1.

INTRODUCTION

1.1 Objective

The objective of this experiment is to:

i. Understand the fundamental concepts of the p-n junction i.e. I-V
characteristics, temperature dependence, ideality factor and reverse saturation
current.
ii. Investigate the light current characteristic of an LED
iii. Study the emission spectrum of an LED

1.2 Background

With the constant development of electronic circuitry and devices, semiconductor

devices have become an essential tool commonly used in manufacturing many unique
and highly specialized products. The key reason to the versatility of semiconductor
materials is that its electrical properties can be easily controlled via doping.

The p-n junction is an example of a semiconductor device enhanced via doping. It has
certain remarkable properties which give rise to many useful functions. In this
experiment, we will investigate these basic properties as well as an example of a p-n
junction, the light emitting diode (LED).

2. THEORY

2.1 Formation of p-n junctions

A p-n junction is formed via the adjunction of a p-type and n-type semiconductor.
Doping is the addition of specific amounts of impurity atoms to vary the conductivity

1
 of a semiconductor by increasing the electron or hole concentrations. A
semiconductor doped with excess holes is known as a p-type, whereas that doped
with excess electrons is known as an n-type.

Figure 1: p-n junction

Applications of a p-n junction include mixing, amplification, rectification, switching

and many other types of electronic operations.

2.2 I-V characteristics of p-n junctions

For an ideal p-n junction, its I-V characteristic is given by the equation:
I = I0 (eqV / kT 1)
(1)
where I0: reverse saturation current
q: electron charge, 1.6 x 10-19C
k: Boltzmanns constant, 1.38 x 10-23 JK-1
T: absolute temperature in K
At room temperature ( 298K), kT / q is approximated to be 0.0252 eV.

For a real diode, a more specific expression is given by:

I = I0 (eqV / nkT 1)
(2)
where n: ideality factor, varies between 1 and 2, depending on the type of material

Under forward bias,

I I0 eqV / nkT
(3)

2
since I0 is extremely small, typically between 10-18 A and 10-9 A
Hence, from (3),
ln I = ln I0 + qV/ nkT
(4)

Figure 2: I-V curve of 1N4001 diode

In semiconductor theory, ni, the intrinsic carrier concentration, is given by:

ni = constant x T 3/2e-Eg / 2kT
(5)
where Eg: bandgap energy
Given that I0 is dependent on ni, it is deduced that
I0 =AT 3 e-Eg / kT
(6)
Hence, ln I0 =ln A + 3 ln T Eg/ kT
(7)
where A is a constant independent of temperature T.
From (4), we get
V =nkT ln(I / I0) / q
(8)
Differentiating (8) with respect to T, we get

3
 (9)
Differentiating (7) with respect to T, we get

(10)
Putting (9) and (10) together,

(11)
For most diodes, on average,

(12)
when a diode is biased by a constant current source. This effect is applicable for
temperature measurements and is commonly observed around 100K.

2.3 Light emitting diodes (LEDs)

The light emitting diode (LED) is a type of p-n junction. The LED operates using the
principle of recombination. Under forward bias, current is injected into the p-n
junction and the LED emits light. In the diode, current develops as holes from the p-
side and electrons from the n-side recombine.

4
 Figure 3: Lamp type LED
- obtained from http://www.microscopy.fsu.edu/

During recombination, an electron from the conduction band is transferred to fill a

hole in the valence band as illustrated in Figure 4. Spontaneous emission may result
and energy is released in the form of a photon. This process occurs in an entirely
random manner.

Figure 4: Forward-biased p-n junction

The transition of electrons from the conduction band to the valence band requires the
conservation of energy and momentum. This is only possible with direct bandgap
semiconductors, whereby electron transitions do not require changes in crystal
momentum, and is hence efficient. Figure 5 shows that the conduction band minima

5
and the valence band maxima occur at the same k value, where k is known as a wave
vector. However, in indirect bandgap semiconductors, they do not coincide. Thus, a
change in crystal momentum is required for the transitions to occur.

Figure 5: Direct and indirect bandgap semiconductors

Emitted photons have energy equivalent to Eg, the bandgap energy. It can be
calculated using the equation:
E hc
=
g (13)
-34
where h: Plancks constant, 6.626 x 10 Js
c: speed of light, 3 x 108 m/s
: wavelength of light

LED is incoherent as the emitted photons have irregular phases. The wavelengths of
the emitted light typically range from 30 to 50 nm at 298 K. The spectral position of
the band is dependent on Eg, the bandgap energy of the semiconductor.

The table below illustrates some examples of direct bandgap semiconductor materials
used in the fabrication of LED.

Semiconductor / nm Eg / eV Colour
Gallium nitride, GaN 450 2.76 blue

6
 Indium gallium arsenide phosphide, 555 2.24 green
InGaAsP 590 2.10 yellow
620 2.00 orange
Gallium Arsenide Phosphide, GaAsP 610 2.03 red
Aluminium gallium arsenide, AlGaAs 650 1.91 red
Gallium arsenide, GaAs 860 1.44 infrared
Indium gallium arsenide, InGaAs 1100 1600 1.13 0.77 infrared

Table 1: Direct bandgap semiconductors

Applications of the LED include LED panels, multi-touch sensing, sterilization of

water, lighting, medical applications and many others.

3. EQUIPMENT AND COMPONENTS

1 Power supply
1 Breadboard
2 Digital multimeter (DMM)
1 Dual channel oscilloscope
2 1N4001 diodes
1 resistor (1 k)
1 step down transformer
1 resistor, 68 , 5 W
1 hot air blower
1 Keithley source measure unit (SMU)
1 RS577-780 ultrabright red LED
1 Black box
1 luxmeter
1 Oriel spectrometer
Connecting wires and cables

7
 Figure 6: Equipment and components
4. EXPERIMENT

4.1 I-V Characteristic

Figure 7: Sweep circuit

On the breadboard, assemble the set-up as shown in Figure 7 using the step down
transformer, 68 resistor, 1N4001 diode and connecting wires.
Take note of where the inputs go to respectively (whether it is to CH1 or CH2 of
the oscilloscope).

8
 Set the oscilloscope for X-Y mode.
Initialise the origin appropriately.
Use D.C. position for both inputs.
Plot the I-V graph on a sheet of graph paper with the axes properly labelled.
Ensure that the current range is greater than 150 mA.

a) What are the turn-on voltage and the differential resistance (dV/dI) after turn on
for this diode?

b) Explain why the I-V characteristic appears negative (i.e. inverted) on the
oscilloscope.

9
 Answers:
I/ V

25
Before heat
After heat 20
(-825, 17.5)

15

10

5
(-700, 2.5)

-800 -600 -400 -200 0 V/mV

Graph 1: I-V characteristic graph

Note that the current (I) axis is plotted using the voltage values read off directly from
the oscilloscope. By inspecting Figure 7, we can convert the value to amperes (A),
I (Amperes) = I (Volts) / 68
a) Turn-on voltage = 560 mV

10
b) The negative end of the diode is connected to CH1 whilst the positive end is
grounded. The values of V obtained are thus negative, reflecting the graph about
the y-axis. This can be clearly observed by comparing Graph 1 with Figure 2.

4.2 Temperature dependence

Using the same experimental set-up from Figure 7, apply heat to the diode using
the hot air blower for 2 minutes.
Avoid heating the 68 resistor in the process.
Plot the I-V graph.
Record the amount of voltage shift (in mV) from the two graphs at a constant
current of 150 mA.

c) Explain the observed shift in the characteristic graph.

d) Calculate the diode temperature change (in K) using the measured voltage shift at
150 mA.

Answers:
A constant current of 150 mA is equivalent to reading off 0.15 x 68 = 10.2 V.
When the y-axis is at 10.2 V,
Voltage magnitude before heating = 800 mV
Voltage magnitude after heating = 730 mV
Voltage shift = 730 800 = -70 mV
c) With increase in temperature, electrons gain more energy to overcome the energy
barrier and move from the n-side to the p-side. This increases the conductibility
of the diode. Hence, the turn-on voltage decreases.
d) Using equation (12),

T = (-70)/ (-2.4)
= 29.1 K (to 3 s.f.)

11
4.3 Reverse saturation current and ideality factor

Figure 9: Circuit for diode parameter extraction

Assemble the circuit as shown in Figure 9.
Vary the input current over 3 decades from 0.01 mA to 10 mA.
Record the voltage and current values from the Digital Multimeters provided.
Plot I vs V on a semilog graph.

e) Explain how the slope of the ln I vs V relationship can be determined from the
semilog graph
f) Determine the reverse saturation current, I0, and the ideality factor n.

Answers:
I/ mA V/ V
0.01 0.407
0.10 0.487
1.00 0.580
10.00 0.684
Table 2: I -V values from Figure 9 circuit

e) From equation (4), ln I = ln I0 + qV/ nkT

The slope can be determined as follows:

= 24.759
= 24.8 (to 3 s.f.)
f) I0 is equivalent to the y-intercept

12
Hence, I0 = 5.5 x 10-10 A
Equating slope to q / nkT at 298 K, Slope = 1/ (0.0252 n)
n = 1.60 (to 3 s.f.)

13
 (0.684, 10)

(0.125, 1 x 10-5)

Graph 2: I/ mA vs V/V

14
4.4 LED light current characteristic

Figure 10: Set-up for light current measurement

The experimental set-up above investigates the light current (L-I) characteristic of
the RS577-780 LED.
Set the source measure unit (SMU) to a constant current source.
Light emitted from the LED is detectable by the luxmeter.
The luxmeter reading is proportional to optical power.
Determine a suitable range and interval of measurements and the maximum
voltage applied to the LED.
Plot the L-I curve on a graph paper and comment on the results.
g) Explain why the experiment is carried out in the black box.
h) When the LED current is 0 mA, record the luxmeter reading and explain why.

For
L/the
luxexperiment, we have chosen maximum voltage of 1.85 V and varied the D.C.
current from 0 to 30 mA. The L-I values are tabulated as follows:

20 Input current/ mA Luxmeter Reading/ lux

0 4
5 4
16 10 5
15 8
20 12
12 25 16
30 19
Table 3: L-I values from Figure 10 set-up
Answers:
8

4
15
 x

x
x x

0 5 10 15 20 25 30 I/ mA

Graph 3: L-I characteristic curve

The experimental data starts with 4 lux even when input current is 0 mA.
Subsequently, a linear relationship is exhibited from 10 mA to 30 mA. The curve is
not perfectly linear probably because of material properties and the limits of the
device configuration.
g) The experiment is carried out in a black box so that most of the light from the
LED can be focused on the luxmeter and minimal light energy is lost to the
surroundings.
h) The luxmeter reading is 4 lux. This is because the black box is not perfectly built,
allowing light to penetrate the interior. Hence, an ideal situation where there is
complete darkness in the box is not realized. This could be also due to the fact that
the luxmeter is not properly zeroed.

4.5 LED emission spectrum

16
 0 0
9
Figure 11: Set-up for measurement of emission spectrum

The spectrometer is a spectrum analyzer which utilizes a reflective grating to

scatter the incident source radiation on the entrance slit into its frequency
components.
Rotating the entrance width adjust dial varies the intensities of the source light.
For this experiment, the width-adjust dials are fixed to 1 mm. Do not adjust the
slit widths.
Set-up your apparatus as shown in Figure 11.
Make sure that the photodetector output is set to LED mode.
Connect the photodetector output port to a digital multimeter.
By varying wavelength at suitable intervals, plot the emission spectrum of the
LED using wavelength against intensity.
The wavelength of interest is known to lie within 600 nm to 700 nm.
The intensity is known to be proportional to the voltage readings of the
multimeter.
i) Calculate the bandgap energy in terms of eV.
j) Linewidth is of an LED radiation is defined by the full width at half maximum
(FWHM) of the emission spectrum. FWHM, in terms of nm, is the total width of

17
 the spectrum taken at 50% from the peak. Determine the linewidth of the LED
and comment on the results.

Answers:
For the experiment, we have chosen a constant D.C. source at 25 mA. The values of
the multimeter reading and the corresponding wavelengths are as follows:

Wavelength/ nm Intensity/ mV
600 84.93
610 84.93
620 84.93
630 84.93
640 84.93
645 93.24
650 261.67
655 500.38
660 502.33
665 207.96
670 105.63
675 84.93
680 84.93
690 84.93
700 84.93
Table 4: Results of intensity at various wavelengths

18
 (658, 540)

(651, 312) (663, 312)

Graph 4: LED Emission Spectrum

i) Using equation (13),

= 3.021 x 10-19 J
= 1.89 eV (to 3 s.f.)
This value is consistent to the bandgap energy of a typical semiconductor, which
is around 2 eV to 3 eV.

j) From the graph, linewidth = 663 651

= 12 nm
The linewidth estimates the range of frequencies of photons that can be emitted
due to spontaneous emission. This shows that LED emission is not
monochromatic.

19
5. DISCUSSION

i. What factors determine the turn-on voltage of a p-n junction diode?

Explain its dependence on the factors that you mentioned.

The turn-on voltage of a p-n junction diode is dependent on temperature, electron

and hole doping concentrations. When temperature is increased, lowering of the
energy barrier as depicted by the slope in Figure 12 occurs. More holes are able to
transit from the p-side to the n-side. Likewise, electrons from the n-side are able
to overcome the energy barrier and move to the p-side.

Figure 12: Flow of charged carriers under forward bias

In addition, with higher concentrations of holes or electrons are on each side,

more electrons will make it to the p-side and more holes to the n-side. These
factors will facilitate overall current flow through the diode and decrease the turn-
on voltage.

ii. In part 4.3, suggest an extension to this experiment by which the

coefficient A of a diode in equation (6) can be determined?

20
 From experiment 4.3, we are able to deduce the ideality factor and I0. From
equation (6), we know that
I0 =AT 3 e-Eg / kT
Hence, by changing the temperature, T, at which the 1N4001 diode is exposed to,
we can obtain a different I0 value. A new semi-log graph can be plotted and a new
value of I0 can be deduced. Using the 2 new values of I0 and T obtained, substitute
into equation (6), solve simultaneously and you will be able to obtain the
coefficient A. A physical method to obtain the new temperature is not required if
we use the hot air blower provided to change the temperature. This is because in
experiment 4.2, we already obtained the temperature difference. Adding this value
to the room temperature of 298 K will give us the new temperature.

iii. Explain the factors that influence the spectral linewidth of the emission
spectrum that you measured in part 4.5.

Spectral linewidth is affected by several factors. Doping of the active region is

one important factor. In surface-emitting LEDs where the active region is more
heavily doped, the emission wavelength moves to longer wavelengths (lower
energies) and the spectral response curve is broadened. Fluctuations in
temperature also affect the response characteristics. The line-width of the
spectrum will increase with increase in temperature due to a larger energy spread
of the carrier distribution and phonon scattering.

iv. Describe in some detail the operating principles of a grating spectrometer.

A spectrometer is an apparatus designed to measure the distribution of radiation of

a particular range of wavelengths from a source. Its primary components are a
diffraction grating and a radiant power detector such as a photo-emissive cell or a
photomultiplier tube. A diffraction grating is an instrument designed for isolating
a narrow portion of the spectrum. Radiant power enters the entrance slit of the

21
 diffraction grating which restricts the radiant power to a narrow spectral band and
transmits it through the exit slit to the photosensitive surface of the detector.

Operation of the grating spectrometer:

LED light first passes through an entrance slit and is reduced to a ray of light
when it hits the reflective grating.
When the light from the LED hits the reflective grating in the spectrometer, it
is spread out into its frequency components. This process is shown in Figure
13.
Scattered
components

LED light

Figure 13: Scattering of LED light

The LED light source passes through a region holding a monatomic gas. Upon
collision with electrons, and with each other, some atoms in the discharge are put
into a higher energy state than normal. Returning to their normal state, the atoms
give up their excess energy through a series of transitions in which the electron
drops from excited states to energy states of successively lower energy, until
finally reaching the ground state. In each transition electromagnetic radiation is
emitted with a wavelength that depends on the energy lost by the electron. Thus
the incident LED ray is scattered into wavelengths, with the scattering angle
depending on the wavelengths.

22
6. CONCLUSION

From the experiment, the I-V characteristic graph plotted from experimental values
coincides with the theoretical graph, the only difference being the inversion of the
graph due to circuit connections. Current through the diode starts to increase rapidly
when the voltage across the diode becomes greater than the turn on voltage.

It is also observed that turn on voltage decreases with increasing temperature. This is
due to the lowering of the energy barrier at the p-n junction, facilitating the flow of
holes down the energy barrier and electrons against the energy barrier. Experimental
results also show that reverse saturation current indeed falls within the range 10-18 A
and 10-9 A. In fact, the value of the saturation current is so small (5.5 x 10-10 A) that it
is not visible from the display of the oscilloscope. Hence, a straight line is observed
from -550 mV to 0 mV on the x-axis. The ideality factor obtained also falls with the
range from 1 to 2.

Through the experiments with the LED, it is discovered that the light current
relationship of an LED is linear, although certain limitations in device set-up
conditions or material properties may cause this relationship to deviate from the
norm. Investigation of the emission spectrum of LED illustrates that linewidth
approximates the range of frequencies of photons that can be emitted via spontaneous
emission and that LED light is not monochromatic. Also, it is shown that the energy
of the LED emission (1.89 eV) is approximately equivalent to the bandgap energy of
a typical semiconductor.

23
7. REFERENCES

John Jackson Sparkes, Semiconductor Devices, Chapman and Hall , 1994

Wikipedia
L227 Lab Manual
EE2003 Lecture Notes
EE2002 Lecture Notes
B.G. Streetman, Solid State Electronic Devices 5th edition, Prentice Hall, 2000
P.Bhattacharya, Semiconductor Optoelectronic Devices, 2nd edition, Prentice
Hall, 1997

24