SEMICONDUCTOR OPTICAL DEVICE WITH IMPROVED EFFICIENCY AND
OUTPUT BEAM CHARACTERISTICS
BACKGROUND OF THE INVENTION
A. Field of the Invention
[0001] The invention relates generally to light emitting devices, more particularly, to
semiconductor optical lasers having an improved design to reduce internal optical loss.
B . Description of the Related Art
[ooo2] Semiconductor lasers are attractive sources for optical power generation because
they are more efficient, smaller, and less expensive than other types of lasers. There are
increasing applications for higher output power single-spatial-mode semiconductor lasers
such as for pumping optical fiber amplifiers/lasers, optical wireless communications,
optical fiber transmitters, and other laser applications. The output optical power from a
semiconductor laser (P) is governed by a simple relationship: P = S x (I - Lh). As the
current (I) injected into the semiconductor laser is increased above the threshold current
value (Lh), the output power increases proportionally with a constant of proportionality S
known as the slope efficiency. High power semiconductor lasers are operated at current
injection levels high above threshold. Therefore, the slope efficiency is the most dominant -
parameter determining the attainable maximum output power. It is at least partially for this
reason that increasing the slope efficiency is critical for increasing the output power of
semiconductor lasers. An additional benefit of increased slope efficiency is an increase in
the net electrical-to-optical conversion efficiency which reduces power consumption and
heat generation. However, for single-mode applications, such as coupling light into optical
fibers, it is crucial that the angular intensity profile of the optical radiation emitted by the
semiconductor laser remain nearly diffraction limited. This requires that the optical mode
at the laser output facet must be single-lobed and preferably have an intensity profile that is
substantially gaussian. Therefore, increasing the slope efficiency of a single-mode laser is most beneficial if the optical mode at the output facet retains a single-lobed intensity
profile.
[ooo3] The slope efficiency is directly influenced by the background optical loss that the
optical mode experiences inside the semiconductor laser. A lower internal optical loss
results in a higher slope efficiency. For conventional semiconductor lasers, more than 50%
of the lasing mode is propagating in the doped cladding region, where the optical loss due
to free-carrier absorption is significant. This loss is particularly large for the p-side
cladding. The internal loss can be reduced by using broadened waveguide structures as
described in Electron. Lett., vol. 32, pp. 1717-1719, 1996. However, a broadened
waveguide may cause adverse effects as well. The increased carrier transport time in the
broadened guiding layer will increase the population of carriers in the guiding layers and
enhance the carrier recombination. Consequently, the temperature sensitivity of the slope
efficiency is deteriorated at high temperatures. The modulation bandwidth will also be
degraded for broadened waveguide lasers since the carrier transit time is increased.
[ooo4] Higher slope efficiency and reduced internal loss can also be achieved by using low
doping concentrations in the cladding layer as suggested by J.W. Pan, et. al. at the 10th Int.
Conf. on Molecular Beam Epitaxy, Cannes, France, 1998. However, electron leakage
current over the low doped cladding layer has been found to become significant with
temperature, and render threshold current and slope efficiency highly temperature sensitive.
[ooo5] Another approach to reduce the internal loss employs an asymmetric transversal
layer design to reduce the confinement factor of the optical mode in the active region
[IEEE Photon. Technol. Letters, vol. 11, no. 2, pp. 161-163, Feb. 1999]. The
confinement factor is the ratio of the portion of optical mode that overlaps with the active
region to the entire optical mode. This device uses an "optical trap" layer on the n-side of
the active region to lower the confinement factor. A fundamental disadvantage of this
approach, however, is that the transverse optical mode is no longer a single lobed gaussian
shape. Instead, it has two distinct lobes. Thus the far field emission pattern from the laser
is no longer diffraction limited. This distortion of the far field pattern significantly
degrades device performance including reduced coupling efficiency of the laser to a single
mode fiber.
SUMMARY OF THE INVENTION [0006] The present invention is directed to overcoming or at least reducing the effects of
one or more of the problems set forth above and other problems with the prior art.
[ooo7] According to one aspect of the present invention, an optical device is provided
comprising a gain section adapted to emit radiation at a radiation wavelength, a coupling
section adjacent to the gain section for transitioning radiation between an active waveguide
and a passive waveguide, and a passive section adjacent to the coupling section supporting a
single-lobed optical mode in the passive waveguide at the radiation wavelength. The
passive waveguide has an index of refraction and dimension such that the confinement of
the radiation within the active waveguide in the gain section is reduced.
[ooo8] According to another aspect of the present invention, a method of generating optical
radiation is provided comprising the steps of generating a multi-lobed lasing mode in a gain
section, one lobe of the lasing mode being substantially in an active waveguide in the gain
section, and coupling the multi-lobed lasing mode to a single lobed optical mode
substantially in a passive waveguide. The passive waveguide has an index of refraction and
dimension such that the confinement of the multi-lobed lasing mode within the active
waveguide in the gain section is reduced. Preferably, the passive waveguide'has an index
of refraction and dimension such that the passive waveguide supports a single lobed optical
mode in the passive section. More preferably, the passive waveguide has an index of
refraction and dimension such that the single lobed optical mode is a substantially
diffraction-limited gaussian mode.
[ooo9] According to another aspect of the present invention, an optical device is provided
comprising a gain section adapted to emit radiation at a radiation wavelength within an
active waveguide, the gain section adapted for supporting a multi-lobed optical mode
comprising a first and second lobe at the radiation wavelength, a coupling section adjacent
to the gain section for transitioning the radiation from the active waveguide to an n-doped
passive waveguide, and a passive section adjacent to the coupling section for supporting a
single-lobed optical mode in the n-doped passive waveguide. The passive waveguide has an
index of refraction and dimension such that confinement of the multi-lobed optical mode at
the radiation wavelength within the active waveguide in the gain section is reduced. The
peak intensity of the first lobe in the gain section occurs within the active waveguide and
the peak intensity of the second lobe in the gain section occurs within the passive
waveguide.
BRIEF DESCRIPTION OF THE DRAWINGS [ooιo] The foregoing advantages and features of the invention will become apparent upon
reference to the following detailed description and the accompanying drawings, of which:
[ooιi] Figure 1 is a cross-sectional view of a semiconductor laser according to an
embodiment of the present invention.
[0012] Figure 2 is a perspective view of a semiconductor laser according to an embodiment
of the present invention.
[ooi3] Figure 3 is a graph of Electric Field v. Transverse Distance in a gain section of a
semiconductor laser according to an embodiment of the present invention.
[0014] Figure 4 is a graph of Electric Field v. Transverse Distance in a passive section of a
semiconductor laser according to an embodiment of the present invention.
[0015] Figure 5 is a beam propagation simulation showing the light or radiation intensity in
the gain, coupling, and passive sections as radiation is coupled between the active
waveguide and the passive waveguide in a semiconductor laser according to an embodiment
of the present invention.
[ooi6] Figure 6 is a graph showing a single lobed output of a semiconductor laser
according to an embodiment of the present invention.
[ooi7] Figure 7 is a graph showing the change in the electric field profile of the optical
mode in the gain section for different passive waveguide parameters in a semiconductor
laser according to an embodiment of the present invention.
[0018] Figure 8 is a graph showing Laser Output Power v. Current curves for various
passive waveguide configurations in a semiconductor laser according to an embodiment of
the present invention.
DETAILED DESCRIPTION OF SPECIFIC EMBODIMENTS [0019] Reference will now be made in detail to presently preferred embodiments of the
invention. Wherever possible, the same reference numbers will be used throughout the
drawings to refer to the same or like parts.
[0020] A semiconductor laser 100 according to a first embodiment of the present invention
is shown in the cross-sectional view of Figure 1, and the perspective view of Figure 2. By
way of example, other optical devices according to the present invention may include light
emittin diodes (LED), semiconductor optical amplifiers, and other optical devices as
would be readily apparent to one skilled in the art.
[0021] In this first embodiment, the semiconductor laser 100 comprises a gain section 9, a
coupling section 10, and a passive section 11 which each may include multiple layers. These layers include a passive waveguide 2 and an active waveguide 4 separated by a
spacer layer 3. Preferably, the passive waveguide 2 spans the length of the laser 100. As
will be described later, the active waveguide 4 preferably spans at least the length of the
coupling section 10 and the gain section 9, but may be omitted in the passive section 11.
The coupling section 10 may be lengthened to include a section (as indicated by the dashed
line in Figure 2) that improves the fabrication tolerances. It should be appreciated that this
section indicated by the dashed line is thus not required for functionality, but is provided
for improved manufacturing of the device. The laser 100 further includes a substrate layer
1 and a bottom metallization layer 8 parallel to the passive waveguide 2 on a side opposite
the active waveguide 4. Also, the laser 100 includes a cladding layer 5, a cap layer 6, and
a top metallization layer 7 positioned parallel to the active waveguide 4 on a side opposite
the passive waveguide 2.
[0022] The function of these layers will now be described in reference to Figure 1 and
Figure 2. Top metallization layer 7 is used at least partially to provide a low resistance
metal contact for applying a current to the gain section 9 of the laser 100. For example,
top metallization layer 7 may comprise layered Ti/Pt/Au or Cr/Au. Similarly, bottom
metallization layer 8 is also used at least partially to provide a low resistance metal contact
for applying a current to the gain section 9 of the laser 100. For example, metallization
layer 8 may comprise layered Ge/Au or Ni/Au/Ge/Ni/Au. Gain section 9 is pumped with
current by applying a voltage between metallization layers 7 and 8, thereby providing the
gain required for lasing. Capping layer 6 is preferably heavily doped with a p-type
material, and is used to assist in forming an ohmic contact to the top metal contact layer 7.
[0023] Cladding layer 5 is configured so as to confine the optical mode within the active
waveguide and to separate the top metallization layer 7, having relatively high optical
absorption losses, from the active waveguide 4. The cladding layer 5 typically has an
optical absorption greater than an optical absorption of the passive waveguide 2, spacer
layer 3, and substrate 1 at the radiation wavelength because the cladding layer 5 preferably
comprises a p-type material. Cladding layer 5 is provided parallel to active waveguide 4 on
a side opposite to the passive waveguide 2.
[0024] Spacer layer 3 is configured so as to confine the optical mode in the passive section
11 to the passive waveguide 2. The spacer layer 3 is configured with an index of refraction
lower than the index of refraction of the passive waveguide 2 or the index of refraction of
the active waveguide 4, thereby acting as a cladding layer for the passive section 11.
Preferably, the spacer layer 3 spans the length of the laser 100.
[0025] Laser efficiency may be improved by reducing internal losses due to absorption, i.e.
internal absorption loss, of the optical mode within the laser 100 itself. Internal absorption
loss is typically greater in the active waveguide 4 and cladding region 5 than in the passive
waveguide 2, because the active waveguide 4 and cladding region 5 typically have a higher
absorption loss at the laser emission wavelength than the passive waveguide 2. The larger
loss in the active waveguide 4 and cladding region 5 is due, in part, to the significant free-
carrier absorption in the active waveguide 4 and cladding region 5. To reduce the internal
absorption loss in the laser 100, the laser 100 is configured such that confmement of a first
optical mode at the radiation wavelength (the emission wavelength for a laser) within the
active waveguide 4 is reduced, thereby shifting some of the first optical mode into the
passive waveguide 2. As would be readily apparent to one skilled in the art, the term
"confinement" and "confinement factor" are used synonymously.
[0026] Passive waveguide 2 is configured to reduce the confinement in the gain section 9
and also to support a single-lobed mode in the passive section 11 of the device. Absorption
in the passive waveguide 2 may be reduced by careful selection of materials with a larger
bandgap than a material of the active waveguide 4. In a preferred first embodiment, the
first material may be an n-type material with a lower optical absorption at the radiation
wavelength than the gain section 9, such as n-doped GaxInα-x)AsyP(i-y) or AlyGaxIno-x-y)As.
The internal loss in the laser 100 is reduced because some of the first mode propagates in
the passive waveguide, which has a lower absorption at the radiation wavelength of interest.
Active waveguide 4 is configured to generate radiation in the gain section of the device. A
preferred material for the active waveguide 4 is GaxInα-x)AsyPu-y) or AlyGaxInα-x-y)As. The
particular material may be altered to promote radiation at a specific desired radiation
wavelength.
[0027] Adjusting the materials to reduce internal absorption loss and reducing confinement
within the active waveguide 4, however, may result in a multi-lobed optical mode which is
not a desirable output mode in many laser implementations. As shown by the graph of
Figure 3, the optical mode in the gain section typically includes a first lobe and a second
lobe at least partially attributable to the reduced confinement in the active waveguide 4. A
multi-lobed optical mode has a low coupling efficiency to a single mode fiber, and is
therefor typically not a preferred output from the laser 100. The multi-lobed optical mode
can be converted to a single-lobed optical mode via coupling section 10. An example
coupling section 10 that couples an optical mode from the active waveguide 4 to the passive
waveguide 2 is a tapered coupler. Coupling section 10 comprises a taper where the optical
mode is transferred from the active waveguide 4 to the passive waveguide 2. The taper can
be a resonant coupled taper as shown in Figure 2, or adiabatic in nature. Resonant coupled
tapers are preferred as they are generally shorter in length than adiabatic coupled tapers. In
the case of semiconductor optical amplifiers, as would be readily apparent to one skilled in
the art, coupling sections adjacent to each end of the gain region could be implemented.
The passive section 11 comprises a facet that reflects radiation resonantly coupled to the
passive waveguide 2 so that the reflected radiation is coupled back to the active waveguide
4 through the taper. In.the case of semiconductor optical amplifiers, as would be readily
apparent to one skilled in the art, the passive section adjacent to the coupling section would
comprise a facet that reflects little or no radiation back to the coupling section.
[0028] Figure 2 shows a coupling section 10 that shifts the optical mode from the active
waveguide 4 to the passive waveguide 2 and converts a multi-lobed optical mode to a
single-lobed optical mode. Preferably, the coupling between the active waveguide and
passive waveguide is accomplished with low optical loss. Preferably, the active waveguide
4 is coupled to the passive waveguide 2 such that the output of the laser 100 will be a
diffraction-limited single-lobed substantially gaussian mode with a high coupling efficiency
to a single mode fiber. Figure 4 shows the electric field as a function of distance for such a
diffraction-limited single-lobed substantially gaussian mode within the passive waveguide 2
in the passive section 11 of a particular device embodiment. The passive waveguide 2
should be of a dimension and index of refraction such that a single-lobed mode exists in the
passive section 11 at the facet. As would be readily apparent to one skilled in the art, the
term "dimension" is intended to describe the physical dimension of the passive waveguide
2. Thus, the present invention achieves improved loss characteristics while maintaining a
highly-efficient single-lobed output mode.
[0029] As shown in the beam propagation simulation of Figure 5, coupling section 10 transfers the optical mode from the active waveguide 2, located between 1.7 μm and 2.0
μm on the vertical axis in Figure 5, to the passive waveguide 4, located between 0.2 μm
and 1.0 μm on the figure. For this simulation, a 200 μm long taper located between 0.1
mm and 0.3 mm on the bottom axis of Figure 5 was used to couple the optical power from
the active waveguide 4 to the passive waveguide 2. Approximately 95 % of the optical
mode was coupled between the active waveguide 4 and the passive waveguide 2. Further, a
single lobed optical mode was observed at an output of the passive waveguide 2 as
evidenced by the graph of Figure 6.
[0030] In this first embodiment, the active waveguide 4 has been completely removed in
section 11 as it is generally not required to stabilize the optical mode after being transferred
between the active waveguide 4 and the passive waveguide 2 in the coupling section 10.
Thus the optical mode is substantially confined within the passive waveguide 2 in the
passive section 11.
[0031] The present invention allows customization of the laser 100 for a particular
implementation by adjusting the properties of the passive waveguide 2. As shown in the
graph of Figure 7, line 12 shows a single lobe optical mode substantially only in the active
waveguide 4, as is common for conventional lasers. As the index of refraction of the
passive waveguide 2 is increased relative to the index of refraction of the active waveguide
4, progressively shown by lines 13, 14, and 15 respectively, the mode becomes two lobed.
This decreases the confinement in the active waveguide 4 and shifts the optical mode to the
passive waveguide 2, resulting in decreased internal absorption loss as previously
described. Thus, the index of refraction of the passive waveguide 2 can be selected to
achieve the desired confinement factor in the active waveguide 4
[0032] The present invention further allows higher output power as the laser 100 may be
operated with higher slope efficiency. As shown in the graph of Figure 8, semiconductor
laser output power is charted versus current for various passive waveguide 2
configurations. As the index of the passive waveguide 2 is increased, the confinement in
the active waveguide is reduced and the differential slope efficiency (change in
power/change in current) of the laser 100 increases. The differential slope efficiency
increase is at least partially attributable to the decreased internal absorption loss by reducing
the confinement in the active waveguide 2, and by pushing the mode, more into the passive
waveguide 4 where the optical absorption losses are lower. Increased differential efficiency
allows higher output power when the laser 100 is operated at high current far above
threshold.
[0033] A preferred implementation as shown in Figure 1 and Figure 2, is further described
in the following table:
[0034] The index and the bandgap of the passive waveguide 2 can be optimized to obtain
high power and low internal loss. The active waveguide 4 in this preferred embodiment is
a three quantum well waveguide in a separate confinement heterostructure. The spacer 3 is
composed of InP. The metal contact layer is preferably an ohmic contact for providing a
good electrical contact to the metal layers of the laser 100.
[0035] Thus, an optical device comprising a gain section 9, a coupling section 10, and a
passive section 11 has been described according to the present invention. Many
modifications and variations may be made to the techniques and structures described and
illustrated herein without departing from the spirit and scope of the invention.
Accordingly, it should be understood that the methods and apparatus described herein are
illustrative only and are not limiting upon the scope of the invention.