RESEARCH DURING GRADUATE STUDIES AT NORTH CAROLINA A&T STATE UNIVERSITY My research project was divided into two

parts, 1) Steam Reforming Catalyst synthesis and validation and 2) Development of Pd-membrane to separate hydrogen form the gas mixture. The specific research activities include but not limited to the following: Task 1. To synthesize novel highly ordered bifunctional multimellic (bi-metallic and tri-metallic) nanocatalysts in mesoporous silica with high surface area using a one-pot procedure for efficient hydrogen (H2) production with minimal catalyst deactivation. Task 2. To understand the fundamental parameters controlling nanocatalyst activity, stability, and H2 selectivity using a designed steam reforming reactor and in situ techniques in order to maximize H2 production and to minimize deactivation of the catalyst due to CO formation. Task 3. To optimize the combination of support and ratio of the metals in the nanocatalyst. Task 4. To carry out the steam reforming reaction in tubular packed bed and membrane reactors. Catalyst Validation and Reactor Design One of the major applications of low-temperature reforming nanoscatalysts is in steam-reforming of different hydrocarbons for production of hydrogen, a green fuel for fuel cells. To demonstrate the application of these bimetallic nanomaterials in catalytic reactions, we investigated steam reforming of methanol (SRM) in a tubular packed-bed reactor and hydrogen-selective membrane-reactor. These reforming reactions are equilibrium limited and the products are H2 and CO2 with minor amounts of CO. The catalysts were tested under different operating conditions. Packed Bed Reactor: We investigated the SRM reaction in a laboratory scale packed-bed reactor using noble and non-noble bimetallic nanoparticles on mesoporous silica support as catalyst. The efficacy of this new class of metal clusters as catalyst in terms of methanol conversion, yield and hydrogen selectivity was evaluated. Membrane Reactor Separator: Inorganic membrane reactors offer the possibility of combining reaction and separation in a single operation at high temperatures to overcome the equilibrium limitations experienced in conventional reactor configurations. In our lab, we developed a new class of Pd-stainless steel (Pd-SS) composite membrane by surfactant induced electroless plating method. The hydrogen transport mechanism is dominated by solution-diffusion process and makes it an excellent candidate for studying equilibrium limited reactions in membrane reactor-separator configuration. We studied the SRM in Pd-SS tubular membrane reactor using the new class of bimetallic nanoparticles on mesoporous silica support. By simultaneous production and separation of hydrogen in the membrane reactor, we will demonstrate that the thermodynamic equilibrium limitation can be shifted to the right and as a result, the methanol conversion; yield and hydrogen selectivity can be further increased.

The membrane reactor setup has been show in Fig 1. Also in Fig-2 the prepared membrane has been shown (tubular and disc membrane).

Figure 1. Membrane Reactor setup (Diffusion Cell).

Figure 2. Tubular and disk type Pd-membrane prepared by SIEP(Surfactant induced Electroless Plating.

Description of the steam methanol reforming experiment The experimental setup is shown in Figure 3. All catalysts were reduced in 4% hydrogen gas environment at 550 C for 8 to 10 hours for activation. The reduced catalysts were placed in the reactor. The length of the catalyst zone in the reactor was about 3 inch. The reactor is a fixed bed type reactor and a very low amount of catalyst (300 mg) was loaded. Both side of the reactor was packed with glass wool. The catalysts were again reduced in situ by flowing 4% hydrogen for 4 hours at 400 C. Before each experiment, whole system was purged thoroughly with helium to remove air and any other gases present in the system. The liquid feed was adjusted to a water/methanol molar ratio of 3. The premix feed solutions were injected to the system by a micropump at flow rates 0.1, 0.2 and 0.3 ml/min. The temperature of the reactor was varied from 250 to 300 C. The operating conditions used in the steam reforming reaction are summarized in Table1.

Figure Error! No text of specified style in document. Schematic of the steam reforming experimental setup

Outer Flange

Inner Flange

Figure Error! No text of specified style in document. Design of the steam reforming reactor.

Using the composition of the reactor effluent stream, methanol conversion, selectivity of different products and hydrogen yield were evaluated.

Table 1 Operating conditions used in steam reforming of methanol. Operating Parameter Temperature (C) Feed flow rate (ml/min) Steam to methanol molar ratio Catalyst Loading Amount of Catalyst (ml) Sand (ml) 1.7 1.7 Range 250 to 350 0.1 to 0.3 3 to 1

Characterization of the Developed Material An extensive fundamental research was done on the characterization of the developed materials in order to establish the relationship between their chemical and physical properties and their performance. The characterization techniques for catalyst include microscopy such as SEM, AFM and TEM, FT-IR, X-ray. Surface area and pore size distribution was determined using Nova Quantachrome Surface Area Analyzer 2200e. The surface morphology of prepared membrane was done by SEM. Also the permeability of different gases was studied using the above diffusion. The He flux was close to 0 STDL/min for the prepared membrane. This is a big success since He is the smallest molecule.

Personal Skills Very good at designing experiment. Built the reforming set-up form the scratch. Method development, operation and calibration of Hp 6890 series GC. Advanced user of the Nova Quantachrome Surface Area Analyzer 2200e. Other laboratory equipment can use: FT-IR, UV-VIS, Contact Angle measurement, HPLC.

Novel Nanocatalysts for Hydrogen Production

Debasish Kuila1, Balaji Tatineni1, Azharul Islam2, Yulia Basova1, Atikur Rahman2, Mizanur Rahman2, Saiful Islam2, Shamsuddin Ilias2, Jasmine Taylor2, and James King1
1

The pH of the solution was measured using an Oakton digital pH meter (100 series) with a glass electrode. Synthesis. The hexagonal mesophase of CTAB was prepared using the combination of ethanolic solution of cobalt chloride or nickel nitrate and aqueous palladium nitrate solution, TMOS and conc. ammonia. First, the cobalt or nickel salt was dissolved in ethanol and palladium nitrate solution was added to it and mixed thoroughly. CTAB was dissolved in water with gentle heating to ~ 30C. The bimetallic solution was added to the CTAB solution very slowly with stirring. The precursor TMOS was added to this solution and mixed thoroughly to hydrolyze the silica precursor. Concentrated ammonia solution was then added with vigorous stirring. The pH of the resulting solution was ~ 10.5 and stirring was continued for 3 h. The resulting material was kept in an oven at 60C for 14 h, filtered, and washed with water till the pH of the filtrate reached 7. It was finally washed with acetone and ethanol to remove any physisorbed particles present inside the material. The filtered material was dried at 100C for 8 h and then calcined at 580C for 16 h. The bimetallic mesoporous silica was reduced using 4% hydrogen and argon mixture at 570C for 4 h. As described below, the bimetallic nanoparticles size varies from 2 to 7 nm and distributed uniformly. MCM-41 was synthesized using the procedure of Grun and Buchel.10,11 Results and Discussion The X-ray diffractograms of synthesized MCM-41, PdCo and PdNi-MCM-41 samples with different Pd/Co/Ni ratios show three peaks that can be indexed to d100, d110 and d200 reflections of a two dimensional, hexagonal unit cell. The observed gradual loss of long-range order of the materials with the increase in PdCo / PdNi concentration could be due to the degradation in ordering and periodicity and their interaction with silica framework as evidenced by the decreasing intensities and smoothness of the peaks. These results indicate that the size of the bimetallic nano structures is less than 7 nm or amorphous in nature.13 The SEM images of the synthesized materials indicate that the particle size of the material decreases with increase in concentration of Pd-Ni and Pd-Co. The average particle size of the material varies from 300 to 600 nm. The TEM images confirm the existence of mesopores, the presence of bimetallic nanostructures, and validate the pore size. In Figure 1, the Pd-Co /bimetallic nanocrystals appear as black spots on the inner pore walls of silica. The image also shows that the orderliness of the pore structure is still maintained despite the use of one-pot synthesis conditions. Elemental analysis data suggest uniform distribution of the bimetallic nanoparticles in mesoporous silica. The surface properties of the bimetallic silica materials are measured by BET adsorption and desorption isotherms. All samples yielded type IV isotherm and the decrease in the adsorption volume per gram of the material is a result of the dilution of MCM-41 by the PdCo oxide phase. The specific surface area of the prepared materials decreases with the increase in PdCo loading (Table 1). The synthesized bimetallic nanocatalysts have been tested for the production of hydrogen. The preliminary results are shown in Figures 2 and 3. The data suggest that the higher concentration of Pd (B6) in the material promotes more selectivity and less carbon monoxide production. The studies with different ratios of Pd-Co and Pd-Ni catalysts will be presented.

Department of Chemistry, 2Department of Chemical Engineering, North Carolina A & T State University, Greensboro, NC 27411 *Corresponding Author; E-mail: dkuila@ncat.edu

Introduction The synthesis of highly ordered nanomaterials containing metallic nanoparticles is a growing area of research in material science. The quantum-size effect of these metallic nanoparticles displays unique magnetic, optical, catalytic and electronic properties compared to the corresponding bulk metal particles.1,2 In general, the nanostructured materials with diameter less than 10 nm have potential applications due to their dimensions.3 These properties can be further adjusted by the addition of another metal as a result of its ligand and ensemble effects.4 One such unique property of the bimetallic nanoparticles is the superparamagnetic transition (with no hysteresis behavior) exhibited by particles of ferromagnetic materials that are smaller than the magnetic domain size of the material.5 These materials are also expected to show improved catalytic activity and selectivity different from those of conventional metal catalysts.6-8 One of our goals is to use bimetallic nanocatalysts to produce H2 from syngas, via biomass gasification, and alcohols, derived from biomass. This is very important for the economy of sustainable energy. The development of robust and reusable catalysts is a major challenge in this regard. Different procedures such as sorption, phase transition, ion exchange, complexation, or covalent grafting have been described in literature to incorporate inorganic metals into mesoporous silica.9-12 Nanostructures and mesoporous silica can also be prepared separately following size-selective adsorption to incorporate the nanoparticles into the mesopores.7 These multi step synthetic routes have draw backs such as adsorption of nanostructures on the external surface, leading to bigger particle size and weakening of the special properties of the materials. To address these problems, we have developed a facile and versatile one-pot synthesis at room temperature for the composite material containing Pd/Co and Pd/Ni alloy nanoparticles embedded in silica. We report here the properties of mesoporous silica containing bimetallic Pd/Ni and Pd/Co nanoparticles that can be used for production of hydrogen from biofuels. For initial studies, we have considered steam reforming of methanol and preliminary results on hydrogen production are included. Experimental Materials. All materials were analytical grade and used as purchased without further purification. Deionized ultra pure distilled water was used throughout the experiments, unless indicated. Tetramethylorthosilicate(TMOS), cetyltrimethylammoniumbromide (CTAB), cobalt chloride hexahydrate (CoCl2.6H2O) and nickel nitrate hexahydrate (Ni(No3)2.6H2O) were obtained from Sigma-Aldrich. Palladium nitrate, Pd(NO3)2, was procured from Alfa Aesar, USA. Ethanol, acetone and ammonia were acquired from Fisher. Instrumentation. Smalland wide-angle powder X-ray diffraction (XRD) patterns were measured using a Bruker AXS D8 Discover XRay Diffractometer System with monochromatic CuK radiation. The particle size, morphology and phase purity of the samples were analyzed using a Hitachi S5500 High Resolution Scanning Tunneling Electron Microscope (HR-STEM). The surface property of the material was monitored using a Nicolet Magna 850 IR spectrometer.

Acknowledgement This work has been supported by North Carolina A & T State University startup and DOE grants. We thank Drs. S. Yarmolenko, D. Kumar and Z Xu for their help with the characterization of the materials.

Prepr. Pap.-Am. Chem. Soc., Div. Fuel Chem. 2009, 54 (2), xxxx

Table 1: Physicochemical Properties of MCM-41 and Bimetallic MCM-41 Composition (g) material Properties Average BET pore surface diameter area (BJH) (m2/g) () 1103 24.5

Total pore volume (BJH) (mL/g) 0.75

d100

Wall thickness
a

Si

Co

Pd

() 37.5

() 18.8 2.6 -

MCM41 MCMPd-CoB1 MCMPd-CoB2 MCMPd-CoB5 MCMPd-CoB6

a

907

21.8

0.55

34.6

18.15

2.6

0.29

0.02

708

21.7

0.43

34.4

18.02

2.6

0.44

0.03

964

24.3

0.61

34.7

15.8

2.6

0.29

0.06

920

21.6

0.57

34.4

18.1

2.6

0.29

0.12

The wall thickness were calculted as a0 - Pore diameter (a0 = 2d100/3).

Fig. 3. Effect of methanol flow rate on H2 selectivity at different temperatures [MCM-Pd-Co materials]

Fig. 1 TEM Image of MCM-Pd-Co material; Resolution 300kx, 7.0866 pixels per nm.

Fig. 2 Effect of methanol flow rate on CO/H2 ratio at different temperatures [MCM-Pd-Co materials]

References. (1) Schmid G. Nanoparticles: From Theory to Application; WileyVCH, Weinheim 2004. (2) Fendler J. H. Nanoparticles and Nanostructured Films; WileyVCH, Weinheim 1998. (3) Yang H.; Lu Q.; Gao F.; Shi Q.; Yan Y.; Zhang F.; Xie S.; Tu B.and Zhao D.; Adv. Funct. Mater. 2005, 15, 1377. (4) Sinfelt J. H. Bimetallic Catalysts, Wiley, New York, 1983. (5) Vondrova M.; Klimczuk T.; Miller V. L.; Kirby B. W.; Yao N.; Cava R.J. and Bocarsly A.B., Chem. Mater. 2005, 17, 6216. (6) T. Abe, Y. Tachibana, T. Uematsu, M. Iwamoto, Chem. Commun. 1995, 1617. (7) T.Hirai, H. Okubo, I. Komasawa, J. Phys. Chem. B 1999, 103, 4228. (8) N. Toshima, Y.Wang, Chem. Lett. 1993, 1611. (9) Zhang L.; Shi J.; Yu J.; Hua Z.; Zhao X.and Ruan M. Adv. Mater. 2002, 14, 1510. (10) Zhu K.; Yue B.; Zhou W. and He H. Chem, Comm. 2003, 98. (11) Gao F.; Lu Q.; Liu X.; Yan Y.and Zhao D. Nano Lett. 2001, 1, 743. (12) Zhu K.; He H.; Xie S.; Zhang X.; Zhou W.; Jin S. and Yue B. Chem. Phys. Lett. 2003, 377, 317. (13) Folch B.; Larionova J.; Guari Y.; Datas L.and Guerin C. J. Mater. Chem. 2006, 16, 4435. (14) Corain B; Schmid G. and Toshima N. Metal Nanoclusters in Catalysis and Materials Science. Elsevier, 2008.

Prepr. Pap.-Am. Chem. Soc., Div. Fuel Chem. 2009, 54 (2), xxxx

Journal of Chemical Engineering, IEB. Vol. ChE. 24, No.1, January 2006 -December 2006

STUDY OF AN EVAPORATION SYSTEM FOR SODIUM HYDROXIDE SOLUTION

Md. Atikur Rahman* and Nymul Ehsan Khan
Department of Chemical Engineering Bangladesh University of Engineering and Technology, Dhaka 1000, Bangladesh

Abstract The textile industry in Bangladesh uses a substantial quantity of caustic soda (sodium hydroxide) to clean and prepare fabrics for dyeing. In all textile mills dealing with finishing, particularly with the mercerizing of cotton and mixed cotton fabrics, dilute caustic soda solutions come out as effluent from mercerizing unit. Caustic soda is expensive and its disposal without neutralization is unacceptable. A considerable amount of caustic soda can be saved if the dilute solution leaving the mercerizing step can be collected, concentrated and recycled. The concentration of the dilute caustic soda solution can be carried out by means of evaporation. A single effect evaporation system to concentrate the dilute sodium hydroxide solution was constructed and operated. The operating results confirm the economic viability of such recovery scheme. Introduction
Evaporation of sodium hydroxide solution is performed on large scale and over a wide concentration range in the production of sodium hydroxide. The textile dyeing and finishing industry in Bangladesh use dilute solution of caustic soda in different operating steps such as desizing, washing and mercerizing. The concentrations of aqueous caustic soda vary from 4 to 20% by weight. At the end of such operations, the caustic solutions are discarded as effluents having lower caustic concentration, which pollutes the water body. The concentration of caustic solution leaving the mercerizing section as effluent is about 4% by weight. A laboratory scale setup capable of producing 25 kg/hr water vapor and resulting in a solution of 20% sodium hydroxide from a 4% by weight of caustic soda effluent was designed, built and tested. This paper deals with the operation of the laboratory scale evaporation setup and its application in the industry for recovering caustic soda from waste effluent. The system was designed to evaporate of 25 kg/hr of water from 31.25 kg 4% feed solution of sodium hydroxide to yield 6.25 kg 20% solution. The flash tank would constantly contain 40 kg of 20% solution. The dilute feed solution, after passing through the preheater, would mix with a circulating stream of concentrated solution from the tank and pass through the evaporator. The mixture of liquid and vapor enters the flash tank where the vapor separates from the liquid and leaves through the top of the flash tank to pass through the preheater to preheat the feed solution. The flow was under natural circulation. The operability of the system was tested by using water as the working fluid.

Experimental Study
A 0.381 m ID and 2.44 m long cylindrical column with conical bottom made from 6.35 mm SS 304 Sheet was used as the flash separator for the boiling solution. A shell and tube exchanger (6 tubes 1.52 m long, 9.14 mm ID, SS 304, 0.10 m ID CS Shell) was used for vaporization and a preheater of shell and tube type (4 tubes 0.76 m long, 9.14 mm ID, SS 304, 0.076 m ID CS shell) was used to preheat the feed solution with vapor from the vapor separator. The steam used in the exchanger was at 16-20 psig. The setup is shown in Figure 1. Details of the setup are available elsewhere1.

Fig. 1: Setup of the process

Results and Discussions

After testing, the setup was run using actual solution of sodium hydroxide. Table 1 lists the data collected and calculated results in a summarized form. The feed rate was calculated using initial and final feed tank readings and intermediate charges. The difference in

Corresponding authors e-mail: mrahman@ncat.edu

35

Journal of Chemical Engineering, IEB. Vol. ChE. 24, No.1, January 2006 -December 2006

flash tank levels (between initial and final) and the amount of solution fed was used to calculate the total water evaporated and the average evaporation rate. Table 2 gives a calculation of the cost savings for a typical textile factory producing 1000 kg of dilute depreciation neither does it include the cost saved for solution leaving mercerizing unit per day (this costs do

not include the cost of electricity, overhead and waste treatment that would have been required). As can be seen a textile factory involved in dyeing and finishing could save around Tk. 1000 for each 1000 kg of mercerizing effluent.

Table: 1: Data Collected and Calculated Results

Steam Run Feed Run Press., Time, Rate, psig hr kg/hr Water Evap. Rate, kg/hr 22 18.76 Water Total Overall Heat Overall Heat evap/sq. ft Cond. Steam Cond. heat Steam Transfer Transfer Rate, Cond, Rate, transfer Cond., Coeff., Coeff., kg/hr kg/hr kg/hr surface kg W/m2.K W/m2.K 145.7 126.3 26.5 25.3 21952425 18752072 24.65 17.9 5.5 3.85 362 319 5.99 5.11 Evaporator Preheater Ratio of water evap to steam consumed 0.83 0.74

1+2 3+4

20 20

4.5+1 2+3, for flow 1.67+3

29 6 24

Heat transfer calculation and data 2,4. Table 2: Cost calculation for a typical textile factory. Feed solution (4%) Sodium hydroxide saved Market price of sodium hydroxide Cost of sodium hydroxide saved Water evaporated Steam consumed (taking water evaporated to steam consumed ratio =0.8) Heat required Heat provided by NG (taking boiler efficiency = 0.8) Heating value of NG (source www titasgas org bd) NG required Price of NG (source www titasgas org bd) Cost of NG Cost of daily labor Cost saved Quantity 1000 kg 40 kg Tk 40/kg 800 kg 1000 kg 2.257 x 106 kJ 2.821 x 106 kJ 1050 Btu/SCF(1108 kJ/SCF) 2546.2 SCF Tk 145.2/1000 SCF 370 200 1030 Cost. Tk

1600

References Conclusion
The recovery of caustic soda from the discarded effluent can be performed by using the system used in this work. This recovery would bring economic benefit to the industry and lead to resource recovery and reduction of treatment cost. 1. Khan, Nymul E. Khan and Rahman, Md. Atikur, Study of an Evaporation System for Sodium Hydroxide Solution, B.Sc. Engg. (Chemical) Thesis, Chem. Eng. Dept., BUET, November 2006. Donald Q. Kern, Process Heat Transfer, Tata McGraw-Hill, New Delhi, 1997. Alan S. Foust, Leonard A. Wenzel, Curtis W. Clump, Louis Maus and L. Bryce Andersen, Principles of Unit Operations (Second Edition), John Wiley & Sons, 1994. Perrys Chemical Engineers Handbook (Seventh Edition), Editors: Robert H. Perry (Late), Don W. Green and James O. Malony (Associate Editor), McGraw-Hill, 1997.

2. 3.

Acknowledgement
The authors are indebted to Dr. AKMA Quader, Professor of Chemical Engineering Department, BUET for his kind supervision of this project 4.

36

Proceedings of the International Conference on Chemical Engineering 2008 ICChE2008, 29-30 December, Dhaka, Bangladesh

A COMPARATIVE STUDY OF MICROFLUIDICS IN T-, Y- AND W-MICROMIXERS

Md. Atikur Rahman#, Debasish Kuila##, and Shamsuddin Ilias*#
Mechanical and Chemical Engineering Department# Department of Chemistry## North Carolina A&T State University Greensboro, NC 27411, USA.

Ravi Shankar
Pfizer Research Corporation Groton, CT 06340, USA.
In recent years, there is a growing interest in understanding the fast reactions (liquid phase) using micromixer as a mixing device. Microfluidic devices offer an opportunity to study these fast competitive reactions in micro-scale. To understand the mixing characteristics of two liquid streams (reactants), we investigated the micromixing of the fluid streams in T-, Y- and W-micromixers using computational fluid dynamics (COMSOL software). The effect of inlet flow conditions on the flow dynamics and the quality of mixing were investigated parametrically. In this paper, we present the simulation results of the fluid mixing characteristics of typical T-, Y-, and W-micromixers.

1. INTRODUCTION
During the past few years the importance of microscale devices in chemical engineering has increased significantly. Microfluidic devices are now used extensively in biomedical applications, including chip-based capillary electrophoresis and the analysis of cells, DNA and proteins (Whitesides et al., 2006; Huber et al., 2003). These devices are also used to study the precipitation kinetics of different competitive fast chemical reactions. Microfluidic devices, such as micropumps, microvalves, microsensors and micromixers have been developed in rapid succession in the past years. The basic requirement for a successful biochemical analysis is rapid and effective mixing of the components. Mixing of two reactants in large scale system is very slow and non-uniform. The understanding of the mixing process is complicated in situations where the rate of reaction is faster compared to the rate of mixing. This makes the process complex and a thorough analysis of the process becomes difficult. In contrast, in micromixers, rapid and effective mixing can be achieved. The mixing process in a typical micromixer is contributed by both heterogeneous and homogeneous mixings. The heterogeneous mixing arises from the convective flow. On the
*

other hand, homogeneous mixing at the molecular level is caused by diffusion between adjacent domains. The micromixers offer mixing of reactants in milliseconds, thereby increasing the residence time and allowing the study of reaction kinetics and biochemical analysis. In the microchannels, high Reynolds number cannot be achieved as mixing mostly relies on diffusion due to the laminar behavior of fluid at low Reynolds numbers. Particular attention has been paid to the geometries of the micromixer to enhance the mixing. Many designs of micromixers have been proposed and are being used in many applications. Micromixers based on active and passive mechanisms have been developed to improve mixing in microfluidic systems. External forces or forms of active control of the flow field by moving parts or varying gradients were used in active mixers (Evans et al., 1997; Moroney et al., 1991; Oddy et al., 2001). There is no energy requirement in passive mixers (Branebjerg et al., 1996; Miyake et al., 1993; Stroock et al., 2002; Chung et al., 2004; Hong et al., 2001; Gobby et al., 2001; Liu et al., 2000, 2001). The staggered herringbone mixer was designed to enhance mixing by chaotic advection (Stroock et al., 2002). A micromixer with

Corresponding Author: Shamsuddin Ilias

E-mail: ilias@ncat.edu

twisted microchannels was developed to enhance the mixing performance (Jen et al., 2003). An inplane passive micromixer that used the Coanda effect was proposed to improve mixing (Hong et al., 2001). Tee micromixers with constant and varying sizes of channels were also investigated (Gobby et al., 2001). A three-dimensional serpentine microchannel device fabricated in a silicon wafer was employed to implement chaotic advection to enhance passive mixing with a Cshaped repeating unit (Liu et al., 2000). A plastic 3D L-shaped serpentine micromixer employing three-dimensional poly-di-methyl-siloxane (PDMS) micro-fabrication and plastic micro-molding technique was designed to enhance the mixing in biological sample preparation (Liu et al., 2001). A passive micromixer using self-circulation in the mixing chamber was proposed to enhance the mixing at high and low Reynolds numbers (Chung et al., 2004). The miniaturized micromixer version of a conventional static Kenics mixer was fabricated, and the mixing mechanism used was a combination of splitting and reorientation (Bertsch et al., 2001). To understand the role of T, Y and W blocks in a passive micromixer, we used computational fluid dynamics to analyze the micromixing flow. To maximize mixing or minimize the pressure drop, system geometry with different blocks were investigated parametrically. The standard deviation in concentration was used as the quantitative measure of mixing for each of the geometries. This quantity varies between 0.5 for a completely unmixed flow and 0 for a completely mixed flow.

given geometry can be solved. In this work, we used COMSOL to solve the microfluidics of T, Y and W- micromixers. In COMSOL, a projection coupling variable is used to compute the mixing efficiency. The quality factor for the mixing is defined by:

Q ( y ) = ( C A C A ) dx
2 0

(4)

where C A is the concentration and C A is the average concentration of the tracer. The projection coupling variable is shown in Figure 1. At every node on the bottom boundary, an integral is taken over the yaxis, as indicated by the red lines.
y

Fig. 1: Integration of projection coupling variable. The dimensions of the micromixer in this study are shown in Fig. 2. The mixer is 2 mm (2000 m) long and 0.8 mm (800 m) wide. It has two inlets and one outlet. The mixing zone is 0.65 mm (650 m). In the mixing zone, inserts (T-, Y- and Wshaped blocks) are placed alternatively to enhance mixing.

2. NUMERICAL SIMULATION
The continuity equation, the momentum equation and the diffusion equation that governs the fluid flow and mass transfer are given as (Bird et al., 2003): = ( V ) (1) t DV = P + 2V (2) Dt DC A = De 2 C A (3) Dt where is the viscosity, is the density of the fluid, and P and V are the pressure and velocity vectors, respectively. The solute concentration, C A is the mole fraction of material A, and De is the effective diffusivity of A in the fluid mixture. The system of equations (Eqns. 1-3), with appropriate initial and boundary conditions for a

Fig. 2: The dimension of the micromixer, T-, Yand W-blocks.

3. RESULTS AND DISCUSSION

3.1 T- Micromixer
The fluid flow in the T-micromixer is shown in Fig. 3. The streamlines and the surface plots are shown here. From this plot, we observe that in most of the fluid path, velocity field is fairly uniform with some localized increased intensity (as shown by light colors). The 3-D plot of the cell Reynolds number of the Tmicromixer is presented in Fig. 4. With the exception of few localized regions, the cell Reynolds number is fairly uniform. For the given

geometry, it is possible to achieve cell Reynolds number as high as 0.0051 in this geometry.

Since the mixing in micromixer mostly due to the molecular diffusion, we find that low velocity results in better mixing. This is indicated by a lower value of Q ( y ) . In the T-micromixer with the lowest inlet velocity of 0.001 m/s (1 mm/s), we obtained Q ( y ) = 0.048.

Fig. 3: Surface plot and streamlines plot of velocity field in T-micromixer.

Fig. 6: Mixing quality plot for T-micromixer.

3.2 Y- Micromixer
The surface and streamline plot of velocity filed in the Y-micromixer is shown in Fig. 7 for a typical inlet velocity of 0.001 m/s. The streamlines in Fig. 7 are significantly different than that of Tmicromixer (Fig. 3) for identical inlet flow conditions. In the central region, we observe some intent interactions between the two fluid streams.

Fig. 4: 3-D plot of Cell Reynolds number in Tmicromixer. The concentration distribution in the T-micromixer is shown in Fig. 5 as surface plot. We observe that the fluid mixing is in the central region of T-blocks. The mixing process here is mostly dominated by diffusive flow.

Fig. 7: Surface plot and streamlines plot of velocity field in Y-micromixer. In Fig. 8, the cell Reynolds number of Ymicromixer is presented in a 3-D plot. If we compare the cell Reynolds number with the Tmicromixer (Fig. 4), we find improved fluid interactions in the y-micromixer. The maximum cell Reynolds number obtained in this design was 0.00639 which is higher than that obtained in Tmicromixer. The concentration distribution in Y-micromixer is shown in Fig. 9 as surface plot. If we compare the concentration distribution in Y- and T-micromixers, we observe that better mixing occurs in this case.

Fig. 5: Concentration surface plot of T-micromixer. The mixing efficiency is greatly affected by the inlet velocity. We used a built in parametric solver in COMSOL to investigate the effect of inlet velocity on fluid mixing in the micromixer. The mixing quality, Q ( y ) , for four different inlet velocities is shown in Fig. 6 for a T-micromixer.

11 for inlet velocity of 0.001 m/s. If we compare flow fields of T- and Y-micromixers (Fig. 3 and Fig. 7) with that of W-micromixer, we observe that presence of W-blocks significantly enhances fluid interactions.

Fig. 8: 3-D plot of Cell Reynolds number in Ymicromixer

Fig. 11: Surface plot and streamlines plot of velocity field in W-micromixer. The 3-D cell Reynolds number is shown in Fig. 12. The maximum cell Reynolds number that can be achieved in W-micromixer is 0.0127 which is more than that obtained in T- and Y-micromixer. The higher Reynolds number is indicative of increased disturbance in the W-micromixer. Fig. 9: Concentration surface plot of Y-micromixer The mixing quality Q ( y ) , is shown in Fig. 10 for four inlet velocities. The lowest value of mixing quality Q ( y ) , was found to be 0.028 for the inlet velocity 0.001 m/s (1mm/s). By comparing Fig. 10 with Fig. 6 we observe that better mixing occurs in Y-micromixer. A lower value of the mixing quality Q ( y ) , is indicative of that phenomenon.

Fig. 12: 3D plot of Cell Reynolds number in Ymicromixer. Fig. 13 shows the concentration surface plot of the W-micromixer. This figure is significantly different than Fig. 5 and Fig. 9 (Concentration plots of Tand Y-micromixer). From the color graph of the figure we can see that complete mixing occurs after the mixing zone. The mixing quality of the W-micromixer is shown in Fig. 14 for inlet velocities from 0.001 m/s to 0.004 m/s. We observe that W-micromixer has the lowest Q ( y ) for a given inlet flow condition when compared with T- and Y-micromixers. For example, with inlet velocity of 0.001 m/s, we obtained Q ( y ) values of 0.048, 0.028, and 0.019 for

Fig. 10: Mixing quality plot for Y-micromixer.

3.3 W-Micromixer
The velocity field is greatly affected by the Wshaped blocks in the micromixer. The surface and streamline plots of the micromixer is given in Fig.

Fig. 13: Concentration surface plot in the Wmicromixer. T-, Y-, and W-micromixers, respectively. This suggests that W-micromixer is a better design compared to Y- and T-micromixers.

Fig. 14: Mixing efficiency plot for W-micromixer.

4. CONCLUSIONS
The effect of T-, Y- and W-blocks in a micromixer was investigated to understand their role in fluid mixing. Mixing quality was used as a measure of mixing efficiency. W-micromixer had the lowest mixing quality compared to T- and Y-micromixers. A lower mixing quality is a measure of better mixing. Using this criterion, we conclude that Wmicromixer is a better design compared to T- and Y-micromixers.

ACKNOWLEDGMENTS
Partial support from Pfizer Corporate Research to Dr. Kuila is gratefully acknowledged.

utilizing microfluidic self-circulation in the mixing chamber, Lab Chip, 4, 7077. 3. Evans, J., Liepmann, D., and Pisano, A.P., (1997), Planar laminar mixer, Proc. IEEE MEMS, Nagoya, Japan, pp. 96101. 4. Gobby, D., Angeli, P., and Gavriilidis, A., (2001), Mixing characteristics of T-type microfluidic mixers, J. Micromech. Microeng., 11, 126132. 5. Hong, C.C., Choi, J.W., and Ahn, C.H., (2001), A novel in-plane passive micromixer using Coanda effect, Proc. TAS 2001 Symposium, California, U.S.A., pp. 3133. 6. Huber, D.L., Manginell, R.P., Samara, M.A., Kim, B.I., and Bunker, B.C., (2003), Programmed Adsorption and Release of Proteins in a Microfluidic Devices, Science, 301, 352-354. 7. Jen, C.P., Wu, C.Y., Lin, Y.C., and Wu, C.Y., (2003), Design and simulation of the gaseous micromixer with chaotic advection in twisted microchannels, Lab Chip, 3, 7781. 8. Liu, R.H., Stremler, M.A., Sharp, K.V., Olsen, M.G., Santiago, J.G., Adrian, R.J., Aref, H., and Beebe, D.J., (2000), Passive mixing in a threedimensional serpentine microchannel, J. MEMS, 9, 190197. 9. Liu, R.H., Ward, M., Bonanno, J., Ganser, D., Athavale, M., and Grodzinski, P., (2001), Plastic in-line chaotic micromixer for biological applications, Proc. TAS 2001 Symposium, California, U.S.A., pp. 163164. 10. Miyake, R., Lammerink, T.S.J., M., and Fluitman, J.H.J., (1993), Micro mixer with fast diffusion, Proc. IEEE MEMS, Fort Lauderdale, U.S.A., pp. 248253. 11. Moroney, R.M., White, R.M., and Howe, R.T., (1991), Ultrasonic induced microtransport, in Proc. IEEE MEMS, Nara, Japan (1991), pp. 277282. 12. Oddy, M.H., Santiago, J. G., and Mikkelsen J.C., (2001), Electrokinetic instability micromixers, Proc. TAS 2001 Symposium, California, U.S.A., pp. 3436. 13. Stroock, A.D., Dertinger, S.K.W., Ajdari, A., Mezic, I., Stone, H.A., and Whitesides, G.M., (2002), Chaotic mixer for microchannels, Science, 295, 647651. 14. Whitesides, G.M., (2006), The Origins and Future of Microfluidics, Nature, 442, 368-374.

REFERENCES
1. Branebjerg, J., Gravesen, P., Krog, J.P., and Nielsen, C.R., (1996), Fast mixing by lamination, Proc. IEEE MEMS, San Diego, U.S.A., pp. 441446. 2. Chung, Y.C., Hsu, Y.L., Jen, C.P., Lu, M.C., and Lin, Y.C., (2004), Design of passive mixers

Accepted Abstracts
American Chemical Society Annual Meeting, Fall 2009, Washington DC Abstract 1 Novel bifunctional nanocatalysts for aqueous-phase reforming. Yulia V. Basova1, Atikur Rahman2, Mizanur Rahman2, Shamsuddin Ilias2 and Debasish Kuila*1. 1 Department of Chemistry, 2Department of Chemical Engineering, North Carolina A & T State University, Greensboro, NC 27411 *Corresponding Author; E-mail: dkuila@ncat.edu Novel mesopore silica supported bi- and multimetallic nanocatalysts, have been synthesized using a hydrothermal procedure. The materials possess high surface area, regular structure and large pore size, which enhance dispersion and reducibility of the catalyst. The performance of nanocatalysts containing palladium, nickel, tin and other transition metals by aqueous-phase reforming of methanol have been studied. These novel materials have demonstrated better catalytic activity and selectivity compared to conventional supported noble metal-based catalysts as well as steady performance with no decrease in hydrogen production over time. Incorporation of less than 7% tin into the nanocatalyst structure reduces CO formation by at least 65% without disturbing the catalysts active sites, thereby producing a bifunctional nanocatalyst. The results obtained on nickel/tin nanocatalyst have shown that palladium can be completely substituted by non-noble metals without affecting the catalyst performance. The morphology, composition, activity and selectivity of the nanocatalysts for hydrogen production will be presented.

American Chemical Society Annual Meeting, Fall 2009, Washington DC Paper 2. The introduction of the pre-print submitted in the Fuel division is given below, Novel Nanocatalysts for Hydrogen Production Debasish Kuila1, Balaji Tatineni1, Azharul Islam2, Yulia Basova1, Atikur Rahman2, Mizanur Rahman2, Saiful Islam2, Shamsuddin Ilias2, Jasmine Taylor2, and James King1 1 Department of Chemistry, 2Department of Chemical Engineering, North Carolina A & T State University, Greensboro, NC 27411 *Corresponding Author; E-mail: dkuila@ncat.edu Introduction The synthesis of highly ordered nanomaterials containing metallic nanoparticles is a growing area of research in material science. The quantum-size effect of these metallic nanoparticles displays unique magnetic, optical, catalytic and electronic properties compared to the corresponding bulk metal particles.1,2 In general, the nanostructured materials with diameter less than 10 nm have potential applications due to their dimensions.3 These properties can be further adjusted by the addition of another metal as a result of its ligand and ensemble effects.4 One such unique property of the bimetallic nanoparticles is the super paramagnetic transition (with no hysteresis behavior) exhibited by particles of ferromagnetic materials that are smaller than the magnetic domain size of the material.5 These materials are also expected to show improved catalytic activity and selectivity different from those of conventional metal catalysts.6-8 One of our goals is to use bimetallic nanocatalysts to produce H2 from syngas, via biomass gasification, and alcohols, derived from biomass. This is very important for the economy of sustainable energy. The development of robust and reusable catalysts is a major challenge in this regard. Different procedures such as sorption, phase transition, ion exchange, complexation, or covalent grafting have been described in literature to incorporate inorganic metals into mesoporous silica.9-12 Nanostructures and mesoporous silica can also be prepared separately following size-selective adsorption to incorporate the nanoparticles into the mesopores.7 These multi step synthetic routes have draw backs such as adsorption of nanostructures on the external surface, leading to bigger particle size and weakening of the special properties of the materials. To address these problems, we have developed a facile and versatile one-pot synthesis at room temperature for the composite material containing Pd/Co and Pd/Ni alloy nanoparticles embedded in silica. We report here the properties of mesoporous silica containing bimetallic Pd/Ni and Pd/Co nanoparticles that can be used for production of hydrogen from biofuels. For initial studies, we have considered steam reforming of methanol and preliminary results on hydrogen production are included.

Abstract submitted in Sixteenth Symposium of Separation Science and Technology. Ethanol Steam Reforming by Silica Encapsulated Bimetallic Pd-Ni Catalyst in a Pd-Composite Membrane Reactor Md. Atikur Rahmana*, Md. Mizanur Rahmana, Md. Saiful Islama, Yulia Basovab, Debasish Kuilab, Shamsuddin Iliasa# a Department of Chemical and Bioengineering, North Carolina A&T State University, Greensboro, NC 27411 b Department of Chemistry, North Carolina A&T State University, Greensboro, NC 27411 Ethanol steam reforming (ESR) reactions are thermodynamically equilibrium limited. For production of hydrogen using ESR is carried out usually in primary and secondary reformers (reactors). To overcome the equilibrium limitations and reduce CO formation, we investigated the application of a Pd-composite membrane as a membrane reactor (MR) for simultaneous production and separation of hydrogen in a single operation. For catalytic ESR reactions, we used silica encapsulated Pd-Ni nanocatalyst which was synthesized in our laboratory by one-pot hydrothermal method. The Pd-composite membrane on microporous stainless steel support was also fabricated in our laboratory by surfactant induced electroless plating (SIEP) method. The reforming reaction was carried out by varying the water to ethanol molar ratio (3:1 to 9:1) in a temperature range of 300-500 C. The influence of the temperature and the steam to ethanol feed ratio on the ethanol conversion, hydrogen yield and the CO-free hydrogen recovery has been evaluated. In this presentation, we will present some of these results along with some H2-permeability data for Pd-composite membrane.
# *

Corresponding author: Shamuddin Ilias (email: ilias@ncat.edu) Presenting author: Md. Atikur Rahman (email: mrahman@ncat.edu)

Abstract Submitted in Annual Meeting of American Institute of Chemical Engineers 2009. Steam Reforming of Methanol and Ethanol in a Pd-Alloy Composite Membrane Reactor Wednesday, November 11, 2009: 3:53 PM Canal E (Gaylord Opryland Hotel) Atikur Rahman, Mechanical and Chemical Engineering, North Carolina A&T State University, Greensboro, NC Mizanur Rahman, Mechanical and Chemical Engineering, North Carolina A&T State University, Greensboro, NC Saiful Islam, Mechanical and Chemical Engineering, North Carolina A&T State University, Greensboro, NC Debasish Kuila, Chemistry Department, North Carolina A&T State University, Greensboro, NC Shamsuddin Ilias, Mechanical and Chemical Engineering, North Carolina A&T State University, Greensboro, NC H2-selective Pd and Pd-alloy membranes in membrane reactor offer a unique opportunity for simultaneous production and separation of hydrogen from many of the industrially important steam reforming reactions by equilibrium shift. In this study, we investigated the steam reforming of methanol and ethanol in a Pd-alloy (Pd-Cu and Pd-Ag) composite membrane reactor. Defect free Pd-alloy thin-film membranes on microporous stainless steel (MPSS) support were fabricated in our laboratory using a newly developed surfactant induced electroless plating (SIEP). The Pd-alloy membranes were tested for thermal stability and H2-perm-selectivity at 4500C. Reforming reactions were studied using: (a) commercially available Ni-based steam reforming catalyst and (b) our in-house synthesized trimetallic Pd-Ni-Sn nano-catalyst supported on mesoporous silica. The effect of steam to alcohol ratio, temperature and space velocity on alcohol conversion, hydrogen yield and carbon monoxide suppression were examined for the two reforming reactions in the membrane reactor. We also investigated the reforming reactions in a packed-bed reactor (non-membrane reactor) under similar operating conditions. In this presentation, we will present some of these experimental results.