Demonstrated fossil-fuel-free energy cycle using magnesium and laser

Article abstract of DOI:10.1063/1.2423320

The authors propose an energy cycle based on a renewable fuel. Magnesium is chosen as an energy carrier and is combusted with water to retrieve energy using many power devices. MgO, the combustion residue, is reduced back to Mg by laser radiation generated from solar and other renewable energy sources. They have achieved an energy recovery efficiency of 42.5% for converting MgO to magnesium, using a laser. Combined with a demonstrated 38% efficiency for converting an artificial sunlight source (metal halide lamp) into laser output energy indicates that the proposed energy cycle is already in a feasible range for practical use.

Full text of DOI:10.1063/1.2423320

Demonstrated fossil-fuel-free energy cycle using magnesium and laser
T. Yabe, S. Uchida, K. Ikuta, K. Yoshida, C. Baasandash et al.
Citation: Appl. Phys. Lett. 89, 261107 (2006); doi: 10.1063/1.2423320
View online: http://dx.doi.org/10.1063/1.2423320
View Table of Contents: http://apl.aip.org/resource/1/APPLAB/v89/i26
Published by the American Institute of Physics.
Additional information on Appl. Phys. Lett.
Journal Homepage: http://apl.aip.org/
Journal Information: http://apl.aip.org/about/about_the_journal
Top downloads: http://apl.aip.org/features/most_downloaded
Information for Authors: http://apl.aip.org/authors
Downloaded 23 May 2012 to 171.67.34.69. Redistribution subject to AIP license or copyright; see http://apl.aip.org/about/rights_and_permissions

APPLIED PHYSICS LETTERS 89, 261107 ͑2006͒
Demonstrated fossil-fuel-free energy cycle using magnesium and laser
T. Yabe,a͒ S. Uchida, K. Ikuta, K. Yoshida, C. Baasandash, M. S. Mohamed, Y. Sakurai,
Y. Ogata, M. Tuji, Y. Mori, Y. Satoh, T. Ohkubo, M. Murahara, and A. Ikesue
Entropia Laser Initiative, Tokyo Institute of Technology, 2-12-1 O-okayama, Meguro-ku,
Tokyo 152-8552, Japan
M. Nakatsuka
Institute of Laser Engineering, Osaka University, 2-6 Yamada-oka, Suita, Osaka 565-0871, Japan
T. Saiki, S. Motokoshi, and C. Yamanaka
Institute for Laser Technology, 2-6 Yamada-oka, Suita, Osaka 565-0871, Japan
͑Received 7 July 2006; accepted 25 November 2006; published online 27 December 2006͒
The authors propose an energy cycle based on a renewable fuel. Magnesium is chosen as an energy
carrier and is combusted with water to retrieve energy using many power devices. MgO, the
combustion residue, is reduced back to Mg by laser radiation generated from solar and other
renewable energy sources. They have achieved an energy recovery efficiency of 42.5% for
converting MgO to magnesium, using a laser. Combined with a demonstrated 38% efficiency for
converting an artificial sunlight source ͑metal halide lamp͒ into laser output energy indicates that the
proposed energy cycle is already in a feasible range for practical use. © 2006 American Institute of
Physics. ͓DOI: 10.1063/1.2423320͔
There is no doubt that we need an energy cycle free of
fossil fuels that otherwise emit greenhouse gases causing
Mg + H2O → MgO + H2 + 86 kcal.
͑1͒
A part of the excess heat of 86 kcal/mol sustains the reaction
temperature. Therefore, once the reaction starts, it will run by
itself without additional heat.
Hydrogen obtained from this reaction under low power
conditions can be used in fuel cells ͑low power reaction͒.
Conversely, hydrogen can be obtained under high power
where it also undergoes combustion ͑high power reaction͒
with the reaction
global warming. Although solar energy is the ultimate re-
newable energy source, it is far from fully realized. Further-
more, since sunlight is available only in the clear daytime, it
cannot be an alternative to thermal power stations unless
effective and large power storage systems are available. Here
we propose the chemical potential of magnesium as such an
energy reservoir. Heat and hydrogen from the reaction of
magnesium with water are used for turbines, reciprocal en-
gines, fuel cells, and so on. The remaining “ash,” MgO, has
to be deoxidized in order to make the energy cycle renew-
able. The deoxidization process is driven by the energy from
a solar-energy-pumped laser,1–5 a laser diode6 powered by a
wind-power generator, or other sources. Since laser radiation
can be focused into a small spot, a very high temperature
͑exceeding 4000 K, needed for MgO deoxidization͒ can eas-
ily be obtained. Once such MgO deoxidization technology is
developed, unsteady solar power can be stored in Mg form to
provide a stable supply of energy.
This letter reports experimental demonstrations of key
technologies: a solar-energy-pumped laser ͑sunlight laser͒, a
magnesium combustion system, and magnesium recovering
͑deoxidization͒ system. We have achieved 38% conversion
efficiency from an artificial sunlight source ͑metal halide
lamp͒ to laser output by Cr-doped Nd-YAG ͑YAG denotes
yttrium aluminum garnet͒ laser,5 42.5% energy recovery ef-
ficiency for deoxidization of MgO, and a method for con-
trolled combustion of magnesium. These three experiments
prove that the energy cycle is already in a feasible range for
practical use.
1
H2 + O2 → H2O + 57.8 kcal.
͑2͒
2
An experimental setup for a low power reaction is shown in
Fig. 1͑a͒, where Mg plates of 20ϫ40ϫ0.3 ͑or 0.6͒ mm3
were placed inside a chamber at the beginning of each ex-
periment run. The Mg plates were ignited by Ohmic heating
fed from an external power supply. Water was added at vari-
ous controlled rates as the reaction proceeded. As the figure
shows, the speed of reaction ͑1͒ can be readily controlled by
changing the thickness of Mg and the rate of water supply.
Hydrogen was produced for 20–30 min in a steady manner
with only 80–150 g Mg supplied. The temperature of the
hydrogen generated was almost constant at about 80 °C
due to the relatively slow Mg combustion. The hydrogen
was generated at 0.28–0.63 g/min, corresponding to
0.57–1.3 kW if the hydrogen was burned as reaction ͑2͒.
By reducing the size of Mg to 2ϫ2ϫ0.6 mm3, reaction
͑1͒ proceeds much faster. As shown in Fig. 1͑b͒, the hydro-
gen gas produced was ejected through a hole of 4 mm diam-
eter and was directed onto a turbine of 80 mm in diameter
having eitht fins of 30ϫ40ϫ1 mm3. In this experiment,
80 2 g of water was initially added to the chamber, with
varying amounts of Mg. It was found that the moles of con-
sumed water were almost twice the initial Mg. Presumably,
some of the water evaporated and was exhausted with H2
driving the turbine together, while most of the MgO re-
As an energy reservoir, magnesium is superior to hydro-
gen because it provides more compact energy storage at
43 GJ/m3, compared to 4.3 GJ/m3 for 70 MPa hydrogen.
When magnesium is heated up to 850 K, it strongly reacts
with water as follows:
a͒
Electronic mail: yabe@mech.titech.ac.jp
0003-6951/2006/89͑26͒/261107/3/$23.00
89, 261107-1
© 2006 American Institute of Physics
Downloaded 23 May 2012 to 171.67.34.69. Redistribution subject to AIP license or copyright; see http://apl.aip.org/about/rights_and_permissions

261107-2
Yabe et al.
Appl. Phys. Lett. 89, 261107 ͑2006͒
FIG. 1. ͑Color online͒ ͑a͒ Steady and controlled hydrogen generation was demonstrated for 20–30 min using about 100 g of Mg composed of 20ϫ40
ϫ0.3 mm3 or 20ϫ40ϫ0.6 mm3 plates. After initial ignition by Ohmic heating, reaction was maintained by water input from the top. Solid lines show results
for various water supply rates with 0.3 mm thick Mg, while dashed line shows results with 0.6 mm thick Mg and 20 ml/min water supply rate. The linearity
shows that the reaction proceeds with constant speed regardless of Mg consumed. ͑b͒ High power reaction with Mg of 2ϫ2ϫ0.6 mm3 generated 8000 rpm
of the turbine shown in the top-left corner. The turbine was driven by exhaust gas from the reaction chamber as shown. Note that the specific area
͑area/volume͒ of Mg fuel is now twice that of the slow reaction fuel. The peak rotation speed and acceleration depend not only on the shape of Mg but also
on the total amount of Mg.
mained in the chamber. Since H2 is 20 times lighter than
MgO, it obtains a much larger velocity and preferentially
escaped the chamber.
A series of experiments demonstrating deoxidization of
MgO has been performed with three different laser condi-
tions. Table I summarizes the results. After evaporation, Mg
was transported to adhesive carbon tape or a copper plate by
Figure 1͑b͒ shows the time evolution of rotational speed
of the turbine, with the initial amount of Mg as a parameter.
The acceleration of the turbine rotation looks similar among
all cases, despite the different initial Mg loading. This sug-
gests that the burning speed was nearly constant, depending
only on the shape of Mg present, so the final rotation speed
was approximately proportional to the total amount of Mg
loaded. A maximum rotation of 8000 rpm was achieved after
4 s using 20 g of Mg. A high speed camera observation re-
vealed that the speed of the exhaust gas was 150 m/s.
Although Mg is abundant ͑the second richest metal ele-
ment in seawater at 1.3 g/l and the eighth richest in earth’s
crust͒, Mg needs to be retrieved through a refinement pro-
cess, which requires a large amount of energy to be drawn
from the energy cycle. With present technology of Mg re-
finement, MgO and CaO are obtained by baking dolomite
that contains MgCO3 and CaCO3. Under reaction with FeSi,
Mg is separated from CaO and MgO at a relatively low tem-
perature, 1500 K. Since FeSi is consumed in the reaction and
is difficult to recover after the process, an alternative tech-
nology that does not require FeSi needs to be developed.
A laser is a very suitable energy source for raising the
temperature well above 4000 K, the temperature required for
complete separation of Mg and O in equilibrium. After
evaporation, Mg vapor moves upward because it is slightly
lighter than air and condenses to form Mg metal as in a
conventional thermal reduction system. However, at atmo-
spheric pressure, Mg atoms collide with O2 within a distance
of only 100 nm and recombine back to MgO. Instead of
using impractically high vacuum to suppress the collision,
we here propose to use an inert gas flow to prevent Mg
atoms from “reoxidization” during the cooling phase, as
shown in Fig. 2͑a͒. As Mg atoms cool down, the rate of Mg
oxidization diminishes.
FIG. 2. ͑Color online͒ ͑a͒ Dissociated Mg from laser illumination is blown
off by Ar flow and is accumulated onto a Cu plate to avoid reoxidization.
Ejected mass rate ͑given by ejected mass divided by irradiation time͒ de-
creases with increasing irradiation time and reaches an asymptotic value.
High speed imaging revealed that intense generation of MgO debris took
place during first 0.2 s followed by steady vaporization. ͑b͒ High speed
camera shows the mass ejection of the ringlike structure at the peripheral
zone surrounding laser spot for 1 kW CO2 with 0.1 s irradiation on the
ceramic MgO. Such mass ejection under evaporation temperature might
increase the efficiency of Mg reduction.
Downloaded 23 May 2012 to 171.67.34.69. Redistribution subject to AIP license or copyright; see http://apl.aip.org/about/rights_and_permissions

261107-3
Yabe et al.
Appl. Phys. Lett. 89, 261107 ͑2006͒
TABLE I. Deoxidization experiments with three different laser conditions. The reduction efficiency signifi-
cantly increases with the aid of gas flow even under atmospheric conditions. Ar+Ice means the copper plate
was cooled by ice. The resulting improvement from cooling implies the importance of cooling speed. The
4 sϫ3 for the irradiation time means three shots of 4 s duration are accumulated onto the plate. The MgO was
in the form of ceramic for YAG laser; otherwise most of the powder was blown off due to extremely high
pressure induced by the pulsed laser. MgO powder was used for the CO2 lasers because the ceramic was broken
after a single shot by thermal stress.
Mole
Irradiation
time
Chamber
pressure
fraction of
Mg ͑%͒
Laser
Gas flow
Collect
CO2
100 W
10 min
10 min
10 min
Ar
Cu plate
Cu plate
Cu plate
0.1 MPa
0.1 MPa
0.1 MPa
27
Ar ͑2 m/s͒
Ar ͑2 m/s͒
29.5
22.8
1000 W
2 s
He ͑2 m/s͒
Ar ͑10 m/s͒
Ar ͑5 m/s͒
Ar ͑15 m/s͒
Ar ͑15 m/s͒+Ice
Carbon
Carbon
Carbon
Cu plate
Cu plate
0.1 MPa
0.1 MPa
0.1 MPa
700 Pa
33
4 sϫ3
32
4 sϫ3
31
0.2 sϫ10
0.2 sϫ10
20
700 Pa
32.6
YAG
1.4 J/10 ns
15 min
͑10 Hz͒
15 min
͑10 Hz͒
60 min
͑10 Hz͒
Ar ͑15 m/s͒
Cu plate
Cu plate
Cu plate
500 Pa
500 Pa
100 Pa
43.4
38.5
0
Ar ͑22 m/s͒
No
gas flow
the inert gas flow and was deposited there. We analyzed the
content of Mg among the deposited material by an electron
probe microanalyzer. The analysis indicated that 20%–43%
of the Mg atoms introduced as MgO were collected as un-
oxidized Mg, as shown in Table I.
Considering latent heats of melting Lm=1.94 kJ/g and
evaporation Lb=8.27 kJ/g and the dissociation energy for
MgO Ld=9.22 kJ/g, we can estimate the maximum evapo-
ration rate X ͑g/s͒ of MgO from ͑C+Lm+Lb+Ld␣͒X=P,
where P ͑kW͒ being the incident laser power and C being the
sensible heat of about 5.4 kJ/g. If the mole fraction ␣ of
dissociated Mg at 0.2 s is used, this gives X=0.054 g/s,
which is smaller than the experimental result. This implies
that some MgO was ejected in the liquid phase from the
peripheral heat-affected zone surrounding the laser spot
͑where high pressure was generated by evaporation͒,7 as
shown in Fig. 2͑b͒.
It is well known that 40% of world’s population will
need new water resources by 2025. If desalination plants are
the only solution, 1500 new plants with 200 000 ton/day ca-
pability will need to be constructed. It is interesting to note
that condensed seawater produced from a desalination pro-
cess of 200 000 ton/day capacity contains 650 ton of mag-
nesium, and thus we can accumulate 3.6ϫ108 ton every year
from 1500 desalination plants.
The authors are greatly thankful to K. Grabowski for his
comments to this letter.
1C. G. Young, Appl. Opt. 5, 993 ͑1966͒.
2V. Krupkin, Y. Kagan, and A. Yogev, Proc. SPIE 2016, 50 ͑1993͒.
3M. Landoa, J. Kagana, B. Linyekina, and V. Dobrusin, Opt. Commun.
222, 371 ͑2003͒.
4A. Ikesue, K. Kamata, and K. Yoshida, J. Am. Ceram. Soc. 78, 2545
͑1995͒.
5T. Saiki, S. Uchida, S. Motokoshi, K. Imasaki, M. Nakatsuka, H. Na-
gayama, Y. Saito, M. Niino and MT. Mori, Proceedings of the Interna-
tional Astronautical Congress 2005, paper No. IAC-05-C3.4-D2.8.09.
6P. A. Crump, M. Grimshaw, J. Wang, W. Dong, S. Zhang, S. Dus, J.
Farner, M. Devito, L. S. Meng, and J. K. Brasseur, Conference on Lasers
and Electro-Optics, Quantum Electronics and Laser Science Conference
͑CLEO/QELS 06͒, 21–26 May 2006, Long Beach, CA, Paper No. JWB24.
7T. Yabe, F. Xiao, and T. Utsumi, J. Comput. Phys. 169, 556 ͑2001͒.
The ejection rate in Fig. 2͑a͒ and the efficiency at 0.2 s
irradiation in Table I indicate that a reduction energy effi-
ciency of 42.5% has been achieved ͑equal to Mg reaction
energy/energy for Mg production͒. This translates to
61 ton/GW h.
Downloaded 23 May 2012 to 171.67.34.69. Redistribution subject to AIP license or copyright; see http://apl.aip.org/about/rights_and_permissions