Chemical Engineering Journal 295 (2016) 384–390

Contents lists available at ScienceDirect

Chemical Engineering Journal

journal homepage: www.elsevier.com/locate/cej

Measuring enthalpy of fast exothermal reaction with infrared

thermography in a microreactor
J.S. Zhang ⇑, C.Y. Zhang, G.T. Liu, G.S. Luo ⇑
The State Key Laboratory of Chemical Engineering, Department of Chemical Engineering, Tsinghua University, Beijing 100084, China

h i g h l i g h t s

A microsystem integrated with infrared camera is developed to measure the reaction enthalpy.

The reaction enthalpy can be determined from the temperature rise of the capillary.

Two reactions including homogeneous and heterogeneous reactions are performed.

It can provide a quick, safe and non-intrusive method to obtain the temperature profiles for fast exothermic reactions.

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

Article history: In this paper, a microsystem integrated with infrared camera is developed to measure the reaction
Received 30 September 2015 enthalpy of fast exothermic reactions. The microsystem is composed of a micromixer and a coiled
Received in revised form 20 January 2016 capillary, which are all placed in a vacuum box to form an adiabatic environment. The fast exothermic
Accepted 22 January 2016
reactions are started in the micromixer and then the temperature profiles of the capillary are measured
Available online 4 February 2016
via the infrared transparent window by the infrared camera. The reaction enthalpy can be determined
from the temperature rise of the capillary. Two reactions which are homogeneous and heterogeneous
Keywords:
respectively are performed with this method. The results show that the data obtained are in good agree-
Microreactor
Infrared thermography
ment with the values reported in literature with the relative experimental errors less than 5%. Compared
Reaction enthalpy with the traditional calorimeter, the microsystem can provide a quick, safe and non-intrusive method to
Exothermal reaction obtain the temperature profiles of the fast exothermic reactions.
Ó 2016 Published by Elsevier B.V.

1. Introduction information with conventional equipment considering the charac-

Over the last two decades microreactors have gained rapid
development in carrying out efficiently numerous chemical reac-
tions, of which the fast and exothermal reactions are thought to
be one of the most suited reactions in microreactors [1–3]. Due
to the characteristics of fast mixing, excellent heat and mass trans-
fer, inherent safety in microreactors, it can bring many advantages:
higher selectivity, faster reaction rate, better process control and
increased safety. A lot of fast and exothermal have been performed
in microreactors such as nitration [4], fluorination [5], rearrange-
ment [6,7] and diazo reaction [8]. For such reactions, how to get
a precise reaction enthalpy and temperature profiles in microreac-
tors are crucial to control the reaction process and perform the
process design. However, it is difficult to get these thermal
⇑ Corresponding authors. Tel./fax: +86 62 795904 (J.S. Zhang). Tel./fax: +86 62
783 870 (G.S. Luo).
E-mail addresses: zhangjisong2005@163.com (J.S. Zhang), gsluo@tsinghua.edu.
cn (G.S. Luo).
teristic dimension smaller than 1000 lm in microreactors.
There have been many methods developed to measure the tem-
perature profiles in microchannels. The thermocouple is simple
and cheap, which is most widely used in microreactors [9,10].
But it has to be embed into the channel and may affect the flow
behavior. Also, one thermocouple can only get one result at distinct
points. As a result, non-intrusive methods have been proposed to
acquire the reliable and accurate characterization of microdevices.
The temperature sensitive tracers such as fluorescent dye [11,12]
and liquid crystals [13] have been applied to measure temperature
profiles in microchannels. However, the method is usually used in
non-reacting systems as the tracers are also sensitive to the react-
ing conditions. In addition, the small measuring range also limits
their application.
As a result, some technologically advanced methods based on
Raman spectroscopy [14,15] and infrared thermometry [16–22]
are attracting more and more attention. The methods can bring
the advantages of recording the complete and continuous

http://dx.doi.org/10.1016/j.cej.2016.01.100
1385-8947/Ó 2016 Published by Elsevier B.V.
 J.S. Zhang et al. / Chemical Engineering Journal 295 (2016) 384–390
temperature field of a specified area. Ewinger et al. [15] used laser
Raman spectroscopy to measure the temperature of water in
microchannels with an accuracy of ±1.2 °C. The procedure had a
lateral local resolution of approximately 15 lm and a depth resolu-
tion of 25 lm. Haber et al. [16] developed a microsystem with
infrared thermometry and the temperature profiles of the fast
exothermal hydrolysis of tetraethoxysilane were measured inside
a microcapillary at different flow rates. Yi et al. [17] presented
the thermal analysis of liquid containing Al2O3 nanoparticles in a
microfluidic platform using an infrared camera, which allows the
non-contact, three dimensional and high resolution capability for
temperature profiling. However, there are two conditions required
for the infrared thermometry, which limit its wide applications.
One is the channel must be covered with infrared transparent
materials, such as sapphire, silicon, germanium or zinc selenide
[18]. The other is that the system needs a careful calibration due
to the radiation losses by absorption or reflection of the IR signal.
In this work, a microsystem, including a micromixer, a coiled
capillary, a vacuum box and the infrared camera was developed
to measure the reaction enthalpy of fast exothermic reactions.
The objective of this work is to provide a quick, safe and non-
intrusive method to obtain the temperature profiles of the fast
exothermic reactions with infrared thermography. The microsys-
tem was first calibrated with a thermocouple. Two reacting sys-
tems which are homogeneous and heterogeneous respectively
were tested at different conditions with the method. The tempera-
ture profiles of the capillary was recorded and then the reaction
enthalpy could be calculated from the temperature profiles along
the length of the capillary.
2. Materials and methods
2.1. Chemicals
Cyclohexanone oxime (C6H11NO, 98%) was purchased from J&K
Scientific Ltd. (Beijing); Sulfuric acid (H2SO4, 98%) and sodium
hydroxide (NaOH, analytical grade) were acquired from Sinopharm
Chemical Reagent Beijing Co. Ltd.
2.2. Experimental setup
The experimental setup used for measuring the temperature
profiles is shown in Fig. 1. It was composed of two pumps, a micro-
mixer, a coiled capillary, a vacuum box and an infrared camera. The
Fig. 1. The experimental setup used for measuring the temperature profiles. (a) Schematic overview, (b) the photograph.
385
microreactor and the coiled capillary were both placed in the vac-
uum box which was vacuumized to a value below 10 kPa during
the experiments. According to the reports in the literature [16],
the value below 10 kPa is enough to suppress the convective heat
transfer. The used infrared camera (PI400, Optris, Germany) had
a resolution of 382  288 pixels with the frequency of 80 Hz. The
measurement temperature range was 20 to 100 °C with a sensi-
tivity of 0.08 °C. The system accuracy was ±2%. An infrared window
(BGIR-75, Beijing Scitlion Technology Co.) with the diameter of
75 mm was fixed on the top of the vacuum box to make the capil-
lary visible to the camera. The window is made of BaF2, which pro-
vides a relatively high and constant transmissivity to the infrared
ray [23]. As shown in Fig. 2, the high transmittance above 90%
can be obtained for the material BaF2 in the spectral range of
0.2–10 lm. When the wavelength is larger than 10 lm, transmit-
tance is greatly decreased to a low value of 10% at 14 lm. Accord-
ing to Wien displacement law, it is more suitable to measure the
temperature above 290 K (<10 lm). During our experiments, the
temperature is usually above 293 K, which means high transmit-
tance above 90% can be obtained using the material BaF2. As a
result, this material is quite suitable as the infrared window in
our experiments. The losses of about 10% of the radiation due to
the infrared window can lead to a slight deviation for the temper-
ature measurement, which can be corrected by calibration as
described in the next section.
The reactants were delivered separately by metering pumps
(Beijing Satellite Co. Ltd.) through the steel capillaries immersed
in a water bath to control the feeding temperature. Two commer-
cial temperature sensors were integrated in the capillaries to cali-
brate the temperature measurement system. The feed streams
were mixed in a T-shaped micromixer (VICI), which is made of
stainless steel with an inner diameter of 0.25 mm. A stainless steel
coiled capillary (1.0 m) with an inner diameter of 1 mm and an
external diameter of 1.6 mm was connected directly downstream
to the micromixer to perform the reaction. The inner diameter is
so small that the temperature profile at radial direction can be
ignored. During the experiments, the temperature profiles of the
coiled capillary at the flow direction were recorded by the infrared
camera via the infrared window. The camera was connected to the
computer and controlled by the software (PI Connect). To get the
variation of temperature versus time, we marked 5 measurement
points every 0.2 m on the capillary. The temperature of these
points would be calibrated as described in the next section. A typ-
ical temperature profile of the coiled capillary is shown in Fig. 3.
386 J.S. Zhang et al. / Chemical Engineering Journal 295 (2016) 384–390

The radiation loss can be estimated according to the Stephan–

Boltzmann law:

U rad ¼ e  5:67  108  ð3034  2984 Þ  aout ¼ 0:077 W ð3Þ

where e is the emissivity of stainless steel, which is assumed to be

0.5.
From the above results, it can be concluded that the main heat
losses come from the convection and radiation. The heat loss due
to conduction can be ignored. For the ideal plug flow reactor
[24], the temperature gradient dT/dz can be calculated with the
equation as follows:

Fig. 2. Spectral transmittance of BaF2. Adapted from the Beijing Scitlion Technology
Co.
Fig. 3. A typical temperature profile of the coiled capillary.
2.3. Calibration of the setup
Though the microreactor and coiled capillary are both placed in
the vacuum box, it is not avoided for the heat loss. Here we try to
estimate the heat losses in the vacuum box from the three mech-
anisms of heat transfer: conduction, convection and radiation. It
is assumed that the temperature rise between the capillary and
environment (DT) is 5 K.
For the conduction heat transfer, the conduction loss (Ucond) can
be estimated using the Fourier law:
dT
u  q  Cp  ¼ hloss  a  DT ð4Þ
dz
where u is the flow rate in the tube, about 0.3 m/s; q is the fluid
density, 997 kg/m3; Cp is the specific heat capacity, 4.18 kJ/(kg K);
hloss is the apparent convective heat transfer coefficient including
the three heat transfer mechanisms; a is specific exchange area
for the capillary, 4/D; D is the capillary diameter.
With the above equation, the dT/dz is determined to be 0.04 K.
Since the capillary length is only 1 m, the temperature change
due to the heat losses is only 0.04 K. As a result, the system with
the vacuum box can be considered as adiabatic when the DT is
5 K. When DT is increased to 10 K, the temperature change due
to the heat losses is still as low as 0.08 K.
The infrared signal received by the camera is influenced by
many factors [25], such as the transmissivity of infrared window
and the emissivity of measured surface, which would lead to a
deviation for the temperature measurement. Especially for the
emissivity, it depends on many factors: material, measuring angle,
geometry of the surface and so on. Therefore, the calibration is
required before accurate temperature measurement is obtained.
During the calibration, the fluids with different temperatures were
pumped into the system, the measured temperatures of every
points of the capillary were all the same with that shown in the
thermocouple. As a result, the measured temperature with infrared
camera could be calibrated with the thermocouple. Theoretically
the resolution of 382  288 pixels can all be calibrated with the
thermocouple. But to simplify the calculation, only 5 measurement
points as shown in Fig. 3 are calibrated.
A typical calibration of the five points is shown in Fig. 4. It is
found that there is a linear relationship between the measured
temperatures and the real temperature for the five points. The
correlation coefficients R2 are all above 0.991, indicating that the
lines fit the results very well.
3. Results and discussion

2
DT pð1:6  103 Þ 5 3.1. Typical temperature profiles
U cond ¼ 2  S  k  ¼2  16 
Dl 4 1:0
¼ 0:0003 W ð1Þ The classic neutralization reaction between dilute acid and base
solution is first tested as the model reaction, which is a typical fast
and exothermal reaction. In the experiments, H2SO4 solution was
where S is the sectional area of the capillary, k is the heat conduc-
used as the acid and NaOH solution was used as the lye, whose
tivity coefficient of stainless steel, about 16 W/(m K), Dl is the
flow rates were 10.6 and 8 mL min1. The concentrations of H+
capillary length.
was kept at 0.52 mol L1 while the concentration of OH was
The convective heat transfer coefficient (hconv) is estimated at
1 mol L1. Fig. 5 shows the typical temperature profiles at different
2 W/(m2 K) according to Krischers correlation [16] for convective
times during measuring process. When the reaction is not started,
heat transfer along horizontal cylinders. So the convection loss
the shape of the coiled capillary can be observed in the window.
(Uconv) can be estimated as follows:
The reason is that the emissivity of steel and air are different
though their real temperatures are the same. When the two fluids
U conv ¼ hconv  aout  DT ¼ 2  p  1:6  103  1  5
are pumped into the system and the reaction is started. With the
¼ 0:05 W ð2Þ time going on, the tube becomes much brighter from the outside
to inside along the capillary. This indicates that the capillary is
where aout is the surface area of the capillary. gradually heat up with the fluids in it. When the time is about
 J.S. Zhang et al. / Chemical Engineering Journal 295 (2016) 384–390 387

Fig. 4. A typical calibration of the five measurement points.

Fig. 5. The typical temperature profiles at different times during measuring process.

9 s, the process is in a steady state. From the temperature profiles, NaOH solution was used as the lye, whose flow rates were 10.6
we can get the temperature rise at different measurement points and 8 mL min1. The concentrations of H+ were kept at 0.43, 0.52
and then the reaction enthalpy can be calculated. and 0.64 mol L1 while the concentration of OH was 1 mol L1.
Excessive OH was used in the measurement, so the molar reaction
3.2. Measuring the reaction enthalpy of homogeneous reaction enthalpy could be calculated based on the feed rate of H+ as
follows:
The classic neutralization reaction between dilute acid and base
C p mDT
solution is still used as the model for the homogeneous reaction. DH m ¼ ð5Þ
nHþ
During the experiments, H2SO4 solution was used as the acid and
388 J.S. Zhang et al. / Chemical Engineering Journal 295 (2016) 384–390

where DHm is the molar reaction enthalpy; Cp is the specific heat Table 1
capacity; m is the mass flow rate of the reacted solution; DT is The results for the neutralization reaction.a

the temperature rise of the reacted solution; nH+ is the molar feed C+H Measurement DT Reaction enthalpy Relative
rate of H+. (mol L1) points (°C) (kJ mol1) error (%)
During the experiments, the NaOH solution was first pumped 0.43 1 3.34 57.1 0.5
into the system and then the acid solution was pumped into the 0.43 2 3.46 59.2 3.1
system. Meanwhile, the plots of temperature versus time at the 0.43 3 3.27 55.9 2.6
0.43 4 3.2 54.7 4.7
five measurement points were recorded by the software as shown
0.43 5 3.44 58.8 2.5
in Fig. 6. The temperatures are all at about 24 °C with the time 0.43 Average of 1–5 3.34 57.1 0.5
below about 10 s. But there is an obvious vibration for all the plots. 0.52 Average of 1–5 4.14 58.1 1.2
To reduce the error, the average value for a certain time range is 0.64 Average of 1–5 5.11 58.8 2.5
adapted. When the acid solution is pumped into the system, the a
The relative error ¼ jmeasuredactual
valueactual valuej
 100%.
value
temperatures at the five points are increased greatly and the order
is the same with that in the flowing direction. After about 15 s, the
temperatures are all turned into a steady state. The temperature 3.3. Measuring the reaction enthalpy of heterogeneous reaction
rise at each point can be obtained. As the solutions used in
the experiments are all diluted, the specific heat capacity used is Besides the homogeneous reaction, there are tremendous
the same with that of water at 25 °C, 4.18 J/(g °C). The calculated heterogeneous reactions in the chemical process. It is also very
results are all listed in Table 1. important to test the performance of this method for heteroge-
As shown in Table 1, the temperature rises for the five points neous reaction. The rearrangement of cyclohexanone oxime
are nearly the same with the H+ concentration of 0.43 mol L1. This (COX) in oleum is an important step to produce caprolactam, the
indicates that the reaction has been finished at the first point. The monomer for nylon-6 [6]. In our previous work, COX is usually dis-
results demonstrate the excellent mixing performance of the solved in octane, an inert organic solvent [7,27]. During the reac-
micromixer. From the temperature rise, we obtain the reaction tion, the oxime will transfer from octane to the oleum and starts
enthalpy at each point. The molar neutralization enthalpy of dilute the reaction. Here oleum is replaced by sulfuric acid (98%). As a
acid and base solutions is 57.3 kJ mol1 in literature [26]. We can result, only the quick protonation of COX in sulfuric acid happens
find that the results are quite close to the literature value and the within the residence time of about 5 s. The rearrangement would
relative errors are all below 5%. When the H+ concentrations not happen within such a short time at the absence of SO3 [28].
are increased to 0.52 and 0.64 mol L1, the obtained results also As a result, the protonation of COX is a good heterogeneous model
agree with the literature value [26] very well. reaction to test the system. The protonation is shown as follows:

Fig. 6. The plots of temperature versus time at the five measurement points with the different H+ concentrations.
 J.S. Zhang et al. / Chemical Engineering Journal 295 (2016) 384–390 389

Fig. 7. The plots of temperature versus time at the five measurement points with the different flow rates.

Compared with the traditional calorimeter, the microsystem

Table 2
The results for the heterogeneous reaction.a integrated with infrared camera can provide a quick, safe and
non-intrusive method to obtain the temperature profiles of the fast
Fo Measurement DT Reaction enthalpy Relative
exothermic reactions. The measuring time is less than 60 s while
(mL min1) points (°C) (kJ mol1) error (%)
the traditional calorimeter usually need 1 h. The consumed reac-
10 1 6.39 74.6 0.6
tants are greatly decreased due to the small dimension of the
10 2 6.33 73.9 1.5
10 3 6.64 77.5 3.3 microsystem. In addition, the safety characteristics of microtech-
10 4 6.77 79.0 5.3 nology make the measuring process much safer and the method
10 5 6.46 75.3 0.4 is quite suitable for performing the fast exothermal reactions with
10 Average of 1–5 6.52 76.1 1.5 very reactive compounds like H2SO4 in this work. Furthermore, the
15 Average of 1–5 7.37 78.3 4.4
method is potential to measure the reaction kinetics of fast
a
Cpo = 2.84  T + 1372 J/(kg K); CpCOX = 3.46  (T  273) + 1943 J/(kg K) [30]; exothermic reactions. However, the method also suffers from the
CpH2SO4 = 2.5  (T  273) + 1816 J/(kg K) [31]. problems such as limited temperature measuring range and diffi-
cult calibration for accurate measurement.
OH
NOH N+ H
ð6Þ 4. Conclusion
H+

In this study, a microsystem integrated with infrared camera

In the experiments, the mass fraction of COX in octane was
2.9 wt%. The flow rates of sulfuric acid was 3 mL min1 while those
of the octane solution were 10 and 15 mL min1. Excessive H2SO4
was used in the measurement, so the molar reaction enthalpy
could be calculated based on the feed rate of COX as follows:
ðC po mo þ C pCOX mCOX þ C pH2 SO4 mH2 SO4 ÞDT
DH m ¼
nCOX
where Cpo, CpCOX and CpH2SO4 are the specific heat capacities of
octane, COX and H2SO4, respectively; mo, mCOX and mH2SO4 are the
mass flow rates of octane, COX and H2SO4, respectively; nCOX is
the molar feed rate of COX.
Fig. 7 shows the plots of temperature versus time at the five
measurement points with different flow rates. Same to Fig. 6, the
temperatures at the five points are also increased greatly when
the acid is pumped into the system. When the flow rate of octane
is increased from 10 to 15 mL min1, the required time into steady
state is decreased.
The calculated results are all listed in Table 2. The temperature
rises for the five points are nearly the same when the flow rate of
octane is 10 mL min1. This indicates that the reaction has been
also finished at the first point. The results demonstrate the fast
mass transfer rate for the microsystem. From the temperature rise,
the reaction enthalpies at each point are calculated. The reaction
enthalpy of COX in H2SO4 is 75 kJ mol1 in literature [29]. We
can find that the results are also quite close to the literature value
and the relative errors are all below 6%. When the flow rate of
octane is 15 mL min1, the obtained results also agree with the
literature value [29] very well.
has been developed to measure the reaction enthalpy of fast
exothermic reactions. The micromixer and coiled capillary have
been placed in a vacuum box to suppress the heat losses forming
an adiabatic system. To ensure the accuracy, the infrared camera
is successfully calibrated with the thermocouple. The reaction
enthalpies of two reactions which are homogeneous and heteroge-
ð7Þ neous respectively are determined at different conditions with
measuring time less than 60 s. The results show that the data
obtained are in good agreement with the values reported in litera-
ture with the relative experimental errors less than 6%. Compared
with the traditional calorimeter, the microsystem can provide a
quick, safe and non-intrusive method to obtain the temperature
profiles of fast exothermic reactions. In the future, the reaction
kinetics of some fast exothermic reactions can be determined with
this method.
Acknowledgements
We gratefully acknowledge the supports of the National Natural
Science Foundation of China (91334201, U1463208, 21506110)
and China Postdoctoral Science Foundation (2015M570111) on
this work.
References
[1] R.L. Hartman, K.F. Jensen, Microchemical systems for continuous-flow
synthesis, Lab Chip 9 (2009) 2495.
[2] P. Plouffe, A. Macchi, D.M. Roberge, From batch to continuous chemical
synthesis—a toolbox approach, Org. Process Res. Dev. 18 (2014) 1286–1294.
390 J.S. Zhang et al. / Chemical Engineering Journal 295 (2016) 384–390
[3] W.N. Lau, K.L. Yeung, R. Martin-Aranda, Knoevenagel condensation reaction
between benzaldehyde and ethyl acetoacetate in microreactor and membrane
microreactor, Microporous Mesoporous Mater. 115 (2008) 156–163.
[4] A.A. Kulkarni, Continuous flow nitration in miniaturized devices, Beilstein J.
Org. Chem. 10 (2014) 405–424.
[5] W. Navarrini, F. Venturini, V. Tortelli, S. Basak, K.P. Pimparkar, A. Adamo, K.F.
Jensen, Direct fluorination of carbon monoxide in microreactors, J. Fluorine
Chem. 142 (2012) 19–23.
[6] K.T. Zuidhof, M.H.J.M. de Croon, J.C. Schouten, Beckmann rearrangement of
cyclohexanone oxime to e-caprolactam in microreactors, AIChE J. 56 (2010)
1297–1304.
[7] J.S. Zhang, K. Wang, Y.C. Lu, G.S. Luo, Beckmann rearrangement in a
microstructured chemical system for the preparation of e-caprolactam,
AIChE J. 58 (2012) 925–931.
[8] Y.C. Lu, D.W. Xin, J.S. Zhang, G.S. Luo, Modeling ethyl diazoacetate synthesis in
an adiabatic microchemical system, Chem. Eng. J. 273 (2015) 406–412.
[9] C. Hany, H. Lebrun, C. Pradere, J. Toutain, J. Batsale, Thermal analysis of
chemical reaction with a continuous microfluidic calorimeter, Chem. Eng. J.
160 (2010) 814–822.
[10] K. Wang, Y.C. Lu, H.W. Shao, G.S. Luo, Measuring enthalpy of fast exothermal
reaction with micro-reactor-based capillary calorimeter, AIChE J. 56 (2009)
1045–1052.
[11] R. Fu, B. Xu, D. Li, Study of the temperature field in microchannels of a PDMS
chip with embedded local heater using temperature-dependent fluorescent
dye, Int. J. Therm. Sci. 45 (2006) 841–847.
[12] J.J. Shah, M. Gaitan, J. Geist, Generalized temperature measurement equations
for Rhodamine B dye solution and its application to microfluidics, Anal. Chem.
81 (2009) 8260–8263.
[13] R. Segura, C. Cierpka, M. Rossi, S. Joseph, H. Bunjes, C.J. Kähler, Non-
encapsulated thermo-liquid crystals for digital particle tracking
thermography/velocimetry in microfluidics, Microfluid. Nanofluid. 14 (2013)
445–456.
[14] R. Kuriyama, Y. Sato, Non-intrusive measurement of microscale temperature
distribution by spontaneous Raman imaging, Microfluid. Nanofluid. 14 (2013)
1031–1037.
[15] A. Ewinger, G. Rinke, A. Urban, S. Kerschbaum, In situ measurement of the
temperature of water in microchannels using laser Raman spectroscopy,
Chem. Eng. J. 223 (2013) 129–134.
[16] J. Haber, M.N. Kashid, N. Borhani, J. Thome, U. Krtschil, A. Renken, L. Kiwi-
Minsker, Infrared imaging of temperature profiles in microreactors for fast and
exothermic reactions, Chem. Eng. J. 214 (2013) 97–105.
[17] P. Yi, A.A. Kayani, A.F. Chrimes, K. Ghorbani, S. Nahavandi, K. Kalantar-zadeh, K.
Khoshmanesh, Thermal analysis of nanofluids in microfluidics using an
infrared camera, Lab Chip 12 (2012) 2520.
[18] G. Hetsroni, A. Mosyak, E. Pogrebnyak, R. Rozenblit, Infrared temperature
measurements in micro-channels and micro-fluid systems, Int. J. Therm. Sci.
50 (2011) 853–868.
[19] C. Ravey, C. Pradere, N. Regnier, J. Batsale, New temperature field processing
from IR camera for velocity, thermal diffusivity and calorimetric non-intrusive
measurements in microfluidics systems, Quant. Infrared Thermogr. J. 9 (2012)
79–98.
[20] C. Pradere, M. Joanicot, J. Batsale, J. Toutain, C. Gourdon, Processing of
temperature field in chemical microreactors with infrared thermography,
Quant. Infrared Thermogr. J. 3 (2006) 117–135.
[21] M. Romano, C. Pradere, F. Sarrazin, J. Toutain, J.C. Batsale, Enthalpy, kinetics
and mixing characterization in droplet-flow millifluidic device by infrared
thermography, Chem. Eng. J. 273 (2015) 325–332.
[22] A.V. Patil, E.A.J.F. Peters, J.A.M. Kuipers, Comparison of CFD–DEM heat transfer
simulations with infrared/visual measurements, Chem. Eng. J. 277 (2015) 388–
401.
[23] D.C. Harris, Materials for Infrared Windows and Domes: Properties and
Performance, The International Society for Optical Engineering, Bellingham,
1999.
[24] J. Haber, M.N. Kashid, A. Renken, L. Kiwi-Minsker, Heat management in single
and multi-injection microstructured reactors: scaling effects, stability analysis,
and role of mixing, Ind. Eng. Chem. Res. 51 (2012) 1474–1489.
[25] M. Ashauer, J. Ende, H. Glosch, H. Haffner, K. Hiltmann, Thermal
characterization of microsystems by means of high-resolution
thermography, Microelectron. J. 28 (1997) 327–335.
[26] Hydroxide, Physical and Chemical Data of Chlor-Alkali Industry, Chemical
Industry Press, Beijing, 1988.
[27] J. Zhang, K. Wang, X. Lin, Y. Lu, G. Luo, Intensification of fast exothermic reaction
by gas agitation in a microchemical system, AIChE J. 60 (2014) 2724–2730.
[28] J.S. Zhang, K. Wang, Y.C. Lu, G.S. Luo, Beckmann rearrangement of
cyclohexanone oxime in a microchemical system: the role of SO3 and
product inhibition, AIChE J. 58 (2012) 3156–3160.
[29] D.F. Chen, Study of Beckmann rearrangement of cyclohexanone oxime in the
production of caprolactam, China Synth. Fibres Ind. 15 (1992) 44–49.
[30] G. Liu, L. Ma, J. Liu, Handbook of Chemistry and Chemical Engineering
Properties, Chemical Industry Press, Beijing, 2002.
[31] S.W. Liu, Y. Qi, D. Liu, Y.P. Liu, Service Manual of Sulfuric Acid, Southeast
University Press, Nanjing, 2001, pp. 1–42.