Efficiency limits of laser power converters for optical power transfer applications

Abstract

We have developed III–V-based high-efficiency laser power converters (LPCs), optimized specifically for converting monochromatic laser radiation at the eye-safe wavelength of 1.55 µm into electrical power. The applications of these photovoltaic cells include high-efficiency space-based and terrestrial laser power transfer and subsequent conversion to electrical power. In addition, these cells also find use in fibre-optic power delivery, remote powering of subcutaneous equipment and several other optical power delivery applications. The LPC design is based on lattice-matched InGaAsP/InP and incorporates elements for photon-recycling and contact design for efficient carrier extraction. Here we compare results from electro-optical design simulations with experimental results from prototype devices studied both in the lab and in field tests. We analyse wavelength and temperature dependence of the LPC characteristics. An experimental conversion efficiency of 44.6% [±1%] is obtained from the prototype devices under monochromatic illumination at 1.55 µm (illumination power density of 1 kW m−2) at room temperature. Further design optimization of our LPC is expected to scale the efficiency beyond 50% at 1 kW m−2.

Full text

This content has been downloaded from IOPscience. Please scroll down to see the full text.
Download details:
IP Address: 81.108.41.185
This content was downloaded on 10/05/2016 at 16:50
Please note that terms and conditions apply.
Efficiency limits of laser power converters for optical power transfer applications
View the table of contents for this issue, or go to the journal homepage for more
2013 J. Phys. D: Appl. Phys. 46 264006
(http://iopscience.iop.org/0022-3727/46/26/264006)
Home Search Collections Journals About Contact us My IOPscience

IOP PUBLISHING JOURNAL OF PHYSICS D: APPLIED PHYSICS
J. Phys. D: Appl. Phys. 46 (2013) 264006 (6pp) doi:10.1088/0022-3727/46/26/264006
Efficiency limits of laser power converters
for optical power transfer applications
J Mukherjee1, S Jarvis1, M Perren2and S J Sweeney1
1Advanced Technology Institute, University of Surrey, Guildford, GU27XH, UK
2EADS Astrium, 12 rue Pasteur, 92152 Suresnes, France
E-mail: j.mukherjee@surrey.ac.uk
Received 6 November 2012, in final form 29 January 2013
Published 11 June 2013
Online at stacks.iop.org/JPhysD/46/264006
Abstract
We have developed III–V-based high-efficiency laser power converters (LPCs), optimized
specifically for converting monochromatic laser radiation at the eye-safe wavelength of
1.55 µm into electrical power. The applications of these photovoltaic cells include
high-efficiency space-based and terrestrial laser power transfer and subsequent conversion to
electrical power. In addition, these cells also find use in fibre-optic power delivery, remote
powering of subcutaneous equipment and several other optical power delivery applications.
The LPC design is based on lattice-matched InGaAsP/InP and incorporates elements for
photon-recycling and contact design for efficient carrier extraction. Here we compare results
from electro-optical design simulations with experimental results from prototype devices
studied both in the lab and in field tests. We analyse wavelength and temperature dependence
of the LPC characteristics. An experimental conversion efficiency of 44.6% [±1%] is obtained
from the prototype devices under monochromatic illumination at 1.55 µm (illumination power
density of 1 kW m−2)at room temperature. Further design optimization of our LPC is
expected to scale the efficiency beyond 50% at 1kW m−2.
(Some figures may appear in colour only in the online journal)
1. Introduction
Most photovoltaic (pv) cells are designed and developed for the
conversion of the broad spectrum of solar energy into electrical
power; however, it is well known that a pv cell demonstrates
the maximum optical to electrical conversion efficiency when
illuminated by monochromatic light at a wavelength that
closely corresponds to the band-gap energy of the absorbing
material [1]. In fact, it is commonly argued that in order to
get maximum conversion efficiency from a solar cell, it should
consist of an infinite number of absorbers cascaded together,
each absorbing at the band-gap equivalent wavelength so as
to cover the entire solar spectrum [1]. Cascading multiple
single-wavelength absorbers inside a solar cell is not practical
and issues of current balancing remain challenging; recently,
however, there has been a renewed interest in the development
of single-junction pv cells for converting laser light into
electrical power for power-by-light (wireless or optical-fibre-
based power delivery) applications [2]. These laser power
converters (LPCs) find use in optical power transfer via optical
fibres leading to potential applications in optical local area
networks, a system that is free from electromagnetic noise
or fire hazards prevalent in conventional electrical power
transfer systems which use copper wires. Other useful
applications include remote powering of subcutaneous sensors
or transducers and several other free-space optical power
delivery applications. One such key free-space optical power
delivery application is conversion of beamed laser power
into electrical energy. Such lasers can either be terrestrially
based or on-board satellites. For laser beaming from space,
high-power lasers, either directly solar pumped or indirectly
pumped by diode lasers, which are in-turn driven by electrical
energy generated by on-board solar cells on a satellite (sun-
synchronous [3]), can be used to beam power down to earth
(or any other target location) where it is converted to electrical
power by LPCs. For optical power transfer via optical
fibres, 1.3 and 1.55 µm wavelengths are most suitable as
radiation at these wavelengths suffers the least attenuation
while propagating inside the fibre, thereby increasing the
overall system efficiency [4]. Laser power beaming from
0022-3727/13/264006+06$33.00 1© 2013 IOP Publishing Ltd Printed in the UK & the USA

J. Phys. D: Appl. Phys. 46 (2013) 264006 J Mukherjee et al
space demands very high-power lasers emitting at wavelengths
which suffer the least attenuation while propagating through
the Earth’s atmosphere. In addition, for terrestrial power
beaming applications and applications in remote powering
of subcutaneous equipment, the laser wavelength has to be
both eye and skin safe. It is now well known that radiation
at wavelengths longer than 1.4 µm is safe for both human
eyes and skin up to a power density of 1 kW m−2[5]. The
minimum attenuation optical transmission windows within the
Earth’s atmosphere are around 1.55 µm and 2.1 µm in the near-
and mid-infrared, respectively [6]. Ultra-high power high
brightness fibre lasers have now been demonstrated at both
1.55 and at 2.1 µm[7,8]. Thus, to cater to the applications
mentioned above, a pv cell optimized for maximum conversion
efficiency at 1.55 or 2.1 µm is required. We will focus on pv
cells for 1.55 µm radiation conversion in this work. It should
be noted that a dedicated pv cell for 2.1 µm will fundamentally
have lower conversion efficiency due to the lower photon
energy at 2 µm compared with 1.55 µm[9].
In the past, both InGaAs- and GaSb-based LPCs
have been designed for the conversion of near- and mid-
infrared monochromatic radiation, with InGaAs-based devices
generally demonstrating higher conversion efficiency mainly
due to the comparative immaturity in the GaSb growth
technology [10–12]. LPCs based on an In0.53Ga0.47As emitter
grown lattice-matched to InP has a band gap corresponding
to 1.65 µm. This LPC demonstrated a maximum conversion
efficiency of 34% at 1.55 µm[10]. GaSb (band edge
at 1.68 µm)-based LPCs have demonstrated a conversion
efficiency of 45% at 1.55 µm; however, at very high power
densities 1kWm
−2[11]. While both these LPCs can
convert light at 1.55µm, (i.e. radiation above the band gap
of the respective pn-junction semiconductor materials), the
conversion efficiency may not be maximized due to the so-
called ‘quantum deficit’ [13], i.e. high-energy electrons excited
above the band gap will have to relax down to the band edge
releasing heat into the lattice, before they can be extracted out
to the external circuit. To the best of our knowledge, there are
no reports on LPCs specifically designed and optimized for
the conversion of monochromatic light at 1.55µm, barring the
only report from NASA [14], which is based on InGaAs/InP
[15], hence likely to have similar efficiency characteristics as
reported in [10].
2. The LPC design and optimization
One of the key design requirements for an efficient photovoltaic
cell is fast extraction of the photo-generated carriers into an
external circuit before they recombine and are lost inside
the device. To this end, one needs to have the anode and
cathode contacts appropriately placed in the design together
with an optimized absorber design. For an LPC incident
with monochromatic radiation, a homo-junction (single-band
gap) absorber design works best, simply because the carrier
transport towards the electrical contacts is unimpeded by band-
offsets unavoidable in a hetero-junction absorber design. Our
LPC design is based on lattice-matched InxGa1−xAsyP1−y/InP
and incorporates a top anti-reflection coating (ARC) to allow
maximum transmission of the incident radiation on the LPC
to reach its absorber region [16]. InxGa1−xAsyP1−y/InP
allows us to achieve a wide range of absorber band gaps by
tuning the alloy composition while maintaining a lattice match
hence avoiding creation of defects and related photo-generated
carrier loss in otherwise lattice mismatched systems [17].
Photon-recycling [18] is achieved using the bottom n-contact
metallization which is a near perfect reflector of 1.55 µm
radiation. As the anode is made in direct contact with the
p-InxGa1−xAsyP1−ythrough a thin layer of lattice-matched
InGaAs, our design also incorporates a p-InP window layer
[19]. This layer blocks the hole-transport to the top surface
(semiconductor-ARC), via the valance-band offset between
p-InxGa1−xAsyP1−y/InP, where they can otherwise get lost
via recombination at surface defects [19] and ensures their
transport directly to the anode. In addition, InP, being a wider
band-gap material compared with InxGa1−xAsyP1−y, does not
block the radiation incident at 1.55 µm from reaching the
absorber region.
The optimal composition of the InxGa1−xAsyP1−y
absorber was decided under a closed iterative cycle of electro-
optic simulations [20] and characterization of prototype
devices [16]. The doping of each layer and the layer thickness
was optimized for maximizing the conversion efficiency. First
the pn-homojunction region of the LPC was optimized by
judiciously varying the absorber layer thicknesses and their
corresponding doping levels while keeping the doping level
and thicknesses of other layers fixed. The thicknesses and
doping levels of the other layers were then optimized. Finally,
the absorber material band edge was optimized for maximizing
the conversion efficiency by varying the material composition
at a fixed illumination power density. Further details may be
found in [16]. Simulated characteristics for the optimized LPC
as a function of incident intensity are shown in figure 1and
are similar to the analytical calculations reported by Pena and
Algora [9]. The difference between our results and those
reported in [9] can be attributed to the more sophisticated
model used here in comparison with a diode-equation-based
analytical model as used in [9]. In our device design and
optimization technique, first the LPC absorption for an un-
optimized design was calculated using the finite-difference
time-domain (FDTD) technique [20] and the results used for
carrier-transport calculation by solving the conventional drift–
diffusion equations [20], both in 2D (lateral and vertical). For
the absorption calculation, Neumann type boundary conditions
were used everywhere on the device edges except the top
surface where the incident radiation is incident on the device,
where a Dirichlet boundary condition was used. For the carrier-
transport calculations, Neumann type boundary condition was
used everywhere on the device edges except at the Ohmic
contacts for which Dirichlet boundaries are more appropriate.
The non-radiative recombination loss mechanisms taken
into account for the simulations were Shockley–Read–Hall
(SRH) and Auger recombination [20,21,24]. The material
parameters used were obtained from the in-built material
property database in the commercial software used for the
simulations, further details of which may be found in [20].
The aforementioned optimization procedure was then used to
2

J. Phys. D: Appl. Phys. 46 (2013) 264006 J Mukherjee et al
Figure 1. Simulated (electro-optical) dependence of the optical to
electrical conversion efficiency, photo-current (sheet) density Jsc
and open-circuit voltage VOC from the LPC as a function of incident
intensity (Pin)at 1.55 µm. Here, Jsc is the photo-current density
(perpendicular to the growth plane) under zero load whereas VOC is
the voltage across the LPC across an infinite load. Jsc is the
dark-current (sheet) density. This particular case is a simulation for
a p-on-n LPC where the absorber band edge is at 0.8 eV and Ohmic
losses are not considered. A flat wave-front of incident radiation is
used for calculating the absorption characteristics. A simulated
p-on-n design keeping the same epitaxial design results in LPC
characteristics closely matching to the ones shown in this figure.
Self-heating is not taken into account for this simulation. All
material parameters used are for 300 K.
arrive at the final cell design. The doping of each layer and
thickness obtained after the design optimization were closely
matched to those reported in [19]. A 0.25 µm thick single layer
SiO2ARC was used, which results in >95% transmission at
1.55 µm[25].
The simulations show that a conversion efficiency of
∼75% can ideally be achieved from the LPC at 1 kW m−2,
beyond which the conversion efficiency begins to saturate
as the open-circuit voltage (VOC)approaches the equivalent
maximum band gap (the quasi-Fermi level separation) of the
absorber material (∼0.8 eV in this case). In a real device,
however, the actual magnitude of temperature-dependent
recombination coupled with Ohmic losses is likely to result
in a lower efficiency than calculated (the simulation does not
account for Ohmic losses). It is clear from figure 1that
the photo-generated current (sheet) density Jsc continues to
increase, as expected, with increasing illumination intensity
(Pin)[22]. However, the open-circuit voltage, which is
directly proportional to the logarithm of the photo-generated
current over the dark current (see figure 1), saturates as the
light intensity increases, thereby saturating the conversion
efficiency [22]. Note that the dark current is a function of the
LPC area. Thus, from a design standpoint, an LPC absorber
scheme (doping, layer thickness, etc) for a fixed device area
should be optimized to maximize the photo-current Ipc at a
given illumination intensity, in order to maximize the optical
to electrical conversion efficiency. However, it should also
be noted that while the dark current increases with device
area, the sheet resistance decreases, thereby resulting in an
optimal device area for which the conversion efficiency will
be maximum. The LPC reported here has a device area of
5mm ×5 mm. Self-heating effects were not included in the
simulations as it required high computational resources when
Figure 2. Derivative (with respect to wavelength ‘λ’) of
photo-current spectra showing the band edge position from some
prototype devices tested are shown. Prototype devices having band
edges shorter and longer than 1.55 µm were tested (1.509, 1.588 and
1.653 µm) as shown.
coupled with electro-optical simulations. The effect of self-
heating was estimated as discussed below.
Under a constant thermal load on the LPC, illuminated
(CW) at a given intensity (hence absorption peak shifted to
a longer wavelength), it was found that an ideal band gap
for maximum optical to electrical conversion of power at
1.55 µm occurs not at the corresponding band edge of 0.8 eV
but at a higher energy (shorter wavelength) where there is
an optimum balance between an increase in absorption [23]
(due to an increased density of states) and carrier loss via
thermalization (due to a finite quantum deficit) [21,24]. This
optimal band gap is dependent on the magnitude of self-heating
(thermal load), but as the absorption for InxGa1−xAsyP1−y
plateaus at higher energies [23], one can still design an
absorber with an optimal band gap which will result in
maximized conversion efficiency in a given range of absorber
temperatures. This important result which should also prove
useful for conventional solar-cell design considerations is
currently being investigated further in a concerted study. Based
on the above considerations, various prototype devices with
absorber band edges shorter and longer than 1.55 µm were
tested, see figure 2. Among these, the LPC which showed the
best efficiency characteristics had a band edge at 1.588 µm and
is reported here. The derivative of the photo-current spectra
from some representative samples tested for comparison is
shown in figure 2.
3. Results: lab and field tests
The LPCs were first tested in the laboratory (henceforth called
lab tests) and then in a large aircraft hangar where long-
range laser power transfer (at 30 m) experiments (henceforth
called field tests) were performed. The lab and the field
test results match well indicating that the LPC characteristics
are reproducible. The square (25 mm2)prototype LPC was
bonded n-down on a copper sub-mount which was temperature
controlled for all the tests, see figure 3.
3

J. Phys. D: Appl. Phys. 46 (2013) 264006 J Mukherjee et al
Figure 3. Top view of the prototype LPC (5mm ×5mm)mounted
n-side down on a copper sub-mount. The p-side contact has the
usual bus-bars the ends of which are connected to a common anode
and multiple wire-bonds from it go into a larger contact pad.
For the lab tests the LPC was illuminated using a
high-power single-frequency diode laser at 1.55 µm (external
cavity tunable laser (TLK-L1550R) from Thorlabs [26], with
maximum output power (fibre coupled) of ∼45 mW at 1.55 µm
emission), whereas for the field tests a high-power fibre laser
(maximum output power ∼50 W, M2∼1.05, wall-plug
efficiency =12.5% [27]) at 1.55 µm was beamed from a
distance of 30 m onto the LPC. For the field tests, the laser beam
was expanded (from beam waist 4 to 12 mm) using a beam-
expander (Thorlabs BE03M-C [26]) in order to reduce and
control the power density incident on the LPC. Two different
LPCs from the same batch were used for the lab and field tests,
respectively. Both the LPCs tested in the lab and field show
very similar efficiency characteristics with increasing incident
illumination intensity at 1.55 µm. However, as the LPC tested
in the lab underwent several tests before the illumination
intensity-dependent tests were performed; it shows a steeper
droop due to ageing and degradation compared with a fresh
device which was used for the field tests. It is evident that both
the devices in the lab and field tests show maximum conversion
efficiencies at a power density of ∼1kWm
−2(see figure 4).
This behaviour is also seen in the non-thermal simulations (see
figure 1), where the conversion efficiency starts saturating at
∼1kWm
−2. Note that unlike the non-thermal simulations,
in practice, the increasing illumination intensity increases the
thermal load on the LPC (as the LPC conversion efficiency
is not increasing further), reducing the conversion efficiency
via various temperature-dependent loss mechanisms [21,24].
The calculated conversion efficiency at 1 kW m−2is ∼75%
compared with 44.6% obtained in practice from the field
tests (42.3% from lab tests). Higher calculated values of
the conversion efficiency result from the lack of accurate
information about the strengths of various loss mechanisms
included in the simulations. This, however, highlights the
achievable conversion efficiency from the LPC design which
requires further optimization and control of the various loss
Figure 4. Dependence of LPC conversion efficiency on incident
power density at 1.55 µm. A maximum conversion efficiency of
44.6% is achieved at ∼1kWm
−2(corresponding fill factor is
75.1%) beyond which the conversion efficiency drops due to an
increased thermal load. Corresponding conversion efficiency from
lab tests (different LPCs from same batch) are 42.3% (fill factor is
75.1%). The LPC sub-mount was temperature controlled at 21 ◦C.
The contact shadowing loss and Gaussian illumination of the LPC is
taken into account for this calculation. An estimated error of ±1%
is attributed to this calculation.
mechanisms. As the optimized LPC design uses a thick
absorber layer, highly susceptible to epitaxial growth-induced
defects, the SRH recombination is likely to be the dominant
photo-generated loss mechanism in our current LPC design. In
addition, as the LPC uses an InGaAsP/InP material system with
a band gap ∼0.8 eV, Auger recombination is likely to become
significant for high photo-generated carrier densities, i.e. under
high intensity illumination [24]. A study to further understand
the fundamental loss mechanisms in the devices is currently
underway and the results will be reported separately. To the
best of our knowledge, we note that that a conversion efficiency
of 44.6% [±1%] is the highest achieved so far in practice from
any LPC at 1 kW m−2at room temperature. Currently further
design optimization is in progress in order to maximize the
conversion efficiency from the LPC.
The temperature dependence of the conversion efficiency
and the fill factor (ratio of maximum obtainable power to the
product of the open-circuit voltage and short-circuit current)
was also measured both in the lab and in field tests by varying
the sub-mount temperature (16 to 26 ◦C for lab tests, 14 to
40 ◦C for field tests) as shown in figure 5. At a power density
of ∼1kWm
−2, the drop in both the conversion efficiency and
fill factor is ∼0.1% per ◦C (by fitting a straight line through
the data points at 0.95 kW m−2and 1kWm
−2)and the results
are repeatable in both the lab and field tests. It should be noted
that the drop in both the cell efficiency and fill factor with
increasing temperature is relatively steeper at higher power
densities (2.37 kW m−2)compared with that 1 kW m−2. This
is attributed to an increased thermal load at higher illumination
intensities.
4

J. Phys. D: Appl. Phys. 46 (2013) 264006 J Mukherjee et al
Figure 5. Temperature dependence of the LPC conversion
efficiency and fill factor showing a drop of ∼0.1%/◦C (four times
lower than commercial amorphous Si solar cells [22]) at
∼1kWm
−2. Sub-mount temperature varied from 16 to 26 ◦C for the
lab tests and from 14 to 40 ◦C for the field tests. As different lasers
were used for the lab and field tests, the achievable power densities
for the lab and field tests are different. Lines in black correspond to
cell efficiency and black squares are data points from field tests. Red
lines show the fill-factor behaviour and red squares are the
corresponding data points from field tests. Blue circles are data
points from the lab tests for both the cell efficiency and fill factor.
Note that the drop in both the cell efficiency and fill factor is
relatively steeper at higher power densities (2.37kWm−2)compared
with that 1 kW m−2. This is attributed to an increased thermal load
at higher illumination intensities.
4. Conclusions
Our III–V-based LPC currently achieves an optical to electrical
conversion efficiency of 44.6% [±1% at 1 kW m−2beyond
which a thermal load-induced efficiency droop is observed.
Theoretical (non-thermal) modelling suggests that the limiting
efficiency at 1 kW m−2could be much higher (up to 75%).
For the current design, together with Ohmic losses, the SRH
non-radiative recombination (which increases with device
temperature) is likely to be the dominant loss mechanism
for photo-generated carriers in the thick absorber region of
the LPC, as thick semiconductor layers are more susceptible
to epitaxial growth-induced defects [21], resulting in the
deviation from the modelled results. Furthermore, as the
LPC design utilizes the InGaAsP/InP material system with
a band gap ∼0.8 eV, Auger recombination is another major
non-radiative temperature-dependent recombination channel,
which becomes significant under high photo-generated carrier-
densities in the device (hence, under high illumination power
densities). Further design optimization will be required
in order to control and reduce loss mechanisms affecting
the conversion efficiency. It should also be noted that our
design optimization has revealed that the optical to electrical
conversion efficiency for a monochromatic pv (i.e. a laser
power converter) is maximum when the absorber band edge
is at a longer wavelength than the radiation illuminating the
device. This optimum band edge depends on the thermal
load on the LPC (i.e. the absorber lattice temperature) for a
given illumination intensity via the temperature-induced band-
gap change effect (hence absorption edge shift) and occurs
when the effects of increased absorption at higher energies is
appropriately balanced via the quantum-deficit-induced carrier
energy loss. This should prove useful in future designs
of efficient photovoltaics. As the absorption characteristics
for most direct band-gap semiconductor materials plateau at
wavelengths shorter that the band gap (i.e. at higher energies)
one can still design an absorber with an optimal band gap which
will result in maximized conversion efficiency in a given range
of LPC temperatures. We are currently investigating this in
detail for LPC optimization which is expected to scale the
efficiency beyond 50% at 1kW m−2. We believe that in order
to maintain the maximum achievable conversion efficiency
1kWm
−2at higher power densities, especially in use for
LPC arrays, excellent thermal management schemes would be
required to keep the thermal load on the LPC to a minimum.
LPC designs with improved thermal management schemes are
now being studied.
Acknowledgments
The authors are grateful to EADS-Astrium for financial support
of this project. They would also like to thank W Wulfken and
H Hartje at EADS ASTRIUM space transportation, Bremen,
Germany, for their help with the field tests. In addition, they
acknowledge useful discussions with CIP Technologies (now
Huawei), UK, during the development phase of the LPC design
prototype reported here.
References
[1] Jain R K 1993 IEEE Trans. Electron Devices 40 1993
[2] Pena R and Algora C 2011 Prog. Photovolt. 20 117 (and
reference therein)
[3] Flournoy D M 2012 Solar Power Satellites (Berlin: Springer)
[4] Gowar J 1993 Optical Communications Systems (London:
Prentice-Hall)
[5] Steen W M and Mazumdar J 2010 Laser Material Processing
(Berlin: Springer) p 521
[6] Rao M M 2001 Optical Communication (Hyderabad: Sangam
Books) p 60
[7] Okhotnikoc O G (ed) 2012 Fibre Lasers (New York: Wiley
VCH) chapter 1
[8] Thulium based CW fibre lasers, www.ipgphotonics.com/
products tlm flm.htm
[9] Pena R and Algora C 2007 Spanish Conf. on Electron Devices
(Madrid) p 274
[10] Wojtczuk S J 1997 Proc. 26th IEEE Photovoltaic Specialist
Conf. (Anaheim, CA) p 971
[11] Andreev V, Khvostikov V, Kalinovsky V, Lantratov V,
Grilikhes V, Rumyantsev V, Shvarts M, Fokanov V and
Pavlov A 2003 Proc. 3rd World Conf. on Photovoltaic
Energy Conversion (Osaka) p 761
[12] Bett A W, Dimroth F, Stollwerck G and Sulima O V 1999
Appl. Phys. 69 119
[13] W¨
urfer P 2005 Physics of Solar Cells (New York: Willey) p 57
[14] www.grc.nasa.gov/WWW/RT/2005/RP/RPV-wilt.html
[15] Wilt D 2012 private communication, AFRL
[16] Sweeney S J and Mukherjee J 2012 Laser Power Convertor
(Patent filed August 2012 in UK)
5

J. Phys. D: Appl. Phys. 46 (2013) 264006 J Mukherjee et al
[17] Luque A 1990 Physical Limitations to Photovoltaic Energy
Conversion ed A Luque andGLAra
´
ujo (Bristol:
Adam Hilger)
[18] Piprek J (ed) 2005 Optoelectronic Devices: Advanced
Simulation and Analysis (Berlin: Springer)
[19] Jain R K 2003 3rd World Conf. on Photovoltaic Energy
Conversion (Osaka) vol1p75
[20] RSoft simulation package (Solar cell utility, FullWave and
LaserMOD) www.rsoftdesign.com/
[21] Ding D, Johnson S R, Yu S-Q, Wu S-N and Zhang Y-H 2011
J. Appl. Phys. 110 123104
[22] Green M 1982 Solar Cells: Operating Principles, Technology
and System Applications (Englewood Cliffs, NJ:
Prentice-Hall) pp 86–92
[23] Dutta N 1980 J. Appl. Phys. 51 6095
[24] Phillips A F, Sweeney S J, Adams A R and ThijsPJA1999
IEEE J. Sel. Top. Quantum Electron. 5401
[25] www.photonics.byu.edu/antireflection.phtml
[26] www.thorlabs.de/thorProduct.cfm?partNumber=TLK-L1550R
[27] IPG Photonics, ELR series
(www.ipgphotonics.com/apps mat single ELR.htm)
6