Journal of Thermal Science and Technology

6 downloads 0 Views 2MB Size Report
transfer coefficient was about 3.9×102 W/(m2·K) when water flowed through the .... Cp2. = Specific heat of external fluid. [J/(kg·K)]. D1. = Inner diameter of tube .... overall heat transfer coefficient, U, was changed between 0 to 1.00×103 W/(m2·K). ... g. ] Distance from the inlet [mm]. Distance from inlet [mm]. 273. 293. 313. 333.
Journal of Thermal Science and Technology

Vol. 7, No. 1, 2012

Isothermal Reactor for Continuous Flow Microwave-Assisted Chemical Reaction* Mitsuhiro MATSUZAWA**,***, Shigenori TOGASHI** and Shinji HASEBE*** **Hitachi Research Laboratory, Hitachi, Ltd., 832-2, Horiguchi, Hitachinaka, Ibaraki, Japan E-mail: [email protected] ***Department of Chemical Engineering, Kyoto University, Kyotodaigaku-Katsura, Nishikyo-ku, Kyoto-shi, Kyoto, Japan

Abstract An isothermal reactor in which reaction solutions can be controlled at constant temperature under constant microwave irradiation was developed. This is useful for investigating microwave effects on chemical reactions that are not observed under conventional heating conditions. We devised a structure in which a heat-transfer medium with a low dielectric loss factor, which hardly absorbs any microwaves, flowed outside a spiral reaction tube and designed the basic structure of the reactor using electromagnetic simulation to optimize the energy absorption rate. The conditions for increasing the temperature controlling ability of the reactor were also investigated theoretically and experimentally by taking into consideration the influences of three elements: the velocity of the internal fluid, the material for the tube, and the velocity of the external fluid. The velocity of the external fluid had the greatest influence on temperature controlling ability and the material for the tube had the least influence under the experimental conditions. The overall heat transfer coefficient was about 3.9×102 W/(m2·K) when water flowed through the quartz reaction tube at 7.1 mm/s and the external fluid flowed outside the tube at 44 mm/s. We also tested and confirmed that the temperature of water used as internal fluid could be controlled to within ±1.5 K at 309.3 K when microwaves at 26 W were irradiated into the reactor, whereas the temperature of water was over 373 K and boiled without the heat-transfer medium flowing outside the reaction tube using a conventional method of microwave heating. In addition, we investigated microwave effects on Suzuki-Miyaura coupling reaction using the developed isothermal reactor and we confirmed that the temperatures were controlled well in the reactor. The yields obtained by microwave heating were almost the same as that obtained by oil-bath heating. Key words: Microwave Heating, Chemical Reactor, Continuous Electromagnetic Simulation, Temperature Control

Flow,

1. Introduction

*Received 13 June, 2011 (No. 11-0324) [DOI: 10.1299/jtst.7.58]

Copyright © 2012 by JSME

Chemical reaction processes using microwaves have attracted a great deal of attention. Numerous effects caused by microwaves in organic synthesis have been reported(1), (2), e.g., reduced chemical reaction times from hours to minutes, reduced side reactions, and improved selectivity. Chemical-synthetic procedures using microwaves are expected to conserve energy, decrease the burden imposed on the environment, and simplify reaction processes. However, the exact mechanism by which microwaves enhance chemical processes is still unknown. It has been particularly controversial whether there is a non-thermal effect

58

Journal of Thermal Science and Technology

Vol. 7, No. 1, 2012

that is not explained merely by a thermal effect. For example, Sun et al.(3) reported that the rate of hydrolysis of adenosine triphosphate (ATP) is 25 times faster under microwave irradiation than with conventional heating at comparable temperatures. The next year, however, Jahngen(4), who was one of the authors, conducted the same experiments and concluded that the rate of hydrolysis solely depends on the temperature and not on the method of heating. Mayo et al.(5) conducted ring-closing metathesis reactions using microwaves and reported that microwaves accelerated the reaction. However, after that, Garbacia et al.(6) conducted the same reaction both by microwave heating and conventional oil-bath heating and reported that they did not find any evidence for a significant non-thermal microwave effect. Herrero et al.(7) reevaluated four reactions in which previous studies had claimed the existence of non-thermal effects under microwave heating, and they found no evidence for the existence of non-thermal microwave effects. In the previous studies, the temperatures of reaction solutions were measured with infrared sensors, but Herrero et al. measured temperatures at three points in a reaction tube by using fiber-optic probes that could accurately measure the temperatures of the reaction solution without being influenced by microwave irradiation. They reported that the temperatures measured with the fiber-optic probes were higher than those measured with the infrared sensors and that temperature gradients were formed within the reaction solution. They pointed out that the non-thermal effects observed in the previous studies might be the result of inaccurate measurements of temperature using external infrared sensors. A major reason for the difficulties with probing microwave effects, especially non-thermal effects, is the difficulty of controlling reaction solutions at a constant temperature under constant microwave irradiation in conventional microwave-assisted chemical reactors. The reaction solution temperature rises while the reaction solution absorbs microwaves, making it difficult to compare the outcomes obtained by microwave heating to those obtained by conventional oil-bath heating at the same temperature. Therefore, the objective of this study was to develop an isothermal reactor in which the reaction solution could be controlled at a constant temperature under constant microwave irradiation to enable microwave-specific effects to be investigated that were not observed under conventional heating conditions, especially the existence of non-thermal microwave effects. We previously developed a continuous flow microwave-assisted chemical reaction system in which the reaction solution could be heated by microwaves while it flowed through a reaction tube(8). A new function by which the reaction solution could be controlled at a constant temperature was added to the system in this study. Figure 1 outlines the procedure we used. First, we designed the basic structure of the reactor using electromagnetic simulation to optimize the energy absorption rate. Next, we estimated the temperature controlling ability of the reactor by calculating heat transfer and simulating the electromagnetic conditions. Next, we theoretically investigated the conditions for increasing temperature control. Finally, we experimentally tested the temperature control of the reactor.

59

Journal of Thermal Science and Technology

Vol. 7, No. 1, 2012

1. Design of basic structure using electromagnetic simulation to optimize energy absorption rate

2. Estimation of temperature controlling ability by heat transfer calculation and electromagnetic simulation

3. Investigation of condition for increasing the temperature controlling ability 4. Validation of the temperature controlling ability Fig. 1

Development process of reactor

2. Nomenclature Cp1 Cp2 D1 D2 Dc h1 h2 K Kp Nu0 Nu1 Nu2 Pr1 Pr2 Qall Q(z) r1 r2 Re1 Re2 T(z) T0 u1 u2 U α1 α2 ε’ ε’’ λ1 λ2 λ3 ρ1 ρ2 ν1 ν2

= Specific heat of internal fluid [J/(kg·K)] = Specific heat of external fluid [J/(kg·K)] = Inner diameter of tube [mm] = Outer diameter of tube [mm] = Diameter of curvature of tube [mm] = Heat-transfer coefficient inside tube [W/ (m2·K)] = Heat-transfer coefficient outside tube [W/ (m2·K)] = Dean Number (= Re1 D1 / Dc ) [–] = K Pr1 [–] = Nusselt number of internal fluid in a straight tube [–] = Nusselt number of internal fluid in a curved tube (= h1D1/λ1) [–] = Surface averaged Nusselt number for a cylinder in cross flow [–] (= h2D2/λ3) = Prandtl number of internal fluid (=ν1/α1) [–] = Prandtl number of external fluid (=ν2/α2) [–] = Microwave energy generated from microwave generator [W] = Microwave energy absorbed per mass of internal fluid at z [W/kg] = Internal radius of reaction tube [mm] = External radius of reaction tube [mm] = Reynolds number of internal fluid (= u1D1/ν1) [–] = Reynolds number of external fluid (= u2D2/ν2) [–] = Temperature of internal fluid at z [K] = Temperature of external fluid [K] = Flow velocity of internal fluid [m/s] = Flow velocity of external fluid [m/s] = Overall heat transfer coefficient [W/(m2·K)] = Heat diffusivity of internal fluid [m2/s] = Heat diffusivity of external fluid [m2/s] = Relative permittivity [–] = Relative dielectric loss factor [–] = Thermal conductivity of internal fluid [W/ (m·K)] = Thermal conductivity of tube [W/ (m·K)] = Thermal conductivity of external fluid [W/ (m·K)] = Density of internal fluid [kg/m3] = Density of external fluid [kg/m3] = Kinetic viscosity of internal fluid [m2/s] = Kinetic viscosity of external fluid [m2/s]

60

Journal of Thermal Science and Technology

Vol. 7, No. 1, 2012

3. Design of Isothermal Reactor 3.1 Configuration Figure 2 has a photograph (top) and a schematic (bottom) of the basic structure of the apparatus. It consists of a microwave generator that generates microwaves at 2.45 GHz (Hitachi, HPP121A-INV-02), a power monitor that measures the power of incident and reflected waves, a stub tuner that matches impedance in the waveguide, and an applicator that the reaction tube penetrates vertically. Figure 3 is a schematic of the structure of the isothermal reactor we developed for this study, which is a double pipe heat exchanger. A heat-transfer medium with a low dielectric loss factor, which hardly absorbs any microwaves, flows outside a spiral reaction tube. The microwaves generated from the microwave generator in this structure were intended to be absorbed mainly by the reaction solution as the microwaves passed through the heat-transfer medium and the reaction tube, whose dielectric loss factors were low, and heat was exchanged between the reaction solution and the heat-transfer medium. We fabricated the reaction tube from quartz (ε’ (permittivity): 2.9, ε’’ (dielectric loss factor): 5.2×10-3), the outer tube from PTFE (ε’: 2.1, ε’’: 4.2×10-4) and used perfluorinated liquid Fluorinert (3M, brand) (ε’: 1.9, ε’’: 1.9×10-4) as the heat-transfer medium. 1026

675

Control unit

Pump unit

Monitor

393 Power Stub tuner monitor

Microwave generator (Hitachi, HPP 121A-INV-02)

Fig. 2

Reaction tube

Applicator (Fig. 3)

Continuous flow microwave-assisted chemical reaction system

64

Reaction tube (Quartz) (Size of tube: Inner diameter, 3.0 mm; Outer diameter, 5.0 mm; Diameter of coil, 20.0 mm; Pitch, 6.5 mm; Number of turns, 13.5 Length along the center line, 967 mm)

37 (L1)

Outlet

Short-circuit plane

Microwave

143 (L2)

55

Heat exchange Applicator

Outer tube (PTFE)

Heat-transfer medium (Low dielectric loss factor)

Fig. 3

Inlet Reaction solution

Structure of an isothermal reactor

61

Journal of Thermal Science and Technology

Vol. 7, No. 1, 2012

It is important in this reactor for microwaves to be absorbed efficiently by the reaction solution without being absorbed by the outer tube, the reaction tube, or the heat-transfer medium. Consequently, we used electromagnetic simulation to design the location and shape of the reaction tube and the shape of the applicator. Since the energy absorbed by the heated material is proportional to the square of the electric intensity of the area, it was necessary to locate the reaction tube in an area where electrical intensity was strong. We used MicroStripesTM (CST) as an electromagnetic simulator. In this device, incident waves from the microwave generator and reflected waves that are reflected at the short-circuit plane encounter interference and standing waves are formed in the waveguide. Accordingly, we designed the diameter of the coil and the position of the reaction tube, L1, and its length, L2, as shown in Fig. 3 for the reaction tube to be located in an area where the electrical field intensity was strong of the standing waves. Table 1 lists the dielectric properties used in the simulation. We calculated the conditions where the reaction tube was filled with water and heat-transfer medium was filled between the outer tube and reaction tube and discretized so that they were about 0.3 mm apart and the dielectric properties were constant. Figure 4 has the electric field intensity distribution calculated for the applicator we developed, which indicates that the reaction tube was located in the area where the electric field intensity was strong. About 95% of the microwaves from the microwave generator were absorbed by the water. We found in a previous investigation that the rate at which energy was absorbed by the heated material decreased as the inner diameter of the reaction tube decreased, e.g., the energy absorption rate was about 95% when a reaction tube with an inner diameter of 3.0 mm and an outer diameter of 5.0 mm was used, but it was about 30% when a reaction tube with an inner diameter of 1.6 mm and an outer diameter of 3.0 mm was used. Therefore, we made the inner diameter of the reaction tube 3.0 mm and its outer diameter 5.0 mm in this study. Table 1

Dielectric properties used in simulation

Component Reaction solution Outer tube Reaction tube Heat-transfer medium Applicator

Material Water PTFE Quartz

ε’ 77.0 2.10 2.90

ε’’ 11.0 4.20×10-4 5.20×10-3

Fluorinert

1.90

1.90×10-4

Aluminum





A Applicator

Reaction tube

Reaction tube

Stub Microwave

0

0.5

1.0

Normalized electric field intensity [–]

0

A

0.5

1.0

Normalized electric field intensity [–]

(a)

Front cross sectional view Fig. 4

(b)

A–A cross sectional view

Electric field intensity distribution in the applicator

62

Journal of Thermal Science and Technology

Vol. 7, No. 1, 2012

3.2 Estimates of temperature controlling ability Next, we estimated the temperature controlling ability of the reactor to optimize the conditions. Figure 5 has a schematic of the heat balance model of the reaction tube. As discussed in Subsection 3.1, this reactor had a double pipe structure in which the heat-transfer medium with a low dielectric loss factor flowed outside the reaction tube and heat was exchanged between the reaction solution and the heat-transfer medium. We used electromagnetic simulation and one-dimensional heat-transfer calculations to find the temperature controlling ability of the reactor. Internal fluid (Reaction solution) h1 (Heat-transfer coefficient inside tube)

h2 (Heat-transfer coefficient outside tube)

Q(z) u1 , T(z)

λ2 (Thermal conductivity of tube)

u1 , T(z+∆z) U·2πr1 ∆z(T(z)-T0)

z

Reaction tube r1 r2

Fig. 5

External fluid (Heat-transfer medium) u2 , T0

Heat balance model of reaction tube

The heat balance in the reaction tube is expressed by Eq. (1), where the position along the direction of flow of the internal fluid is z, the temperature of the internal fluid at z is T(z) [K], and the temperature of the external fluid is T0 [K]. Here, the internal radius of the reaction tube is r1 [m], the external radius of the reaction tube is r2 [m], and the overall heat transfer coefficient at the inner surface of the reaction tube is U [W/(m2·K)]. Further, the microwave energy absorbed per mass of the internal fluid at z is Q(z) [W/kg], the specific heat of the internal fluid is Cp1 [J/(kg·K)], and the density of the internal fluid is ρ1 [kg/m3].

dT ( z ) 1 2U Q(z ) − = {T (z ) − T0 } dz C p1u1 C p1 ρ1 r1u1

(1)

It was assumed that reaction solution was irradiated by microwaves and controlled at 308 K for about 2 min. The increasing-temperature curve T(z) was calculated from Eq. (1) when the microwave output power from the microwave generator, Qall, was 26 W and the overall heat transfer coefficient, U, was changed between 0 to 1.00×103 W/(m2·K). Table 2 lists the values used in the simulation. The Q(z), the microwave energy absorbed per mass of the internal fluid at z, was calculated by electromagnetic simulation. Figure 6 plots the Q(z) calculated with simulation. We found that the microwave energy absorbed by water was highest around the center in this reactor. Figure 7 plots the calculated results for T(z) when U was 10.0, 1.00×102, 3.00×102 W/(m2·K) as examples. It was assumed that T(z) with U=10.0 W/(m2·K) represented the results under natural convection conditions in air, and T(z) with U=1.00×102 and 3.00×102 W/(m2·K) represented the results obtained for the reactor in this study as examples. The vertical axis indicates the temperature, and the horizontal axis indicates the distance from the inlet of the reaction tube. Here, the temperature of the internal fluid was highest at z=524 mm and lowest at z=0 mm when U was 3.00×102 W/(m2·K). It was possible to control the temperature of the internal fluid from the inlet to the outlet within ±4.7 K at 312.8 K whereas the temperature of the internal fluid reached 353 K around z=385 mm without the heat-transfer medium flowing when U was 10.0 W/(m2·K). Half the difference

63

Journal of Thermal Science and Technology

Vol. 7, No. 1, 2012

between the maximum and minimum temperatures was defined as ∆T, as shown in Fig. 7. The ∆T indicates the temperature controlling ability of the reactor. Therefore, ∆T was calculated when the overall heat transfer coefficient U was changed between 0 to 1.00×103 W/(m2·K). Figure 8 plots the calculated results. The ∆T decreased, meaning that temperature controlling ability increased as the overall heat transfer coefficient, U, increased. Table 2

Values used in calculations

4.18×103 26.0 1.50×10-3 2.50×10-3 308 (Natural convection) 10.0, U [W/(m2·K)] 1.00×102, 3.00×102 u1 [m/s] 7.07×10-3 (= 3 cm3/min) 3 3 ρ1 [kg/m ] 1.00×10 Q(z) was calculated by electromagnetic simulation

Cp1 Qall r1 r2 T0

[J/(kg·K)] [W] [m] [m] [K]

8.0

353

U=10.0 W/(m2·K) U=1.00×102 W/(m2·K)

Temperature [K]

6.0 5.0 4.0 3.0 2.0 1.0

333 (z=524 mm)

U=3.00×102 W/(m2·K)

313

312.8 ±4.7 K

∆T (z=0 mm)

293

0.0 0

200

400

600

800

1000

273

Distance from thefrom inletinlet [mm] Distance [mm]

0

200

400

600

800

1000

Distance fromfrom the inlet [mm] Distance

Fig. 6

Q(z) calculated by simulation

Difference in temperature ∆T [K]

Q(z) [×103 W/kg]

7.0

Fig. 7

Inceasing-temperrature curve when U was 10.0, 1.00×102 and 3.00×102 W/(m2·K)

30 25 20 15 10 5 0 0

200

400

600

800

1000

Overall heat transfer coefficient U [W/(m2·K)] Fig. 8

Calculation of difference in temperature ∆T when U was changed

64

Journal of Thermal Science and Technology

Vol. 7, No. 1, 2012

3.3 Investigation into conditions for increasing temperature control Next, we theoretically investigated the conditions for increasing the overall heat transfer coefficient, U. The overall heat transfer coefficient, U, at the inner surface of the reaction tube is generally expressed by Eq. (2), where h1 is the heat-transfer coefficient inside the tube, h2 is the heat-transfer coefficient outside the tube, and λ2 is the thermal conductivity of the tube. Here, r1 is the internal radius of the tube and r2 is the external radius of the tube, as shown in Fig. 5.

r 1 1 r1 1 r1 = + Ln 2 + U h1 λ2 r1 h2 r2

(2)

As seen in Eq. (2), the overall heat transfer coefficient, U, consists of three elements: the first term of the right side in Eq. (2) is the thermal resistance inside the tube, the second term is the thermal resistance of the tube, and the third term is the thermal resistance outside the tube. It is necessary to increase the overall heat transfer coefficient, U, by decreasing the value of right side in Eq. (2) to increase temperature control. We estimated the overall heat transfer coefficient when the inner and the external diameters were 3.00 and 5.00 mm, respectively. The Nusselt number, Nu1, of fully developed laminar flow in a curved tube is suggested by Eq. (3), which was derived by Ishigaki(9), where Nu0 is the Nusselt number with a constant wall temperature gradient in a straight tube, Nu0= 3.657, K is the Dean number, and Dc is the diameter of the curvature. Nu1 =

(

h1 D1 -0.726 = 0.180 Nu 0 K p 1 + 2.72 K p λ1

)

(3)

   K p = K Pr1 = Re1 Pr1 D1 ( 0.7 < Pr1 < 10 )   Dc   (Nu1/Nu0 is 1 when Nu1/Nu0 is smaller than 1) As seen in Eq. (3), h1 is dependent on the Reynolds number of the internal fluid, Re1, viz., on the velocity of the internal fluid, u1, when the inner and outer diameters are fixed. On the other hand, the surface-averaged Nusselt number for a cylinder in a cross flow, Nu2, is suggested by Eq. (4), which was derived by Churchill and Bernstein(10). 1

1 3

4

5 5   0.62 Re 2 Pr2   Re 2 8  1+  Nu 2 = 0.3 +  1  2.82 × 10 5   2 4      1 +  0.4  3    Pr     2     h D  Nu 2 = 2 2 (Re2 ⋅ Pr2 > 0.2) λ2   2

(4)

As seen in Eq. (4), h2 is dependent on the Reynolds number of the external fluid, Re2, viz., on the velocity of the external fluid which flows perpendicularly to the reaction tube, u2 , when the inner and outer diameters are fixed. Consequently, as shown in Eqs. (2)–(4), the overall heat transfer coefficient, U, depends on three parameters: the flow velocity of the internal fluid, u1, the thermal conductivity of the tube, λ2, and the flow velocity of the external fluid, u2, when the inner and outer diameters of the tube are fixed.

65

Journal of Thermal Science and Technology

Vol. 7, No. 1, 2012

Overall heat transfer coefficient U [W/(m2·K)]

Figures 9, 10, and 11 plot the results for the overall heat transfer coefficient calculated from Eqs. (2), (3), and (4) when the velocity of the internal fluid, the thermal conductivity of the tube, and the velocity of the external fluid were changed. Table 3 summarizes the values used in the calculations. The overall heat transfer coefficient in Fig. 9 was constant in a range from 0 to about 5 mm/s when Nu1/Nu0 was 1. This means that the influence of secondary flow is small because the velocity of the internal flow is small and the overall heat transfer coefficient is equal to that of a straight tube. Figures 9, 10, and 11 indicate that the velocity of the external fluid had the greatest influence on the overall heat transfer coefficient, and the material for the tube had the least influence on the overall heat transfer coefficient. This suggests that increasing the velocity of the external fluid is the most effective way of increasing the temperature controlling ability of the reactor.

600

λ2=1.30 W/(m·K) u2=17.0 mm/s

500 400 300

Nu1 /Nu 0 is 1

200 100 0 0

5

10

15

20

25

30

Flow velocity of of theinternal internalfluid fluidu1u1[mm/s] [mm/s] Flow velocity

Influence of flow velocity of internal fluid on overall heat transfer coefficient

Overall heat transfer coefficient U [W/(m2·K)]

Fig. 9

600 u1=7.07 mm/s u2=17.0 mm/s

500 400 300 200 100 0 0

50

100

150

200

Thermalconductivity conductivityofofthe reaction tube λ2λ2 Thermal reaction tube [W/(m·K)] [W/(m·K)]

Influence of thermal conductivity of reaction tube on overall heat transfer coefficient

Overall heat transfer coefficient U [W/(m2·K)]

Fig. 10

600 u1=7.07 mm/s λ2=1.30 W/(m·K)

500 400 300 200 100 0 0

10

20

30

40

50

Flow velocity of of theexternal externalfluid fluidu2u[mm/s] Flow velocity 2 [mm/s]

Fig. 11

Influence of flow velocity of external fluid on overall heat transfer coefficient

66

Journal of Thermal Science and Technology

Vol. 7, No. 1, 2012

Table 3

D1 D2 Dc u1 u2 α1 α2 λ1 λ2 λ3 ρ1 ρ2 ν1 ν2 Cp1 Cp2

[m] [m] [m] [m/s] [m/s] [m2/s] [m2/s] [W/(m·K)] [W/(m·K)] [W/(m·K)] [kg/m3] [kg/m3] [m2/s] [m2/s] [J/(kg·K)] [J/(kg·K)]

Values used in calculations

3.00×10-3 5.00×10-3 20.0×10-3 (7.07×10-3) *1 (17.0×10-3) *2 1.44×10-7 3.39×10-8 0.600 (1.30) *3 0.0670 1.00×103 1.88×103 8.93×10-7 2.80×10-6 4.18×103 1.05×103

(0–30 in Fig. 9) (0–50 in Fig. 11)

(0–200 in Fig. 10)

*1 value used in Figs. 10 and 11, *2 value used in Figs. 9 and 10, *3 value used in Figs. 9 and 11

4. Experiment 4.1 Overall heat transfer coefficient of reaction tube We experimentally investigated what influence the velocity of the internal fluid, the thermal conductivity of the tube, and the velocity of the external fluid had on the overall heat transfer coefficient. Figure 12 has a schematic of the experimental apparatus and Table 4 lists the experimental conditions. Microwaves were not generated in this experiment because its purpose was to investigate the heat-transfer property of the reaction tube itself. We used water as the internal fluid. We measured the temperature of the water at points (a)–(f) in Fig. 12, using fiber-optic probes when the water at 308.2 K at inlet (a) flowed through the tube at 1.0–30.0 mm/s and the external fluid at 333.2 K flowed outside the reaction tube at 20.0–50.0 mm/s.

(f)

Reaction tube (Quartz)

(e) (d) u2

55

u1 (c) (b)

Outer tube (PTFE)

(a)

Heat-transfer medium 333.2 K u2: 20.0–50.0×10-3 m/s Fig. 12

Reaction solution water, 308.2 K u1: 1.0–30.0×10-3 m/s

Experimental apparatus

67

Journal of Thermal Science and Technology

Vol. 7, No. 1, 2012

Table 4

Experimental conditions for measurements of overall heat transfer coefficient

Temperature of water at inlet (a) Temperature of external fluid at inlet u1 u2

[K]

308.2

[K]

333.2

[m/s] [m/s]

(7.10×10-3) *4 (17.0×10-3) *5

(1.0–30.0 in Fig. 14) (20.0–50.0 in Fig. 16)

*4 value used in Figs. 15 and 16, *5 value used in Figs. 14 and 15

Then, we calculated the overall heat transfer coefficient, U, from the measured temperatures using Eq. (5), which was equal to Eq. (1) when Q(z) was 0. For example, Fig. 13 plots the temperatures measured when the water at 308.2 K at inlet (a) flowed through the tube at 23.6 mm/s and the external fluid at 333.2 K flowed outside the reaction tube at 17.0 mm/s. Since Cp1, ρ1, r1, u1, and T0 were known, the overall heat transfer coefficient, U, was determined with the least-squares method by fitting U in Eq. (5) to the temperature data. In this case, U was 3.3×102 W/(m2·K). All the overall heat transfer coefficients were calculated with this method. dT ( z ) 2U =− {T (z ) − T0 } dz C p1 ρ1r1u1

(5)

Temperature [K]

343 333 323 313 303 293

Measurements Eq. (5) when U was 3.3×102

283 273 0 Fig. 13

200 400 600 800 Distance from Distance from theinlet inlet[mm] [mm]

1000

Temperature data (u1=23.6 mm/s, u2=17.0 mm/s).

Figure 14 plots the results obtained from measurements and calculations of the overall heat transfer coefficient from the measured temperatures when the velocity of the internal fluid was changed. In this reactor, we assumed reaction in which residence time was longer than about 40s, which was equal to 23.6 mm/s of u1. Accordingly, we conducted the experiments with changing the velocity of the internal fluid u1 between 0 and 23.6 mm/s. The triangles plot the measurements and the solid line indicates the overall heat transfer coefficient calculated from Eqs. (2), (3), and (4). Figure 14 indicates that the overall heat transfer coefficient increased as the flow velocity of the internal fluid increased, and the calculated results were within 30% of the measured results. However, the flow velocity of the internal fluid should actually be determined by having an adequate residence time for a chemical reaction, e.g., a long tube is necessary to increase the internal flow velocity because this internal flow velocity in a long tube is faster than that in a short tube when the residence times are the same. Other devices, such as a prolonged reaction tube, are necessary for increasing the overall heat transfer coefficient by increasing u1 separately.

68

Journal of Thermal Science and Technology

Vol. 7, No. 1, 2012

Overall heat transfer coefficient U [W/(m2·K)]

Figure 15 plots the measured and calculated results for the overall heat transfer coefficient when the thermal conductivity of the reaction tube was changed. We used quartz, alumina, and aluminum nitride as the materials for the tube. The thermal conductivity, in order, was 1.3, 25, and 1.7×102 W/(m·K). Since it was difficult to fabricate a spiral reaction tube from alumina and aluminum nitride, we did an experiment using straight tubes with 3.0-mm inner diameters and 5.0-mm outer diameters. The triangles plot the measurements for the straight tube and the solid line indicates the overall heat transfer coefficient for a spiral tube calculated from Eqs. (2)–(4) for reference. The overall heat transfer coefficients were, respectively, 2.8×102, 3.1×102, and 3.1×102 W/(m2·K) when we used quartz, alumina, and aluminum nitride as the tube materials, and U was improved by about 11% from quartz to alumina. However, the U of the aluminum nitride was only slightly different from that of alumina. These results indicate that the influence of the tube material on the overall heat transfer coefficient was small under the experimental conditions. Figure 16 plots the measured and calculated results for the overall heat transfer coefficient when the flow velocity of the external fluid was changed. The triangles plot the measurements and the solid line indicates the overall heat transfer coefficient calculated from Eqs. (2)–(4). The overall heat transfer coefficient increased as the flow velocity of the external fluid increased, and U was 3.9×102 W/(m2·K) when the velocity of the external fluid was 44 mm/s. In addition, the calculated results were within 31% of the measured results. Equations (3) and (4) express the average Nusselt number at one part of a curved tube. However, this reaction tube actually formed a complex straight section and curved section. In addition, the heat-transfer medium may not have flowed uniformly. These may have caused the difference between the measured and calculated results for the overall heat transfer coefficients. We found from these experiments that the influence of the external fluid was greatest and that of the tube material was least on the overall heat transfer coefficient in this reactor, as established from the calculations presented in Subsection 3.3. 600

λ2=1.30 W/(m·K) u2=17.0 mm/s

500 400 300 200

Eqs. (2)–(4)

100 0 0

5

10

15

20

25

30

Flow velocity of of theinternal internalfluid fluidu1u1[mm/s] [mm/s] Flow velocity

Influence of flow velocity of internal fluid on overall heat transfer coefficient Overall heat transfer coefficient U [W/(m2 ·K)]

Fig. 14

600

u1 =7.07 mm/s u2 =17.0 mm/s

500

Quartz

400

Alumina

Aluminum nitride

300 200

Eqs. (2)–(4)

100 0 0

50

100

150

200

Thermalconductivity conductivityofofthe reaction tube λ2λ2 Thermal reaction tube [W/(m·K)] [W/(m·K)]

Fig. 15

Influence of thermal conductivity of reaction tube on overall heat transfer coefficient

69

Journal of Thermal Science and Technology Overall heat transfer coefficient U [W/(m2·K)]

Vol. 7, No. 1, 2012

600

u1=7.07 mm/s λ2=1.30 W/(m·K)

500 400 300 200

Eqs. (2)–(4)

100 0 0

10

20

30

40

50

Flow velocity of of theexternal externalfluid fluidu2u[mm/s] Flow velocity 2 [mm/s]

Fig. 16

Influence of flow velocity of external fluid on overall heat transfer coefficient

4.2 Validation of temperature control through experiments The experiment described in Subsection 4.1 revealed that the overall heat transfer coefficient was about 3.9×102 W/(m2·K) when water flowed through the quartz spiral reaction tube at 7.1 mm/s and the external fluid flowed outside the tube at 44 mm/s. Under these conditions, we investigated the temperature controlling ability of the reactor when it was irradiated with microwaves. We measured the temperature of the water at points (a)–(f) in Fig. 12 with fiber-optic probes when the water at 308.2 K at inlet (a) flowed through the quartz tube at 7.1 mm/s, and the external fluid at 308.2 K flowed outside the reaction tube at 44 mm/s. Microwaves were irradiated at 26 W and 57 W, respectively. For comparison, we also conducted experiments without external fluid flowing outside the tube, in other words, with natural convection in air. Figure 17 plots the results obtained from the measurements. When the heat-transfer medium flowed outside the reaction tube, the temperature of the water was controlled to within ±1.5 K at 309.3 K when the microwaves at 26 W were irradiated and within ±4.6 K at 313.2 K when the microwaves at 57 W were irradiated. However, the temperature of the water was over 373 K between (c) and (d) and boiled without the heat-transfer medium when the microwaves at 26 W were irradiated. This indicated that the temperature controlling ability of the reactor under constant microwave irradiation was improved significantly compared to temperature control using the conventional method of microwave heating. This ability is sufficient for investigating effects that are specific to microwaves.

373

(a) (b)

(c)

(e) (f)

boiling

353 Temperature [K]

(d)

26 W (without heat-transfer medium)

333

57 W 313.2±4.6 K

313

309.3±1.5 K

26 W 293

273 0

200

400

600

800

1000

Distance from theinlet inlet[mm] [mm] Distance from

Fig. 17

Increasing-temperature curve measured by experiments

70

Journal of Thermal Science and Technology

Vol. 7, No. 1, 2012

4.3 Investigation of microwave effects on Suzuki-Miyaura coupling reaction Decreases in reaction times and the obtaining of good yield have been reported by using microwaves in Suzuki-Miyaura coupling reaction(11)–(13). We investigated whether specific microwave effects existed in the homogeneous reaction that were not observed in the conventional heating method using the developed isothermal reactor. Equation (6) shows the reaction formula. Solvent: DMF (N,N-dimethylformamide) Catalyst: Pd(PPh3)4

B(OH)2 Phenylboronic acid

+

(6)

Br 4-Bromotoluene

4-Phenyltoluene

Figure 18 shows the schematic of experiment apparatus. Two reaction solutions were preheated at 353 K in an oil-bath and mixed using a T-shaped mixer 1.25 mm in diameter. The mixed reaction solution flowed into the developed isothermal reactor, and microwave power and the temperature of the heat-transfer medium were controlled to keep the reaction solution at about 353 K. As shown in Fig. 19, we conducted experiments in which microwave power was changed to 14 W and 61 W to investigate the effect of microwave power, and we controlled the temperature of the heat-transfer medium at 346 K and 313 K, respectively to keep the temperature of the reaction solution at about 353 K. On the other hand, in the oil-bath heating experiment, we controlled the temperature of the heat-transfer medium to keep the reaction solution at about 353 K without irradiating microwaves. We measured the temperature of the reaction solution at points (i)–(viii) in Fig. 19 using fiber-optic probes. The residence time between (i) and (viii) was about 240 s. In comparison, we also investigated the temperature rise curve without the heat-transfer medium flowing outside the reaction tube when the reaction solution was heated by microwaves at 14 W from room temperature.

Isothermal reactor Sample Microwave Phenylboronic acid: 0.26 mol/L KOH: 0.60 mol/L 0.50 mL/min in DMA and water Syringe Pump A Syringe Pump B 4-Bromotoluene : 0.50 mL/min 0.20 mol/L Pd(PPh3)4: 5.0mol% in DMA

Fig. 18

Thermostat bath

Stainless tube I. D. 1.0 mm Gear pump Reaction solution Oil bath T-shape mixer Heat-transfer (353 K) (I. D. 1.25 mm) medium

Apparatus used in Suzuki-Miyaura coupling reaction experiment

71

Journal of Thermal Science and Technology

Vol. 7, No. 1, 2012

(viii)

(viii) (vii)

(vii)

Microwave (14, 61 W)

Heat-transfer medium (346, 313 K)

Fig. 19

(vi) (v)

(iv) (iii)

(iv) (iii)

Heat-transfer medium (359 K)

(ii) (i)

Oil bath (353 K)

(a)

(vi) (v)

(ii) (i)

Oil bath (353 K)

Microwave heating

(b)

Oil-bath heating

Experimental apparatus and conditions in microwave heating and oil-bath heating

Figure 20 shows the temperature profile. The temperatures obtained by microwave heating and oil-bath heating were controlled well within ±2.4 K between (i) and (v), but the difference in the temperature at (vi) was ±9.6 K. When the microwave energy absorbed by the reaction solution and the transfer energy from the reaction solution to the heat transfer medium were the same, the temperature of the reaction solution had to be kept constant. However, the transfer energy from the reaction solution to the heat transfer medium was more than the microwave energy absorbed by the reaction solution at (vi), so the temperature at (vi) differed greatly. On the other hand, the temperature increased from room temperature (300 K) to 368 without the heat-transfer medium when microwaves at 14 W were irradiated.

373

(i)

(ii)

(iii)

(iv)

(v)

(vi) (vii) (viii)

Temperature [K]

A (Oil-bath)

353 B (MW, 14 W)

333 C (MW, 61 W)

313

MW, 14 W Without heat-transfer medium

293 273 0

100

200

300

400

500

600

Distance from inlet [mm] Fig. 20

Temperature profile

Figure 21 shows the result of yields. The yields obtained by microwaves were almost the same as that obtained by oil-bath heating. In addition, no difference in the yields was found when microwave power was changed. In this experiment, we did not find any evidence of non-thermal effects of microwaves.

72

Journal of Thermal Science and Technology

Vol. 7, No. 1, 2012

40 35

Yield [%]

30 25 20 15 10 5 0 A Oil bath

Fig. 21

B Microwave 14 W

C Microwave 61 W

Results of yields

5. Conclusions We developed an isothermal reactor that could control reaction solutions at constant temperature under constant microwave irradiation. This was useful for investigating microwave effects, especially the existence of non-thermal microwave effects. We devised a structure in which a heat-transfer medium with a low dielectric loss factor flowed outside the spiral reaction tube, and we designed the basic structure of the reactor using electromagnetic simulation to optimize the energy absorption rate. We also investigated the conditions necessary to increase the temperature controlling ability of the reactor theoretically and experimentally by taking into account the influences of three elements: the velocity of the internal fluid, the material for the tube, and the velocity of the external fluid. The velocity of the internal and the external fluid had the large influence on controlling the temperature and the material for the tube had the least influence under the experimental conditions. However, since the flow velocity of the internal fluid should actually be determined by having an adequate residence time for a chemical reaction, other devices, such as a prolonged reaction tube, are necessary for increasing the overall heat transfer coefficient by increasing the velocity of the internal fluid separately. Accordingly, we tried to increase the overall heat transfer coefficient by increasing the velocity of the external fluid in this study. The overall heat transfer coefficient was about 3.9×102 W/(m2·K) when water flowed through the quartz reaction tube at 7.1 mm/s and the external fluid flowed outside the tube at 44 mm/s. We also controlled the temperature of the water acting as the internal fluid within ±1.5 K at 309.3 K when the microwaves at 26 W were irradiated into the reactor, whereas the temperature of the water was over 373 K and boiled without the heat-transfer medium flowing in the conventional method of microwave heating. In addition, we investigated microwave effects on Suzuki-Miyaura coupling reaction using the developed isothermal reactor and we confirmed that the temperatures were controlled well in the reactor. The yields obtained by microwave heating were almost the same as that obtained by oil-bath heating. We did not find any effects of microwaves in the homogeneous Suzuki-Miyaura coupling reaction experiment.

Acknowledgements This research was financially supported by the Project of “Development of Microspace and Nanospace Reaction Environment Technology for Functional Materials” of the New Energy and Industrial Technology Development Organization (NEDO), Japan.

73

Journal of Thermal Science and Technology

Vol. 7, No. 1, 2012

References (1) (2) (3)

(4)

(5) (6)

(7)

(8)

(9)

(10)

(11)

(12) (13)

Hoz, A., Ortiz, A. D. and Moreno, A., Microwaves in organic synthesis. Thermal and non-thermal microwave effects, Chemical Society Reviews, Vol. 34 (2005), pp. 164–178 Lidström, P., Tierney, J., Wathey, B. and Westman, J., Microwave assisted organic synthesis—a review, Tetrahedron, Vol. 57 (2001), pp. 9225–9283 Sun, W. C., Guy, P. M., Jahngen, J. H., Rossomando, E. F. and Jahngen, E. G. E., Microwave-Induced Hydrolysis of Phospho Anhydride Bonds in Nucleotide Triphosphates, The Journal of Organic Chemistry, Vol. 53 (1988), pp. 4414–4416 Jahngen, E. G. E., Lentz, R. R., Pesheck, P. S. and Sackett, P. H., Hydrolysis of Adenosine Triphosphate by Conventional or Microwave Heating, The Journal of Organic Chemistry, Vol. 55 (1990), pp. 3406–3409 Mayo, K. G., Nearhoof, E. H. and Kiddle, J. J., Microwave-Accelerated Ruthenium-Catalyzed Olefin Metathesis, Organic Letters, Vol. 4 (2002), pp. 1567–1570 Garbacia, S., Desai, B., Lavastre, O. and Kappe, C. O., Microwave-Assisted Ring-Closing Metathesis Revisited. On the Question of the Nonthermal Microwave Effect, The Journal of Organic Chemistry, Vol. 68 (2003), pp. 9136–9139 Herrero, M. A., Kremsner, J. M. and Kappe, C. O., Nonthermal Microwave Effects Revisited: On the Importance of Internal Temperature Monitoring and Agitation in Microwave Chemistry, The Journal of Organic Chemistry, Vol. 73 (2008), pp. 36–47 Matsuzawa, M., Togashi, S. and Hasebe, S., Fundamental Study on Development of a Continuous Flow Microwave-Assisted Chemical Reactor, Transactions of the Japan Society of Mechanical Engineers. (B) (in Japanese), Vol. 75 (2009), pp. 316–322 Ishigaki, H., Analogy of Laminar Convective Heat Transfer in Pipe Flows with Secondary Stream (Curved Pipe and Orthogonally Rotating Pipe), Transactions of the Japan Society of Mechanical Engineers. (B) (in Japanese), Vol. 61, (1995), pp. 1782–1789 Churchill, S. W. and Bernstein, M., A Correlating Equation for Forced Convection from Gases and Liquids to a Circular Cylinder in Crossflow, Journal of Heat Transfer, Vol. 99 (1977), pp. 300–306 Comer, E. and Organ, M. G., A microcapillary system for simultaneous, parallel microwave-assisted synthesis, Chemistry– A European Journal, Vol. 11 (2005), pp. 7223–7227 Namboodiri, V. V. and Varma, R. S., Microwave-accelerated Suzuki cross-coupling reaction in polyethylene glycol (PEG), Green Chemistry, Vol. 3 (2001), pp. 146–148 Leadbeater, N. E. and Marco, M., Ligand-Free Palladium Catalysis of the Suzuki Reaction in Water Using Microwave Heating, Organic Letters, Vol. 4 (2002), pp. 2973–2976

74