtained by normalizing the photocurrent to the 
incident flux. To suppress the dc components all the 
measurements were performed using the lock-in 
technique.  

400 500 600 700 800
0.0

0.1

0.2

0.3

0.4 pi(a-SiC:H)n/ITO/pi(a-Si:H)n

 

-10 V

+3 V

P
ho

to
cu

rr
en

t (
μ

a)

a) 

400 500 600 700 800

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35
Back cellFront cell

ITO/pi´(a-SiC:H)n/ITO/pi(a-Si:H)n/ITO

 

 

P
ho

to
cu

rr
en

t (
μ

a)

Wavelength (nm)

0.0

0.5

1.0

1.5

2.0-5V

+1V

b) 
Figure 2.  a) p-i’-n-p-i-n spectral photocurrent under different 

applied voltages b) Front, p-i’ (a-SiC:H)-n, and back, p-i (a-
Si:H)-n spectral photocurrents under different applied bias.  

Figure 2a displays the spectral photocurrent of the 
sensor under different applied bias (+3V<V<-10V), the 
internal transparent contact was kept floating in all 
measurements. In Figure 2b the spectral photocurrent, 
under different electrical bias is displayed for the front, p-
i’ (a-SiC:H)-n, and the back p-i (a-Si:H)-n, photodiodes.  

Results confirm that the front and back photodiodes 
act, separately, as optical filters. The front diode, based 
on a-SiC:H heterostructure, cuts the wavelengths higher 
than 550nm while the back one, based on a-Si:H, cuts the 
ones lower than 500nm. Each diode, separately, presents 
the typical responses of single p-i-n cells with intrinsic 
layers based on a-SiC:H or a-Si:H materials, respectively. 
Since the current across the device has to remain the 
same, in the stacked configuration, it is clearly observed 
the influence of both front and back diodes modulated by 
its serial connection through the internal n-p junction. 
 
 
4 SELF OPTICAL 
AMPLIFICATION  

4.1 Self bias amplification under 
uniform irradiation  

When an external electrical bias (positive or 
negative) is applied to a double pin structure, its main 
influence is in the field distribution within the less photo 
excited sub-cell [6]. The front cell, under red irradiation; 
the back cell, under blue light and both, under green 
steady state illumination  In comparison with 
thermodynamic equilibrium conditions (dark), the 
electric field under illumination is lowered in the most 
absorbing cell (self forward bias effect) while the less 
absorbing reacts by assuming a reverse bias configuration 
(self reverse bias effect).  

In Figure 3, the spectral photocurrent at different 
applied voltages is displayed under red (a), green (b) and 
blue (c) background irradiations and without it (d). 
Results confirm that a self biasing effect occurs under 
unbalanced photogeneration. As the applied voltages 
changes from positive to negative the blue background 
enhances the spectral sensitivity in the long wavelength 
range. The red bias has an opposite behavior since the 
spectral sensitivity is only increased in the short 
wavelength range. Under green background the spectral 
photocurrent increases with the applied voltage 
everywhere. 
 

 

M. Vieira et al. | i-ETC -  CETC2011 Issue, Vol. 2, n. 1 (2013) ID-12

i-ETC: ISEL Academic Journal of Electronics, Telecommunications and Computers   http://journals.isel.pt/index.php/IAJETC



a)

b)

c)

d)
Figure 3.  Spectral photocurrent under reverse and forward 

bias measured without and with (λL) background illumination. 

In Figure 4 the ratio between the spectral 
photocurrents @ +1V and -8 V, under red, green and 
blue steady state illumination and without it (dark), are 
plotted.  

400 450 500 550 600 650 700

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5 Blue background

Green background

Red background

+1 V
-8 V

 

 

I a
c (
λ L

)/
I(

da
rk

)

Wavelength (nm)  

Figure 4.   Ratio between the photocurrents under red, green 
and blue steady state illumination and without it (no 

background).  

As expected, results show that the blue background 
enhances the light-to-dark sensitivity in the long 
wavelength range and quenches it in the short 
wavelength range. The red bias has an opposite behavior; 
it reduces the ratio in the red/green wavelength range and 
amplifies it in the blue one. The sensor is a wavelength 
current-controlled device, that make use of changes in the 
wavelength of the optical bias to control the power 
delivered to a load, acting as an optical amplifier. Its 
gain, defined as the ratio between the photocurrent with 
and without a specific background depends on the 
background wavelength that controls the electrical field 
profile across the device. If the electrical field increases 
locally (self optical amplification) the collection is 
enhanced and the gain is higher than one. If the field is 
reduced (self optical quench) the collection is reduced 
and the gain is lower than one. This optical nonlinearity 
makes the transducer attractive for optical 
communications. 

4.2 Self bias amplification under 
transient conditions and uniform 
irradiation  

A chromatic time dependent wavelength combination 
(4000 bps) of R (λR=624 nm), G (λG=526 nm) and B 
(λB=470 nm) pulsed input channels with different bit 
sequences, was used to generate a multiplexed signal in 
the device. The output photocurrents, under positive (dot 
arrows) and negative (solid arrows) voltages with (colour 
lines) and without (dark lines) background are displayed 
in Figure 5. The bit sequences are shown at the top of the 
figure.  

Results show that, even under transient input signals 
(the input channels), the background wavelength controls 
the output signal as in Figures 3 and 4. This nonlinearity, 
due to the transient asymmetrical light penetration of the 
input channels across the device together with the 
modification on the electrical field profile due to the 
optical bias, allows tuning an input channel without 
demultiplexing the stream. 

400 450 500 550 600 650 700

0.0

2.0

4.0
Green Background

P
ho

to
cu

rr
en

t (
nA

)

Wavelength (nm)

 +1 V
 0 V
 -3 V
 -5V
 -8 V

400 450 500 550 600 650 700

0.0

2.0

4.0

6.0

8.0

10.0

12.0

Blue Background

 

P
ho

to
cu

rr
en

t (
nA

)

Wavelength (nm)

 +1 V
 0 V
 -3 V
 -5V
 -8 V

400 450 500 550 600 650 700

0.0

2.0

4.0
No background

P
ho

to
cu

rr
en

t (
nA

)

Wavelength (nm)

 +1 V
 0 V
 -3 V
 -5V
 -8 V

400 450 500 550 600 650 700

0.0

0.5

1.0

1.5

2.0 Red Background

 

P
ho

to
cu

rr
en

t (
nA

)

Wavelength (nm)

 +1 V
 0 V
 -3 V
 -5V
 -8 V

M. Vieira et al. | i-ETC -  CETC2011 Issue, Vol. 2, n. 1 (2013) ID-12

i-ETC: ISEL Academic Journal of Electronics, Telecommunications and Computers   http://journals.isel.pt/index.php/IAJETC



0.0 0.5 1.0 1.5 2.0 2.5
0

1

2

3

4 ΦL=0
BRR

V=+1V

V=-8V

Red

Green

Blue
P

ho
to

cu
rr

en
t (
μA

)

Time (ms)  
Figure 5.   Single and combined signals @-8V; without (solid 

arrows) and with (dotted arrows) green optical bias. 

In Figure 6 the green channel is tuned through the 
difference between the multiplexed signal with and 
without green bias. 

0.0 1.0 2.0 3.0 4.0 5.0
-1

0

1

2

3

4

5

I
ON

-I
OFF

bias off

B

GR

V=+1V

V=-8V

Red
Green
Blue

P
ho

to
cu

rr
en

t (
μ
A

)

Time (ms)  
Figure 6.  Multiplexed signals @-8V/+1V (solid /dot lines); 

without (bias off) and with (R, G, B) green optical bias. 

 

5 WAVELENGTH-DIVISION 
(DE)MULTIPLEXING DEVICE 

In this section we report the use of the same devices 
to combines the demultiplexing operation with the 
simultaneous photodetection of the signal.  

The device makes use of the fact that the optical 
absorption of the different wavelengths can be tuned by 
means of electrical bias changes or optical bias variations 
(Figure 5 and Figure 6). This capability was obtained 
using adequate engineering design of the multiple layers 
thickness, absorption coefficient and dark conductivities 
[7]. The device described herein operates from 400 to 
700 nm which makes it suitable for operation at visible 
wavelengths in optical communication applications.  

Monochromatic pulsed beams together or one single 
polychromatic beam (mixture of different wavelength) 
impinge in the device and are absorbed, according to 
their wavelength (Figure 7). By reading out, under 
appropriate electrical bias conditions, the photocurrent 

generated by the incoming photons, the input information 
is electrically multiplexed or demultiplexed. 

0.0 1.0 2.0 3.0 4.0 5.0
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5 R+G+B

B[10101010]
G[10011001]
R[01111000]

V=+1V

V=-8V R&G&B

G

 

B
R

P
ho

to
cu

rr
en

t (
μ

A
)

Time (ms)
 

Figure 7.  Single (R, G and B) and combined (R&G&B) 
signals under -8V (solid arrows) and +1V (dotted arrows). 

In the multiplexing mode the device faces the 
modulated light incoming together (monochromatic input 
channels). The combined effect of the input signals is 
converted to an electrical signal, via the device, keeping 
the input information (wavelength, intensity and 
modulation frequency). The output multiplexed signal, 
obtained from the combination of the three optical 
sources, depends on both the applied voltage and on the 
ON-OFF state of each input optical channel. Under 
negative bias, the multiplexed signal presents eight 
separate levels. The highest level appears when all the 
channels are ON and the lowest if they are OFF. 
Furthermore, the levels ascribed to the mixture of three or 
two input channels are higher than the ones due to the 
presence of only one (R, G, B). Optical nonlinearity was 
detected; the sum of the input channels (R+B+G) is lower 
than the correspondent multiplexed signals (R&G&B). 
This optical amplification, mainly on the ON-ON states, 
suggests capacitive charging currents due to the time-
varying nature of the incident lights. Under positive bias 
the levels were reduced to one half since the blue 
component of the combined spectra falls into the dark 
level, the red remains constant and the green component 
decreases. 

In the demultiplexing mode a polychromatic 
modulated light beam is projected onto the device and the 
readout performed by shifting between forward and 
reverse bias.  
Figure 7 displays the input and multiplexed signals under 
negative (-8V) and positive (+1V) electrical bias. As 
expected, the input red signal remains constant while the 
blue and the green ones decrease as the voltage changes 
from negative to positive. 

To recover the transmitted information (8 bit per 
wavelength channel) the multiplexed signal, during a 
complete cycle, was divided into eight time slots, each 
corresponding to one bit where the independent optical 
signals can be ON (1)or OFF (0). Under positive bias, the 
device has no sensitivity to the blue channel, so the red 

M. Vieira et al. | i-ETC -  CETC2011 Issue, Vol. 2, n. 1 (2013) ID-12

i-ETC: ISEL Academic Journal of Electronics, Telecommunications and Computers   http://journals.isel.pt/index.php/IAJETC



and green transmitted information are tuned and 
identified. The highest level corresponds to both channels 
ON (R=1, G=1), and the lowest to the OFF-OFF stage 
(R=0; G=0). The two levels in-between are related with 
the presence of only one channel ON, the red (R=1, G=0) 
or the green (R=0, G=1). To distinguish between these 
two situations and to decode the blue channel, the 
correspondent sub-levels, under reverse bias, have to be 
analyzed. The highest increase at -8V corresponds to the 
blue channel ON (B=1), the lowest to the ON stage of the 
red channel (R=1) and the intermediate one to the ON 
stage of the green (G=1). Using this simple key algorithm 
the independent red, green and blue bit sequences were 
decoded as: R [01111000], G[10011001] and 
B[10101010 which are in agreement with the signals 
acquired for the independent channels. 
 
6 THE OPTOELECTRONIC 
MODEL.  

6.1 Two connected transistor model  

Based on the experimental results and device 
configuration an optoelectronic model was developed [8] 
which is displayed in Figure 8. The optoelectronic block 
diagram (a) consists of four essential elements: a double 
pin device for detection, a voltage supply to dc voltage 
bias; optical connections for light triggering and optical 
bias to dc light bias control. These four elements when 
connected together form the essentials of the 
optoelectronic circuit.  
 

      

 

Optical
bias

Double
PIN

Photodiode
structure

Voltage
Supply

Ligth
Triggering

1

2

3

4

Load

Optical
bias

Double
PIN

Photodiode
structure

Voltage
Supply

Ligth
Triggering

1

2

3

4

Load

Double
PIN

Photodiode
structure

Voltage
Supply

Ligth
Triggering

1

2

3

4

Load

a) 

p

p
Q1

Q2

n

p

p
i

n

p
i

n

n
p

p
i
p
i’

n
p

I1,3

I2,4

p

p
Q1

Q2

n

p

p
i

n

p
i

n

n
p

p
i
p
i’

n
p

I1,3

I2,4

p

p
Q1

Q2

p
Q1

Q2

n

p

p
i

n

p
i

n

n
p

p
i
p
i’

n
p

I1,3

I2,4

b) 

I4 

Q2 

I1

I2 

Q1 

C2 R2 
R1 (V) 

C1 I3 

c) 
 

Figure 8.  a) Optoelectronic block diagram. b) Compound 
connected phototransistor equivalent model. c) ac equivalent 

circuit. 

The monolith device is modelled by the two-
transistor model (Q1-Q2) (Figure 8b) that can be 
considered to constitute pnp (Q1) and npn (Q2) 

phototransistors separately. The ac circuit representation 
is displayed in Figure 8c.  

Its operation is based upon the following principle: 
the flow of current through the resistor connecting the 
two transistor bases is proportional to the difference in 
the voltages across both capacitors (charge storage 
buckets). Two optical gate connections ascribed to the 
different light penetration depths across the front (Q1) 
and back (Q2) phototransistors were considered to allow 
independent blue (I1), red (I2) and green (I3, I4) channels 
transmission. Four square-wave current sources with 
different intensities are used; two of them, I1 and I2, with 
different frequencies to simulate the input blue and red 
channels and other two, I3 and I4, with the same 
frequency but different intensities to simulate the green 
channel due to its asymmetrical absorption across both 
front and back phototransistors. The charge stored in the 
space-charge layers is modelled by the capacitor C1 and 
C2. R1 and R2 model the dynamical resistances of the 
internal and back junctions under different dc bias 
conditions.  

Once the ac sources are connected in the load loop 
an ac current flows through, establishing voltage 
modifications across the two capacitors. During the 
simultaneous transmission of the three independent bit 
sequences, the set-up in this capacitive circuit loop is 
constantly changing in magnitude and direction. This 
means that the voltage across one capacitor builds up 
until a maximum and the voltage across the other builds 
up to a minimum. The system collapses and builds up in 
the opposite direction. It tends to saturate and then leave 
the saturation because of the cyclic operation. This 
results in changes on the reactance of both capacitors. 
The use of separate capacitances on a single resistance R1 
results in a charging current gain proportional to the ratio 
between collector currents. The dc voltage, according to 
its strength, aids or opposes the ac currents.  
 

6.2 Optoelectronic simulation 

The multiplexed signal was simulated by applying 
the Kirchhoff’s laws for the simplified ac equivalent 
circuit (Figure 8c) and the four order Runge-Kutta 
method to solve the corresponding state equations. 

 

)(
1

1

)(
111

11

2,1

2

1
2,1

222121

11112,1 ti

C

Ctv

CRCRCR

CRCR
dt

dv

⎥
⎥
⎥
⎥

⎦

⎤

⎢
⎢
⎢
⎢

⎣

⎡

+

⎥
⎥
⎥
⎥

⎦

⎤

⎢
⎢
⎢
⎢

⎣

⎡

−−

−
=

  

(1) 

( ) ( )tv
R

ti 2,1
2

1
0 ⎥

⎦

⎤
⎢
⎣

⎡
=  

(2) 

with v1(t) and v2(t) the emitter-base voltages of Q1 and Q2 
transistors and i(t) the generated photocurrent under 
transient conditions. MATLAB was used as a 
programming environment and the input parameters 

M. Vieira et al. | i-ETC -  CETC2011 Issue, Vol. 2, n. 1 (2013) ID-12

i-ETC: ISEL Academic Journal of Electronics, Telecommunications and Computers   http://journals.isel.pt/index.php/IAJETC



chosen in compliance with the experimental results 
(Figure 7). 

In Figure 9 the simulated currents (symbols) under 
negative and positive bias are compared. In Figure 10 and 
under negative bias the simulated current under green (a) 
and red backgrounds are displayed. The current sources 
are also displayed (dash lines). To simulate the red and 
the green backgrounds, current sources intensities were 
multiplied by the on/off ratio between the input channels 
with and without optical bias. The same bit sequence of 
Figure 7 was used. To validate the model the 
experimental multiplexed signals are also shown (solid 
lines). 

Good agreement between experimental and 
simulated data was observed. Under negative bias and if 
no optical bias is applied, the expected eight levels are 
detected, each one corresponding to the presence of 
three, two, one or no color channel ON. Under positive 
bias or steady state irradiation the levels are amplified or 
reduced depending on the external control. The expected 
optical amplification is observed due to the effect of the 
active multiple-feedback filter when the back diode is 
light triggered. 

 

0.0 1.0 2.0 3.0 4.0 5.0
0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5  i(R
1
=1KΩ,R

2
=5kΩ)

 i(R
1
=10MΩ,R

2
=5kΩ)

 I
1
=0.32μΑ

 I
2
=1.1μΑ

 I
3
=1.1μΑ

 I
4
=0.82μΑ

V=+1V

 

V=-8V

P
ho

to
cu

rr
en

t (
μ

A
)

Time (ms)  
Figure 9.  Multiplexed simulated (symbols), current sources 

(dash lines) and experimental (solid lines) under negative 
(R1=1KΩ; -8V) and  positive (R1=10MΩ; +1V) dc bias  

without any background. 

Under forward bias (high R1) the device remains in 
its non conducting state unless a light pulse (I2 or I2+I4) is 
applied to the base of Q2. This pulse causes Q2 to conduct 
because the reversed biased n-p internal junction behaves 
like a capacitor inducing a charging current (I2+I4) across 
both collector junctions. The collector of the conducting 
transistor pulls low, moving the Q1 base toward its 
collector voltage, which causes Q1 to conduct. The 
collector of the conducting Q1 pulls high, moving the Q2 
base in the direction of its collector. This positive 
feedback (regeneration) reinforces the Q2 already 
conducting state and a current I2+I4 will flow on the 
external circuit.  

Green irradiation (Figure 10a) moves asymmetrically 
the Q1 and Q2 bases toward their emitter voltages, 
resulting in lower values of I3 and I4 when compared 
without optical bias. I1 and I2 slightly increase due to the 

increased carrier generation on the less absorbing 
phototransistors. Under negative bias During the duration 
of the red and blue pulses (I1 or I2 ON), as without optical 
bias, the internal junction remains forward biased and the 
transferred charge between C1 and C2 reaches the output 
terminal as a capacitive charging current. During the 
green pulse (I3 and I4 ON) only residual charges are 
transferred between C1 and C2. So, only the charges 
generated in the base of Q2 (I4) reaches the output 
terminal as can be confirmed by the good fitting between 
simulate and experimental differences of both 
multiplexed signals without and with optical bias. Under 
red background the expected optical amplification in the 
low wavelength range is observed (Figures 3a and 4) due 
to the effect of the active multiple-feedback filter when 
the back diode is light triggered 
 

0.0 1.0 2.0 3.0 4.0 5.0
0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5
 i(OFF)

R
1
=1KΩ

R
2
=5kΩ

  i(ON)
 I

1
=0.32μΑ

 I
2
=1.21μΑ

 I
3
=0.54μΑ

 I
4
=0.82μΑ

C
1
/C

2
=2.5

φ=0

 

φ=73mWcm-2

λ=554 nm

C

P
ho

to
cu

rr
en

t (
μ

A
)

Time (ms)

i
ΦOFF

-i
ΦON

a) 

0.0 1.0 2.0 3.0 4.0 5.0
0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5
 i(OFF) R1=1KΩ

R
2
=5kΩ  i(ON)

 αI
1
=0.93

 αI
2
=2.7

 αI
3
=0.54

 αI
4
=0.5φ=0

 

φ=73mWcm-2

λ=624 nm

P
ho

to
cu

rr
en

t (
μ

A
)

Time (ms) b) 
Figure 10.  Multiplexed simulated (symbols), current sources 

(dash lines) and experimental (solid lines): under negative 
(R1=10MΩ; -8V) and  positive (R1=1kΩ; +1V) dc bias  and 

green (a) and red (b) backgrounds. 

 

6.3 Transfer characteristics 

To illustrate the transfer characteristics effects due to 
changes in steady state light, dc control voltage or 
applied light pulses the following steps in the operation 
and control of the optoelectronic circuit are considered: 

1 Positive voltage bias control. Only an ac current 
is flowing through the load capacitance C2. The device 
has high ohmic resistance (high R1) and remains in its 

M. Vieira et al. | i-ETC -  CETC2011 Issue, Vol. 2, n. 1 (2013) ID-12

i-ETC: ISEL Academic Journal of Electronics, Telecommunications and Computers   http://journals.isel.pt/index.php/IAJETC



non conducting state unless a light pulse (I2 or I2+I4) is 
applied to the base of Q2. This pulse causes Q2 to conduct 
as the reversed biased n-p internal junction behaves like a 
capacitor inducing a charging current across R2. No 
amplification effect was detected since Q1 acts as a load 
and no charges are transferred between C1 and C2 
(Figure 6a).  

2 Negative dc voltage bias control. Under 
negative bias the device has low ohmic resistance (low 
R1) the base emitter junction of both transistors are 
inversely polarized and conceived as phototransistors, 
thus taking advantage of the amplifier action of 
neighboring collector junctions, which are polarized 
directly. This results in a charging current gain 
proportional to the ratio between both collector currents 
(C1/C2). The device behaves like an optoelectronic 
controlled transmission system that stores, amplifies and 
transports the minority carriers generated by the current 
pulses, through the capacitors C1 and C2. Here, the dc 
voltage control creates a voltage across one or both 
capacitors which, when superimposed with an ac pulse, 
collectively saturates the circuit. No additional change in 
the voltage across the capacitors occurs. This results in 
maximum power transfer to the load.  

3 Optical bias control. Depending on it 
wavelength (Figures 3 and Figure 4) the optical bias 
changes the amplitude of the ac current sources by an α 
factor, and so the voltages across one or both capacitors. 
Blue, red or green irradiations move asymmetrically the 
bases of Q1, Q2 or both toward their emitter voltages 
(self-forward effect), resulting, respectively in lower 
values of I1, I2, I3 and I4 when compared with no optical 
bias. The circuit can leave the saturation resulting in a 
wavelength controlled power transfer to the load that 
allows tuning input channel from the stream (Figure 10).  

Results also show that the two-transistor model 
explains the difference between the conduction 
mechanisms, under positive and negative bias, helping to 
understand the signal decoding algorithm. Under positive 
bias the red and green channels are immediately decoded 
since the collected current is due exclusively to the 
generated carriers at the back diode. Under negative bias, 
depending on the ratio between C1 and C2, different 
charging currents across the reversed junctions have to be 
considered. The balance between them depends on the 
presence of three, two or one channel ON (Fig. 2-3). So, 
by comparing the different sublevels signals under 
positive and negative bias it is possible to recover the 
blue, the green and the red input channels. 
 
 
 
 
 
 
 

7 CONCLUSIONS  

A double pi’n/pin a-SiC:H heterostructure with optical 
gate connections was presented. Multiple monochromatic 
pulsed communication channels were transmitted 
together and the multiplexed signal analyzed at different 
electrical and optical bias. Results show that the output 
signal has a strong nonlinear dependence on the light 
absorption profile. This effect is due to the self biasing of 
the junctions under unbalanced light generation profiles. 
Self optical bias amplification under uniform irradiation 
and transient conditions is achieved. An optoelectronic 
model based on four essential elements: a voltage supply; 
a double pin photodiode; optical connections for light 
triggering and optical power sources for light bias 
explains the operation of the optical system. 
 

REFERENCES 

                                                 
[1] M. J. Connelly, Semiconductor Optical Amplifiers. 
Boston, MA: Springer-Verlag, 2002. ISBN 978-0-7923-7657-6. 
[2] P. Louro, M. Vieira, Yu. Vygranenko, A. Fantoni, M. 
Fernandes, G. Lavareda, N. Carvalho Mat. Res. Soc. Symp. 
Proc., 989, A12.04 (2007). 
[3] Michael Bas, Fiber Optics Handbook, Fiber, Devices and 
Systems for Optical Communication, Chap, 13, Mc Graw-Hill, 
Inc. 2002. 
 [4] M. Vieira, A. Fantoni, M. Fernandes, P. Louro, G. 
Lavareda, C. N. Carvalho, Journal of Nanoscience and 
Nanotechnology, Vol. 9 , Number 7, July 2009 , pp. 4022-
4027(6). 
[5] M. Vieira, M. Fernandes, P. Louro, A. Fantoni, M. A. 
Vieira, J. Costa, M. Barata, Mat. Res. Soc. Symp. Proc.Volume 
1245, A08-02 (2010) . 
[6] M. Vieira, A. Fantoni, P. Louro, M. Fernandes, R. Schwarz, 
G. Lavareda, and C. N. Carvalho, Vacuum, Vol. 82, Issue 12, 8 
August 2008, pp: 1512-1516. 
[7] P. Louro, M. Vieira, Yu. Vygranenko, A. Fantoni, M. 
Fernandes, G. Lavareda, N. Carvalho Mat. Res. Soc. Symp. 
Proc., 989 (2007) A12.04. 
[8] M. A. Vieira, M. Vieira, M. Fernandes, A. Fantoni, P. 
Louro, M. Barata, Amorphous and Polycrystalline Thin-Film 
Silicon Science and Technology 2009, MRS Proceedings Vol. 
1153, A08-0. 

M. Vieira et al. | i-ETC -  CETC2011 Issue, Vol. 2, n. 1 (2013) ID-12

i-ETC: ISEL Academic Journal of Electronics, Telecommunications and Computers   http://journals.isel.pt/index.php/IAJETC