Investigating the Modal Dependencies of Beam Quality
Via Spectrally-resolved Imaging of the Mode Structure in Diode Lasers
Stephen M. Misak, James A. Beil, Rebecca B. Swertfeger and Paul O. Leisher
Rose-Hulman Institute of Technology, 5500 Wabash Ave., Terre Haute, IN 47803, U.S.A.
Keywords: Beam Quality, Tapered Diode Laser, Broad-Area Diode Laser, Echelle Grating Spectrometer, Spectrally-
Resolved Modes, Modal Power Distribution, Astigmatism.
Abstract: Laser beam quality is an important factor for free-space communication and other high power applications.
To achieve the power requirements for such applications, there is a trade-off with the M
2
Beam Quality factor.
While direct diode lasers offer higher efficiency in a smaller footprint compared to solid-state and fiber laser
systems, beam quality is poor due to multi-mode operation. M
2
measurements compliant with ISO 11146
standards require numerous measurements, especially for multimode lasers. It is possible to use faster, modal
decomposition methods for measuring M
2
by employing a spectrometer to spatially separate the modes of a
laser. This work presents a custom Echelle Grating Spectrometer for spatially separating laser modes. This
tool provides the basis for an alternative method of M
2
measurements via Modal Power Distribution analysis.
1 MOTIVATION
With improvements to spatial and spectral brightness,
diode lasers continue to gain interest in fiber pumping
and industrial cutting. Single emitter broad-area laser
(BAL) diodes have become a core part of many fiber
laser modules due to continued improvements in
reliability and efficiency (Kanskar et al, 2013).
However, limitations in brightness and beam quality
have prevented diode lasers from being directly used
in many applications, such as free space optical
communication. Mode-control of tapered diode lasers
has enabled increased spatial brightness using a
Master-Oscillator Power-Amplifier (MOPA)
structure, seen in Fig. 1, to achieve high power
without a highly multimodal beam degrading M
2
(Kelemen et al, 2009). However, the tapered profile
degrades the beam quality with an astigmatic beam
indicated by the virtual waist in Fig. 1. Comparing the
changes in mode structure and M
2
would enable
engineers to improve their understanding of
multimode operation on beam quality. Improving
beam quality in multimode operation of diode lasers
is advantageous due to their superior wall-plug
efficiency, optical power, and footprint size. Spatially
examining the impact of multimode operation on
beam quality provides a novel approach that can
procure more information to improve diode lasers for
industrial and pumping applications.
Figure 1: MOPA structure with ridge length (L
MO
), tapered
length (L
PA
), and astigmatism (L
a
).
2 EXPERIMENT
Three diode lasers were examined to investigate the
relationship between mode structure and beam
quality—a BAL, an offset BAL, and a tapered laser.
The BAL and offset BAL have a specified emitter
width of 95 μm with a 1.5 mm cavity length. The
“offset” BAL is labelled as such due to non-uniform
current injection – current is injected along one edge
of the back side of the chip only, resulting in a non-
uniform gain distribution in the lateral direction. The
tapered laser uses a MOPA structure with a 1.5 mm
L
MO
L
PA
L
a
vir tualwaist
Misak S., Beil J., Swertfeger R. and Leisher P.
Investigating the Modal Dependencies of Beam Quality - Via Spectrally-resolved Imaging of the Mode Structure in Diode Lasers.
DOI: 10.5220/0006152702450251
In Proceedings of the 5th International Conference on Photonics, Optics and Laser Technology (PHOTOPTICS 2017), pages 245-251
ISBN: 978-989-758-223-3
Copyright
c
2017 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved
245
ridge length and a 4.5 mm tapered region at a full
aperture angle of 6° leading to a 450 μm emitter. A
light-current-voltage (LIV) and spectral analysis
were performed on the lasers to compare wall-plug
efficiency, slope efficiency, and thermal wavelength
drift. For the tapered laser, beam waist measurements
were also performed to characterize the astigmatism.
The beam quality of the lasers was characterized
using the moving slit technique based on the ISO
11146 standard. Beam quality was measured with
respect to current, illustrating the trends in the beam
quality of the fast and slow axis. The lasers were also
examined using spectrally-dispersed mode imaging
with a custom Echelle spectrometer to demonstrate
the viability of using this tool for modal power
distribution analysis.
2.1 LIV and Spectrum
The LIV measurements were performed with a
thermopile and a digital multimeter; spectral
measurements were performed by fiber coupling light
into an optical spectrum analyser. Spectral
measurements were recorded after the LIV data in
order to measure the spectrum at the operating
conditions for peak wall-plug efficiency. To obtain an
estimate of the thermal wavelength drift, the spectrum
was measured at 20°C and 25°C.
2.2 Beam Profiler
In order to measure the M
2
value of the lasers and the
astigmatism in the tapered laser, the beam was
collimated with an 8 mm focal length asphere and
focused with a f = 40 mm plano-convex lens. To
minimize lens aberrations, the curved sides of the lens
were facing the collimated region of the beam. A
reflective neutral density (ND) filter was also placed
in the collimated region and an absorptive ND filter
was placed after the lens to reduce the intensity at the
detector, preventing saturation. Figure 2 shows the
diagram of the system and the experimental setup
used for measurements. When using the system to
measure the beam caustic, the 40 mm focal length
lens was swapped out for a f = 75 mm lens to provide
a higher magnification. The astigmatism of the laser
can be measured with greater precision at high
magnification because the distance between the focal
points increase with the magnification squared.
Equation 1 expresses the relation between beam
caustic (Δz) and astigmatism (ΔL) based on the
moving beam profiler method (Sun, 1997).
ΔΔ
(1)
When using the beam profiler for M
2
measurements,
the 4σ beam waist was measure at more than 10
points. The M
2
value was determined by fitting Eq. 2
to the measured data (Crist and Nelson, 2012).
(2)
To ensure an accurate fit with the data, measurements
were taken near the minimum beam waist and in the
linear region of propagation.
Figure 2: System design diagram of the beam quality
measurement system and overhead view of the
experimental setup with labelled parts.
2.3 Custom Echelle Spectrometer
A custom Echelle spectrometer shown in Fig. 3 was
used to image the spectrally-resolved modes of the
diode lasers. The first section of the spectrometer
focuses the beam through an f = 13.86 mm asphere
and reimages it with 10 times magnification using an
f = 100 mm best form lens. For high power lasers, a
plate beam splitter is place in the region between the
apshere and best form lens. The second part of the
system reimages the beam 1:1 with spectrally
dispersed modes. The f = 500 mm lens collimates the
beam and acts as the lens for 1:1 imaging. In
collimated space, the beam double passes the Echelle
grating using a dielectric mirror with 99% reflectivity
from 980 - 1025 nm to reflect the single pass back to
the grating with minimal aberrations. To achieve high
(2) Asphere
(1) Diode
(4) Plano-convex
Lens
(5) ND filter
(3) ND filter
(6) Beam
Profiler
PHOTOPTICS 2017 - 5th International Conference on Photonics, Optics and Laser Technology
246
Table 1: Laser Wavelength and Power Characteristics.
Laser
Wavelength at
peak efficiency
Peak Wall-Plug
Efficiency
Turn-on
Current
Slope
Efficiency
Power at peak
efficiency
Thermal
Wavelength Drift
BAL 972 nm 62.4% 0.29 A 1.02 W/A 2.29 W ~0.7 nm/K
Offset BAL 813 nm 52.9% 0.41 A 1.14 W/A 2.76 W ~0.3 nm/K
Taper 979 nm 33.8% 1.65 A 0.85 W/A 5.74 W ~0.1 nm/K
Figure 3: System design diagram of the custom Echelle spectrometer and overhead view of the experimental setup with
labelled parts.
dispersion in the Littrow configuration, the grating
operates with peak intensity in the 6
th
diffraction
order with 316 lines/mm and a blaze angle of 63°. To
achieve an optimum magnification for image clarity
and size, an f = 50 mm best form lens magnifies the
dispersed mode image by a factor of four.
3 EXPERIMENTAL RESULTS
LIV and spectral analysis were in good agreement
with results from BAL and tapered laser diode
measurements in prior work by Kelemen et al,
showing that tapered lasers offer higher maximum
output power at lower efficiency. M
2
increased with
current as expected for multimode diode lasers
(Kelemen et al, 2004). The images from the Echelle
Spectrometer illustrated the cause of the increasing
M
2
with higher order modes lasing as the current
increased.
3.1 Basic Laser Characteristics
The results of the LIV and spectral measurements are
shown in Table 1 and Fig. 4 – 5 to compare the
characteristics of the three diode lasers. These results
are typical of tapered and broad area lasers. BAL have
higher slope efficiency than tapered lasers due to
additional losses and higher internal laser temperature
(Kelemen et al, 2004). While efficiency is lower in
tapered lasers, higher powers are achievable from a
single emitter. The tapered laser used in this work is
rated to 6 W at a 300 mA ridge current and a 10 Amp
taper current. The tapered laser also has a smaller
spectral bandwidth compared to the BAL and the
offset BAL. For the offset BAL, its wall-plug
(2) Asphere
(3) Best Form
Lens
Magnified
Image
(4) Beam
Splitter
(5) Collimating
Lens
(6) Echelle
Grating
(1) Diode
Magnified Image
with Dispersion
Image and
Magnify
Re-image with
Dispersion
(8) Best Form
Lens
Image and
Magnify
(7) Mirror
Investigating the Modal Dependencies of Beam Quality - Via Spectrally-resolved Imaging of the Mode Structure in Diode Lasers
247
efficiency is lower than the BAL, but it yields more
power and a higher slope efficiency.
Figure 4: LIV measurements for the (a) BAL, (b) offset
BAL, and (c) tapered laser diodes.
Figure 6 depicts the astigmatism in the tapered laser
diode which followed the expected increasing trend
at higher operating currents based on prior work by
Kelemen et al 2004. The increasing astigmatism
occurs as a result of Snell’s law and the changing
refractive index of the laser due to temperature
variations with increasing current. As the current
heats the tapered region, the distance increases
between the output facet and the virtual waist with the
astigmatism inversely proportional to the effective
refractive index (Kelemen et al, 2004).
Figure 5: Spectral measurements for the (a) BAL, (b) offset
BAL, and (c) tapered laser diodes using fiber coupled light
into an OSA with 0.02 nm resolution.
3.2 Laser M
2
Beam Quality Factor
Beam propagation theory predicts that for multimode
lasers, the presence of higher order modes will cause
the M
2
beam quality factor to increase as the modes
are larger than the diffraction limited beam. In this
sense, M
2
serves as a “times diffraction limited”
factor (Siegman, 1993). Figures 7 and 8 detail the
0
10
20
30
40
50
60
70
0.0
0.4
0.8
1.2
1.6
2.0
2.4
2.8
3.2
0.0 0.5 1.0 1.5 2.0 2.5 3.0
Efficiency
Voltage (V) & Power (W)
Current (A)
(a)
0
10
20
30
40
50
60
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
Efficiency
Voltage (V) & Power (W)
Current (A)
(b)
0
5
10
15
20
25
30
35
40
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
012345678910
Efficiency
Voltage (V) & Power (W)
Current
(
A
)
(c)
0.0
0.2
0.4
0.6
0.8
1.0
964 968 972 976
Relative Intensity (a.u.)
Wavelength (nm)
20 C
25 C
(a)
0.0
0.2
0.4
0.6
0.8
1.0
809 811 813 815
Relative Intensity (a.u.)
Wavelength (nm)
20 C
25 C
(b)
0.0
0.2
0.4
0.6
0.8
1.0
978.0 978.5 979.0 979.5
Relative Intensity (a.u.)
Wavelength (nm)
20 C
25 C
(c)
PHOTOPTICS 2017 - 5th International Conference on Photonics, Optics and Laser Technology
248
Figure 6: Tapered laser astigmatism measurements
recorded using the beam profiler setup.
increasing trend in M
2
as a function of current for the
two broad-area lasers. We expect from theory that the
M
2
value increases with high drive currents due to the
additional lasing modes. The measurements appear to
follow a linear trend which suggests that the modal
power distribution is non-uniform at these lower
current values based on Eq. 3 (Siegman and
Townsend, 1993). For constant normalized power
coefficients (c
n,m
) and mode indices (n), M
2
is
proportional to n
2
when adding more modes at higher
current. Based on this equation, the M
2
value for the
beam can be measured if the modes of the laser can
be isolated for modal power distribution analysis.
,
,
21
(3)
Figure 7: Slow axis M
2
measurements for the BAL and
Offset BAL with increasing drive current.
3.3 Spectrally-Resolved Modes
The mode structure images (Fig. 9 – 11) from the
spectrometer confirm the additional higher order
modes with increasing current from 0.9 - 1.0 amps.
Variations in the irradiance of individual lateral
modes is observed at the camera, so the non-uniform
Figure 8: Fast axis M
2
measurements for the BAL and
Offset BAL with increasing drive current.
power distribution is plausible theory. The images
shown below are a small section of the full mode
structure visible on the camera. The BAL modes have
the same structure as those observed in previous work
with repeating groups of longitudinal modes, each
with several lateral modes (Stelmakh, 2009) (Crump
et al, 2012). Figure 9 demonstrates the viability for
the spectrometer system to measure M
2
via the modal
power distribution method because the lateral modes
are spatially separated enough to use image analysis
for determining the normalized power coefficients.
Figure 9: Spectrally-resolved modes of the BAL.
The longitudinal modes of the tapered laser were
resolved with a ridge current of 100 mA and a tapered
amplifier current of 3.25 amps, but the astigmatism in
the laser prevented simultaneous focus in both axes.
A cross-cylinder pair could be added to correct this
issue in future work. Based on the separation between
the modes in Fig. 10, the modal power distribution
method for M
2
appears feasible with the addition of
optics to correct the astigmatism.
Figure 10: Spectrally resolved modes of the tapered laser.
0.70
0.75
0.80
0.85
0.90
0.95
1.00
02468
Astigmatism (mm)
Taper Current (A)
y = 32.00x + 23.60
R² = 0.95
y = 28.29x + 1.91
R² = 0.99
0
10
20
30
40
50
60
0.0 0.5 1.0 1.5
Slow Axis M
2
Current (A)
Offset BAL
BAL
y = 0.0531x + 1.1377
R² = 0.9541
y = 0.0491x + 1.1355
R² = 0.9693
1.160
1.165
1.170
1.175
1.180
1.185
1.190
1.195
0.0 0.5 1.0 1.5
Fast Axis M
2
Current (A)
Offset BAL
BAL
Investigating the Modal Dependencies of Beam Quality - Via Spectrally-resolved Imaging of the Mode Structure in Diode Lasers
249
For the offset BAL, the mode structure in Fig. 11
resembles the modes of the standard BAL, but
individual lateral modes are not able to be resolved.
The blurred mode profile could result from the non-
uniform current injection. The indistinct quality of the
modes demonstrates the limitations for isolating
lateral modes. Lasers with irregular mode structures
and mode spacing smaller than 3 pm cannot be
properly resolved with this spectrometer, limited by
the resolution of the Echelle grating (Misak, 2015).
Figure 11: Spectrally dispersed modes of the offset BAL.
4 MODAL DECOMPOSITION
While using a scanning beam profile to determine M
2
via the ISO standard method produces reliable and
consistent results, other methods exist that provide
accurate measurements in less time. Schmidt et al
demonstrated a system that measures the modal
amplitudes to measure the beam quality in real-time
while maintaining agreement with ISO11146 beam
quality measurements. In order to acquire the modal
amplitudes, the system employed computer generated
holograms and complex analysis (Schmidt et al,
2011). With the Echelle spectrometer shown in this
work, similar measurements of the modal power can
be made, provided that the mode separation is large
enough to spatially isolate the modes on the camera.
With the appropriate numerical analysis, BALs and
tapered lasers with similar characteristics to those in
this work can be used for modal decomposition M
2
measurements with the spectrometer. Multimode
VCSELs have also been shown to have enough
spatial separation in previous work (Misak, 2015).
5 CONCLUSIONS
Beam quality remains an important factor in many
high power applications. With more in-depth analysis
of the multimode structures and the impact of higher
order modes on beam quality, engineers can develop
new techniques to improve the performance of high
power laser diodes. Further analysis can be performed
by utilizing the spatial separation between modes
with the Echelle spectrometer described in this work.
This tool provides as basis for additional research into
the impact mode power distribution on the beam
quality of diode lasers.
ACKNOWLEDGEMENTS
The authors acknowledge the support provide by
NASA with award number NNX16AD20G.
REFERENCES
Crist, J. and Nelson, C. (2012) ‘Predicting laser beam
characteristics’, Laser Technik Journal, Vol. 9, No. 1,
Available from: doi.org/ 10.1002/latj.201290006.
Crump, P. et al. (2012) ‘Experimental and theoretical
analysis of the dominant lateral waveguiding
mechanism in 975 nm high power broad area diode
lasers’, Semiconductor Science and Technology, Vol.
27, No. 4, Available from: doi.org/10.1088/0268-
1242/27/4/045001.
International Organization for Standardization 2005,
Lasers and laser-related equipment – Test methods for
laser beam widths, divergence angles and beam
propagation ratios – Part 1: Stigmatic and simple
astigmatic beams, ISO 11146-1:2005, International
Organization for Standardization, Geneva.
Kanskar, M. et al. (2013) ‘High Reliability of High Power
and High Brightness Diode Lasers’, nLight
Corporation, Available from: http://www.nlight.net/
news/technical-papers.
Kelemen, M. T. et al. (2009) ‘High-Power High-Brightness
Lasers’, m2k-laser GmbH, Available from: http://
www.optosolutions.com/doc/TaperedLaser_intro_070
209.pdf (Accessed 08 August 2016).
Kelemen, M. T. et al. (2004) ‘Astigmatism and Beam
Quality of High-brightness Tapered Laser Diodes’,
Semiconductor Lasers and Laser Dynamics,
Strasbourg, France, SPIE, Vol. 5452, No. 233.
Available from: doi.org/10.1117/12.545221.
Misak, S. M. et al. (2015) ‘Spectrally Resolved Imaging of
the transverse modes in multimode VCSELs’, Vertical-
Cavity Surface Emitting Lasers, San Francisco, CA,
SPIE, Vol. 9381, No. 93810L. Available from:
doi.org/10.1117/12.2076629.
Schmidt, O. A. et al. (2011) ‘Real-time determination of
laser beam quality by modal decomposition’, Optics
Express, Vol. 19, No. 7, pp. 6741-6748. Available
from: doi.org/10.1364/OE.19.006741.
Siegman, A. E. (1993) ‘Defining, measuring, and
optimizing laser beam quality,’ Laser Resonators and
Coherent Optics, Los Angeles, CA, SPIE, Vol. 1868,
No 2. Available from: doi.org/10.1117/12.150601.
Siegman, A. E. and Townsend S. (1993) ‘Output beam
propagation and beam quality from a multimode stabe-
PHOTOPTICS 2017 - 5th International Conference on Photonics, Optics and Laser Technology
250
cavity laser’, IEEE Journal of Quantum Electronics,
Vol. 29, No. 4, pp. 1212-1217. Available from:
doi.org/10.1109/3.214507.
Stelmakh, N. (2009) ‘External-to-cavity lateral-mode
harnessing devices for high-brightness broad-area laser
diodes: concept, realizations, and perspectives’ Novel
In-Place Semiconductor Lasers, San Jose, CA, SPIE,
Vol. 7230, No. 72301B. Available from:
doi.org/10.1117/12.808681.
Sun, H. (1997) ‘Measurement of laser diode astigmatism’,
Opt. Eng., Vol. 36, No. 4, pp. 1082-1087 Available
from: doi.org/10.1117/1.601147.
Investigating the Modal Dependencies of Beam Quality - Via Spectrally-resolved Imaging of the Mode Structure in Diode Lasers
251
