Applied Energy 88 (2011) 1–16

Contents lists available at ScienceDirect

Applied Energy
journal homepage: www.elsevier.com/locate/apenergy

Development of micro power generators – A review

S.K. Chou a,*, W.M. Yang a, K.J. Chua a, J. Li a, K.L. Zhang b
a
Department of Mechanical Engineering, National University of Singapore, 9 Engineering Drive 1, Singapore 117576, Singapore
b
Department of Manufacturing Engineering and Engineering Management, City University of Hong Kong, 83 Tat Chee Avenue, Hong Kong

a r t i c l e i n f o a b s t r a c t

Article history: The demand for energy sources that are compact, lightweight and powerful has significantly increased in
Received 7 May 2010 recent years. Traditional chemical batteries which are highly developed are unable to meet the demand
Received in revised form 5 July 2010 for high energy intensity. This gap is expected to widen in the future as electronic devices need more
Accepted 6 July 2010
power to support enhanced functionalities. Hydrocarbon fuels have energy densities much greater than
Available online 8 August 2010
the best batteries. Therefore, taking advantage of the high energy density of chemical fuels to generate
power becomes an attractive technological alternative to batteries. To address the growing demand for
Keywords:
smaller scale and higher energy density power sources, various combustion-based micro power genera-
Micro-scale combustion
Micro power generators
tors are being developed around the world. This review paper provides an update on recent progresses
Micro-thermophotovoltaic system and developments in micro-scale combustion and micro power generators. The paper, broadly divided
Micro thrusters for propulsion into four main sections, begins with a review of various methods to enhance and stabilize the combustion
Direct energy conversion at micro-scale, subsequently improving the efficiency. This is followed by a description of various micro-
and PV cells thermophotovoltaic power generators. The third section focuses on MEMS based solid propellant micro-
propulsion system. Lastly, a brief review is made to other micro power generators.
Ó 2010 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2. Micro-scale combustion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2.1. Key issues and major challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2.2. Progress made so far . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.3. Practical micro-combustors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.3.1. Swiss-roll micro-combustors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.3.2. Cylindrical tubes with backward facing steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.4. Future works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.4.1. Catalyzed micro-combustion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.4.2. Porous media micro-combustion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3. Micro-thermophotovoltaic (TPV) power generator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3.1. Optimization of micro cylindrical combustor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.2. Experimental results of the micro-TPV power generator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.2.1. Effect of backward facing step height . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.2.2. Effect of wall thickness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3.2.3. Effect of flow rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.2.4. Effect of combustion chamber . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.2.5. Effect of fuel/oxidant mixture type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
4. MEMS based solid propellant micro-propulsion systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
4.1. Three-layer sandwich design of solid propellant microthruster. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
4.2. Two-layer building block design of solid propellant microthruster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
4.3. Fabrication of the two-layer building block microthruster. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
4.4. Combustion and thrust tests of the two-layer building block microthruster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
4.5. Ignition study of the two-layer building block microthruster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

* Corresponding author. Tel.: +65 65162215; fax: +65 67791459.

E-mail address: [email protected] (S.K. Chou).

0306-2619/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.apenergy.2010.07.010
2 S.K. Chou et al. / Applied Energy 88 (2011) 1–16

5. Other power MEMS systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

5.1. Micro gas turbine engine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
5.2. MEMS rotary engine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
5.3. Micro P3 heat engine. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
5.4. Thermoelectric micro power generator and micro cooler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
5.5. Mechanical energy scavengers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
5.6. Nano energetic material based power MEMS systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
6. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

1. Introduction Supposing the chemical energy contained in the fuel can be uti-
lized efficiently, the power source package is expected to shrink
The innovative approaches to convert available energy into significantly without compromising the power output. So far, a ser-
usable forms using micro and nanotechnologies can contribute to- ies of combustion-based micro power systems have been success-
wards sustainable energy development and meet the growing need fully prototyped, including the micro gas-turbine [3], the micro
for power in small scale applications. The dominant power sources thermoelectric device [4] and the micro thermo-photovoltaic
for portable electronics are currently batteries. Their limited en- (TPV) system [5]. Despite the fact that their current efficiency
ergy density and adverse effects on the environment upon disposal was relatively low, these micro systems have demonstrated the
are driving the emergence of a new class of micro power sources or feasibility of generating power in the order of a few Watts within
power MEMS. Power MEMS are defined as microsystems for power a volume of some cubic centimeters [6]. Therefore, the study on
generation and energy conversion, including propulsion and cool- micro-combustion engines is still regarded as one of the promising
ing, which can be categorized as: (1) micro thermodynamic ma- frontiers of MEMS [7].
chines including microturbines, miniature internal combustion One of the challenges to micro-combustion is how to sustain
engines and micro coolers; (2) solid-state direct energy conversion flames in micro-combustors with the size in millimeter and even
including thermoelectric and photovoltaic microstructures; and (3) sub-millimeter scale [8]. The ratio of heat loss to heat generation
vibration energy harvesting devices including piezoelectric, mag- is inversely proportional to the characteristic length of the mi-
netic or electrostatic micro generators, and micro thrusters for cro-combustor. Thus, a smaller combustor implies a larger propor-
propulsion. tion of heat loss from the flame zone, which may potentially lead to
The foci of this paper are micro-scale combustion, micro-ther- thermal quenching [9,10]. On the other hand, the absorption and
mophotovoltaic (TPV) power generator, solid propellant micro- destruction of active species (radicals) generated during combus-
propulsion systems. Combustion of hydrocarbon or hydrogen fuels tion also have impact on the flame stability, and this effect is inten-
provides a much higher power density on a per unit mass or unit sified as the combustor size decreases [11]. Some research groups
volume basis compared with most conventional energy systems. applied combustion-enhancing methods such as external heating
Therefore, harnessing energy from micro-combustion is an ideal [12,13] to overcome the heat loss problem, while others [14,15]
way to realize the miniaturization of a power source package that experimented on catalyzed combustion with the intention to min-
will meet the requirements of MEMS. MEMS power devices have imize the radial quenching effect.
been prototyped in the past years. However, ensuring stable and
efficient combustion remains a challenge. The reduced combustor
2.1. Key issues and major challenges
volume experiences significant heat loss and radical destruction
on the combustor wall. Understanding the fundamental character-
In contrast to the conventional-sized counterpart, combustion
istics of combustion in the micro-scale is the key to improving the
in a micro-combustor depends heavily on the heat recirculation
system efficiency and optimizing the design. The micro-TPV power
through the solid structure [16,17]. Heat transferred from the
generator is one type of power MEMS attracting increased atten-
tion because it has no moving parts, is highly robust and reliable,
and is suitable for use in commercial electronics and personal mi-
cro devices. Microspacecraft is one of many applications of MEMS
technology. The low cost, reliable and versatile clusters of micro-
spacecraft have more advantages than a conventional spacecraft
in fabrication, launch and operation. In a microspacecraft, a mi-
cro-propulsion system is required for high-accuracy station keep-
ing, altitude control, and orbit adjustment. Other power MEMS
systems besides the above-mentioned micro-TPV and microthrus-
ter are briefly introduced in one section.

2. Micro-scale combustion

The commencement of sustained micro-combustion research

may be traced back to about two decades ago, mainly attributed
to the proliferation of the micro-electromechanical systems
(MEMS) and their demand for miniaturized power sources [1]. It
is well known that power systems employing hydrogen or hydro-
carbon fuels offer much higher energy density on a per unit mass
basis, compared to conventional systems [2], as shown in Fig. 1. Fig. 1. Energy density of different energy systems.
 S.K. Chou et al. / Applied Energy 88 (2011) 1–16 3

flame zone (high temperature zone) through upstream conduction fore, entropy generation can provide an indication of the efficiency
preheats the unburned mixture, which may lead to a broadening of of a micro-combustion process [30].
the reaction zone in micro-combustors [18]. The thermal coupling
between the reacting gases and the combustor wall requires a 2.2. Progress made so far
careful selection of the combustor wall material (the thermal con-
ductivity, in particular) in order to balance the heat recirculation to In the past few years, significant progress in the micro-combus-
the unburned mixture and the heat loss on the ambient. tion research has been achieved through both experimental studies
Due to the size constraint, measurement of flow parameters and numerical investigations. These studies have helped deepen
such as temperature and velocity inside a micro-combustor is ex- the understanding of fundamental issues such as flame stability
tremely difficult. Despite the difficulty in doing measurement, ef- and flammability limits associated with micro-combustion, which
forts have been made to embed fine temperature sensors to contributes to the improved combustion efficiency. Table 1 sum-
measure the temperature of the reacting gases [19,20] and to ana- marizes some experimental studies on combustion flows in mi-
lyze the fuel conversion rate at the combustor exit [21,22], though cro-combustors with simple geometries such as cylindrical tubes,
these results are more suitable to be taken as ‘lumped’ properties. rectangular and radial channels, from which it is safe to conclude
In recent years, the application of laser technology for temperature that self-sustained combustion of hydrogen and hydrocarbon fuels
measurement has gained more attention especially when used (mainly methane and propane) can be achieved in micro-combus-
with conventional sized combustors. Bessler and SchuLz [23] tors with the characteristic length down to sub-millimeter scale.
developed a novel temperature imaging technique based on la- Compared to methane and propane, hydrogen demonstrates the
ser-induced fluorescence (LIF) of NO which can be used to measure remarkable advantage of sustaining the flame in smaller micro-
the temperature of premixed and partially premixed ethylene/air combustors, despite some disadvantages such as its storage
Bunsen flames. Lee et al. [24] investigated this technology for stea- requirements.
dy and high pressure premixed methane/air flames application. On the other hand, numerical methods have been implemented
Muller et al. [25] also measured the concentration profiles of OH, to study the fundamental characteristics of micro-combustion,
O2 and NO as well as temperature fields in diffusion flames of a partly owing to the difficulty of measuring parameters in such con-
length of approximately 300 mm and 40 mm in diameter by planar fined spaces. As the combustor size shrinks to millimeter or even
LIF. It is believed that this technology can be used to measure the sub-millimeter scale, it is expected that the combustion process
temperature profile of a flame in the micro-combustor. will differ from that in the macroscale in terms of flame thickness,
Through the studies on flame propagation in narrow channels, flammability, flow field and heat loss intensity. In order to study
flame instabilities (together with acoustic emission for some cases) the scale effects on the above mentioned issues, an analytical study
were noted for both premixed flames by Zamaschikov [26] and (scale analysis) was carried out by Li et al. [33]. For the premixed
Mellish et al. [27] and non-premixed flames by Richecoeur and combustion flow in a cylindrical micro-combustor with the inner
Kyritsis [28] Prakash et al. [29]. With the combustor wall being radius of r0, the energy equation is given by
heated by external heating source, Maruta’s group [12,13] ob-    
served various unstable flame modes such as cyclic oscillatory mo- @ 1 @ @ @T 1 @ @T
ðqucp TÞ þ ðqv cp TrÞ ¼ k þ k r þq ð1Þ
tions (in a cylindrical tube) and rotating pelton-like flames (in @x r @r @x @x r @r @r
radial channels). Main reasons leading to flame instabilities in mi-
cro-combustors may include flame-flow and flame-structure inter- where u is the axial flow velocity, T is the temperature, k is the ther-
actions, heat loss induced high-frequency extinction–reignition mal conductivity of gas, v is the radial velocity, q is the volumetric
and mass transfer limitations (diffusion of species) for non-pre- heat generation rate, r is radial coordinate, q is the density of gas.
mixed flames. The scale of each variable in the domain may be expressed as u  SL,
Heat loss from a combustion system decreases the entropy of x  d, r  r0, T  DT, and q  q0, respectively (d is the flame thick-
that system and thus the level of molecular disorder or random- ness). Therefore, Eq. (1) can be rewritten as
ness. Since the volumetric heat loss depends on the surface-to-vol-    2
q0 cp SL d q0 cp v d d d q 2
ume ratio, a decrease in the combustor size (implying an increase þ ¼1þ þ d ð2Þ
k k r0 r0 k DT
of the surface-to-volume ratio) increases the entropy transfer
through the combustor wall. In a thermodynamic sense, entropy where SL is the flame speed. The two terms on the left-hand side
generation leads to the loss of available work (energy loss). There- represent axial convection: enthalpy change rate and the radial con-

Table 1
Experimental studies on micro-combustors with simple geometries.

Author(s) Fuel-oxidizer Combustor Geometry Combustor size Combustor material

Zamashchikov [26] C3H8/Air Cylindrical tubes D = 2.7 mm Stainless steel
CH4/Air Ceramic
H2/Air Quartz
Mellish et al. [27] CH4/O2 Cylindrical tubes D = 0.5 mm Stainless steel
Alumina
Quartz
Miesse et al. [31] H2/O2 Rectangular slots H = 0.75 mm Stainless steel
C3H8/O2 Alumina
CH4/O2 Quartz
Cordierite
Prakash et al. [29] CH4–O2 Rectangular slots H = 0.75 mm Alumina
Maruta et al. [12] C3H8/Air Cylindrical tubes D = 2 mm Quartz
Fan et al. [13] CH4/Air Radial micro-channels H = 0.5–3 mm Quartz
Boyarko et al. [14] H2/O2 Cylindrical tubes D = 0.4–0.8 mm Platinum
Yang et al. [32] H2/Air Cylindrical tubes D = 3 mm Silicon carbide coated with platinum
4 S.K. Chou et al. / Applied Energy 88 (2011) 1–16

vection: heat loss rate, respectively. The three terms at the right- 2.3.1. Swiss-roll micro-combustors
hand side represent the axial conduction: heat gain rate, the radial There are a large number of studies dealing with this particular
conduction: heat loss rate and the heat generation: heat input rate, configuration, with one typical representation shown in Fig. 2. The
respectively. It can be seen that the ratio of the radial heat conduc- basic idea of Swiss-roll micro-combustors is to utilize the thermal
tion to the axial heat conduction is (d/r0)2. An asymptotic analysis of energy in the combustion products to preheat the unburned mix-
this ratio implies that as d is far smaller than r0 (macroscale com- ture in order to extend the flammability limits. The counter-cur-
bustion), the heat loss in the radial direction is negligible, while, if rent heat exchanger model is a simplified treatment of this
d is comparable to r0 (micro-scale combustion), the radial heat flux configuration. Both non-catalytic and catalytic [40] Swiss-roll mi-
becomes significant. cro-combustors with the channel widths ranging from a few milli-
Another interesting case is the Knudsen number effects on the meters to the sub-millimeter scale have been fabricated and tested.
temperature and flow field in micro-combustion. It has been ar- It was shown that in the presence of catalytic reactions, combus-
gued that although a micro-combustor (millimeter to sub-millime- tion can be sustained even at very low Reynolds number and lower
ter) is generally larger than a micro fluidic channel (normally tens temperature, compared to pure gas-phase combustion. The flame
microns), the gas temperature in a micro-combustor is much high- position (center and off-center) is affected by the Reynolds num-
er. As a result, the mean free path of the gases in the micro-com- ber, thermal properties of the combustor wall and the heat loss
bustor becomes larger. In view of this, Li et al. [37] investigated from the combustor to the ambient. With proper catalyst (such
the effect of the slip-wall boundary condition on micro-combus- as platinum) coated on the combustor wall, self-ignition and even
tion. Results show that the effect is negligible. In another words, ‘flameless’ combustion were observed under certain flow condi-
the combustion flow in a micro-combustor is still regarded as con- tions. A full two-dimensional numerical model [41] which incorpo-
tinuum flow. Therefore, the governing equations for conventional rates the effects of temperature-dependent gas and solid
combustion still apply. properties, viscous flow, surface-to-surface radiation heat transfer
Table 2 lists some numerical investigations done in recent and one-step chemical reaction was developed to predict the
years. Major conclusions made by these studies include: (1) mix- extinction limit of a Swiss-roll micro-combustor. Federici and Vla-
ture pre-heating effectively extends the range within which self- chos [42] studied the flame stability in a single-pass heat recircu-
sustained flames can be achieved; (2) heat recirculation through lation micro-burner which was simplified from the multiple-turn
the combustor wall is key in determining the flammability limits Swiss-roll micro-combustors, noting that heat recirculation pro-
(flashback, blow-off and extinction) in micro-combustors; (3) a foundly affects blow-off limit due to pre-heating of the incoming
catalytic micro-combustor offers a wider range of flammability gases, but has a minimal effect on extinction. Studies on this inter-
limits, compared to that with only gas-phase reactions; and (4) esting topic are still going on.
flame temperature in micro-combustors are affected by the com-
bustor size, geometry and mixture flow rate.
2.3.2. Cylindrical tubes with backward facing steps
Cylindrical tubes were widely used to study the fundamental
characteristics of micro-combustion, as shown in Table 1, because
2.3. Practical micro-combustors

Depending on the specific application, there are a large variety

of configurations for practical micro-combustors. In terms of mate-
rials used to construct micro-combustors, there are basically three
groups. One is the silicon and silicon-based materials which have
been widely used in MEMS devices. Modern micro-fabrication
technologies (both chemical and mechanical) enable the realiza-
tion of sophisticated design with sufficient precision. The second
one is metals and alloys such as stainless steel, iron and copper
which can survive the high temperature without degradation.
The last group is some non-metallic materials such as ceramic
and alumina which have good mechanical strength and high melt-
ing point. Schubert et al. [39] introduced some advanced micro-
machining technologies that can be applied to metals and alloys
with accuracy up to the micron level. Considering the typical
dimensions encountered in practical micro-combustors, there are
no significant technical difficulties in the fabrication process. In
the following part, two types of micro-combustors with relatively
simple geometries (but other than straight cylindrical tubes or Fig. 2. A typical representation of a 3D Swiss-roll micro-combustor.
rectangular channels) will be reviewed.

Table 2
Numerical investigations of micro-combustors with simple geometries.

Author(s) Fuel-oxidizer Combustor geometry Combustor size Reaction mechanism

Karagiannidis et al. [34] CH4/air Fuel-lean Plane channel H = 1 mm, L = 10 mm Detailed gas-phase and surface reactions
Norton and Vlachos [10] C3H8/air Parallel plates H = 600 lm, L = 10 mm One-step global reaction
Norton and Vlachos [9] CH4/air (U = 1.0) Parallel plates H = 600 lm, L = 10 mm One-step global reaction
Hua et al. [35] H2/air (U = 1.0) Cylindrical chamber Down to d = 0.1 mm Detailed gas-phase reactions
Kaisare and Vlachos [36] CH4/air C3H8 + air (U = 1.0) Parallel plates H = 600 lm One-step global reaction
Li et al. [37] H2/air (U = 0.5) Cylindrical tubes and parallel plate d = 0.4–0.8 mm Detailed gas-phase reactions
Li et al. [38] CH4/air (U = 0.9) Cylindrical tubes and parallel plate d = 1–2 mm Detailed gas-phase reactions
 S.K. Chou et al. / Applied Energy 88 (2011) 1–16 5

the flames can be simplified to two-dimensional cases. Flame sta- fect of porous media on flame temperature and flammability at the
bilization in a straight cylindrical tube is essentially difficult. The micro-scale both by numerical simulation and experimental test.
NUS research group led by Professor S.K. Chou employed a simple
yet effective way, that is, a backward facing step in the millimeter-
sized cylindrical tubes to control the flame position. This configu- 3. Micro-thermophotovoltaic (TPV) power generator
ration has been used as heat source (emitter) for the micro-TPV
system. Experimental results [43] showed that compared to the The micro-TPV power generator is a typical direct energy con-
straight cylindrical tubes, the backward facing step is effective in version device, which uses PV cells to convert heat radiation, from
stabilizing flames in micro-combustors. Furthermore, Li and the combustion of fossil fuels, into electricity. It does not include
coworkers [44,45] confirmed that the position of the peak wall any moving parts; its fabrication and assembly are relatively easy.
temperature can be localized within a narrow zone over a wide As a result, it can be more commonly used for commercial elec-
range of flow conditions (flow velocity and fuel–air equivalence ra- tronics and micro devices. TPV generation of electricity dates back
tio). In order to maximize the output of the micro-TPV generator, a to more than 40 years [50]. However, it is only in recent years that
high wall temperature and uniform distribution is desirable. technological improvements in the field of low band gap photovol-
Therefore, the wall temperature distribution (and radiation power taic cells and high temperature selective emitting materials have
through the combustor wall) is of primary concern. It was noted evoked a renewed interest in TPV generation of electricity [51]. A
that reducing wall thickness [46] and decreasing step height [47] series of conventional and small TPV system have been developed
give rise to higher wall temperature. Recently, a comprehensive [52,53]. However, all of these TPV systems feature a combustor
experimental study to characterize the wall temperature and radi- more than 100 cm3 in volume. In 2002, a micro cylindrical TPV sys-
ation power through the cylindrical micro-combustors was con- tem with a combustor in the order of 0.1 cm3 in volume was initi-
ducted, concluding that the expansion ratio (before and after the ated in the Department of Mechanical Engineering, National
backward facing step) has a remarkable impact on the performance University of Singapore [5]. In 2003, Nielsen et al. [54] also pre-
of the emitter [48]. sented a micro planar TPV system. Different from other micro
power generators being developed around the world, the high sur-
face area-to-volume ratio of micro-combustor holds promise for
2.4. Future works
the micro-TPV power generator to achieve a high power density.
Because the micro-TPV system depends on the radiation energy
2.4.1. Catalyzed micro-combustion
from the surface of the micro-combustor to generate electrical
As generally known, by applying a proper catalyst, the activa-
power, at the same heat flux density per unit surface, the heat flux
tion energy of the reaction can be lowered significantly. As such,
via the wall in terms of per unit volume will increase drastically for
combustion can be initiated and sustained at lower temperature
a micro-combustor. Therefore, the output power of the micro-TPV
than that with only pure gas-phase reactions. Unlike the gas-phase
system will increase as the package size reduces, which makes the
reactions, surface reactions are not very sensitive to the reduced
study of micro-TPV system particularly attractive.
size of micro-combustors, which implies the possibility of further
Fig. 3 shows the schematic of the micro cylindrical TPV system
shrinking of micro-combustors. Boyarko et al. [14] tested combus-
developed by NUS [55]. The system mainly comprises a cylindrical
tion in platinum micro-tubes with an inner diameter of 0.4–
SiC emitter (i.e. micro-combustor), a simple nine layer dielectric
0.8 mm. Okamasa et al. [22] developed a catalytic micro-combus-
filter, and a GaSb photovoltaic cell array. Furthermore, cooling fins
tor by coating palladium on nano-porous alumina. Yang et al.
are employed to remove the waste heat of PV cells. H2/O2 mixture
[32] also investigated the effects of catalytic wall on combustion
is burned in the micro SiC combustor releasing heat. As the emitter
in the cylindrical micro-combustors for the micro-TPV application.
is heated to a high temperature, it emits a stream of photons. The
It will be interesting to find out more about the catalytic effects on
spectrums of SiC emitters operating at temperatures 1000–1600 K
the flame temperature and flammability limits. Moreover, the
contain significant proportions of sub-band gap photons with ener-
respective importance of gas-phase reactions and surface reactions
gies not sufficient enough to generate charge carriers in the PV
as a function of micro-combustor size would be very useful for real
cells. This portion of energy will be absorbed by the PV cells and
engineering applications.
results in a destructive heat load on the generator components,
subsequently lowers the conversion efficiency of the system dras-
2.4.2. Porous media micro-combustion tically. In order to improve the overall efficiency of the micro-TPV
Porous media combustion is a technology widely used in large system, it is very important to recycle these photons. So a simple
scale applications. It is different from conventional combustion nine layer dielectric filter is employed in the micro-TPV system.
which is characterized by a free flame, thin reaction zone and high The filter is fabricated with alternating layers of silicon and silicon
temperature gradients. In conventional combustion, the combus- dioxide, which is able to recycle a portion of the photons with
tion process occurs in a mostly gaseous environment, convection energies lower than the band gap of the PV cells back to the emit-
is the main mode of pre-heating. The gases themselves are non- ter, and transmit most of the photons with energies greater than
conducting as they have very low thermal conductivity. Whereas the band gap of the PV cells. Fig. 4 shows the reflectance of the
in porous media, the combustion takes place in a three dimen- filter.
sional solid porous matrix having interconnected pores. Apart from It is well known that only photons with energy greater than the
convection, the conduction and radiation modes of heat transfer bandgap of PV cells can evoke free electrons and generating elec-
are also activated. This enhances the heat transfer (pre-heating) tricity under the function of p/n junction formed in PV cells. Silicon
from the burned hot gases to unburned mixture. These character- cell (whose bandgap is 1.1 eV) is usually employed to convert solar
istics seem favorable to micro-combustion because the pre-heating radiation into electricity and heat due to its easy fabrication and
is useful in increasing the flame temperature, and therefore sus- low cost. However, compared to solar radiation, the photons emit-
taining flames in a smaller space. Chou et al. [49] simulated the ted from a combustion heat source at 1000–1600 K are distributed
combustion process in a micro cylindrical combustor filled with at much lower energy levels and longer wavelengths (1.5–3.5 lm).
porous media and found that a higher and more uniform temper- Therefore, Low bandgap and high quality PV cells should be devel-
ature distribution can be achieved along the wall of the micro- oped instead of Silicon solar cells, so that both the power density
combustor. It will be very interesting to find out more about the ef- and the efficiency can be maximized. Here, GaSb PV cell is em-
6 S.K. Chou et al. / Applied Energy 88 (2011) 1–16

Fig. 3. Sketch of a micro cylindrical TPV power generator.

cro-combustor. It is found that the performance of the micro-

combustor is significantly affected by the mass flow rate, fuel/oxi-
dant equivalence ratio and the diameters of the micro-combustors,
etc. [46,55–57].
Figs. 6 and 7 show the direct photos of combustion in micro
cylindrical combustors under different conditions. It can be ob-
served from Fig. 6 that the wall temperature of the micro-combus-
tor with a diameter of 3 mm is much higher than that with a
diameter of 4 mm. This is because, when the volumetric flow rate
is the same, the flow speed is higher for the smaller tube, resulting
in a higher convective heat transfer coefficient, and thereby a high-
er wall temperature. Furthermore, with the increase of H2/air
equivalence ratio (U), the wall temperature increases significantly
due to more fuels taking part in combustion. From Fig. 7, we can
Fig. 4. Reflectance of the nine layer dielectric filter [56]. also see that the wall temperature will increase with the increase
of the flow rate.

3.2. Experimental results of the micro-TPV power generator

A prototype micro-TPV power generator is shown in Fig. 8,

where the cylindrical SiC emitter is not incorporated into the sys-
tem [55]. The overall volume is 3.1 cm3. The performance of the
micro-TPV system is evaluated thoroughly.

3.2.1. Effect of backward facing step height

The experiments are carried out in a micro-TPV power genera-
tor with a combustor of 3.5 mm in diameter to study the effect of
backward facing step height on the performance of the system.
Fig. 9 shows the structure of micro-combustor. Three kinds of step
heights are designed by changing the diameter d1, i.e. 0.4 mm,
0.5 mm and 0.6 mm respectively. It was found that flame-flows
back to the upstream of the backward facing step when the step
height is 0.4 mm, however, stable flame can be obtained in the
combustion chamber when the step height are 0.5 mm and
0.6 mm. The results indicate that the backward facing step should
Fig. 5. Quantum efficiency of GaSb PV cells [56].
be high enough to avoid back flow of flame.
Fig. 10 shows the performance of the micro-TPV system with
ployed to convert radiation energy into electricity, since it is the different backward facing steps, when the flow rate at the inlet is
most mature PV cells available. Fig. 5 shows the quantum effi- 0.0476 l/s [56]. From the figure we can observe a slight increase
ciency of GaSb PV cells. for the short-circuit current and the maximum output power when
the step height increases from 0.5 mm to 0.6 mm, especially at
high H2/air ratio. When H2/air ratio is 1.0, the increases are 6.2%
3.1. Optimization of micro cylindrical combustor and 6.7% for short-circuit current and maximum power output
respectively. This is because that a proper step height can facilitate
As one of the most important components of micro-TPV system, recirculation of combustion mixture near the wall, enhancing the
the major demand for micro-combustor is to produce a high and mixing process of combustion around the rim of the tube and
uniform temperature distribution along the wall. Numerical simu- ensuring a complete and stable combustion. So, a higher tempera-
lation and experimental testing was carried out to optimize the mi- ture can be achieved along the wall, thereby increasing the electri-
 S.K. Chou et al. / Applied Energy 88 (2011) 1–16 7

Fig. 8. A prototype micro cylindrical TPV power generator without cooling fins.

Fig. 6. Photos of micro-combustion under the same volumetric flow rate H2/air
equivalence ratio U = 0.6, 0.8, 1.0 and 1.2 (left to right). (a) Diameter = 4 mm (b)
diameter = 3 mm.

Fig. 9. Sketch of micro-combustor with different backward facing step heights.

Fig. 7. Photos of micro-combustion under different flow rates, u0 = 6, 8 and 10 m/s

(left to right).

cal power output of the micro-TPV system. It should be noted that Fig. 10. Performance of the micro-TPV system with different backward facing steps
the step height cannot be increased too much, or a dead zone of [56].
combustion will be existed, which will decrease the performance
of the micro-TPV power generator. combustion chamber is kept as the same for the three combustors:
Furthermore, from the figure, it can be observed that for the two 3 mm in diameter and 18 mm in length.
kinds of step heights, with an increase in the H2/air ratio, both the Fig. 11 shows the maximum electrical power output of the mi-
short-circuit current and the maximum electrical power output in- cro-TPV system under different wall thickness [46], when the flow
crease drastically. This is because more fuels take part in combus- rate at the inlet is kept as 12 m/s, and H2/air ratio varies from 0.5 to
tion, subsequently increases the temperature along the wall of 1.0. They clearly indicate: with a decrease in the wall thickness of
micro cylindrical combustor, improving the spectral distribution the micro SiC combustor, the maximum electrical power output in-
and radiation energy density, thereby, generating higher current creases drastically, especially at low H2/air ratio. For example, as
and electrical power output. the wall thickness decreases from 0.8 mm to 0.6 mm, the increase
in the maximum electrical power output is more than 36%. Espe-
cially, at H2/air ratio of 0.5, the increase in the maximum electrical
3.2.2. Effect of wall thickness power output is 68%. When the wall thickness further decreases
Three kinds of micro cylindrical SiC combustors with different from 0.6 mm to 0.4 mm, the increase in the maximum electrical
wall thickness (0.4 mm, 0.6 mm and 0.8 mm) are fabricated. The power output is also evident but relatively lower.
8 S.K. Chou et al. / Applied Energy 88 (2011) 1–16

Fig. 11. Maximum electrical power output of the micro-TPV system with different
wall thicknesses [46].

Fig. 13. Performance of the micro-TPV system with different combustion chamber
diameters.

Table 3
Performance of a micro-TPV system with different fuel/oxidant mixtures.

Type of fuel/ Open- Short- Maximum Mean wall

oxidant circuit circuit output power temperature
mixture voltage (V) current (A) (W) (K)
H2/air 2.28 0.52 0.92 1325
CH4/O2 2.29 1.25 2.18 1501
H2/O2 2.31 1.42 2.50 1530

compare the effect of different fuel on the performance of the mi-

Fig. 12. Maximum electrical power output of the micro-TPV system at different cro-TPV system at the same input energy. It can be seen from the
flow rates [46]. table that the micro-TPV system has the lowest performance when
H2/air mixture is employed, this is because 79% of the air is N2,
which significantly decrease the flame temperature of H2/air com-
3.2.3. Effect of flow rate bustion. As a result, the mean wall temperature of micro-combus-
Fig. 12 shows the performance of the micro-TPV system with tor with H2/air mixture is much lower than that with the other two
different wall thickness [46], when the H2/air ratio is kept as 0.9, kinds of fuel/oxidant mixtures, subsequently lowers the output
and flow rate changes from 8 m/s to 12 m/s. With the increase of power of the micro-TPV system. The performance of the micro-
flow rate, the maximum electrical power output of the system un- TPV system with H2/O2 mixture is slightly better than that with
der different wall thickness increases almost linearly. This is be- CH4/O2 mixture. This can be explained partly by the higher com-
cause the increase in flow rate enhances the heat release rate, bustion rate of H2/O2 mixture, so that the high temperature com-
subsequently increases the temperature along the wall of micro- bustion products have more time to complete heat transfer with
combustor. the wall of the micro-combustor. These results indicate that the
type of fuel/oxidant has a significant impact on the performance
3.2.4. Effect of combustion chamber of the micro-TPV system.
The performance of the micro-TPV power generator with differ- It should be noted, just like the other micro power generators
ent combustion chamber diameters (3.0 mm and 3.5 mm) are being developed around the world, the development of micro-
shown in Fig. 13 [56], the flow rate is the same. It can be observed TPV system is still at its early stage. To fully exploit the potential
that there is a significant increase both for the short-circuit current of the micro-TPV, further work on the system is required. For
and the maximum output power of micro-TPV system as combus- example, in the present design, hydrogen is employed as the fuel.
tion chamber decreases from 3.5 mm to 3.0 mm. When H2/air ratio The storage and transport is a challenge. In the future, liquid fuel
is 1.0, the increases are 52% and 57% for the short-circuit current and natural gas could be alternatives. For simplicity of fabrication,
and the maximum output power respectively. This is because the SiC is employed as the emitter in the present design, which is a
increase in combustion chamber diameter lowers the flow speed, typical broadband emitting material. The spectrums of broadband
decreasing the heat exchanging rate between high temperature emitters operating at temperatures of 1000–1600 K contain signif-
combustion product and the combustor wall drastically, resulting icant proportions of sub-band gap photons with energies not suffi-
in a lower wall temperature. The reduced wall temperature not cient to generate charge carriers in the PV cells. A proper selective
only decreases the radiation power density, but also worsens the emitter or filter will need to be developed to reshape the radiation
spectral distribution, reducing the percentage of useful radiation spectrum so as to improve the quantum yield of the PV cells. Por-
photons. ous media combustion can also be employed to further improve
the radiation efficiency of the micro-combustor and the output
3.2.5. Effect of fuel/oxidant mixture type power density of the micro-TPV system.
Three kinds of fuel/oxidant mixtures are employed in the test, A novel modular micro-TPV system consisting of a series of pla-
namely, CH4/air, CH4/O2 and H2/O2 mixture [57]. The performance nar combustors, emitters, filters and PV cells is being developed in
of the micro-TPV system is shown in Table 3, where the flow rate of NUS. Fig. 14 shows the schematic of the system. Compared to the
H2 is 4.2 g/h, while the flow rate of CH4 is 10 g/h, so that we can single cylindrical design, this kind of design has several advanta-
 S.K. Chou et al. / Applied Energy 88 (2011) 1–16 9

There are different kinds of micro-propulsion systems being

investigated including micro pulsed plasma thruster, colloid mic-
rothruster, field emission electric propulsion microthruster, hall
microthruster, cold gas microthruster, vaporizing microthruster,
field ionization microthruster, and digital microthruster [60]. As
for the propellant, there are several fuels available for microthrus-
ter. Liquid or gaseous fuels can flow for arbitrarily long times and
the microthrusters can be refueled. However, the disadvantages of
liquid or gaseous fuels are leakages, high pressure storage, techni-
cal complication, and contamination problems. Usually solid fuel
Fig. 14. Schematic of the modular micro-TPV system. microthruster is a single-use device. Nevertheless, this disadvan-
tage can be partially compensated by developing microthruster ar-
rays. Even if some of the individual microthrusters fail to work, the
ges. Firstly, the planar design will simplify the fabrication and array with some redundant microthrusters can still deliver the de-
assembly of the system, most importantly; it is much easier for signed thrust and impulse. In addition, solid propellant microthr-
us to fabricate planar selective emitter, filters and PV cells. Sec- uster might be easier to fabricate and with lower fabrication
ondly, it is not easy to incorporate a recuperator into a single mi- cost. It requires no elaborate system of pumps, fuel lines, and
cro-TPV power generator, which will increase the overall volume valves. Consequently, there are no moving parts and the leakage
of the system significantly. However, for the modular design, this possibility of the propellant is very low. High level of integration
increase is relatively much smaller. The incorporation of a recuper- can be possibly achieved because of the minimization 22 of the to-
ator will not only improve the efficiency of the system, but also tal system volume and complexity. Integrated with MEMS technol-
make it possible for us to use liquid fuel or liquefied gas. Thirdly, ogy, the solid propellant microthruster is specifically suitable for
it is easy to adjust the amount of micro-TPV units according to dif- applications in microspacecraft with high level of integration and
ferent power requirement. miniaturization.
Preliminary studies indicate that a high and uniform tempera-
ture can be achieved in a micro planar combustor, as shown in 4.1. Three-layer sandwich design of solid propellant microthruster
Fig. 15, the direct photos of combustion in a planar micro-combus-
tor under different H2/air equivalence ratios. From the figure, we The solid propellant microthruster is a relatively new class of
can also see that the wall temperature increases rapidly with the microthruster. It is becoming a world-wide active field of research.
increase of H2/air equivalence ratio. It is expected that a prototype In US, research work was carried out on MEMS microthruster dig-
micro modular TPV system to be developed in the near future. ital propulsion system [61]. Arrays of ‘‘Digital Propulsion” mic-
rothruster were fabricated and tested. According to their design,
4. MEMS based solid propellant micro-propulsion systems a three-layer sandwich microthruster was fabricated containing
micro resistors, thrust chambers, and rupture diaphragms. Solid
Microspacecraft is one application of MEMS technology in the propellant was loaded into each individually sealed chamber.
aerospace field. One focus of the aerospace industry is to reduce When energizing the resistor, the propellant was ignited, raising
spacecraft life-cycle cost while still delivering a spacecraft that the pressure in the chamber and rupturing the diaphragm. An im-
can perform useful science or commercial service. One of the pulse was imparted as the high-pressure fluid was expelled from
objectives is the development of microspacecraft. The reduction the chamber. A total of 106 thrusters were fabricated on a single
in complexity of microspacecraft will lower cost by facilitating sys- wafer. The initial tests, employing lead styphnate as the propellant,
tem integration. Additionally the small sizes allow the selection of produced 104 N s of impulse and about 100 W of power. In Eur-
a smaller, less expensive launch vehicle, or the integration of mul- ope, a solid propellant microthruster on silicon was proposed
tiple microspacecrafts per vehicle. Future spacecraft missions may [62]. Its principle is based on the integration of energetic material
be composed of many microspacecrafts flying in formation rather within a silicon micro machined system. The operational concept is
than a single larger spacecraft. Clusters of microspacecrafts may based on the combustion of an energetic propellant stored in a mi-
be employed to increase the reliability of the system, form a large cro machined chamber. Each microthruster contains three main
sparse aperture, or simply to provide greater coverage of an area. In parts (a heater, a chamber, and a nozzle). Generally speaking, the
each case, the micro-propulsion system is required in microspace- basic design concepts of previous approaches are the three-layer
craft for station keeping, attitude control, gravitation compensa- sandwich configurations, which normally contain three parts con-
tion, and orbit adjust [59]. sisting of a propellant combustion chamber, a micronozzle (or

Fig. 15. Photos of combustion in a planar micro-combustor, H2/air equivalence ratio U = 0.6, 0.8, 1.0 (left to right) [58].
10 S.K. Chou et al. / Applied Energy 88 (2011) 1–16

Fig. 17. Schematic of the two-layer building block design of solid propellant
microthruster [64].

which improves the fabrication process efficiency. Second, for the

three-layer sandwich design, the chambers are etched vertically
from the wafer surface. Consequently, longer chamber length is
difficult to fabricate due to the depth limitation of micro-fabrica-
tion technology. Furthermore, the nozzle divergence angle, length
and throat dimension have great impact on the performance of
the microthruster. It is very difficult for the sandwich design to
realize arbitrary nozzle dimensions. However for building block
design, the chamber and convergent–divergent nozzle can be
Fig. 16. (a) Three-layer sandwich design of solid propellant microthruster and (b) etched horizontally along the wafer surface. As a result, longer
Fabricated microthruster chip in a ceramic package [61]. chambers and arbitrary nozzle dimensions can be easily achieved.
Third, anodic bonding is employed in the building block design to
bond the glass and silicon wafers together. Its bonding quality is
burst diaphragm), and an igniter as shown in Fig. 16a. Fig. 16b
better than that of the bonding methods (adhesive and epoxy)
shows the fabricated three-layer sandwich microthruster chip in
adopted in the sandwich design.
a 24-pin ceramic package.
One drawback of the solid propellant microthruster is that it is a
one-shot device. Therefore it lacks the restart ability. Nevertheless,
4.2. Two-layer building block design of solid propellant microthruster this can be partially redeemed by realizing microthruster arrays.
The microthruster array can contain an arbitrary number of single
A building block design different from the above-mentioned microthrusters. Even if some of the single microthrusters fail to
sandwich configuration was proposed by Zhang et al. [63–65] as work, the array with some redundant microthrusters can still
shown schematically in Fig. 17. In this new design, a metal thin deliver the designed thrust and impulse. Moreover, the single mic-
film igniter is deposited onto a Pyrex-7740 glass wafer. A silicon rothrusters in the array can be addressed and ignited individually,
wafer is fabricated using MEMS technologies to contain a combus- several together or in controlled sequences, to produce controlled
tion chamber and a convergent–divergent nozzle. The glass wafer and vectored thrust. In the building block design, the microthruster
and the silicon wafer are bonded together to form a three-dimen- array can be fabricated by bonding single microthrusters and/or
sional microthruster. The chamber is loaded with the solid propel- microthruster layers together, like in building block. A schematic
lant. Once the propellant is ignited, the combustion gas expands of one way to address and ignite single microthrusters individually
through the nozzle and produces the desired thrust and impulse in the microthruster array to produce the controlled, vectored
bit. A larger piece of glass wafer is used and connection is made thrust and impulse bit is shown in Fig. 18.
to the end of the glass protruding out from the silicon part of the
microthruster. The protruding glass part of the microthruster is in-
stalled into a special micro-connector to realize the connection 4.3. Fabrication of the two-layer building block microthruster
with the motherboard that contains the power source, addressing
electronics and communication ports. The fabrication of the two-layer building block microthruster is
The building block design has several advantages over the performed using standard MEMS technologies. A Pyrex-7740 glass
three-layer sandwich design. First, the propellant chamber and wafer is used as the substrate for fabricating the metal thin film
the convergent–divergent nozzle are fabricated simultaneously, igniter. The Au/Ti metal layers are deposited by e-beam evapora-
 S.K. Chou et al. / Applied Energy 88 (2011) 1–16 11

Fig. 18. Schematic of addressing system for single microthrusters in a building block microthruster array [64].

tion and patterned by photolithography and lift-off processes. A

deep reactive ion etching (DRIE) process is performed on a silicon
wafer to yield the propellant chamber and the convergent–diver-
gent nozzle. A thermal oxidation process is then performed to cov-
er the wafer with a 246 nm thick SiO2 layer to improve the
electrical insulation property of the silicon, which can increase
the ignition efficiency due to the minimization of the possible cur-
rent leakage. The glass wafer with the Au/Ti thin film igniter and
the silicon wafer with the chamber and nozzle are bonded together
using an anodic bonding process. The fabricated 3-D microthruster
with the Ti igniter and solid propellant is shown in Fig. 19. The
microthruster array is installed on a custom-built electronic cir-
cuitry for addressing, ignition, and thrust testing as shown in
Fig. 20 [66].

4.4. Combustion and thrust tests of the two-layer building block

Fig. 20. Microthrusters on circuit board.
microthruster

It is known that the chemical and thermal quenching due to the

tire surface rapidly. The ejected plume from the convergent–diver-
small size of the microthruster may cease the micro-combustion.
gent nozzle exit can also be visually seen from Fig. 18.
To validate the feasibility of the two-layer building block solid pro-
A specific setup is developed to test the thrust and impulse of
pellant microthruster design, micro-combustion experiment is car-
the solid propellant microthruster both at sea level with the ambi-
ried out as shown in Fig. 21, where the frames are taken from a
ent temperature of approximately 300 K and in vacuum with a
high-speed digital video camera. The combustion process of the
back pressure of 80 Pa. The stand consists of a high-speed digital
gunpowder-based solid propellant can be visually observed in
video camera to capture the propellant micro-combustion, a piezo-
the design due to the transparent property of the Pyrex-7740 glass.
electric force sensor, a charge amplifier, a data acquisition system
The propellant is ignited initially at some point close to the thin
and a DC power supply to measure the produced thrust precisely.
film igniter. Then the combustion occurs and propagates to the en-
Fig. 22 shows a typical measurement signal in vacuum. The com-
bustion lasts for around 0.66 ms and the peak value of thrust is
about 0.13 N. At 0.00 ms, the thin film igniter is energized and
the combustion of the propellant is initiated. The generated thrust
increases quickly and reaches the peak value at 0.25 ms. The thrust
maintains a high value for around 0.05 ms and decreases gradually.
At 0.66 ms the combustion of the propellant is complete and the
impulse has been delivered. Total impulse, defined as the integra-
tion of thrust over combustion time, is 3.52  105 N s for this sin-
gle microthruster.

4.5. Ignition study of the two-layer building block microthruster

One key challenge in developing the solid propellant microthr-

uster is to control the propellant transient ignition process. The
ignition power, ignition delay, and ignition energy are crucial
parameters for the applications of the microthruster. A finite-ele-
ment based electro-thermal modeling has been performed to pre-
dict and optimize the transient ignition process of the solid
Fig. 19. 3-D microthrusters [64]. propellant [66]. The model is able to determine the evolution of
12 S.K. Chou et al. / Applied Energy 88 (2011) 1–16

Fig. 21. Microcombustion of solid propellant [64].

Fig. 22. Variation of the thrust with combustion time [64].

the propellant temperature with time, space and voltage, and also
to derive ignition power, ignition delay, and ignition energy, which
can determine an optimum ignition voltage to minimize ignition Fig. 23. Ignition power and delay variations with ignition voltage [67].
energy and to improve ignition efficiency. Fig. 23 shows the igni-
tion power and ignition delay variations with ignition voltage. Igni-
tion power increases and ignition delay decreases while the trostatic micro generators, nano energetic material based micro
ignition voltage increases. Ignition delay changes more drastically power systems. This section gives a brief introduction to these
than ignition power. Thus ignition delay is the main factor that af- power MEMS systems.
fects ignition energy and ignition efficiency. Fig. 24 shows ignition
energy variation with ignition voltage. Ignition energy decreases
rapidly with voltage when ignition voltage is lower than 15 V. 5.1. Micro gas turbine engine
When the voltage is higher than 15 V, ignition energy decreases
slowly and reaches the lowest at 22 V. The ignition energy in- One of the first power MEMS devices was the gas turbine engine
creases slowly when voltage is over 22 V. It can be seen that opti- studied at MIT [68]. It was designed to generate power at the 10–
mum is quite flat at 15–22 V. The lower limit (15 V) is 50 W scale. The engine design is shown in Fig. 25. The diameters of
recommended to be used due to the power source consideration the centrifugal compressor and radial turbine rotor are 8 mm and
in practical applications. 6 mm, respectively. The compressor discharge air wraps around
the outside of the combustor to cool the combustor walls, capture
the waste heat, increase the combustor efficiency and reduce the
5. Other power MEMS systems external package temperature. The rotor is supported on a bearing
on the periphery of the compressor by thrust bearings on the rotor
Besides the above-discussed micro-TPV and microthruster, centerline. The axial loads are supported by the thrust bearings and
there are also other power MEMS systems being investigated such a thrust balance piston behind the compressor disk support. The
as micro turbines and miniature internal combustion engines, balance piston is the air source for the hydrostatic journal bearing.
thermoelectric power generators and micro coolers, vibration en- The thrust bearings and balance piston are supplied from external
ergy harvesting devices including piezoelectric, magnetic or elec- air sources. The design peripheral speed of the compressor is
 S.K. Chou et al. / Applied Energy 88 (2011) 1–16 13

larger-scale ‘‘mini-rotary” engines have been fabricated from steel.

The mini rotary engine chambers are approximately 1000–
1700 mm3 in size and their displacements range from 78 mm3 to
348 mm3. Testing has been performed using hydrogen–air mix-
tures and range of spark and glow plug designs as the ignition
source. Preliminary testing has shown net power output of up to
2.7 W at 9300 RPM.

5.3. Micro P3 heat engine

Whalen et al. have developed a heat engine (known as the P3)

based on the compression and expansion of a saturated two-phase
fluid that is driven by an external heat source [72]. As shown in
Fig. 27, an external heat source is used in this engine to convert
the working fluid from liquid to gas. The expansion of the fluid
bends a silicon membrane. A thin piezoelectric film on the bending
membrane produces electric power. The engine would operate
close to the resonant frequency (1000 Hz) of the membrane to
achieve peak power. This engine could be used in an energy-scav-
Fig. 24. Ignition energy variation with voltage [67].
enging situation or with an external combustor.

5.4. Thermoelectric micro power generator and micro cooler

Thermoelectric power generators have also been widely inves-

tigated to use thermoelectric elements to generate electricity.
Thermoelectric power generator benefits at small scale because
of the increase in surface area-to-volume ratio. Ronney et al. devel-
oped a ‘‘Swiss Roll” counter flow heat exchanger and combustor
using electrochemical fabrication processes as shown schemati-
cally in Fig. 28. Both gas phase and catalytic combustion have been
demonstrated in the ‘‘Swiss Roll” device with an enthalpy flux of
order 50 W at micro-scale. Such a ‘‘Swiss Roll” device could pro-
Fig. 25. H2 engine with conduction-cooled turbine constructed from silicon wafers duce of order 1 W of electric power with reasonable conversion
[68]. efficiency from thermoelectric elements [15].
Advance in MEMS fabrication technique has made it possible to
fabricate various micro devices. Some of these devices require the
500 m/s and the rotation rate is 1.2 Mrpm. External air is employed
precise thermal management and the compact cooling system for
to start the engine and 11 g thrust or 17 W shaft power is expected
effective cooling within a small volume. The thermoelectric micro
from the engine.
cooler has become a promising candidate due to its cooling power
Other groups including Singapore Institute of Manufacturing
density higher than that of the conventional bulk cooler. Thermo-
Technology, The University of Tokyo and Tohoku University, etc.
are also focusing on the study of a micro gas turbine engine
[69,70].

5.2. MEMS rotary engine

The University of California at Berkeley has been developing a

MEMS rotary engine [71]. Fig. 26 shows the engine design. The ulti-
mate goal of the project is to develop a liquid hydrocarbon fueled
MEMS-size rotary engine capable of delivering power on the order
of milli-watts. To investigate engine behavior and design issures, Fig. 27. Schematic of a P3 vapor cycle heat engine [72].

Fig. 26. micro rotary engine [71].

14 S.K. Chou et al. / Applied Energy 88 (2011) 1–16

Fig. 28. Schematic plan view of a Swiss Roll burner [15].

electric devices use the Peltier effect to achieve micro-scale cool-

ing. The thermoelectric device contains over hundreds of n–p cou-
ples connected electrically in series, but thermally in parallel
between two planar substrates. Numerical analysis has been per-
formed to figure out the performance of the thermoelectric micro Fig. 29. Schematic of a thermoelectric cooler [73].
cooler with a 3D model as shown in Fig. 29. A small size and col-
umn-type thermoelectric cooler is studied and Bi2Te3 and Sb2Te3
are selected as the n- and p-type thermoelectric materials, respec-
tively. The effect of parameters including the temperature differ-
ence, the thickness of a thermoelectric element, the current, and
the number of thermoelectric pairs on the performance of the mi-
cro cooler has been investigated. The predicted results show that
the performance can be improved for the thick element with the
large number of thermoelectric pairs or the small cross-sectional
area of the element [73].

5.5. Mechanical energy scavengers

Power MEMS also include devices that convert mechanical en-

ergy into electric power. Several mechanically-based energy scav-
enging systems have been developed. These systems tend to
produce electric power in the order form lW to mW, depending Fig. 30. 3D cut-away schematic of a piezoelectric actuator device [74].
on size and excitation level. Most efforts aim to convert vibration
kinetic energy to electric power, where the kinetic energy is trans-
ferred to a mass within the device that is normally supported on face-to-volume ratio of the reactants [75]. Recently, nano Al/CuO
springs. The electric generator can be piezoelectric or it can be nanowire based nEM has been realized by integrating nano Al with
electromagnetic by having the vibrating mass be magnetic. A novel CuO nanowires grown from copper thin film deposited onto a sili-
micro hydraulic transducer driven by a piezoelectric actuator was con substrate [76]. The Cu thin film is converted into CuO nano-
developed by Roberts et al. as shown in Fig. 30. High pressures wires and CuO/Cu2O thin film after the thermal oxidation. After
could be achieved within the transducer by exciting the actuator. Al deposition by thermal evaporation, nano Al is integrated with
Similarly, by applying pressure to the transducer (for example by CuO nanowires to form a core–shell nanostructure and also depos-
a heel strike), one should be able to extract electric power from ited onto the CuO/Cu2O thin film beneath the nanowires to form a
the piezoelectric layer, possibly as much as 1 W from a person with layered structure as shown in Fig. 31. Experimental results suggest
a weight of 100 kg [74]. that nano Al reacts with CuO nanowires with an onset temperature
of about 500 °C. The reaction is based on the solid–solid diffusion
5.6. Nano energetic material based power MEMS systems mechanism. After melting, the remaining Al reacts with the CuO/
Cu2O thin film beneath the CuO nanowires. The total heat of reac-
Nano energetic materials (nEMs) are believed to be one key to tion of the thermite reactions is estimated to be around 2950 J/g.
the great advance in power MEMS systems such as micro actua- The realized Al/CuO based nEM has the advantages of enhanced
tion, micro igniter, micro-propulsion and other energy-demanding interfacial contact area, reduced impurities, and tailored dimen-
microsystems. NEMs have shown improved performances in terms sions compared to previous approaches. Most interestingly, the
of energy release rate, ignition, stability, and mechanical properties Al/CuO based nEM is realized onto a silicon substrate, a basic mate-
compared to their bulk or micro counterparts due to increased sur- rial for microelectronics and microsystem. Therefore, this will
 S.K. Chou et al. / Applied Energy 88 (2011) 1–16 15

Fig. 31. Cross-sectionview of SEM images of (a) CuO nanowires and (b) Al/CuO based nano energetic materials [76].

of work that needs to be done before its application in practice.

For example, direct energy conversion power generators such as
micro-TPV system and micro thermoelectric system, face the chal-
lenge of low energy conversion efficiency. Indirect energy conver-
sion systems such as micro turbine engine and micro rotary engine
face even more challenges in fabrication, assembly and efficiency.
The application of MEMS based solid propellant micro-propulsion
systems is also hindered by the lack of suitable solid propellants.
Even so, the great progress achieved during the past years indicates
that the future of power MEMS is bright.

References

[1] Epstein AH, Senturia SD. Macro power from micro machinery. Science
1997;276:1211.
[2] Ohadi MM, Buckley SG. High temperature heat exchangers and microscale
combustion system: application to thermal system miniaturization. Exp
Thermal Fluid Sci 2001;25:207–17.
[3] Waitz IA, Gauba G, Yang ST. Combustors for micro-gas turbine engines. ASME J
Fluids Eng 1998;120:109–17.
[4] Sitzki L, Borer K, Schuster E, Ronney PD, Wussow S. Microscale combustion
research for applications to MEMS rotary IC engine. In: The 3rd Asia-Pacific
Conf Combustion, Seoul, Korea; 2001. p. 1–4.
[5] Yang WM, Chou SK, Shu C, Li ZW, Xue H. Development of
microthermophotovoltaic system. Appl Phys Lett 2002;81:5255–7.
Fig. 32. Al/CuO nano energetic material based micro/nano initiator. [6] Fernandez-Pello AC. Micropower generation using combustion: issues and
approaches. Proc Combust Inst 2002;29:883–925.
[7] Ko WH. Trends and frontiers of MEMS. Sensor Actuat A 2007;136:62–7.
[8] Aghalayam P, Vlachos DG. Roles of thermal and radical quenching in emissions
of wall-stabilized hydrogen flames. AIChE J 1998;44:2026–34.
probably open the door to integrate the nEM into microsystem, [9] Norton DG, Vlachos DG. Combustion characteristics and flame stability at the
thus leading to functional devices. microscale: a CFD study of premixed methane–air mixtures. Chem Eng Sci
Very recently, Zhang et al. [77] developed an innovative micro/ 2003;58:4871–82.
[10] Norton DG, Vlachos DG. A CFD study of propane–air microflame stability.
nano initiator by integrating the Al/CuO nanowire based nano Combust Flame 2004;138:97–107.
energetic material with a Au/Pt/Cr micro heater on a substrate as [11] Kim KT, Lee DH, Kwon S. Effects of thermal and chemical surface-flame
shown in Fig. 32. This functional device was fabricated using interaction on flame quenching. Combust Flame 2006;146:19–28.
[12] Maruta K, Kataoka T, Kim NI, Minaev S, Fursenko R. Characteristics of
standard MEMS technologies and nano manufacturing, which is combustion in a narrow channel with a temperature gradient. Proc Combust
suitable for batch fabrication and high level of integration. Com- Inst 2005;30:2429–36.
bustion of the nano energetic material is accompanied by a bright [13] Fan A, Minaev S, Kumar S, Liu W, Maruta K. Regime diagrams and
characteristics of flame patterns in radial micochannels with temperature
flash of light. This is due to the high temperature of the combus-
gradients. Combust Flame 2008;153:479–89.
tion. The ejected high temperature products can ensure the suc- [14] Boyarko GA, Sung C, Schneider SJ. Catalyzed combustion of hydrogen–oxygen
cessful ignition of the attached reactive materials even if the in platinum tubes for micro-propulsion applications. Proc Combust Inst
attachment between the reactive materials and the initiator is 2005;30:2481–8.
[15] Ahn J, Eastwood C, Sitzki L, Ronney PD. Gas-phase and catalytic combustion in
not intimate. The micro/nano initiator has many potential applica- heat-recirculating burners. Proc Combust Inst 2005;30:2463–72.
tions such as triggering the inflation of airbags in automobiles, mi- [16] Ronney PD. Analysis of non-adiabatic heat-recirculating combustors. Combust
cro-propulsion, arm fire/safe devices, and many other ordnance Flame 2003;135:421–39.
[17] Ju Y, Choi CW. An analysis of sub-limit flame dynamics using opposite
systems. propagating flame in mesoscale channels. Combust Flame 2003;133:483–93.
[18] Leach TT, Cadou CP, Jackson GS. Effect of structural conduction and heat loss
on combustion in micro-channels. Combust Theory Model 2006;10:85–103.
6. Conclusion [19] Norton DG, Wetzel ED, Vlachos DG. Fabrication of single-channel catalytic
microburners: effects of confinement on the oxidation of hydrogen/air
mixtures. Ind Eng Chem Res 2004;43:4833–40.
Although significant progress has been achieved, the develop- [20] Zhang X, Mehra A, Ayon AA, Waitz IA. Igniters and temperature sensors for a
ment of power MEMS is still at the early stage. There is still a lot micro-scale combustion system. Sensor Actuat A 2003;103:253–62.
16 S.K. Chou et al. / Applied Energy 88 (2011) 1–16

[21] Kim NI, Kato S, Kataoka T, Yokomori T, Maruyama S, Fujimori T, et al. Flame [50] White DC, Wedlock BD, Blair J. Recent advance in thermal energy conversion.
stabilization and emission of small Swiss-roll combustors as heaters. Combust In: 15th Annual power sources conference, Atlantic city, NJ; 1961. p. 125–32.
Flame 2005;141:229–40. [51] Coutts TJ. A review of progress in thermophotovoltaic generation of electricity.
[22] Okamasa T, Lee G-G, Suzuki Y, Kasagi N, Matsuda S. Development of a micro Renew Sustain Energy Rev 1999;3:77–184.
catalytic combustor using high-precision ceramic tape casting. J Micromech [52] Durisch W, Bitnar B, Mayor J-C, Fritz VR, Sigg H, Tschudi HR, et al. Small self-
Microeng 2006;16:S198–205. powered grid-connected thermophotovoltaic prototype system. Appl Energy
[23] Bessler WG, SchuLz C. Quantitative multi-line NO-LIF temperature imaging. 2003;74:149–57.
Appl Phys B 2004;78:519–33. [53] Fraas LM, Samaras JE, Huang HX, Minkin LM, Avery JE, Daniels WE, et al. TPV
[24] Lee TH, Wolfgang GB, Helmut K, Christof S, Jay BJ. Quantitative temperature generators using the radiant tube burner configuration. In: 17th European
measurements in high-pressure flames with multiline No-LIF thermometry. photovoltaic solar energy conference and exhibition, Munich; 22–26 October
Appl Opt 2005;44:6718–28. 2001. p.1–4.
[25] Muller D, Triebel W, Bochmann A, Schmidl G, Eckardt D, Burkert A, et al. Two- [54] Nielsen OM, Arana LR, Baertsch CD, Jensen KF, Schmidt MA. A
dimensional concentration and temperature measurements in extended thermophotovoltaic micro-generator for portable power applications. In:
flames of industrial burners using PLIF. Proc SPIE 2003;5191:66–74. 12th International conference on solid state sensors, actuators and micro
[26] Zamashchikov VV. Experimental investigation of gas combustion regimes in systems, Boston; 8–12 June 2003. p. 714–7.
narrow tubes. Combust Flame 1997;108:357–9. [55] Yang WM, Chou SK, Shu C, Li ZW, Xue H. Design, Fabrication and testing of a
[27] Mellish BP, Miller FJ, Ti’en JS, Dietrich DL, Struk PM. Premixed flames stabilized prototype micro thermophotovoltaic power generator. J Microelectromech
on or propagating inside micro tubes. In: The 4th joint meeting of the US Syst 2004;13:851–6.
sections of the combust inst, Philadelphia; 2005. p. 1–4. [56] Yang WM, Chou SK, Shu C, Li ZW, Xue H. Experimental study of micro-
[28] Richecoeur F, Kyritsis DC. Experimental study of flame stabilization in low thermophotovoltaic systems with different combustor configurations. Energy
Reynolds and Dean number flows in curved mesoscale ducts. Proc Combust Convers Manage 2007;48:1238–44.
Inst 2005;30:2419–27. [57] Chou SK, Yang WM, Li ZW, Li Z. A promising energy source for portable MEMS
[29] Prakash S, Armijo AD, Masel RI, Shannon MA. Flame dynamics and structure devices. In: The 2nd joint international conference on sustainable energy and
within sub-millimeter combustors. AIChE J 2007;53:1568–77. environment, Bangkok, Thailand; 21–23 November 2006 [A-040-1-6].
[30] Li ZW, Chou SK, Shu C, Yang WM. Entropy generation during microcombustion. [58] Li J, Chou SK, Li ZW, Yang WM. A potential heat source for the micro-
J Appl Phys 2005;97:84914-1-8. thermophotovoltaic (TPV) system. Chem Eng Sci 2009;64:3282–9.
[31] Miesse CM, Masel RI, Jensen CD, Shannon MA, Short M. Submillimeter-scale [59] Bayt RL. Analysis, fabrication, and testing of a MEMS-based micropropulsion
combustion. AIChE J 2004;50:3206–14. system. Ph.D. Dissertation, Massachusetts Institute of Technology; 1999.
[32] Yang WM, Chou SK, Shu C, Li ZW, Xue H. Study of catalytic combustion and its [60] Choueiri EY. Overview of US academic programs in electric propulsion. In:
effect on microthermophotovoltaic power generators. J Phys D: Appl Phys 35th AIAA joint propulsion conference; 1999 [AIAA Paper 99-2163].
2005;38:4252–5. [61] David Jr HL, Siegfried WJ, Ronald BC, Erik KA. Digital micropropulsion. Sensor
[33] Li ZW, Chou SK, Shu C, Xue H, Yang WM. Characteristics of premixed flame in Actuat A Phys 2000;80:143–54.
microcombustors with different diameters. Appl Thermal Eng 2005;25: [62] Rossi C, Dumonteui M, Camps T, Briand D, Phuong QP, de Rooij NF. Matrix of
271–281. 1010 addressed solid propellant microthrusters: review of the technologies.
[34] Karagiannidis S, Mantzaras J, Jackson G, Boulouchos K. Hetero-/homogeneous Sensor Actuat A Phys 2006;126:241–52.
combustion and stability maps in methane-fueled catalytic microreactors. [63] Zhang KL, Chou SK, Ang SS. Development of a solid propellant micro-thruster
Proc Combust Inst 2007;31:3309–17. with chamber and nozzle etched on a wafer surface. J Micromech Microeng
[35] Hua JS, Wu M, Kumar K. Numerical simulation of the combustion of hydrogen– 2004;14:785–92.
air mixture in micro-scaled chambers. Part I: fundamental study. Chem Eng Sci [64] Zhang KL, Chou SK, Ang SS, Tang XS. A MEMS-based solid propellant
2005;60:3497–506. microthruster with Au/Ti igniter. Sensor Actuat A Phys 2005;122:113–23.
[36] Kaisare NS, Vlachos DG. Optimal reactor dimensions for homogeneous [65] Zhang KL, Chou SK, Ang SS. Performance prediction of a novel solid-propellant
combustion in small channels. Catal Today 2007;120:96–106. microthruster. J Propul Power 2006;22:56–63.
[37] Li J, Chou SK, Li ZW, Yang WM. A comparative study of H2–air premixed flame [66] Zhang KL, Chou SK, Ang SS. A wireless addressing interface circuitry for
in micro combustors with different physical and boundary conditions. microthruster array applications. Aircr Eng Aerosp Technol 2007;79:628–34.
Combust Theory Model 2008;12:325–47. [67] Zhang KL, Chou SK, Ang SS. Investigation on the ignition of a MEMS solid
[38] Li J, Chou SK, Yang WM, Li ZW. A numerical study on premixed micro- propellant microthruster before propellant combustion. J Micromech
combustion of CH4–air mixture: effects of combustor size, geometry and Microeng 2007;17:322–32.
boundary conditions on flame temperature. Chem Eng J 2009;150:213–22. [68] Epstein AH. Millimeter-scale, micro-electro-mechanical systems gas turbine
[39] Schubert K, Brandner J, Fichtner M, Linder G, Schygulla U, Wenka A. engines. J Eng Gas Turb Power 2004;126:205–27.
Microstructure devices for applications in thermal and chemical process [69] Shan XC, Wang ZF, Maeda R, Sun YF, Wu M, Hua JS. A silicon-based micro gas
engineering. Microscale Thermophys Eng 2001;5:17–39. turbine engine for power generation. In: Proceeding of DTIP of MEMS &
[40] Vigan J, Gajdeczko BF, Dryer FL, Milius DL, Aksay IA, Yetter RA. Development of MOEMS, Stresa, Italy; 2006. p. 1–5.
a microreactor as a thermal source for micro-electromechanical systems [70] Isomura K, Murayama M, Teramoto S, Hikichi K, Tndo Y, Togo S, et al.
power generation. Proc Combust Inst 2002;29:909–16. Experimental verification of the feasibility of a 100 W class micro-scale gas
[41] Kuo CH, Ronney PD. Numerical modeling of non-adiabatic heat-recirculating turbine at impeller diameter 10 MM. In: Proceeding of power MEMS 2005, The
combustors. Proc Combust Inst 2007;31:3277–84. University of Tokyo, Tokyo, Japan; 2005. p. 25–8.
[42] Federici JA, Vlachos DG. A computational fluid dynamics study of propane–air [71] Fu K, Aaron JK, Fabian CM, David CW, Carlos FP, Al PP, et al. Design and
microflame stability in a heat recirculation reactor. Combust Flame experimental results of small-scale rotary engines. In: Proceedings of ASME
2008;153:258–69. international mechanical engineering congress and exposition, New York,
[43] Yang WM, Chou SK, Shu C, Li ZW, Xue H. Combustion in micro-cylindrical USA; 2005.
combustors with and without a backward facing step. Appl Thermal Eng [72] Whalen S, Thompson M, Bahr D, Richards C, Richards R. Design, fabrication and
2002;22:1777–87. testing of the P3 micro heat engine. Sensor Actuat A Phys 2003;104:290–8.
[44] Li J, Chou SK, Yang WM, Li ZW. Experimental and numerical study of the wall [73] Lee KH, Kim OJ. Analysis on the cooling performance of the thermoelectric
temperature of cylindrical micro combustors. J Micromech Microeng micro-cooler. Int J Heat Mass Transfer 2007;50:1982–92.
2009;19:015019-1–015019-11. [74] Roberts DC, Li H, Steyn JL, Turner KT, Mlcak R, Saggere L, et al. A high-
[45] Li J, Chou SK, Huang G, Yang WM, Li ZW. Study on premixed combustion in frequency high-stiffness piezoelectric actuator for microhydraulic
cylindrical micro combustors: transient flame behavior and wall heat flux. Exp applications. Sensor Actuat A Phys 2002;97–98:620–31.
Thermal Fluid Sci 2009;33:764–73. [75] Rossi C, Zhang K, Esteve D, Alphonse P, Tailhades P, Vahlas C. Nano energetic
[46] Yang WM, Chou SK, Shu C, Li ZW, Xue H. Effect of wall thickness of micro materials for MEMS: a review. J Microelectromech Syst 2007;16:919–31.
combustor on the performance of micro-thermophotovoltaic power [76] Zhang K, Rossi C, Tenailleau C, Alphonse P, Rodriguez GAA. Development of a
generators. Sensor Actuat A Phys 2005;119:441–5. nano Al/CuO based energetic material on silicon substrate. Appl Phys Lett
[47] Li ZW, Chou SK, Shu C, Yang WM. Effects of step height on wall temperature of 2007;91:113117.
a microcombustor. J Micromech Microeng 2005;15:207–12. [77] Zhang K, Rossi C, Petrantoni M, Mauran N. A Nano initiator realized by
[48] Li J, Chou SK, Li ZW, Yang WM. Characterization of wall temperature and integrating Al/CuO-based nanoenergetic materials with a Au/Pt/Cr
radiation power through cylindrical dump micro-combustors. Combust Flame microheater. J Microelectromech Syst 2008;17:832–6.
2009;156:1587–93.
[49] Chou SK, Yang WM, Li J, Li ZW. Porous media combustion for micro
thermophotovoltaic system applications. Appl Energy 2010;87:2862–7.