Journal of Computer and Communications, 2013, 1, 46-49
Published Online December 2013 (http://www.scirp.org/journal/jcc)
http://dx.doi.org/10.4236/jcc.2013.17011
Open Access JCC
The Study of Relaxation Time in Test of 940 nm
Semiconductor Laser*
Jiachun Li, Jianjun Li, Tao Liu, Bifeng Cui, Jun Deng, Jun Han, Linjie He, Shengjie Lin
Key Laboratory of Opto-Electronics Technology, Beijing University of Technology, Beijing, China.
Received September 2013
ABSTRACT
Conventional test of the peak wavelength of a laser used to be applied immediately after a device is injected current.
However, the results can not be considered as an accurate description to temperature characteristic. This passage puts
forward a concept of relaxation time in wavelength texts, mainly based on the experiment of 940 nm strain quantum
well laser, confirming that under constant current, wavelength will get through a process of rising, and then, reach the
limit. This process brings th e effect on sp ectral characteristics of a d evice which cannot be ignored and the accumulated
heat in relaxation time will gradually impact the emission wavelength of the laser, even crest split to form bimodal phe-
nomenon.
Keywords: Semiconductor Laser; Test; Temperature Characteristics; Relaxation Time
1. Introduction
With the development of fiber-optical technology, the
fiber pump sources manufacture has been the critical te-
chnology to improve the signal transmission efficiency.
The strongest absorption peak of Yb-doped optical fiber
is around 976 nm [1]. However, the absorption peak is
sharp; as a result, the requirements of wavelength and
bandwidth ar e very strict, which makes it suitable for the
application of pulse signal. While the absorption peak
near 940 nm has high absorption bandwidth, it will not
appear concentration quenching phenomenon, which is
beneficial to the application of continuous signal. As a
consequence, the research of laser pump sources used by
940 nm semiconductor laser with reliable and stable high
power output has a gr e a t signifi c a nc e .
In fact, output power of laser is constantly improving,
and the problem of heating effect becomes more serious.
The quantity of heat produced by a working laser makes
the temperature of active layers rise sharply, which re-
sults in the wavelength red shift and the cavity surface or
interior to burn down [2]. So it is an important issue to
study the temperature character of the device. Nowadays,
the study on testing the heat character of the device by
controlling the heat or changing its current is major. But
it is rare to study the effect on the device itself made by
the heat which is accumulating with time and produced
by the device when it works under constant current. This
paper will study and analyze this effect, and makes it
clear that the influence on the spectral characteristic of a
device. Moreover, we put forward a concept of relaxation
time that wavelength could test after it, which can esti-
mate a device’s spectral character more accurately.
2. Put Forward the Inference
In conventional test of semiconductor laser, the test is
used to be applied immediately after a device is injected
current. However, in practical applications, the device often
works under constant current conditions, Understanding
the effect on wavelength characteristics by the device’s
heat accumulation over time will help to accurately as-
sess the thermal properties.
Wavelength change is due to heating of the device
causing the band gap widens, thereby causing the de-
crease of photon energy, and the increase of wavelength.
Heating of a device operates primarily from the P-clad-
ding layer, P-cladding layer is heavily doped, the resis-
tance rate is high, the device generates a lot of heat while
continue to work. In certain process conditions, the vo-
lume resistivity will generate heat in a certain range, with
certain cooling capacity of the package, we predicted that
if a device work under a certain current conditions, after
some time, heating degree will reach a state of equili-
brium. As a response, the temperature and wavelength of
the device will reach a state of dynamic equilibrium the
*
Supported by the advanced technology fund of Beijing University of
Technology, No.002000514312003.

The Study of Relaxation Time in Test of 940 nm Semiconductor Laser
Open Access JCC
47
same. As this process related to the change of tempera-
ture, we called it as dynamic thermal relaxation process,
the corresponding time is defined as the thermal relaxa-
tion time τT.
In order to verify the existence of relaxation time, we
carried out a set of validation experiments.
3. Experimental Verification
3.1. Chip Materials and St ructural Design
An epitaxial wafer used in the test had been grown b y the
MOCVD system, the epitaxial structures are shown in
Figure 1, the epitaxial layer were mainly in turn: GaAs
substrate; GaAs buffer layer; Al0.3GaAs, up and down
cladding layers, 700 nm; the two Al0.1GaAs/InGaAs com-
posed of strained quantum well optical waveguide layer;
GaAs ohmic contact thickness 450 nm. Center wavelength
is designed to 940 nm.
The electrode structures of the device are designed to
ridge, the ridge width of 100 μm. The current blocking
layer is grown by using PECVD systems, 200 nm of SiO2
material are obtained, P-type electrode material is Ti an d
Au, N-type electrode material is AuGeNi alloy and Au,
TO3 package with seat tube, adapter copper heat sink for
the heat sink, the package will be P type electrode lay a flip
process, the final packaged chip cavity length is 2000 μm.
3.2. Experimental Design and Installation
Experiment is using the spectral tester to test the spectral
characteristics of the instrument, first set up the device
test injection current I0 and projected peak wavelength λp,
then began spectral measurement, the instrument will
automatically λp’ centered at 40 nm spectral range scan,
the spectrum to be combined with precise measurements
to complete the rotation of the grating, spectral measure-
ments to each scanning time 25 s, 25 s intervals so expe-
riment a peak wavelength for data recording, the device
is continuously administered during the current I0.
Figure 1. Epitaxial structure.
3.3. Results and Discussion
Figure 2 shows a spectrum when the injection current I0
is 1.5A. Four curves mean four devices which fabricated
by the same process and packaged. Abscissa represents
for the time of duration injection current I0, and ordinate
represents the output peak wavelength of light λp, injec-
tion current always remains the same. The change of
peak wavelength would be obtained by the curves in
Figure 2 The peak wavelength gets to an extreme value
with the time increasing. This verifies the existence of
the thermal relaxation time. The period, which from the
beginning of the test to the peak wavelength gets a steady
data, is called the test relaxation time.
The following equation which relates to peak wave-
length λp and the relaxation time τT can be got.
λp = a/(1 + c*exp(−b*τT)) (1)
In Equation (1), a, b and c are model parameters, and
different parameters corresponding to different devices.
In this paper, it based on device No.1 as an example for
fitting. The fitting curve showed in Figure 3 and gets
that a = 950.7，b = 0.01114，and c = 0.009262.
Figure 2. I0 = 1.5A Red shift curve of peak wavelength in
relaxation time.
Figure 3. Simulation of peak wavelength changing in relax-
ation time.
Al0.1GaAs
Al0.1GaAs
GaAs
Al0.3GaAs
GaAs
Al0.1GaAs
GaAs
Al0.3GaAs
InGaAs

The Study of Relaxation Time in Test of 940 nm Semiconductor Laser
Open Access JCC
48
4. Further Study on Relaxation Time
In order to study the variation of spectral characteristics
with time in the relaxation time, we carried out the fol-
lowing experiments:
4.1. The Change of Peak Wavelength during
Relaxation Time
One device set to continuous testing with change the in-
jection current I0. Firstly, the inj ection current of 1.5A is
set. Each 25 seconds record its emission peak wavelength
and until the peak wavelength remains a steady value and
dynamic stability condition. Then set the injection cur-
rent to 1.5A and repeat the step above to measure the
data. Later in the same method the peak wavelengths is
measured at 2A and 3A Units
Figure 4 shows the spectrum chart when the injection
current at 1A, 1.5A, 2A and 3A. After relaxation time,
which keep applying test current at 1A, a limit peak wa-
velength of 944 nm and 2.5 nm red shift are obtained. At
I0 = 3A, red s hift gets to 5 nm during the relaxation time.
Figure 3 shows when larger injection current is meas-
ured the larger limit value of red shift should be got. And
increased the current value by a step 1A, the change of
wavelength most gets to steady state within 300 seconds.
It cannot ignore the heat accumulated with the increase
of time even run at a constant current. A deviation is
taken into the test results by the heat effecting and when
increasing injection current the heat effect on emission
wavelength of the device is also gradually increased.
Predictably, when high power laser continue to run at a
large current the heat produced during relaxation time
will have great influence on the output wavelength. There-
fore, to get a more accurate measure of the characteristics
the influence of relaxation time would be considered.
4.2. Device Temperature Varia tion in Relaxation
Time
In order to study the relationship between device temper-
Figure 4. Redshift curve of peak wavelength in different
current conditions.
ature and peak wavelength in relaxation time, we did a
test to observe temperature changes. We use a cooler to
keep heat sink at room temperature 25˚C, the measured
temperature with instrument is the highest temperature of
the central area of the device. Inject a constant current of
3A to the device, record the temperature and the peak
wavelength every 25 s. As shown in Figure 5, curve
slowly down after, and gradually stabilized after 200 s.
So, the temperature change as we predicted in the pre-
vious experiment. From the curve, in 3A constant current,
we can see that the device temperature changes from 302
K to 307 K, the average rate of change over time as
0.025 K/s in the relaxation time.
Figure 6 is the relationship curve between the device
temperature and the peak wavelength, the wavelength
increases with increasing temperature. As shown in data
distribution, data corresponding to the abscissa increa-
singly dense, indicating that changes of temperature the
more backward the smaller it is. The rising trend has be-
come increasingly slow curve, corresponding to the wa-
velength change it is getting smaller and.
Figure 5. I0 = 3A The temperature of device change in re-
laxation time.
Figure 6. I0 = 3A Device temperature and the peak wave-
length of the output light of the curve.

The Study of Relaxation Time in Test of 940 nm Semiconductor Laser
Open Access JCC
49
5. Summary
Through the test of the wavelength of 940 nm strain
quantum well laser under the constant current, the rise of
temperature and the temperature will reach the limit.
With the constant current increased, the limit value will
be greater. At room temperature, under 3A continuous
current condition, the peak wavelength of the laser will
reach steady state after 300 s. Compared with the imme-
diately measured data, the wavelength has increased by 4
nm and the temperature increased by 5˚C. And as the
current increases, the accumulated heat in relaxation time
will gradually impact the emission wavelength of the
laser, even crest split to form bimodal phenomenon. Un-
der the constant current, in order to get a well output
spectral characteristics, the accumulated heat in relaxa-
tion time must be considered. Selecting suitable operat-
ing current should be based on the device performance
after the relaxation time.
REFERENCES
[1] R. Pathak, J. M ine lly , J. Haapamaa, J. Watson, D. Schleu-
ning, H. Winhold, et al., “915 nm Laser Bar-Based High-
Performance Sources for Fiber Laser Pumping,” 2009,
Article ID: 719808
[2] X. Z. Ma, J. Huo, Y. Qu and S. L. Du, “8 Temperature
Characteristics of 808 nm Semiconductor Lasers,” Infra-
red and Laser Engineering, Vol. 12, 2010, pp. 1306-
1309.
[3] G. R. He, W. J. Shen, Q. Wang, W. H. Zheng and L. H.
Chen, “Temperature Characteristics of 980 nm High Pow-
er Vertical Ca vity Surface Emi tting Lasers,” Infrared and
Laser Engineering, Vol. 1, 2010, pp. 57-60.
[4] H. W. Qu, X. Guo, L. M. Dong, J. Deng, X. L. Da, Z. T.
Xu and G. D. Shen, “St udy on the Temperature Characte-
ristics of Vertical Cavity Surface Emitting Laser,” Infra-
red and Laser Engineering, Vol. 2, 2005, pp. 83-86.