Academic

year 2011-2012

Erasmus Mundus Master in Membrane Engineering Semester S1

Report on the six-month project

A review of Alumina: Most abundant and productive material of the mother nature.

KAYAALP Umay PAPOUTSOGLOU Dimitra

January 2012

Supervisor: AYRAL Andre; andre.ayral@iemm.univ-montp2.fr

BACCHIN Patrice ; bacchin@chimie.ups-tlse.fr

TABLE OF CONTENT
1. INTRODUCTION ........................................................................................................................... 3 2. ALUMINA AS A MATERIAL ....................................................................................................... 4 2.1. NOMENCLATURE ..................................................................................................................... 4 2.2. STRUCTURE AND MINERALOGICAL PROPERTIES ........................................................ 8 2.2.1. STRUCTURE OF ALUMINA PHASES ..................................................................................................... 8 2.2.2. PSEUDOMORPHOSIS .............................................................................................................................. 9 2.2.3. SURFACE AREA OF ALUMINA ............................................................................................................. 12 2.2.4. POROSITY .............................................................................................................................................. 13 2.2.5. SORPTIVE CAPACITY ........................................................................................................................... 15 2.3. MECHANICAL PROPERTIES OF ALUMINA ..................................................................... 16 2.4. THERMAL PROPERTIES OF ALUMINA ............................................................................ 20 2.5. CHEMICAL PROPERTIES OF ALUMINA ........................................................................... 22 2.5.1. WET CHEMICAL REACTIONS OF SINTERED ALUMINA .................................................................. 22 2.5.2 REACTION OF CHEMICAL ELEMENTS WITH ALUMINA ................................................................... 23 1.8. COLLOIDAL PROPERTIES OF ALUMINA ......................................................................... 24 3. ALUMINA MEMBRANES .......................................................................................................... 26 3.1. INTRODUCTION ..................................................................................................................... 26 3.1 PREPERATION OF ALUMINA MEMBRANES .................................................................... 29 3.2.1. MACROPOROUS ALUMINA MEMBRANE PREPARATION ................................................................ 29 3.2.2.MESOPOROUS ALUMINA MEMBRANES ............................................................................................. 31 3.2.3 MICROPOROUS ALUMINA MEMBRANES ........................................................................................... 35 4.DESIGN OF THE MEMBRANE MODULES ............................................................................. 38 4.1 DIFFERENT TYPES OF MODULES ...................................................................................... 38 4.1.1 ALUMINA MEMBRANE MODULES ....................................................................................................... 39 4.1.2. COMMERCIALIZED MODULES OF MEMBRANE ALUMINA ............................................................... 42 4.2. SEPARATION CHARACTERISTICS FOR ALUMINA MEMBRANES ............................. 45 5. APPLICATIONS ........................................................................................................................... 49 5.1. CERAMIC MEMBRANES ........................................................................................................ 49 5.2. ALUMINA MEMBRANES APPLICATIONS ........................................................................ 52 5.2.1. LIQUID PHASE SEPARATION APPLICATIONS ............................................................................... 52 5.2.2. GAS PHASE SEPARATION ................................................................................................................. 63 6.SUMMARY AND CONCLUSIONS .............................................................................................. 74 APPENDIX A : MEMBRANE MATERIAL SHEET ..................................................................... 75 APPENDIX B: CHEMICAL INTEREST OF ALUMINA .............................................................. 77 REFERANCES ................................................................................................................................... 85

1. INTRODUCTION Ceramists are not close agreement as to the substances included in the term of ceramics, nor do they seem to have devised as simple, consistent definition of the term that is entirely satisfactory. Kingerly defined it as the art and science of making and using solid articles which have as their essential component, and are composed in large of inorganic nonmetallic materials. L. Mitchell defined ceramics as all high-temperature chemistry and physics of nonmetallic materials, and the techniques of forming products at high temperatures. The first definition allows inclusion of materials having melting points below room temperature, as, for example ice; while the second does not exclude certain organic substances that may been produced at high temperatures, such as carbon disulfide. Although materials of all kinds, including organic substances, are involved in the preparation of ceramics, it is believed that these definitions are too broad to cover the ceramic applications of alumina. The investigation of alumina as a ceramic material was undertaken to provide information under the following specifications: a review dealing with alumina both from a theoretical and a practical point of view, and including information on the nomenclature, properties of alumina, alumina as a membrane material and finally industrial alumina membrane applications. The following information has been gathered: A general review tried to be gathered to understand deeply about alumina material and also alumina membranes. occurrence in nature crystal or mineralogical characteristics mechanical, thermal, chemical and colloidal properties. alumina membranes fabrication, modules and industrial applications.

2. ALUMINA AS A MATERIAL

2.1. NOMENCLATURE De Morveau suggested the word alumine in 1786 as the proper name for the basic earth of alum. This was Anglicized to Alumina in England while Germany tornerde is still used, meaning clay earth. The term alumina is presently used rather indefinitely in ceramic literature to denote; 1. aluminous material of all types taken collectively 2. the anhydrous and hydrous aluminum oxides taken indiscriminately 3. the calcined or substantially water free alimunium oxides without distinguishing the phases present 4. corundum or alpha alumina, specifically. It is often used interchangeably with the molecular formula Al2O3. The true meaning is sometimes hard to determine from the context. More than 25 alumina solid phases is defined in recent years. But it is doubtful if all of them really exist or not. These phases includes, amorphous hydrous and anhydrous oxides, crystalline hydroxides and oxides, and aluminas containing small amounts of oxides of alkalies or alkaline earths, designated as beta aluminas. The phases found in the nature, and few of artificial types, have common or mineralogical names. Most of them also are defined by greek letter formulas. Corundum, emery, sapphire, and ruby are more or less pure forms found in the nature and known for antiquity as abrasives and gem stones. All consist of the the phase designated alpha alumina (! Al2O3). Figure 1 shows the dehydration sequence of alumina hydrates in air. Diaspore Another native mineral, described by Hay in 1801, was named diaspore by him, from the Greek for scatter because it flew apart upon heating. 4

 Figure 1: Dehydration Sequence Of Alumina Hydrates In Air

Gibbsite Vaquelin in 1802 gave its formula as Al2O3 3H2O. Dewey named a well- crystallized mineral gibbsite for G. Gibbs an American mineralogist. It corresponded with the formula Al(OH)3 or Al2O3 3H2O. It is the principal phase of the trihydrate bauxites. Bauxites Berthier examined a mineral from Les Baux in southern France, containing about %52 Al2O3 and 20% bound water, from which it was supposed that the mineral was Al2O3 2H2O. The mineral was named bauxite by St. Clair Deville. Bohemite After X- Ray diffraction became generally used for analysis of chemical components a new pattern of bauxite is discovered. This bauxite has %15 bound

water Al2O3 H2O. This component named bohemite for both natural and artificial products. These phases are the izomers of diaspore. In USA, the word bauxite has come to mean any highly aluminous ore composed mainly of one or more of phases, gibbsite, boehimite and diaspore. In reality no phase corresponding to Al2O3 2H2O has been found. In 1925, Haber devised a system of nomenclarature for trivalent alumina phases the known. The alpha series included diaspore and corundum (alpha alumina); the gamma series included hydrargillite (gibbsite), bauxite (bohemite), and gamma alumina. The classification was obviously based on the end product of calcination, but this is somewhat arbitrary because gamma alumina also transforms to alpha alumina. Bayerite A new phase is identified by Bhm in aluminum hydroxide precipitates that had aged moist for several months at room temperature. The water content was about that of gibbsite, but the X-Ray pattern was different, indicating an isomer of gibbsite. This phase called Bayerite by Fricke in 1928 on the erroneous supposition that it was normal product of the Bayern process. L. Milligan had already shown in 1922 that the Bayer product is predominantly gibbsite. Bayerite was claimed to have been found in nature by Gedeon, and more recently (1963) by Gross and Heller. The failure of Haber classification system to distinguish between bayerite and gibbsite prompted the devising of the Alcoa system of nomenclature (Frary). In this system, the choice of Greek letters was initially based on the relative abundance of phase in nature. Gibbsite was called alpha alumina trihydrate; bohemite, alpha alumina monohydrate; bayerite beta alumina trihydrate; and diaspore, beta alumina monohydrate. Gamma alumina and beta alumina had the same significance as in the Haber system.

Table 1: Nomenclature of Crystalline Aluminas (WALTER , Alumina as a Ceramic Material, 1970, page 5) Mineralogical Name Phase of Form Name Symposi um (1) Alcoa (2) ! !"! !! 3!! ! ! !"! !! 3!! ! ! !"! !! !! ! ! !"! !! !! ! Chi Eta Gamma Kappa Ioata (17) Alpha Haber (3) ! !"(!")! ! !"(!")! ! !"(!")! ! !"(!")! Aluminas Corundum, Sapphire Zeta Alumina Chi Eta Gamma Kappa Alpha Gamma Alpha Chi + Gamma Gamma Delta Kappa + Theta Alpha Rho Chi + Gamma Eta Gamma Kappa + Delta Alpha 2 Xi , Xi (13)
1

British (4) ! !"! !! 3!! ! ! !"! !! 3!! ! ! !"! !! !! !

French (5)

Other

Hydroxides Gibbsite (6) Al(OH)3 Hydrargillite Bayerite (7) Nordstrandite (1) Randomite (8) Bayerite II (8) Bauxite (9) Bohemite (10) Diaspore Tohdite (11) Al(OH)3 Al(OH)3

Al(OH)2 5 !"! !! !! ! ! !"! !! (12)

AlOOH AlOOH AlOOH

AlO Al2O M2O 11 Al2O3 (14) M2O 6 Al2O3 (15) MO 6 Al2O3 Li2O 5 Al2O3 (10)de Lapparent (11)Yamaguchi (12)Steinheil (13)Cowley (14)Rankin and Merven (15)Scholder (16)Barlett (17)P.A. Foster

(1) Ginsberg, Huttig, Strunk-Lichtenberg (2) Edwards, Frary, Stumpf, et al. (3) Haber, Weiser, and Milligan (4) Rooksby, Day, and Hill (5) Thibon, Tertian, and Papee (6) Dewey (7)Fricke (8)Teter, Gring, and Keith (9)Bhm

Standardization of the nomenclature for aluminas is very desirable particularly to avoid the confusion in the hydrous phases. Gingsberg reported the conclusions of a symposium held in 1957, in which an attempt was made to devise a universal standard nomenclature. Some features of the proposed system are improvements, for example the substitution of hydroxide instead of hydrate, namely, aluminum trihydroxide for alumina trihydrate; aluminum oxide hydroxide for alumina trihydrate; aluminum oxide hydroxide for alumina monohydrate. Also, it was agree to use the Alcoa nomenclature for the transition aluminas, but to designate some of them as forms rather than phases, to imply the present uncertainty about them. The confusion in naming the hydroxide phases has not been resolved, however.

2.2. STRUCTURE AND MINERALOGICAL PROPERTIES The remarkable range of properties of the hydrous and non hydrous crystalline properties of alumina has been interesting for researchers, and its structure has been induced much scientific curiosity. Examples of these structural peculiarities are the factors determining the phenomena of transition phases, and the exceptional strength and hardness of corundum. Besides the ideal crystal structures, which are a rarity in actual ceramic systems, defect crystal structure and microstructure are significant. Gross structure beyond the crystal lattice beyond the crystal lattice is also of ceramic interest. 2.2.1. Structure of Alumina Phases Crystal structure is the main factor controls the properties of aluminas. In general, the phases of most significance in alumina are those produced by pseudomorphic dehydration. The crystal structures of alumina phases are shown in Table 2. Mineralogical properties of the various phases are shown in Table 3.

2.2.2. Pseudomorphosis Achenbach (1931), Damerell (1932), and Teritan (1950) have shown that the dehydration of gibbsite crystals is pseudomorphic, that is, external shape of the crystals is retained and there is an orientation relationship of the crystal axes of the new phases to those of the original. The crystals lose transparency and smoothness upon heating, and fine-grained fibers develop parallel to the hexagonal surface. Void space resulting from the loss of water from the gibbsite and increasing density of the transition phases is distributed in microporosity of very high surface area in the porous skeleton (Weitbrecht and Fricke). Pseudomorphosis is of considerable importance because of its effect on surface area of the intermediate phase structures, and on crystal size and size distribution of the fully calcined aluminas for ceramic forming processes. With the basis of alumina structure, all the transition aluminas have oxygen ions in approximately cubic close packing. The differences in their patterns represent changes in intensities of reflections resulting from differences in distribution of the aluminum ions. The initial cationic disorder of the low-temperature phases depends upon the source of the alumina. The transitions become more ordered with increasing heat treatment.

Table 2: Crystal Structure of the Aluminas (WALTER , Alumina as a Ceramic Material, 1970, page 30)

Phase Formula Crystal System Mole- Space cules Group
! !! ! !!

Unit Cell Parameters Angstroms a b c 8.641 5.07

Angle

Ref.

Hydrated Aluminas Gibbsite ! !"! !! 3!! ! ! !"! !! Bayerite 3!! ! Nordstrandite !"! !! 3!! ! Bohemite ! !"! !! !! ! Diaspore ! !"! !! !! ! Transition Aluminas Chi Eta Gamma Delta Iota Theta Kappa Corundum ! !"! !! Al2O !"# !"! !! Beta Aluminas (21) !"! ! 11!"! !! !! ! 11!"! !! !"# 11!"! !! !"# 6!"! !! !"# 6!"! !! !"# 6!"# Zeta Alumina !"! ! 11!"! !!

Monoclinic Monoclinic Monoclinic Orthorhombic Orthorhombic Cubic Cubic (Spinel) Tetragonal Tetragonal Orthorhombic Monoclinic Orthorhombic

4 2 8 2 2 10 10 32 4 4 32 2 1 1 1 2 2 2 2

9.72 5.06 19.12 3.7 2.844

1 3 4 5,12 5 9 9,8 6,7 10,11 9 13 9 2 20 22 14 14 16 15 15 17 18,19

!" !! !" !! ! !! ! !!

4.716 8.679 8.63 5.01 2.868 12.227 4.396 9.426

7.95 7.9 7.95 7.95 7.79 7.967 7.967 23.47 7.73 7.78 2.92 5.63 2.95 11.86 8.49 12.73 13.96 4.758 4.98 7.915 5.58 5.58 5.56 5.54 5.56 5.58 12.991 22.45 22.67 22.55 21.83 21.95 22.67

! Rhombohedral !! ! Cubic ! Cubic (Spinel) !!

Hexagonal Hexagonal Hexagonal Hexagonal Hexagonal Hexagonal Cubic

! !! ! !! ! !! ! !! ! !! ! !! ! !!

7.9

(1) Megav (2) Swanson, Coo Isaacs, & Evans (3) Unmack (4) Lippens (5) Swanson & Fuyat (6) Saalfeld (7) Brindey & Nakahira (8) Verwey

(9) Stumpf (10) Teritan & Papee (11) Rooymans (12) Reichertz & Yost (13) Kohn (14) Beevers & Brouhult (15) Laderqvist (16) Bragg (17) Adelskld

(18) Kordes (19) Braun (20) Hoch & Johnson (21) Scholder & Mansmann (22) Filonenko, Larov, Andreeva & Prevszer

10

Table 3 : Mineralogical Properties of the Alumina (WALTER , Alumina as a Ceramic Material, 1970, page 31) Phase ! Hydrated Aluminas Gibbsite 1.568 Bayerite 1.583* Bohemite 1.649 Diaspore 1.702 ! Transition Aluminas Chi Eta Gamma Delta Iota Theta Kappa Corundum Al2O AlO ! Al2O3 Beta Aluminas Sodium Beta 1.7604 1.635 - 1.650 1.642 Index of Refraction nd Cleav- erge

Mohs Hard

Micro Hard 2 kg/mm 2150 2070

Dens. Measured g/ml 2.42 2.53 3.01 3.44 3.0 ! 2.5-3.6 3.2 ! 3.2 ! 3.71 ! 3.56 3.3 3.96-3.98 3.84

Ref.

1.568 1.659 1.722 ! 1.7686

1.587 1.665 1.75 * 1.59 -1.65 1.604 1.66-1.67 1.67-1.69

(100) (010) (010) none

2.5 -3.5 3.5 - 4 6.5 - 7 9.0

1,2 3,2 4,5,14 1,14 19 6 20 7 6 6 21,15 13

1.77 - 1.80 1.676 1.675

Potassium Beta 1.640 1.668 Magnesium Beta 1.629 1.665-1.680 Calcium Beta 1.752 1.759 1.754 1.763 Barium Beta 1.694 1.702 Lithium Zeta 1.735 *Average ! Estimate (1) Dana (8) Rankin & Mervin (2) Roth (9) Kordes (3) Montoro (10) Toropov (4) Ervin (11) Filonenko (5) Bonshtedt (12) Wisnyi Kupletskkaya (13) Filonenko, Larov, (6) Thibon Andreeva & Pevzner (7) Foster, P. A. (14) Fricke & Severin

3.731 3.69 3.61

8,10,18 17,18 16,8 11,12 10 9

(15) Coble (16) Bragg, Gottfried, West (17) Kato & Yamauchi (18) Beevers & Brohult (19) Stumpf (20) Gingsberg (21) Biltz & Lemke

11

2.2.3. Surface Area of Alumina Alumina is widely used as a catalyst or catalyst support in many heterogeneous catalytic processes owing to its high surface area, superior chemical activity and low cost. In order to prepare the thermal-stable alumina with high surface area and large pore volume, two ways have been adopted. One is that some additives including silica, phosphoric oxide, barium oxide, cerium oxide and lanthanum oxide have been added to alumina. But the presence of these additives will modify the original properties. The other way is using some new methods and techniques, such as sol-gel method and supercritical drying techniques. Table 4 lists the specific surface areas, pore volumes, and pore diameters measured for samples of Al2O3. The results show that the alumina samples prepared using -cyclodextrin template had the higher surface areas (124-484 m2/g), larger pore volumes (0.7-1.27 mL/g) and more thermal stability than samples prepared without using -cyclodextrin. The sample A-773 exhibits the highest SBET (484 m2/g) among the alumina samples calcined at different temperatures. When the temperature exceeds 773 OK, the SBET decreases rapidly, but the pore volume changes a little. After calcination at 1273 OK, the A-1273 maintains surface area of 124.2 m2/g and pore volume of 0.70 mL/g. However, B-773 and B-1273 have the surface areas of 348 and 98.3 m2/g, respectively. The pore volume of B-1273 is only 0.54 mL/g.
Table 4: Specific surface areas, pore volumes, and pore diameters measured for samples of Al2O3 Sample A-773 A-923 A-1073 A-1273 B-773 B-923 B-1073 B-1273 SBET (m /g) 484.34 285.27 220.86 124.22 348.01 228.28 201.2 98.3
2

Pore Volume (mL/g) 0.98 1.27 0.87 0.7 1.08 1.14 0.74 0.54

Pore Diameter (A) 60.85 24.41 20.49 72.23 24.54 28.15 24.67 72.16

Harris and Sing study on gels formed by hydrolysis of aluminum isopropoxide. 1100 m2/g of surface area is detected. Storage of these unstable gels in the

12

presence of water vapor caused loos in surface area to about 500 m2/g. The adsorption isotherms of nitrogen, determined at -196 OC on the outgassed products were of the reversible S-type, characteristic of physical adsorption on nonporous solids. Gels that have been dehydrated at room temperature approximately to the formula Al2 . 3H2O showed no X-ray evidence of crystalline structure. 2.2.4. Porosity Porosity, and special case, permeability significantly affect the properties of alumina ceramics, and in a wide range of magnitude. Porosity is generated in sintered alumina structures for various reasons, some of which are to improve permeability to gases and liquids for porous diaphragms and diffuser plates, to increase the thermal insulation of refractories, and to improve the fuel combustion in radiant heaters. Volatile or combustible burn-outs (sawdust, naphthalene) have been used to generate pores. Gas generators include: hydrogen peroxide and aluminum powder with acids or alkalies. Calcining mixtures of ground and unground Bayer alumina at high temperatures can develop gross porosity beyond 50% by volume. Uniformly distributed porosity is attained by bisque firing fine-ground alumina in the undersintered range 1000 to 1400 OC. Barrett, Joyner and Hallenda applied adsorption - desorption on sintered alumina to have further information about pore shape and pore distribution. In literature five general types of hysteresis loops are defined, from which fifteen capillary shapes could be deduced. The adsorption isotherms of the activated forms of alumina fit the three main types, A, B and E, all of which have steep desorption curves. Type A has a steep sorption branch, type B a gradual sorption branch, with a broad hysteresis range, and type E a gradual sorption branch with a narrow hysteresis range. The pore shapes are mainly open and closed tubular capillaries, ink bottle shapes, and slit shapes. In figure 2 the pore size distribution curves were derived from the N2 physisorption isotherms according to the B-J-H method, Figure 2 a and b show, respectively, the pore size 13

distributions for the commercial Al2O3; and the ceramic foams obtained from the same aluminas.
Figure 2: a) N2 physisorption isotherm of commercial -alumina; (b) Pore size distribution of commercial -alumina; (c) N2 physisorption isotherm of commercial -alumina ceramic foam; (d) Pore size distribution of commercial ?-alumina ceramic foam.

Hayes, Budworth, and Roberts; investigated the permeability of sintered aluminum tubes (total porosity 4 to 9 %, purity 99.3 to 99.8% Al2O3). These tubes at first were impermeable to oxygen, nitrogen and argon at temperatures below 1500 OC. At 1500 to 1750 OC, the specimens showed appreciable permeation to oxygen, presumably by a surface diffusion process. The diffusion coefficient was about 100 !"2/sec. very slight or no permeation was found for nitrogen, and none for argon. After continued exposure for 100 hours at 1700 OC, permeation by normal channel-flow developed suddenly and swamped the earlier phenomena. The permeation of nitrogen through hot-pressed sintered alumina (4 to 14% total porosity) was predominantly by Knudsen flow.

14

2.2.5. Sorptive Capacity The strong desiccating action of activated alumina has been known at least since 1879. The properties that make the activated aluminas particularly suitable for desiccant use are: the ability to develop high surface area during formation of dehydration; a high degree of chemical inertness; resistance to softening, swelling, and disintegration when immersed in water or other liquids; high resistance to shock and abrasion; and the ability to return to the original highly adsorptive from by a suitable thermal regenerative treatment. Alumina is also use for special type of adsorption called chromatography, in which the identification and separation of adsorbed ions are usually based on a visual, spatial order of adsorption. Typical properties of desiccant, commercial alumina samples are shown in Table 5. A partial list of gasses and liquids that can be dried by activated alumina (Alcoa brochure, June 1, 1967) includes the following. Gasses Acetylene, air, ammonia, argon, carbon dioxide, chlorine, cracked gas, ethane, ethylene, freon, furnace gas, helium, hydrogen, hydrogen chloride, hydrogen sulfide, methane, natural gas, nitrogen, oxygen, propane, propylene, and sulfur dioxide. Liquids Benzene, butadiene, butane, butene, butyl acetate, carbon tetrachloride, chloro benzene, cyclohexane, ethyl acetate, freon, gasoline, heptane, n hexane, jet fuel, kerosene, lubricating oils, naphtha, nitrobenzene, pentane, pipe-line products, propane, propylene, styrene, toluene, transformer oils, vegetable oils and xylene.

15

 Table 5: Typical Properties of Desiccant, Chromatographic And Catalytic Aluminas (WALTER , Alumina as a Ceramic Material, 1970, page 39) Typical properties Al2O3 % Na2O % Fe2O3 % SiO2 % Loss on ignition (1100 OC) % SO3 % CaO % Nickel Formate % Form Surface area m /g Bulk density, loose kg /m Bulk density, packed, kg/m Spesific Gravity Static sorption at 60% RH Crushing Strenght Pore volume ml/gm pH Sieve Analysis On 80 mesh % Through 270 mesh %
3 3 2

F-1 92.00 0.90 0.08 0.09 6.50 Granular 210.00 832.9620 881.0175 3.30 14-16 55.00

H-151 90.00 1.60 0.13 2.20 6.00 Ball 390.00 816.9435 848.9805 3.1-3.3 22 - 25 75.00

F-20 92.00 0.90 0.08 0.09 6.20 Granular 210.00 929.0730 3.30 9.00 2 max 5 max

T-71 99.5 + 0.01 0.06 0.04 0.00 0.09 0.06 Granular 0.50 1217.4060 0.15-0.20

F-110 92-94 0.08 0.03 0.01 6.0-8.0

F-7 84.00 0.90 0.08 0.09 12.10 2.50 Granular

Ball
180-280

800.9250 832.9620 881.0175 881.0175 0.38

1089.2580 1361.5725

2.3. MECHANICAL PROPERTIES OF ALUMINA Alumina has remarkable mechanical properties in comparison with conventional porcelains and other single oxide ceramics. None of the likely refractory single oxide contenders approaches pure sintered alumina in bending and tensile strength, and is exceed only by stabilized ZrO2 in compressive strength. Many if the advantageous strength characteristics are retained to lesser extent by the high and low alumina porcelains. The interest in mechanical properties stems from several applications such as possible substitution of alumina ceramics for refractory metal parts in air-bone equipment, or fabrication forms in which high mechanical strength, membrane, hardness or thermal shock resistance is important.

16

The mechanical tests of particular significance include: flexural, compressive, tensile torsional, and impact strengths; modulus of elasticity and rigidity; Poissons ratio and bulk modulus; fatigue, creep, internal friction, thermal shock resistance, and flaw detection; and hardness. Structural applications of aluminum oxide in the high temperature field require knowledge of the effect of temperature on mechanical properties. Data on mechanical properties of alumina is collected in Table 6 and 7. The data in the table include information taken from ceramic literature as well as average values for commercial production, taken from the standards of alumina ceramic manufacturers association and the literature of several studies. Table 6 is belonging to typical properties and specifications of commercial grades of 3 kinds of alumina. Hydrated aluminas, prepared in a modern Bayer plant, is indicated the specimen of C. And specimen A belongs to calcined aluminas, which specimen T indicates tabular alumina. Data on the mechanical properties of alumina is collected in Table 7. The data in the table include information taken from the ceramic literature, as well as average values for commercial production, taken from the standarts of the Alumina Ceramic Manufacturers Association.
Table 6: Typical Properties and Spesifications of Hydrated Aluminas-Series C-30 (WALTER , Alumina as a Ceramic Material, 1970, page 21) Typical Properties Al2O3 % SiO2 % Fe2O3 % Na2O % Moisture (110 C) % Specific Gravity Sieve Analysis on 100 mesh % on 200 mesh % on 325 mesh % through 270 mesh %
O

C 65 0.01 0.003 0.16 0.04 2.42 0-1 5-10 30-55 45-70

A 98.9 0.02 0.03 0.45 1.0 3.6-3.8 4-15 50-75 88-98 2-12

T 99.5+ 0.06 0.06 0.20 3.65 -3.8

17

Table 7: Mechanical Properties of Alpha Alumina Oxide (WALTER , Alumina as a Ceramic Material, 1970, page 45) Sapphire Bending Strength (Modulus of Rupture) O Temp C Flame-fused, O oriented 0 between optic axis a and bar axis Flame-fused, oriented 45 between optic axis and a ,b bar axis Flame-fused, oriented 45 between optic axis and a ,b bar axis S25OC=142500 e
-11.83PG-0.60+3.33P O O

Sapphire

Ruby Polycrystalline Alumina

25 600 1000 25 600 1000 25 600 1000

MPa 680 180 300 480 313 567 333 220 567

(1)

(18) S1200OC=73000e Polycrystalline Alumina (99.9% Al2O3 98% theoretical density, hot pressed) (2)
-11.33PG-0.60+3.33P

Crystal Size (microns) 1 2 10 15 O Temperature C O 25 C 447 320 O 400 C 347 247 O 1000 C 327 247 O 1350 C 247 107 Commercial Grades of Polycrystalline Alumina Nominal % Al2O3 %99.9 (5) %99 (3) %94 (3) O 25 C 413 347 307 O 980 C 153 113 Compressive Strength (MPa) Sapphire Polycrystalline O Temperature C 100 % Al2O3 (7) 99% Al2O3 (3) 94 % Al2O3 (3) O 25 C 295 3733 2000 2000 O 25 C 3300 2840 O 400 C 1420 O 800 C 1233 O 1000 C 853 O 1200 C 473 O 1400 C 237 O 1600 C 47 Mohs Scale Hardness on the 9

40 50 233 227 207 93 %85 (3) 307 80

85 % Al2O3 (3) 1600

18

Table 7 continued Single Crystal Orientation 45 O to optic axis (9) 473 350 350 587 Tensile Strength MPa Filaments Uncoated (10) 467 Coated (10) (e) Polycrystalline 94 % Al2O3 (3) 85 % Al2O3 (3)

Temperature C 30 C O 300 C O 800 C O 1050 C O 1100 C O 1200 C O 1400 C

O

173 117 1400 251 243 227 225 209 63 57 123 28 Modulus of Elasticity (E) X 108 MPa 6 -3.95P E (polycrystalline)=59.49 X 10 e , where P = fractional pore volume (8) Single Crystal (12) Polycrystalline Ruby O Temperature C Sapphire (0.75%Cr (11) 94 % Al2O3 (3) 85 % Al2O3 (3) 2O3) O 25 C 35.07 36.07 39.53 26.80 21.27 O 500 C 32.07 33.00 38.18 O 1000 C 29.00 30.07 36.59 O 1200 C 27.93 29.00 35.77 ompliances Elastic constants Elastic C ( X10 MPa) 48.03 48.16 14.27 15.82 10.72 -2.27 Modulus of Ridity(G) X 10 MPa
8 8

( X10 MPa) 2.35 2.17 6.94 -0.72 -0.36 0.49 Polycrsytallined Alumina Cold-Pressed 94 % Al2O3 (3) Zero Porosity Density 3.62 (3) (14) 15.93 11.33 (13)

O Temperature C 25 C
O

Sapphire (13) 15.53 (Reuss) 16.05 (Vogit) Hot-Pressed Zero Porosity (14) 15.51

85 % Al2O3 (3) Density 3.42 (3) 8.67

a Loading rate 94.7 MPa/minute; b minimum creep at 45O; c zero porosity; d less than 5% porosity; e rupture

time less than one minute. (1) Wacthman &Maxwell (1959) (2) Springs Mitchell & Vasilos (1964) (3) Coors Prorcelain Data Sheet 0.001, August 1964 (4) Diamonite Products Manufacturing Company (1963) (5) Frenchtown Porcelain Co., Bull. 5462 (7) Ryshewitch (1941) (9) Wachtman & Maxwell (1954) (8) Knudsen

(10) Berezhkova &Rozhanskii (11) Crandall, Chung, & Gray (1961) (12) Wachtman &Lam (1959) (13) Wahtman, Tefft, Lam & Stinchfield (1960) (14) Lang (1960) (15) Ryshewitch (1951) (16) Spriggs & Brisette (17) Kingery & Pappis (18) Passmore, Spriggs & Vasilos (1965)

19

2.4. THERMAL PROPERTIES OF ALUMINA Chemical and thermal stability, relatively good strength, thermal and electrical insulation characteristics combined with availability in abundance have made alumina attractive for engineering applications. Thermal properties of alumina are listed below on the Table 8.
Table 8: Thermal properties of alumina. (WALTER , Alumina as a Ceramic Material, 1970, page 64) Melting Point Boiling Point Vapor Pressure !Al2O3 !Al2O3 T K 2309 2325 2370 2393 2399 2459 2478 2487 2545 2565 2605 Heat of Formation at 298.16 O K (kcal/mole) -612.8 -609.4 -304.2 -417.8 -400.4 -138 -248 0.03550922 ! 4.0884 10 23.15 8.43 12.16 48.967 59.75 (9) (16) (15) (8) (9) (9)
O

2051.0 9.7 3530 (3800 200 K) Atm 8.7 X 10 1.03 X 10 1.66 X 10 1.68 X 10 2.15 X 10 3.78 X 10 5.81 X 10 9.1 X 10 2 X 10 1.29 X 10 1.91 X 10 O Entrophy at 298.16 K (kcal/mole) 33.51
-6 -6 -6 -6 -6 -6 -6 -6 -6 -6 -6 O

(1) (2)

(5,3) (4) (3) (4) (3) (7,6) (2,14) (2,14) (2)

! Al2O3 3H2O ! Al2O3 3H2O Amorphous ! Al2O3 H2O ! Al2O3 H2O ! Al2O3 AlO Al2O

Enthalpy (E+PV) in kcal/mole ! Al2O3 (!! !"#.!" ! ) !! !

! 11.23206 !"!!" ! + 19.63341 (a)

0.035549846 ! 3.9085 10!! ! ! 11.2306 !"!!" ! + 17.23778 (b) 0.03031602 ! + 8.3979 10!! ! ! 2.81406 ! 10! / ! 12.87764 (c) Specific Heat (cal/g K) ! Al2O3 3H2O 0.2694 + 6.43 X 10!! t (d) 0.2855 at 25 0.348264 8.019 X 10 T 47.8423/T (d,e)
-6

20

Table 8: continued ! Al2O3

O O O O

cal/g K 0.22545 0.24857 0.26372 0.27431 0.28205 0.2783 0.3364 0.3814

cal/g K 0.28789 0.2924 0.29595 0.29877 0.4196 0.468

-6 O

400 500 600 700 800 1318 1510 1660 ! Al2O3 Temperature Range
O

900 1000 1100 1200 1787 2575

(9)

(10) (11)

Thermal Expansion, Linear (X 10 / C) Single Crystal (12) Orientation 0 1.95 4.39 6.26 7.31 7.96 8.65 8.84 8.98 9.08 9.18
O

Polycrystalline
O

C-axis 90 1.65 3.75 5.51 6.52 7.15 7.8 7.96 8.12 8.2 8.3
O

(12) 1.89 4.1 6.03 6.93 7.5 8.08 8.25 8.39 8.49 8.58

-273.16 to 0 -73 to 0 0 to 127 327 527 927 1127 1327 1527 1727

Thermal Conductivity (cal/sec cm C) O O cal/sec cm C Temperature C -263 -253 -233 -223 25 100 300 500 700 900 1100 1300 1500 1700 1900 3 9 14 12 0.086 0.069 0.038 0.025 0.018 0.015 0.014 0.014 0.013 0.014 0.015

(13)

21

(a) Range 400 K to 1200 K; (b) Range 678 K to 1330 K; (c) Range 1290 K 1673 K; (d) T= Kelvin; (e) Range of equation 400 to 1200 K. (1) Schneider, Nat. Bur. Stds. Private communication, Nov , 1968 (2) Brewer & Searcy (3) Rossini, Wagman, et. Al. N. B. S. Circ. 500, 1952 (4) Russel et. Al. (1955) (5) Barany & Kelley (6) Kerr, Johnson, & Hallett (7) Mah (8) Roth, Wirths, & Berendt (1942) (9)Furukawa et. Al. (10) Shomate &Naylor (11) Sheindlin(1964) (12) Wachtman, Scuderi, & Cleek (13)Coors Porcelain Co. (AD 995) (14) Panyushkin &Maltsev (15) Banashek, Sokolov, Rubinchik & Fomin (16) Sokolov, Banashek, & Rubinc
O

2.5. CHEMICAL PROPERTIES OF ALUMINA Chemical reactions of alumina of general ceramic interest include the resistance to attack of sintered alumina by various reagents, particularly at high temperatures. High temperature chemistry includes those chemical phenomena, which occur above 1000 OC. Such temperatures are attained by combustion, by electrical heating, or by chemical explosions and nuclear reactions (Margrave, 1962). Some example studies are listed below due to understand further chemical behavior of alumina. 2.5.1. Wet Chemical Reactions of Sintered Alumina In aqueous solution, aluminum oxide exhibits an amphoteric behavior. It can be expressed by the following equilibra:
! !! !! + 3 !!! !"(!")! !"!! + !! !!

Impermeable alumina has been marked resistance to wet chemical corrosion. As a rule the lower phases of alumina and the hydrous aluminas show increasing chemical reactivity as they decrease in density. Early experiments on the resistance of sintered alumina to attract were intended to demonstrate its suitability as a container, crucibles, etc., for thermal reactions (Winzer, 1932). Concentrated H2SO4, HCl, HNO3, H3PO4, and 20% NaOH dissolved no more than 0.02% of a 30 X 35 mm crucible within six hours, as indicated by loss in weight of the crucible. This is not necessarily indicative of chemical inertness, as for example; phosphoric acid readily reacts even with coarsely crystalline tabular alumina to form slowly soluble phosphate bonds at temperatures below the

22

boiling point of the acid. Dawihl and Klingler (1967) state that sintered alumina containing 3% silicates is far more resistant to corrosion by HCl, HNO3, and H2SO4 in concentrations from 10 to 95% acid and up to 100 OC than titanium, cast silicon and Cr Ni steels. Finely divided alumina is rapidly dissolved by HF, hot concentrated H2SO4, mixtures of these acids, ammonium fluoride, molten alkali bisulfates or pyrosulfates, and by concentrated HCl, especially when under pressure. All these reagents have been used to dissolve alumina in analytical procedures. Sintered alumina dissolves in concentrated H2SO4 faster than some high alumina porcelains containing clay binders (85% Al2O3). Karpacheva and Rozen found that the densest sintered alumina reacts with heavy water, H218O, the following reaction rates, as percent reaction, within 80 minutes, were observed. Temp C Rate (% 80 min) 200 20 400 30 600 50 900 97

Hot water solutions of the free alkali hydroxides and carbonates cause perceptible reaction, the rate being correspondingly faster and higher temperatures and under pressure. 2.5.2 Reaction of Chemical Elements with Alumina Some reactions of interest between the chemical elements and alumina are shown in the Table 9. Reaction of molten alumina with some refractory metals and allots are included. These are presented in alphabetical order, in general, for convenience of reference. Detailed information is given with references in Appendix A related to element interactions with alumina. For references see Appendix A. 23

Table 9: Interest Between The Chemical Elements And Alumina Aggressively attraction Attraction Molten Lithium Potassium Carbon Florine Copper BaO SRO Bishmut Cerium Chromium Iron Manganse Molybdenum Nickel Niobium Palladium Platinum Tantalum Titanium Zirconium Tungusten Sodium Antimony Arsenic Beryllium Chlorine Cobalt Gallium Lead Mercury Nitrogen Phosphorus Silver Sulfur Selenium Tellurium Tin Uranium

No Attraction

Limited attraction

1.8. COLLOIDAL PROPERTIES OF ALUMINA Surface of pure alumina is defined alkaline. By electrophoretic methods Fricke and Keefer (1949) determined the isoelectric point of zero point charge (zpc) of gamma alumina to be at 9.0 pH, that of amorphous Al(OH)3 at pH 9.4, that gibbsite at pH 9.20, and that of bohemite at pH 9.40 to 9.45. The potential determining ions were considered to be H+ and OH-, which enter into electrochemical reaction at the surface in the case of aluminum oxide. The essential part of the surface reaction schematically is as follows:

O (+) + H 2O !## OH !## O($) + H 2O " "

Positive Surface Uncharged surface Negative Surface Isoelectric point of hydrated alumina is listed in Table 10:

H H

H 3O+

OH $

24


Table 10 : isoelectric point of hydrated alumina (OConor 1956) Aging Conditions Precipitation Excess Alkali Equivalent Alkali Deficent Alkali Fresh Preparation Rapid Aging by Heating Suspension Aged 4 Months 5.08 6.63 7.29 6.79 7.28 7.43 5.78 7.06 7.32

Zeta potential of natural corundum reported positive in water 17-20 OC, but it changes to negative on heating to 1000 OC. (OConor 1956) The interaction between the liquid and the surface can be estimated by contact angle (q) measurements. The details of the interactions between the surface and the solvents were explored by analysis of the effect the properties of the liquid on varying the equilibrium contact angle with water (!!" ) with the different anodic alumina surfaces. These values were determined from the intersection of the fits for advancing and receding angles versus contact angle hysteresis with ordinate at ! q=0, one finds the equilibrium angle !!" . For different samples results vary between 62.6 - 73.3. Results obtained Wihelmy Method. (Redon et. al., 2005)

25

3. ALUMINA MEMBRANES 3.1. INTRODUCTION In recent years there has been a growing interest in utilizing inorganic membranes to address a variety of separation problems in many industries Various inorganic membranes made from metals, inorganic polymers and ceramics have been proposed for liquid and gas separation applications. Membrane technologies play an increasingly important role in pollution prevention, resource recovery and waste treatment activities. Due in large part to cost considerations, polymeric membranes dominate these applications, however, the use of polymeric membranes in separations involving aggressive materials such as many organic solvents, acids, bases and oxidants is often limited by the tolerance of the polymeric material to extreme conditions. The interest in utilizing such membranes in separations has increased since the advent of consistent-quality, commercially available ceramic membranes with narrow pore size distributions. Ceramic membranes are noted for their excellent mechanical strength and tolerance to solvents, as well as pH, oxidation, and temperature extremes. For example, they can be used at significantly higher temperatures, have better structural stability without the problems of swelling or compaction, generally can withstand more harsh chemical environments, are not subjected to microbiological attack, and can be backflushed, steam sterilized or autoclaved. An ideal ceramic membrane must be highly selective, permeable and durable. For aqueous applications, or aqueous/organic separations it is desirable for the ceramic to be hydrophilic to maximize flow and minimize fouling. The membrane selectivity is primarily dependent upon the pore size distribution; the narrower the pores size distribution, the more selective the membrane. Mechanical integrity is, enhanced in such applications by slip-casting a relatively thin selective membrane onto a thicker, more permeable support yielding an asymmetric membrane.

26

The former forms of dense (or nonporous) membranes of palladium or its alloys, silver and zirconia have shown to be permeable only to certain gases (e.g., hydrogen and oxygen). These materials are used in sensors, electrodes and coatings. Their industrial use as a separation tool, however, is limited primarily by low permeability compared to microporous metal or ceramic membranes. Currently, microporous stainless steel, silver and ceramic membranes such as alumina, zirconia and glass are available commercially. In Table 11 the selected commercial alumina membranes are listed. Data collected from web based information. Among the ceramic membranes, alumina is gaining acceptance for liquid phase separations. Like zirconia membranes, alumina membranes for liquid phase separations came to commercial fruition from uranium isotope enrichment work. (Hsieh, Bhave, and Fleming, 1988)

Table 11 : Selected commercial Alumina Membranes Manufacturer Trade Name Application Membrane Material ! !"!! ! Al2O3 Support Material ! Al2O3 Membrane pore diameter 40-100 ! 0.2-5 m 0.2-1 m 0.2-5 m Membrane support configuration Tube and multichannel element

Alcoa / SCT

Membralox

UF

MF

! Al2O3

Norton

Ceraflo

MF

! Al2O3

! Al2O3

Tube Tube and Plate

NGK

MF

Al2O3

Al2O3

Alcan/Anotec

Anopore

UF

Al2O3

Al2O3

250 ! 0.2 m Disk

MF

Al2O3

Al2O3

Alumina membranes are constantly growing area. In the Figure 3, it can be seen that, the publication numbers are highly increasing parallel with the membrane research especially during recent years. This data is collected from sciencedirect.com.

27

Figure 3 : Number of publications on membranes and alumina membranes based on years.

In the first figure publications on alumina membranes is evaluated. Second figure evaluates the trend It can be concluded that; Second figure evaluates the trends of alumina membranes trend It can be concluded that; trends of alumina membranes are rising correlated with general trends with of membrane publications during years. of membrane publications during years.

membranes. Alumina The main reasons why alumina is still trending research topic can be explained membranes are promising by following reasons. During the last several decades, many investigations have and developing area in been focused on porous anodic alumina (PAA) due to its benefits of tunable membrane society. nanopore diameter and long-range ordered feature of the porous nanochannels in macroscopic domain. The pore diameter of the PAA nanochannel can precisely be controlled from a few nanometers to several hundreds of nanometers by applying regarding electrolyte, voltage (or current), and reaction temperature during the electrochemical anodization reaction of aluminum substrate. Moreover, reaction time provides a tunability in the thickness of the porous nanochannels from a hundred of nanometers to a hundred of micrometers. Such easy control ability of the pore diameter and the thickness makes the PAA one of the interesting materials, which are frequently being applied in nanoscience. So

28

far, the studies to utilize the PAA have been performed in a wide range of research fields such as nanomaterial design, molecular sieving, photonic and optical device, and catalysis.

3.1 PREPERATION OF ALUMINA MEMBRANES 3.2.1. Macroporous Alumina Membrane Preparation Porous ceramic membranes with pore size ranging from 0.1 to 50 mm and porosity above about 40% are used for filtration (e.g., hot gas filtration), diffusion (e.g., waste water treatment), dispersion rolls, inkpads for fingerprinting and numerous other applications. In particular, filtration is important for the petrochemical, mining and chemical industries. The anodizing of pure aluminum in various acidic electrolytes has attracted considerable attention in recent years since it opens up the possibility of preparing films with well-controlled, uniform pores from tens to several hundreds of nanometers in diameter. The basic cell structure containing cylindrical pores has long been known and methods for preparing regular pore arrangements in a hexagonal pattern with interpore distances of between 50 nm and 500 nm have been reported. The preparation of regular pore arrays typically involves electrolytic polishing and multiple anodising steps or even mechanical pre-texturing. As-prepared porous anodic alumina (PAA) membranes are amorphous to X-ray diffraction (XRD). Their chemical composition is not stoichiometric Al2O3 but incorporates a considerable quantity of anion impurities and hydroxyl groups incorporated from the electrolyte into the alumina structure or bound to the alumina surface. (Kirchner, et. al)

PAA has proved to be useful for fabricating many materials, especially as a template for the synthesis of metallic or semiconductor nanometer -scaled wires and particles. After chemical functionalizing, PAA membranes can be utilized for catalytic or optical purposes. Further, the porous film itself may be employed for filtration, gas separation or as a photonic crystal. Because it is a ceramic oxide,

29

PAA has considerable potential in high-temperature applications, although severe problems can occur when certain types of PAA membranes are heated. Macroporous alumina membranes also can be made from particles or discontinuous fibers by the use of a binder or by sintering. The sintering of ceramic particles is perhaps the simplest approach to forming a porous ceramic filter, however, the sintering of bulk ceramics is a very energy expensive process due to the high temperatures required. The pore size is controlled by the starting particle size, sintering time and temperature. This method is generally used to produce alumina microfiltration filters, which contain larger pores and supports for ultrafiltration membranes, which contain smaller pores. (Avci et. al. ) Binders are most commonly in the form of fine particle dispersions (colloid). The silica colloid is an example. After application by wet forming (as in paper making), drying and appropriate heat treatment is needed. The binder technology is central to the technology of membrane fabrication. A binder is necessary to hold the ceramic particles or fibers together to form a membrane. It must be used in a sufficient quantity in order for the membrane to have acceptable mechanical strength. However, it must not block the pores in the membrane, as is the case if it is used excessively. Thus, an effective binder should be able to bind the particles or fibers together and result in a mechanically strong membrane, even when it is used in a very small proportion. In addition, the binder must be able to withstand high temperatures, as encountered in hot gas filtration. Silica and vitreous glass are widely used binders in the refractory and ceramic industry. It is often used in the form of an aqueous dispersion. The silica colloid is particularly attractive due to the high temperature resistance of silica compared to vitreous glass and the good binder dispersion enabled by the small particle size of the silica in the colloid. The average size of the silica particles in the colloid can range from under 10 nm to over 80 nm. The silica binder is easy to use, but it tends to fill the open or continuous porosity of the filter membrane.

30

Phosphate has been utilized as a binder in the refractory industry for many years. Pirogov et al. used a phosphoric acid (H3PO4) binder in a mullite-corundum body in their study to determine the optimum content of graphite and SiC additives. Birchall et al. studied the mechanical properties of an unsintered SiC compact bonded by aluminum phosphate (AlPO4) glass. Toy and Whittemore evaluated the reactivities of several calcined aluminas with phosphoric acid and demonstrated that a glassy AlPO4 phase and aluminum metaphosphate (Al(PO3)3) are effective bonding phases. 3.2.2.Mesoporous Alumina Membranes Of the present technologies, solgel is the best method for making ceramic ultrafiltration membranes. However, the pore size is generally limited to the sizes of the ceramic precursor particles prior to sintering. For solgels, the particle size distribution is difficult to control, and they must be used immediately after preparation to avoid aggregation or precipitation. (Johns et. al) Commercially available membranes are currently ! -Al2O3 membranes are currently thermally stabilized at 650 OC, with a mean pore size equal to 5 nm. Such membranes can be used for Nanofiltration of aqueous solutions containing inorganic salts or amino acids. They can also be used for gas permeation applications. However these membranes present a rather low chemical stability in aqueous media at very high pH values. More over, their structural evolution has to be taken into account for high temperature applications. (Ayral) Sol-gel process can be applied generally; Solution must be prepared An example of Boehmite ( ! -AlOOH) sol from an inorganic precursor was prepared as in shown figure 4. Pretreatment alumina supports (synthesized or provided). Heating, or solution applications. Membrane solution contacts with support. Penetration of the solution into the voids of the support occurs by capillary action. The support was then shaken to remove any excess solution, and dried at room temperature.

31

The coated support was heated up to 600-1000 OC step by step and held remained in temperature range for stabilization. Multiple coatings were obtained by treating a previously coated filter.

It was reported the unsupported membrane was obtained by dying the sol layer:
!"#$% ! !"##$ !"#
!"#$%& !"#$%& !"#$%&

!!" ! !"##$ !"# !"#$%

!!" ! !"##$ !"# !"#$% ! !!! !! !"#!$$%&'() !"!#$%&"

!!"#! !!"#$%!"&

!!" !"#$! ! !"##$ !"#$%"& !"#$%

the sol layer was formed by dipping the layer was dried at room temperature, the drying step removes excess water, cracks formed easily during drying step.
Fig.4. Preparation procedure of boehmite sol

In order to obtain the Al2O3 membrane without crack, a method, rapid gelation processing for preparation of gel layer can be used. The sols of ! !"##$ are atomized firstly then sprayed onto a substrate. The gel layer was obtained from sols directly. No crack was observed. 32

Figure 5 is the schematic diagram of the rapid gelation processing. The sols are atomizing by a high-velocity gas, and sprayed into the substrate (glass or ceramics plate coated with cellulose acetate film. After drying, the resulting gel is placed in acetone to dissolve the cellulose acetate and to obtain the unsupported gel film). The 0.1- 200 m of thickness of the membrane can be obtained without cracking at conventional drying rate. When the thickness of membrane is higher than 200 ~ m, cracking can be prevented at controlled drying rate. The transparent boehmite gel membranes obtained by rapid gelation were fired at a temperature (>350C) to form ! -alumina membrane.
Figure 5: Schematic drawing of the rapid gelation processing, 1 - nozzle, 2 - atomizing sol and 3 - substrate.

Mesoporous -alumina membranes are formed by dip-coating a porous substrate in a Boehmite (-AlOOH) precursor sol, will be treated by heat and sintering steps. The quality and properties of the membrane depend on the dispersion rheology and quality of the Boehmite sol and the dip-coating process as such. A high quality Boehmite sol is prepared by hydrolysis and condensation of an aluminum alkoxide: ATSB ([C4H9O]3Al), followed by one or more purification

33

steps. Five steps in total are necessary for a single layer mesoporous -alumina membrane preparation: 1. Boehmite sol preparation 2. Boehmite sol purification 3. Dip-coating sol preparation 4. Dip-coating procedure 5. Heat treatment/sintering In the Figure 6 an example of dip coating process is shown.
Figure 6: Filter ring assembly for mesh filtering: top filter ring (left), bottom filter ring (middle) and assembly diagram with nylon mesh (right). The thickness of the top filter clamp ring can be increased to increase the filtration volume. Currently it can hold ~10 ml.

In laboratory scale experiments to the synthesis of mesoporous alumina membranes using slip-casting technique has been tried. Catalytic membranes with precise active layer width and location were prepared by sequential slip casting of Pt/Al and alumina sols. Different distributions of catalyst within the membrane can be easily obtained by varying the thickness of the slip-cast layers as well as their arrangement. The layer widths are controlled by the slip-casting parameters, including slip-casting time, alumina content and particle size. The catalyst loading and dispersion of the active layer can be controlled precisely. The effects of sintering temperature (600-1200C) on the membrane's pore

34

structure, morphology, phase structure, surface area and gas permeability are observed. ( Yeung et. al; 1996) 3.2.3 Microporous Alumina Membranes In the slipcasting method, a porous support is usually made first by conventional ceramic processing techniques to provide rigid structure with relatively large pore size for slip deposition. Since particle size directly related with pore size, the slip used, as the membrane precursor needs to contain well-dispersed particles of uniform size. Depending on the desired pore size of the membrane, the membrane precursor particles may be prepared by precipitation, classification, sol-gel method, etc. In the sol-gel techniques, ultrafine particles of a few nm in diameter can be prepared by polycondensation or redox reactions of aluminum salts or hydrolysis and condensation of aluminum alkoxides. After treatment with a peptizing agent such as an acid and optionally with a viscosity modifier, the slip is deposited on the porous support by the dipping or slipcasting procedure. This procedure for filtration based on capillary pressure drop created by the contact of the slip with the support. (Leenaars, Burggraaf 1985) This pressure drop forces the dispersion medium (e.g water to flow into the dry pores of the support while the slip particles are retained and concentrated at the surface forming a thin membrane. The membrane precursor is then dried and calcined to provide the required pore size for specific applications and the needed bonding between the membrane and the support. These steps must be done with a great sensibility to avoid cracks, which can occur due to shrinkage upon drying and/or calcining. (Hsieh, Bhave, and Fleming, 1988) The slipcasting method is commonly used for making commercial alumina membranes. Method can be replied to organize intermediate layers between thin, permeable, selective membrane and the thick, porous support, which provides the needed mechanical strength. T Thus, alumina membranes, like many other porous inorganic membranes, are composite in nature. The cross-

35

section of a multi-layered alumina membrane composite can be exemplified by the scanning electron micrograph of Figure 7. In general, the selective membrane layer has a thickness of 2-10 !m. This thickness is a trade-off between high flux and mechanical stability. The intermediate layers are generally 10-20, !m thick. The bulk support constitutes the majority of the pore volume and thickness. If the membrane precursor particles are so fine that they penetrate the relatively large pores in the support, the permeability of the membrane/sup port composite will decrease due to partial or complete blockage of the pores in the support. To prevent fine particles from penetrating into the support, one or more intermediate layers of graded particle size are used.
Figure 7: Cross-sectional scanning electron micrograph of a three-layered alumina membrane/support composite (pore diameter of 0.2, 0.8 and 10 m, respectively, in each layer)

The ability to consistently produce high quality alumina membranes on a commercial scale has been the key to wider acceptance of ceramic membranes as a separation tool. The surface morphology of a high quality alumina membrane, shown in Figure 8, displays uniform particle size.

36

Figure 8: Top-view scanning electron micrograph of an alumina membrane (pore diameter of 0.2 !m)

Typically, membranes and supports have a porosity of 30-60% by volume. This range provides a good compromise between permeability and strength. Membrane science has a leading role in innovative processes and is considered one of the main strategic axes of research activities in all developed countries. Advanced technology programs in the USA or Japan involve them, and with an annual growth rate of 10 to 20 %, and a total world market above 10 billion Euros around 2010, membranes are likely to become more and more important in the future. Besides flat membranes as motivated by literature, anisotropic membranes consist of an extremely thin surface layer supported on a much thicker, porous substructure. The surface layer and its substructure may be formed in a single operation or separately. Multi layer membranes and their production will be motivated in next chapter.

37

4.DESIGN OF THE MEMBRANE MODULES The separation science and technology using all the kinds of different membranes is highly influenced by the membrane module configuration as far as the efficiency of a membrane process depends on the design of its module. The cost reduction of membrane module has led to the commercialization of membrane process some decades before depends on the sort of application and the different materials used. In practice, the commercial available membrane modules are assembled in units consisting usually of several membranes, sometimes from many thousands of them.

4.1 DIFFERENT TYPES OF MODULES Currently, four main types of membranes modules are in the market: (a) planar, (b) spiral, (c) tubular (d) hollow fibers modules and (e) honeycomb. The modules are closely related to the geometry of the membranes and therefore of the materials. Each commercial module is consisted of a number of membranes, placed in different ways. The structure of membrane can be (a) symmetric, (b) anisotropic and (c) composite and according to the sort of use, also two main categories can be distinguished: (a) vibrating membrane and (b) submersible membranes. Besides, in terms of geometry, two types of different geometry can exist: (a) planar and (b) tubular [Remigy, 2007]. From these basic configurations, many other secondary modes can also occur such as vibrating hollow fiber modules and many others. In Table 12, main parameters of membrane modules are gathered. The Table 1 is presented in Filtration Membranaire published by CNRS in 2007 and it is only representative for actual membrane technology trends. All the information concerning membrane technology needs frequent updating because membrane engineering is a state-of-the-art research and market field.

38


Table 12 Comparison of performances for different modules and membranes [Remigy, 2007] Different perfomance of membranes modules Geometry Cost of inventissement (U$$/m2, 2000) Energy cost Hydraulic diameter Comptability Membranes replacement Pretreatment Moyen 1 to 5 100 to 400 Membrane by membrane Moyen Medium 0,8 to 1,2 300 to 1000 Entire module Medium Important 12 to 20 10 to 300 Tube by tube Low Low 0,1 to 1 1000 to 15000 Entire module Medium (entire filtration internal/exte rnal) Medium to low (external/int ernal) Cleaning Material Good All the materials Difficult to medium Polymers All the materials Polymer Excellent Medium Planar 50 to 200 Spiral 5 to 100 Tubular 50 to 200 Hollow fibers 5 to 20

4.1.1 Alumina membrane modules Alumina membranes characterized as inorganic, ceramic membranes are usually fabricated in tubular or less often planar module, composite or anisotropic, in some cases immerged, monolith, disk mode. On the other side, vibrant or immerged modules in hollow fibers and planar or spiral geometry are commonly polymer (organic) membranes modes. The most recent trend appeared in alumina membrane market is honeycomb module which combines some of the most important advantages of previous modes.

39

Planar module: In general, membranes of planar module are less met in current market trends and represent either older membrane systems either pilot or patent scale applications. They are based on conventional filter press design and membrane feed spacers and product spacers are layered together between two end plates. The most common uses of plate and frame modules concerns electrochemical applications as ion exchange, electrodialysis, pervaporation systems, electrosensors and membrane electrode assembly. These applications concern all the different kinds of membranes as metal or polymer membranes. Figure 9 a planar aluminium module is given as a patent for an improved proton exchange membrane. The membrane could be used in membrane module unit as electrochemical cells having internal passages parallel to the membrane surface. The passages in the membrane extend from one edge to another and allow fluid flow through the membrane and give access directly to the membrane. The invention, awarded by NASA in 1997, is related to applications in as electrode assemblies for fuel cells, electrolyzes, electrochemical hydrogen and oxygen pumps, and related devices [Gonzalez-Martin et al., 1997].
Figure 9 Membrane electrode assembly containing internal passages [Gonzalez-Martin et al., 1997]

Tubular module: Alumina membranes most of the times have tubular mode which is composed of a number of tubular membranes with an internal relatively

40

small diameter. They may be unique tubes assembled in a module using joints or monolithic composed of several tubes (Figure 10 and 11). The selective part is located inside the tube or tubes. Feed is made at one end of the module and the fluid flows inside the tubes or, the retentate springs at the other end. The permeate pass through the body of the tube or out of the monolith to the external part.
Figure 10 Schematic side-view of membrane module consisting of multi-channel elements

Figure 11 Cross-section of a monolithic multi-channel membrane element [Hsieh et al., 1998]

In general, the main advantages and disadvantages of tubular noticed in the literature are because of ceramic material properties, the most common substance for which they are made of. (a) Advantages:

41

No pre-treatment, accepts slurries in large particles (about one-tenth the diameter of the tube) Easy to clean, use of mechanical cleaning Very low compactness of the order of 10 to 300 m2/m3 Relatively high investment Energy costs can be significant, due to a high rate of circulation Fragility in the case of ceramic membranes

(b) Disadvantages:

As an example for diameter difference, in laboratory scale of Stoitsas et al. (2005) research work, alumina and silica membrane specimens for gas separation are of tubular geometry with a length 340103m, an internal diameter of 8103m and an external diameter of 14103m when the membrane surface per specimen is 8.5103m2. In this case, the ceramic membrane is an asymmetric 4- layer system. The first layer operates as a support, the third as a microltration layer (pore sizes of 100 or 200nm depending on the ring temperature) and the fourth as an ultraltration layer with a pore size of 35nm. The second layer (pore size of 500nm) serves to bridge the gap between the macroporous support and the microltration layer. Honeycomb: Honeycomb pore size membrane is part of the new generation membranes for microfiltration and ultrafiltration applications. It is available for ceramic membranes either of alumina or carbon supports. The term honeycomb concerns the microstructure of porous in porous layer and because of this geometry, the membrane is suitable for special applications. 4.1.2. Commercialized modules of membrane alumina Tubular mode The figure below (Figure 12) is from Veolia Water Solutions & Technologies (VWS) manual for new ceramic membrane modules.

42

Figure 12 Large diameter monolith membrane element with permeate conduits

The industry utilizing the CeraMem technology provides a variety of inorganic microfiltration (MF) and ultrafiltration (UF) membranes. In CeraMem technology, typical multi-channel ceramic membrane has multiple parallel passageways that run from a feed inlet end face to an opposing outlet end face. The surfaces of the passageways are coated with permselective membrane. A feed stream is introduced under pressure at the inlet end face, flows through the passageways over the membrane, and is withdrawn at the downstream end face as retentate. Permeate flux passes through the membrane flows into the porous monolith material. Under an applied pressure, the combined permeate from all the passageways flows through the porous monolith support to the periphery of the monolith and is removed at the monolith exterior surface [veoliawaterst.com/ceramem]. Honeycomb module The Anopore inorganic membrane (Anodisc) in honeycomb module is a novel material with precise, no deformable honeycomb pore structure (Figure 13) with no lateral crossovers between individual pores that filters at precisely the stated cut-off, allowing no larger sized particles to pass through the membrane.

43

Figure 13 AnoporeTM alumina membrane with honeycomb pore size distribution [whatman.com]

The AnoporeTM inorganic membrane is composed of a high purity alumina matrix that is manufactured electrochemically [whatman.com]. This kind of membrane shows a high thermal permeability and selectivity and it can be applied for a wide range of special applications as pharmaceutical or laboratory ultrafiltration process. Generally, commercial alumina membranes have an asymmetric structure with membrane deposition on the inside surface of a tube (or channel). The diameter of module depends on the kind of application. Large scale units can be used eg. in wastewater treatment plants and small scale units can be used eg. in gas separation. A typical industrial installation will have several of these modules arranged in series and/or parallel configuration [Sondhi et al., 2007].

44

4.2. SEPARATION CHARACTERISTICS FOR ALUMINA MEMBRANES For future looking through membrane uses it is important to summarize the most crucial separation characteristics as crossflow mode, flux or rejection properties. Crossflow Crossflow filtration is a pressure-driven separation process in which a stream flows parallel to the filter surface. For microfiltration applications such a parallel flow creates shear forces and/or turbulence to sweep away particulates and prevent blinding of the filter surface. The feed flows past the inside surface of the tube (or channel) and the permeate (filtrate) flows through the porous support to the shell side of the module. This is generally true for all crossflow membrane modules such as tubular, hollow fiber, spiral wound, or plate and frame modules. Inorganic membranes used in commercial large-volume applications are, however, available only in single tubes or multi-channel elements as described earlier. Such membrane-based separation devices have to satisfy certain performance cri-membranes used in commercial large-volume applications are, however, available only in single tubes or multi-channel elements as described earlier. Such membrane-based separation devices have to satisfy certain performance criteria to justify a cost-effective process as flux or filtration rate, long term flux stability and separation or rejection properties of the membrane. Flux This is one of the most important performance criteria for cost effective membrane technology, particularly for large-scale separations. In membrane separation processes the (permeate) flux may be influenced by such factors as crossflow velocity, transmembrane pressure differential, temperature and feed characteristics.

45

Crossflow velocity This is the average rate at which the process fluid flows parallel to and across the membrane surface. Due to continuous removal of filtrate (permeate) through the membrane, the crossflow velocity is somewhat higher at the inlet than at the outlet of the membrane element. For many particulate filtration applications (e .g., MF) it is generally recommended to be in the velocity range of 1-5 m/sec depending on process stream properties such as viscosity, particulate loading, etc ., and on constraints imposed by pressure drop limitation In MF with an alumina membrane, for instance, a crossflow velocity in the range 3-5 m/sec is recommended. An increase in crossflow velocity generally results in an increase in flux, since at higher shear rates the removal of particulates at the membrane surface is more effective. A higher crossflow velocity, however, results in a substantially higher tube side pressure drop which may pose a problem to the membrane material. For most polymeric hollow fibers, the allowable transmembrane pressure difference is limited to about 30 psig, since the typical fiber burst pressure is only about 50-60 psig. In contrast, alumina membranes do not suffer from this limitation and usually have a pressure rating of 250 psig or higher. In Figure 14, a simplified circuit of crossflow membrane is designed, with feed and retentate and pressure application points illustrated.
Figure 14: Simplified schematic of crossflow membrane filtration [Hsieh et al., 1998]

46

Feed It has been found that the flux behavior for pure water (e.g., filtered tap or deionized water) through a single tube is similar to that through a multi-channel alumina membrane element (such as Membralox). However, for scale-up studies, alumina membranes with tube bundles or multi- channel monolithic elements should be used over single tubes to obtain hydrodynamic and fluid management data. Flux stability Critical factor determining process and economic viability of separation applications with membranes is also flux stability. In MF and OF applications, flux decay can be a serious problem. Flux decay is usually a direct result of an increase in the hydraulic resistance of the membrane due to fouling, eg. excessive accumulation of debris/particulates at the membrane surface or in the pores. The phenomenon of membrane fouling is only partially understood. In general, fouling is more severe with symmetric membranes. The majority of alumina membranes have an asymmetric pore structure with a pore size variation across the entire thickness of membrane/support composite. Such asymmetric membranes are thus expected to be less susceptible to internal pore fouling. Many liquid streams contain extraneous matter in the form of small particles, macromolecules, etc. During the filtration operation, these substances are often concentrated near the membrane surface and can precipitate or agglomerate on the surface as well as in the pores. Flux decay was observed in the filtration of tap water with a single-tube alumina membrane when the backflushing technique for cleaning was not employed. This may be due to membrane fouling by trace quantities of particulate and colloidal matter, since prefiltered water showed a higher flux in comparison to unfiltered tap water. A stable flux for extended operation can be obtained if flux decay due to fouling can be controlled.

47

Rejection characteristics One of the important parameters determining membrane rejection characteristics is the mean pore diameter of MF and OF membranes having narrow pore size distributions. In general, the smaller the pore diameter (i.e., the tighter the membrane), the higher will be the rejection coefficient to particulates/ solutes of greater nominal size. However, as pore size decreases, permeability to liquid also decreases rapidly.

48

5. APPLICATIONS 5.1. CERAMIC MEMBRANES There are a variety of industrially important liquid or gas phase systems where the general membrane technology has been successfully applied for separation purposes. Such membrane processes as MF, UF, and reverse osmosis (RO) using polymeric membranes have been in commercial practice for over two decades. However, in recent years ceramic membranes are also being considered for crossflow MF and OF applications and also gas separation application [Laitinen, 2004]. Ceramic membranes are usually composed by metal oxides as aluminum, titanium or zirconium oxide as they are or mixed. They are formed from different layers of particular length and the support is consisted of porous - alumina in tubes unit. The active layer is consisted from alumina (Al2O3, or ), from zirconium (ZrO2) or titanium oxide (TiO2). The main advantages or ceramic membranes can be listed as: sufficient resistance at high temperature ( to 300 C) chemical resistance pH range from 0 to 14 compatibility of organic solvents and ionizing radiation support of sterilization

Organic membranes are hardly resistant to solvents eg. polymeric membranes including PTFE used in wastewater treatment process and on the other side inorganic membranes are more readily used for such kind of application [Remigy et al., 2007]. In applications of industrial wastewater treatment as solvent recycling from polluted solution, conventional reclamation technologies, such as distillation and standard filtration, suffer significant limitations in terms of technical viability, cost, and user friendliness when ceramic membrane technology performs important advantages, as listed in Table 13 [Ciora et al., 2003].

49

Table 13 Advantages of ceramic membrane technology for the recovery of spent high flash solvents [Ciora et al., 2003] M&P Ceramic Membrane Technology Advantage 1. Excellent solvent resistance 2. Excellent recoverd product quality 3. Low temperature operation 4. Good product recovery ratios 5. No additional waste disposal problem 6. Low tech 7. Implemented on small scale Technology is easily implemented. No special training required. Minimal maintenance etc. Most high flash solvent waste is highly segmented with numerous small-scale generators of waste solvent Waste volume necessary for disposal is <10% of original volume >90% solvent recovery can be aschieved No thermal degradation of solvent Finished product quality similar to virgin material Comments Can be used to treat entire range of high flash solvents

At the same moment, considerable disadvantages often limit ceramic membranes in pilot or small scale applications and render polymeric membranes as the most widely applied membranes. Ceramic membranes are related to high cost of preparation, fragility and high cost of installation and as also limited variety of modules (only tubular and more rarely planar). In addition, some ceramic active layers are sensitive to basic attacks when they are deemed to be generally very resistant to alkaline pH [Remigy et al., 2007]. Inorganic membranes can be classified according to the principal component they consist of apart from other layers or additives to carbon, metal and composite membranes [Remigy et al., 2007]. Analytically, the most important applications of these categories are gathered below: 50

Carbon

If the membranes are absolutely consisted of carbon they are used for gas separation. For the filtration field, carbon membranes comprise a carbon porous layer on which a metal oxide layer is deposited (zirconium or alumina). A wide variety of metals can be used for the manufacture of filtration membranes. Two metals, however, stand out: steel (including stainless steel) and aluminum. The stainless steel membranes are composed of sintered particles while those of aluminum are formed of a film of anodized aluminum. The metal membranes are exclusively microfiltration membranes or high UF with a pore size of not less than 20 nm. In Table 13, the most important advantages and drawbacks are revised for the different kinds of materials. Note that the comparisons in terms of thermal and chemical resistance ignore the other materials in the manufacture of modules (seals, adhesives). Combination of different materials Metals

51

5.2. ALUMINA MEMBRANES APPLICATIONS In the recent years of membrane technology development, alumina ceramic or composite membranes are of great interest for a wide range of applications and new structures compositions. They offer a largely independent control of pore size, pore topology, pore orientation, molecular functionalization and pore wall composition. A developing research field focuses on nanosize of porous alumina membranes creating a potential for molecular separators, materials for the inclusion of conducting or semiconducting nanostructures. The use of porous ceramic membranes often offers a solution for aggressive environments as corrosive solutions or high temperature process [Levanen et al., 2004]. The following sections will give more information on applications or possible application areas of alumina ceramic membranes divided in two main categories: 1. Liquid phase separation 2. Gas phase separation Remarkable is that in the analysis of applications below, alumina is one the components of membranes. In many cases composite membranes are studied with many different layers and at least one alumina coat. 5.2.1. Liquid phase separation applications Most of the commercial liquid phase applications for ceramic membranes are either for microfiltration or ultrafiltration. The first large-scale commercial success of ceramic membranes has been in the food and beverage industries. However, significant applications are found also in other areas, such as biotechnology, pharmaceutical, petrochemical, and other process industries as well as in environmental control. In recent years also ceramic nanofiltration membranes have been developed. Microfiltration is mostly applied in cases

52

where liquid streams contain particulates, and ultrafiltration is used when smaller molecules are removed. Some of the important liquid phase applications where alumina membranes are employed are: treatment of industrial and municipal water, sterilization of liquids in the pharmaceutical industry, clarification and sterilization of beverages, e .g. wine, beer and fruit juices, cell harvesting and sterilization, cheese-making by ultrafiltration and wastewater treatment. A number of studies on crossflow filtration can be found in the recent literature [Laitinen et al., 2004]. Although many of them deal specifically with polymeric membranes, the principles of crossflow filtration, in general, can also be applied to membranes made of inorganic materials such as alumina. In all these application areas the ceramic membranes have to compete with the polymeric membranes which are becoming more and more stable and which have an advantage of much lower prices. The use of ceramic membranes can only be justified in cases, where they give much better performance, or in cases, where there are no suitable polymeric membranes available [Hsieh et al., 1998]. In the following part, the applications of alumina membranes in gas phase separation process are gathered in Table 14. i. Food processing Generally, the major advantages of ceramic membranes in the area of food and beverages are their resistance to alkaline cleaning solutions, their stability in steam sterilization, better capability to withstand higher operating pressures, and a consistent pore size. In crossflow microfiltration applications, the major advantage of the ceramic membranes has been the possibility to use uniform transmembrane pressure, which for example in dairy industry has enabled high fluxes and low fouling. The market share of inorganic membranes in the food and dairy industry was about 10% at 1990s and tomorrow this percentage is 53


Table 14 Liquid phase separation process of alumina membranes and composite alumina membranes Porous alumina membrane appications Liquid phase separation applications Food processing Clarification of natural fruit juices Filtration of sugar cane juice Alcoholic beverages Environmental applications fruit juice industry (orange, apple, grape etc.) sugar concenration breweries, wineries

Petrochemical industry wastewater treatment petrochemical industry, oil recovery Pulp and paper industry wastwater treatment Municipal wastewater treatment Activated sludge treatment Nuclear industry wastewater treatment Fluoride removal Arsenic removal from water resources Arsenic and fluoride removal Chromium and arsenic removal Biotechnology and Pharmaceutical applications Fermentation broths Fungal cells separation Penicillin recovery Lysozyme ultrafiltration recovery of valuable antibiotics diafiltrationm polysaccharide production recovery of a water-soluble water reuse in production line municipal wastewater treatment plants, water reuse depollution of wastewater sludge, compost municipal wastewater, chemical industry wastewater depollution of water sources and groundwater depollution of water sources and groundwater metal industry wastewater, valuable metals recovery

supposed to be doubled [Bhave et al., 1991]. The main applications can be roughly divided into (a) concentration of soluble molecules and suspended solids and (b) clarification by removing suspended solids. As already mentioned, the main usage of inorganic membranes is in the dairy industry, in which the key to success has been the invention of transmembrane pressure mode. Some studies have also been made concerning the usage of inorganic membranes for the concentration of other vegetable and animal proteins as well as for the processing of starch and sugar. 54

(a) Clarification of natural fruit juices Clarification or fruit juice such as apple, cranberry and grape is one of the most successful and widely practiced industrial applications of ceramic membranes. Ceramic membrane filtration provides a particularly attractive alternative, replacing such conventional treatments as gelatin addition, holding/decanting, diatomaceous earth, cake filtration, and polishing, with a single unit operation. Membranes produce superior clarity juice and deliver higher yields compared with conventional clarification processes. (b) Filtration of sugar cane juice In the filtration of sugar cane juice, ceramic membranes as carbon/carbon or alumina/alumina [Jiratananon et al., 1997] can be used in several different stages in the raw and refined sugar production. One interesting opportunity is in the microfiltration/ultrafiltration of clarified juice (714 Brix) and/or pre-evaporated juice (20-25 Brix) as a pretreatment prior to ion exchange or chromatographic separations. Pretreated and filtered juice is softened, evaporated and purified using ion exchange and chromatographic processes leading to a better quality refined sugar. Typical operating conditions include feed temperatures of 90100oC, a high crossflow velocity (47 m/s) and transmembrane pressures up to 5bar. The need to purchase, use and dispose of filter aids is eliminated. In some applications, the cost of equipment necessary to dewater filter-aid sludge may be comparable to the cost of the membrane system. [Sondhi et al., 2005]

(c) Alcoholic beverages In the area of alcoholic beverages, alumina membranes are used for the clarification and sterilization of the final products. Composite multilayer ceramic membranes membranes applications exist for the filtration of wine and beer [Gillot et al., 1986] as well as for sake and vinegar [Hagasewa et al., 1991]. By membrane filtration the amount of waste generated and chemicals used during the production can be considerably decreased. The use of membrane filtration instead of a conventional process is advantageous also since membrane filtration produces a sterile product and further sterilization is not necessary. For Membralox membranes these applications have existed since 1988 [Gillot et al., 1990]. In breweries as well as in wineries membrane filtration can be used to 55

replace clarification, stabilization and sterilization steps. The alumina membrane Membralox is for example used to further concentrate tank bottoms or centrifuge concentrates [Gillot et al., 1990]. ii. Environmental applications
The increasing concern for the environment and tighter legislation on emissions are increasing widely the applications of ceramic membranes. In many chemical process applications there is a growing need to treat not only the waste streams, to meet the increasingly stringent environmental regulations, but also to recover and reuse the chemicals. The nature of these process streams can vary and in some cases the process may require aggressive operating and/or cleaning conditions. Examples of such applications include the filtration of chemical solvents, dye and pigment wastewater from dye processing and colouring plants, and highly variable wastewater containing detergents, polymers and organic solvents [Sondhi et al., 2007]. Environmental applications can be roughly divided into wastewaters from various sources, as well as oily wastes and sludges. In the following part the process industry wastewaters from the chemical, textile, and pulp and paper industry are discussed as well as wastewaters from the nuclear industry and sludge from municipal wastewater dehydration. (a) Petrochemical industry wastewater The wastewater applications in the petrochemical industry include the treatment of acidic and alkaline process wastewaters from a vinyl chloride monomer manufacturing facility. Lahier and Goodboy (1993) have shown that the high content of heavy metals from acidic wastewater can be removed by using two-stage membrane filtration with alumina microporous membranes. As a pretreatment the pH of the wastewater was adjusted from 0.7 to 12. In the first filtration stage, the precipitated metal hydroxides were separated from the water and in the second filtration stage the filtered metals were concentrated to 17-20 w.w%. The water recoveries from the filtrations were 80% and 99.5%. In Lahier and Goodboy research work, alumina microfiltration membranes were evaluated for treatment of 3 aqueous streams containing heavy metals, oils, and solids at petrochemical manufacturing facilities. They have also shown that an alkaline waste stream containing dichloroethane, water, and calcium/iron hydroxides can be treated with ceramic membranes.

56

(b) Pulp and paper industry In pulp and paper industry the main interest in studying the possible applications of ceramic membranes has been in bleach plant effluent treatments. The wastewater sources in paper industry comes from disk filtration, flotation and close loop water circulation as well as the clarification of coating colour basin. In Laitinen et al. (1998) as well as in many 90s references [Petterson et al., 1988; Gaddis et al., 1995], alumina membranes are used for wastewater ultrafiltration for suspended solids (SS), COD, color and colloidal material removal. Laitinen tried flat sheet - and -alumina membranes and modified -alumina membranes and concluded in quite low COD removal is achieved (under 22%), iron and odor from color wastewater treated sufficiently and SS and colloids was almost totally remover. Fouling of membrane seems to be the most considerable problem for large scale application of the method. Great advantage of wastewater treatment of pulp industry without using chemical compounds is recycling feasibility, In some cases [Laitinen et al., 2001] existing results have showed that ultrafiltration permeate was satisfactory for continuous recycling and the pulp produced with recycling had better strength properties than when recycling was not used. Hence, toxic discharges were reduced, waste management became more effective, and the need for maintenance was reduced. (c) Municipal wastewater Ultrafiltration/microfiltration is suitable for tertiary treatment of municipal wastewater after standard treatment steps. The final product can be reused as the level of pollutants and microorganisms falls short of environmental legislation limits. In Vera et al. (1997) a study on the filtration of the secondary clarified suspension is studied using alumina microfiltration membranes of 0.14m pore size. The membrane system proved to serve as a total barrier for suspended solids, total coliform, fecal coliform and fecal streptococci. This fact together with the reductions of turbidity, COD and phosphorus removal from secondary treatment made the microfiltered water perfectly adapted to irrigation. A critical flux of 100 l/(mh) was achieved at 1bar pressure and 3 m/s crossflow velocity. Other possible application studied is ceramic membrane microfiltration of hotel wastewater in order to reuse it in secondary purposes, such as toilet flushing [Laitinen, 2004]. Nevertheless, Because of the high cost of membrane installation and performance, the wide scale application seems to be not yet necessary in this kind of municipal wastewater and activated sludge treatment.

57

(d) Activated sludge Xing et al. (2001) has studied the use of a bioreactor combined with -alumina and zirconia ceramic ultrafiltration membranes for the reclamation of urban wastewater. According to results, the treated water could be reused directly for municipal purposes, such as toilet flushing and car washing, and after softening treatment for industrial purposes such as cooling supply and process water. Removal of ammoniac nitrogen and suspended solids was product of membrane separation however COD removal efficiency was attributed more than 90% by bioreactor and no by ceramic membranes. (e) Nuclear industry In the nuclear industry, composite alumina ceramic membranes have been studied for the treatment of radioactive waste streams. AEA Technology in England has done excessive pilot plant studies with ceramic membranes and the results have shown that, for a low-level radioactive general site waste, good alpha and beta/gamma removal can be achieved. [Laitinen,2004]. Cadmium adsorption from silica-alumina membrane is simulated by Pacheso et al. (2006) from nuclear and chemical wastewater treatment. Sol-gel structured nanoparticles of silica and alumina are used for multilayer composite alumina membranes fabrication.

(f) Fluoride removal Nanofiltration (NF) is an attractive technique for reducing fluoride anions (F-) concentrations to acceptable levels in drinking water, however commercial membranes show minimal chloride of fluoride selectivity. In laboratory scale, multilayer organic membranes (polystyrene sulfonate) with porous alumina supports exhibited Cl, F and Br high selectivity, three times more than the commercial available membranes [Pagana et al., 2007]. Moreover, chloride/fluoride selectivity is essentially constant over Cl-/F- feed ratios from 1 to 60, so these separations will be viable over a range of conditions [Seong et.al, 2007].

58

(g) Arsenic removal from water resources Alumina ceramic membranes can be used for As(V) from water resources as contaminated water. For such kind of application, -Al2O3/-Al2O3 can be used. The concentration of arsenic ions decreased from 1ppm, which was originally to power in 5ppb. Therefore, the concentration of ions As (V) in output was lower than the maximum permissible limit set for drinking water [Pagana et al., 2009].In Laitinens doctorate thesis, published in 2004 from Finland Technical University, from all the membrane separation process studied applied in a wide range of wastewaters, alumina membranes assumed to be suitable for arsenic removal from bore well water if flocculation was used as a pretreatment as well as for the treatment of the stone cutting wastewater. The fluxes achieved in short-term experiments for the bore well water were promising and the optimization of operation conditions should be considered [Laitinen, 2004]. (h) Arsenic and fluoride removal In a similar recent study, Wu et al. (2011) highly ordered mesoporous aluminas and calcium-doped aluminas were synthesized. Their uoride adsorption characteristics, including adsorption isotherms, adsorption kinetics, the effect of pH and co-existing anions were investigated. These materials exhibited strong afnity to uoride ions and extremely high deuoridation capacities. The highest deuoridation capacity value reached 450 mg/g. These materials also showed superb arsenic removal ability: 1 g of mesoporous alumina was able to treat 200 kg of arsenic contaminated water with a pH value of 7, reducing the concentration of arsenate from 100 ppb to 1 ppb.

59

(i) Chromium and arsenic removal In laboratory scale, separation processes have been developed for ion Cr (III) removal from water using -alumina membranes [Pagana et al., 2000; Sklari et al., 2005]. In Pagana et al. (2000) is described how two pilotic system have been developed with concentration of chromium ions. In the conclusion of this study, the concentration of ions Cr (III) in output is assumed to be significantly lower than the maximum permissible limit set for drinking water (greek standars). By the same research group, a combined process of removing ions As (V) and Cr(III) from simulated wastewater using nanostructruded alumina membranes provided also sufficient results. The concentrations of arsenic and chromium ions decreased from 1ppm and 0.5ppm, which was originally to power in 5ppb and 3ppb, respectively. Therefore, the concentrations of ions As (V) and Cr (III) in output was lower than the maximum permissible limits set for drinking water. In the figure below (Figure 15) the pilot system is illustrated [Pagana et al., 2008].
Figure 15 Flow diagram of the Cr (III) removal process [Pagana et al., 2008]

In this work, asymmetric multilayer porous ceramic membranes are developed. Composite -Al2O3 membranes made by solgel method are prepared for chromium and aersinc ions removal from water solutions. As(V) removal is achieve d by a two stage adsorption ultraltration processes in series. Moreover, Cr(III ) removal is achieve d by an adsorptionult raltrat ion parallel

60

process (Figure 15) using membrane ultrafiltration in a pilot system (see characteristics inTable 15 ).
Table 15: Basic characteristics of membrane ultrafiltration process applied in Pagana et al., 2007 pilot system for chromium removal (Pagana et al., 2007)

Chromium removal: basic characteristics of the membrane ultrafiltration process Feed volume Final permeate volume Average permeate flux Pressure difference Membrane surface A (A=2rl) 100% 100% 60ml min-1 3.105 Nm-2 0.05m2

In general, the adsorption-ultraltration ion process using ceramic membranes may oer a low cost eective alternative arsenic and chromium purication technology basically in terms of membrane stability, applied pressure and product ux with the additional advantage of being suitable for small local and decentralized units for point-of-use. However, further experiments should be carried out in order to provide more accurate an d widely accepted conclusions [Pagana et al., 2007]. iii. Biotechnology and Pharmaceutical applications

In general framework, in biotechnology and pharmaceutical applications, the biocompatibility of the membranes is important [Shackleton et al., 1987]. In such kind of applications, the main interest has been on the filtration of fermentation broths. In biotechnology, ceramic membranes have been used for primary extraction, purification, and concentration of biomass, antibiotics, vitamins, amino acids, organic acids, enzymes, biopolymers, and biopesticides [Cueille et al., 1990]. There are also applications in the treatment of vaccines, recombinant proteins, cell cultures, and monoclonal antibodies as well as in continuous fermentation. The alumina membranes have several features that indicate the practical benefit in analytical and diagnostic separations. They are transparent, allowing to view the retained materials from either side of the membrane.

61

The capillary pore structure allows no lateral diffusion of liquids along the membrane. Nanoporous alumina membranes are usually prepared electrochemically in sulfuric or phosphoric acid bath. After preparation they are used for microfiltration of biological media, such as human red blood cells and bovine serum albumin [Laitinen, 2004]. (a) Fermentation broths Alumina membranes are used for fermentation broth clarification at numerous installations worldwide, successfully competing with other technologies such as polymeric membranes, vacuum filtration and centrifugation. These systems are often used for the recovery of valuable antibiotics from dilute solutions. Typical operating conditions include feed temperatures of 3060C, high crossflow velocity (58 m/s) and trans-membrane pressures up to 6bar [Laitinen, 2004]. (b) Fungal cells separation Haarstrick et al. [1990] have studied microfiltration using alumina microporous membrane in the separation of fungal cells and the purification of the produced polysaccharide. They concluded that crossflow microfiltration could be an alternative to conventional separation including purification processes and suggested that a diafiltration mode should be used in order to avoid the problems arising from the high viscosity of the broths. (c) Penicillin recovery An alumina microfiltration membrane has been used in laboratory scale [Adikane et al., 2001] to optimize penicillin G recovery from fermentation broth. Reduction over 40% has been achieved in processing time by optimizing the crossflow velocity. Moreover, it is showed that the operating fluxes could be regenerated and a 98% recovery of penicillin G could be obtained over 12 cycles.

62

(d) Lysozyme ultrafiltration Group of Baudry et al. (2001) has studied the use of modified ceramic membranes by alumina and zirconium oxide used for lysozyme and lactoferrin ultrafiltration. It is concluded that as other studies had prouved, adsorbed protein on the membrane surface was able to highlight UF mechanism. The enhanced selectivity was observed when the protein to be retained had additional interactions with the membrane surface. Conrad et al. (1998) studied microfiltration for the recovery of a water-soluble, chiral compound of therapeutic interest from a bioconversion broth of whole cells and soybean oil. It is assumed that that it was possible to use microfiltration to harvest an aqueous product stream from the oil-water bioconversion broth. 5.2.2. Gas phase separation The use of membranes in gas separations has grown at a very rapid pace in recent times. One particularly interesting application of gas separation with membranes is the removal of dilute heavy organics from light gas streams such as the removal of solvents from the exhaust of different process industries [Javaid et al., 2005]. Gas mixtures can be separated by either dense or porous membranes. Microporous inorganic membranes have often been modied using in many cases different types of alumina and studied as possible materials for achieving solubility based separation. The dense ceramic membranes are mostly impermeable to all other gases, giving extremely high selectivity towards oxygen or hydrogen. Oxygen permeation through a dense ceramic membrane is due to a large number of oxygen vacancies that are generated by doping and the electron holes produced by the defect reaction exist in the solid electrolyte. Under a gradient of oxygen partial pressure imposed on the membrane at a high temperature, the oxide ions are transported along with holes from the high partial pressure side to the low partial pressure side. Similarly, when hydrogen is exposed to a mixed proton conducting membrane, it may be transferred through the membrane under a hydrogen partial pressure gradient. Again, apart from the membrane bulk diffusion, the

63

surface reactions are also important and need to be taken into consideration for the hydrogen permeation. In microporous ceramic membranes, the gas permeation behaviour may be dominated by Knudsen diffusion, surface diffusion, multilayer diffusion, capillary condensation or molecular sieving (i.e. congurational diffusion) and is strongly dependent on the pore size and pore size distribution of the membrane, operating temperature and pressure, and the nature of the membrane and the permeating molecules [Laitinen, 2004]. Gas separation using porous ceramic membranes is one of the important research topics [Wiley, 2007]. Recently, most of the development membrane separation activities have been focused on gas separations, particularly as ionic conductors for oxygen transport and as molecular sieve membranes for hydrogen separations. Their use in environmental applications has been very limited due to cost considerations, although they offer several unique advantages in this area, such as chemical and thermal stability and rugged structural stability [Sondhi et al., 2007].
Table 16 : Gas separation by porous alumina membranes [Kaldis et al., 2004] Pore diameter () ~100 ~100 ~100 1020 200-400 ~10 ~10 ~10 ~10 ~30 ~30 ~30 100-200 100-200 Operating temperature (oC) 25-75 25-75 25-75 800 10 31-77 65-70 78-84 82-88 65-70 78-89 82-88 196-197 196-197 Applied pressure differenece (kPa) 3.03 - 16.2 3.03 - 16.2 3.03 - 16.2 343 152-252 4.04-39.4 23.2-106 23.2-106 He/N2 He/C2H6 He/Kr H2/H2S H2/H2S H2O/air H2O/CH3OH H2O/C2H5OH H2O/(CH3)2CHOH H2O/CH3OH H2O/C2H5OH H2O/(CH3)2CHOH H2/He H2/CO Gas mixture

64

100-200 100-200 100-200 100-200 100-200 100-200 100-200 100-200 100-200 100-200 300-400

196-197 196-197 196-197 196-197 196-197 196-197 196-197 20 20 20 65

23.2-106 23.2-106 23.2-106 23.2-106 23.2-106 23.2-106 23.2-106 13.1-92.9

He/N2 H2/CO2 H2/CO He/N2 H2/CO2 H2/C2H6 H2/C3H8 H2/He He/N2 H2/CO2
235U/238U

Sustainable membranes in high temperatures are strongly needed for energy production application as biomass burning (organic waste gasification, coal gasification etc.). Ceramic membranes are appropriate for such kind of applications because of their high thermal stability and corrosion resistance. In installations as coal gasification plants, gas separation is needed for emission treatment after combustion or other processes. Usually, carbon dioxide is the substance which must be isolated in order to fit environmental requirements. Membrane technology is one of the promising tools in these applications and many research works has been realized in recent years. In the following part, the applications of alumina membranes in gas phase separation process are gathered in Table 17.
Table 17 : Gas phase separation process of alumina membranes and composite alumina membranes Porous alumina membrane appications Gas phase separation applications Carbon dioxide capture CO2/N2 separation H2/CO2 separation Hydrocarbons separation Acetone recovery Propane separation Catalytic reactors VOCs oxidation Methane to ethane reaction Industrial/Scientific applications Nox control, gas phase streams from energy plants gazification, combustion emissions paint, ink, oil industry natural gas processing, petroleum refining air pollution, VOCs recovery gas treatment

65

 i. Carbon dioxide capture

There is growing consensus among the scientific community that the rising atmospheric levels of CO2 as a result of human activities (e.g., fossil fuel burning, cement production). Membrane processes appear to be an attractive option to carry out gas separations in terms of their lower environmental impact and energy costs compared to more conventional separation technologies (e.g., distillation, absorption, adsorption, crystallisation). On the guidance of a recent study recently published by Lito et al. (2011), different adsorption models have been screened to account for CO2 and N2 adsorption in MFI-alumina samples [C.- H. Nicolas et al., 2011]. (a) CO2/N2 separation In Sang H. Hyun et al., (1995) top layers of -Al2O3 composite membranes have been modified for improving the separation factor of CO2 to N2. The separation factor through the TiO2 supported -Al2O3membrane was found to be fairly enhanced by silane coupling, but in case of the -Al2O3 supported membrane was not. The CO2/N2 separation factor through the modified -Al2O3 / TiO2 composite membrane is 1.7 at 90C. The separation factor is proportional to the CO2 concentration in the gas mixture, and the modified membrane is stable up to 100C. The main mechanism of the CO2 transport through the modified -Al2O3 layer is known to be a surface diffusion. (b) H2/CO2 separation
Coal gasification is called any process of converting coal into gas for many different uses as illuminating, heating etc. In the figure above (

Figure ) an integrated gasification combined cycle is shown (IGCC). In IGCC (a) air separation and gasification, (b) gas cleaning, (c) gas conditioning and (d) gas separation are combined for total air treatment after coal gasification are aligned (Figure 16). Using conventional technology, it is possible to build the IGCC plant

66

above, technologies such as cryogenics, solvent extraction, adsorbents in pressure or temperature swing adsorption, and low temperature polymeric membranes. Due to compliance with sorption mechanisms, these technologies operate best at low temperatures (<50oC). The problem here is that gas separation follows downstream from gas conditioning. The gasification of coal predominantly produces syngas (CO and H2) with some remaining hydrocarbons, CO2 and water. Hence, the syngas requires further processing.
Figure 16: A conventional air treatment process with carbon capture [Diniz et al., 2007]

The syngas requires further processing through the WGS reaction (see equation below), in order to maximise H2 production. CO + H2O CO2 + H2 H =-41. 2 kJ.mol-1 The WGS reaction is exothermic and the conversion is limited by thermodynamic equilibrium as the conversion to H2 and CO2 decreases with increasing temperature. The reaction is therefore carried out in two stages, in high (350- 400oC) and low temperature (250-300oC) shift reactors with interstage cooling. In addition, further cooling to <50oC is required to reduce the temperature to the temperature acceptable by conventional gas separation technologies. Cooling large stream of hot gases is capital intensive, and incurs a loss of power production; furthermore, these processes are likely to deliver CO2 at reduced pressures, which will have to be re-compressed to >100atm for transportation and storage. Hence, conventional processes will attract large energy penalties. In order to reduce efficiency losses, an alternative is to separate gases at higher temperatures. In this case, inorganic membranes derived from ceramics, silica, 67

metal and by further doping or alloying showed preferential H2 selectivity over CO2 at high temperatures (>200oC) [see e.g. recent work of Joe Da Costa in Australia]. These technologies can also operate in membrane reactor arrangements for the water gas shift reaction, which allow for shifting the reactions to higher conversions due to the extraction of hydrogen from the reaction chamber. The advantage here is two: 1. CO2 is kept at high pressure thus reducing requirements for CO2 compression downstream. 2. As H2 is selectivity taken, the syngas stream is reduced by up 30-35% depending on the recovery rate. Although CO2 will have to be cooled down prior to compression, the volumes are reduced thus requiring a lower cooling duty. Therefore, inorganic membranes and their incorporation in MRs are foreseen to be the technology of choice for advanced IGCC plants as depicted in Figure 17.
Figure 17: An Advanced scheme for IGCC with carbon capture [Diniz et al., 2007]

The scheme shown above (Figure 18) for gas separation and conditioning is being developed in Australia by the University of Queensland as the research provider under the R&D program directed by the Centre for Low Emission Technology (www.clet.net). The use of a higher temperature membrane allows further simplification of the gas cooling, conditioning and separation step, but this scheme requires a high temperature membrane tolerant to the syngas.

68


Figure 18: Schematic of a multi tube membrane module for H2 and CO2 separation [Diriz et al., 2007]

In Diriz et al. (2007) for high temperature gas separation membrane, commercial D-alumina tubes coated with a top -alumina layer, dip-coated with cobalt and selective silica layer. The best membrane performance delivered H2 purity in excess of 98% at 300o C. The positive energy of activation for H2 permeation coupled with the negative energy of activation for CO2 permeation of metal doped silica membranes provide a favourable fundamental property for engineering design of gas separation at higher temperatures (up to 500o C) such as those required in a coal gasification process. The process integration provided by Diriz et al. is potentially beneficial for the next generation of high temperature processing unit operations in low emissions coal gasification. 69

ii.

Hydrocarbons separation

(a) Acetone recovery Huang et al. (1997) modied alumina membranes for recovery of acetone from nitrogen by reducing the pore size to enhance multilayer diffusion and capillary condensation transport mechanisms [Huang et al. 1997]. Depending on the temperature and the feed composition, the separation factor varied from being less than 10 to as high as 200 is dominant. The modied membranes showed higher acetone permeability and higher separation factors as compared to polymeric membranes. However, the performance was strongly inuenced by temperature and feed composition. (b) Propane separation Hydrocarbons as propane or butane are usually by-products of natural gas processing and petroleum refining and they are commonly used as a fuel for engines, heating appliances etc. In Randon et al. (1995) alumina membrane are modied and gas permeation experiments were conducted using nitrogen and propane. The modied membrane exhibited signicantly high propane permeance. The modication had made the membrane hydrophobic and improved the membranes solubility-based separation characteristics. Other attempt for alumina membrane modification using different alkyl trichlorosilanes showed a signicant increase in propane/nitrogen selectivity accompanied, by a loss in permeance, after modication. In comparison to PDMS, one of the best-known polymers for solubility-based separations, the hybrid membranes exhibited equal or greater propane/nitrogen selectivity although at lower propane permeance [Javaid et al., 2005]. McCarley and Way (2001), in a similar study, modied 5nm alumina membranes with C18 trichlorosilane. They conducted both single gas and mixed gas permeation experiments at xed transmembrane pressures. The treated membrane showed a signicant increase in ideal selectivity for heavier gases (n-butane) over lighter gases (nitrogen and methane).

70

iii.

Catalytic reactors

The application of porous ceramic membranes as catalytic reactors starts in the 1980s. The driving force for this change was the possibility of integrating reaction and separation, which had already been achieved in the field of biochemical reaction engineering using polymeric membranes. These, however, were not applicable at the temperatures used in most of the processes of interest in the chemical process industry [Lafarga et al., 1998]. Since the materials used in the manufacture of ceramic membranes are also commonly used as conventional catalyst supports (as alumina), there has been a strong interest in the development of membrane reactors by researchers with previous experience on heterogeneous catalysis, who adapted many of the preparation and characterization techniques used in this field. The most important and valuable fact is that membrane itself is the catalyst [Coronas et al., 1999] and it can be used for a wide range of applications. Catalytic membranes are used to improve contact efficiency with the objective of attaining higher conversions by decreasing mass transfer resistances. Membrane reactor with catalytic surface aims to improve the contact in gasliquidsolid systems by providing a well-defined contact region: the liquid-filled pores of the catalytic zone of the membrane in close proximity to the gas interphase. This does not require a permselective membrane and avoids the problem of catalyst recovery that appears in slurry reactors.

Poorly permselective inert porous membranes such as mesoporous alumina, mesoporous glass etc. can reveal attractive for a number of dehydrogenation reactions. Inltrated MFI/alumina composite membranes have been used as O2 distributors but also as barriers for dehydrogenation reactions in MRs [Julbe et al., 2001]. (a) VOCs removal Volatile Organic Compounds (VOCs) are among the most common air pollutants emitted from chemical, petrochemical, and allied industries. VOCs are one of the main sources of photochemical reaction in the atmosphere leading to various environmental hazards and on the other hand, these VOCs have good commercial value. Hence, growing environmental awareness has put up stringent regulations to control the VOCs emissions. In such circumstances, it becomes mandatory for 71

each VOCs emitting industry or facility to opt for proper VOCs control measures. There are many techniques available to control VOCs emission (destruction based and recovery based) with many advantages and limitations [Khan et al., 2000]. One of these methods for removal and no so often for VOCs recovery is gas separation with ceramic membranes - zeolithe and alumina membranes in most of the cases. By the University of Zaragoza in Spain team and other research groups this concept has already been demonstrated eg. using hydrogenation reactions over Pt/Al2O3 catalysts in membrane module. In Figure 19, it is illustrated a flow-through membrane in which the permeation of a premixed feed stream takes place. This was the approach employed in works by Saracco et al. (1999), who used catalytically modified fly ash filters for alcohol dehydration and for the reduction of nitrogen oxides with NH3. Pina et al. (1999) used Pt/Al2O3 and perovskite-containing membranes operating in the Knudsen regime for the purification (by catalytic combustion) of air streams containing volatile organic compounds (VOCs) in low concentrations. According to this study, since in the Knudsen diffusion regime the probability of collisions between the molecules and the wall of the pores is maximised, this type of membrane would be expected to give high contact efficiency in the reaction of diluted streams, such as those commonly encountered in VOC removal. The results of Pina et al. (1999) showed that the membrane could perform very efficiently in the combustion of VOCs at low temperatures, although at the expense of a significant pressure drop.
Figure 19: Applications of membrane reactors [Coronas et al., 1999]

72

The industrial application of this type of reactor surely requires optimization of the membrane structure aimed to reducing the pressure drop. Otherwise, its use would be restricted to applications involving reaction simultaneous with gas filtration, where the pressure drop is already present. (b) Methane to ethane reaction In Lafarga D. and Santamaria J. projects published in international literature [Lafarga et al., 1994; Sebastian et al., 2009] a new reactor concept has been developed and tested for the conversion of methane into ethane, ethylene and higher hydrocarbons via oxidative coupling. The basic idea consists of providing a low oxygen concentration in the reactor in order to increase hydrocarbon selectivity. To this end, oxygen is supplied to the reacting mixture by permeation through a porous wall as a modified alumina membrane, which is capable to withstand the temperatures and pressures involved, and at the same time give adquate values of oxygen flux [Lafarga et al., 1994]. Concluding, despite the plethora of studies appeared in the open and patent literature focusing on the development of ceramic materials for gas mixture separations [Koutsonikolas et al., 2010] no many industrial processes have been commercialized. Although alumina membrane is the most used in large scale, polymer membranes are still dominating in commercial field of separation process. The lack of selectivity combined with their high unit cost most often ascribed to the support, as well as their lack of hydrothermal stability in some cases, act as key economical and technological barriers for the industrial implementation.

73

6.SUMMARY AND CONCLUSIONS The term alumina includes large number of chemical compounds: amorphous, crystalline trihydroxides, monohydroxides, transition aluminas and stable anhydrous alumina. Alumina takes attention with good thermal, mechanical and chemical stabilitiy, as a material for industrial applications. With wider applications of membrane, alumina is trending topic. Alumina pore size distributions can be controlled easily modifications on operation conditions from a hundred of nanometers to a hundred of micrometers. Besides traditional pathways like sintering and anodic oxidation, alternative methods - sol-gel methods, dip coating, rapid gelation- are introducing with alumina membrane synthesis. The optimum range that will provide a good compromise between permeability and strength is being improved also with additives. According to literature overview, alumina membranes are applied especially in the areas where the advantages of ceramic membranes can be exploited. In biotechnology and pharmaceutical industry their resistance to microbial attack and biological degradation is really very important and their good thermal and chemical stability are often the factors that have made them beneficial in wastewater treatment and gas separation applications. However, they suffer from poor steam stability, fouling and cracking and because of this alumina is often mixed with additives such as metals or oxides to improve such kind of drawbacks [Laitinen, 2004]. Modifications can also be used to increase selectivity by surface adsorption or molecular sieving effects and a wide range of research field in ceramic membrane is oriented to this direction.

74

APPENDIX A : Membrane Material Sheet

MEMBRANE MATERIAL NAME Alumina

1. INTRINSIC PROPERTIES Nature 1. 1. General physicochemical properties Bulk density at RT kg/m3 1. 2. Mechanical properties Tensile Strength (MPa) Bending Strenght Mpa Modulus of Elasticity (E) X 108 MPa Compressive Strenght Mpa Modulus of Ridity(G) X 108 MPa Hardness on the mohs scale 1. 3. Thermal properties Melting point OC Boiling Point OC 1. 4. Surface properties Zero point of charge Water contact angle SBET (m2/g) Pore Diameter (A) Pore Volume (mL/g) 1. 5. Chemical properties Operating pH range Compatibility with solvent 1 Compatibility with solvent 2 a polycrystalline alumina 94% b polycrystalline alumina 85% c polycristalline alumina 99.9% d polycristalline alumina 85.5%

High

Low

1217.406

800.925

173 a 413 c 26.8 a 3733 e 11.3 a

117 b 307 d 21.27 b 1600 b 8.67 b 9

2051.0 9.7 3530 200 f 9.45 9 g 73.3 62.6 484.34 98.3 72.23 20.49 1.27 0.54

phosphoric acid - HF well HCl- HNO3 weak e polycristalline alumina 100% f gamma alumina g bohemite

75

2. CHARACTERISTICS OF COMMERCIAL MEMBRANES BASED ON THIS MATERIAL Membrane name Memrbalox Membrane manufacturer Pall Membrane shape capillary tube, multichannel, monolithic Membrane structure symmetric Support nature ultrapure -alumina, zirconia, titania Membrane type porous Average pore size 5m Molecular weight cut-off 1-5kg/mol Water permeance at RT 300 l/(mh bar) Membrane name Anopore Membrane manufacturer Myriad Membrane shape disc Membrane structure symmetric Support nature polypropylene Membrane type porous Average pore size 0.02-0.1m Molecular weight cut-off 0,1m Water permeance at RT 1 m/m Membrane name Cram inside Membrane manufacturer TAMI Membrane shape Disc Membrane structure Porous, plate, tubular, multichannel Support nature Alumina/titania/zirconia Membrane type dense, porous Average pore size 0,14-1,40m Molecular weight cut-off 1-300kg/mol Water permeance at RT 800 -1200 l/(mh bar)

76

APPENDIX B: Chemical Interest Of Alumina Alkali Metals Ryshkewitch (1960) stated that molten lithium attracts alumina aggressively, potassium to lesser degree, and sodium the least. Reed claimed that the corrosion resistance of sintered alumina to both liquid and vaporized sodium is good at 900 OC. A commercial sintered alumina article disintegrated within 168 hours at 940 OC, however. A synthetic sapphire cylinder (1/4 in. diameter by 1 in. length lost about 1% in weight in 168 hours ay 900 OC but remained clear. Kelman, Wilkinson, and Yaggee found that sodium potassium did not attract below 500 OC, but caused corrosion at 600 OC. Kolosova et al. found a weight loss of only 0.01 to 0.06% in 35 hours at 400 OC for sintered alumina immersed in molten alkali (79%K, 21%Na). The behavior of vaporized alkali metals on oxides and other dielectric materials has been of interest for thermoelectric converters. Wagner and Coriell (1959) tested Al2O3, BN, ZrO2, MgO, HfO2, ThO2, CaO, and NbC to exposure to cesium vapor at temperatures as high as 1475 OC. Only fused alumina remained unaffected by the test. It was concluded that cesium probably reacts with impurities in the contacting material, but the effect can ben observed only when the surface areas available for reaction are relatively large, as is the case with sintered specimens. Higgings attributed pitting of single crystals by cesium vapor at 600 OC to silicon. And barium impurities; neutron irradiation increased the attack. C. E. Addams (1959) observed that rubidium effectively condensed at high temperatures only on oxides with which it could form stable complex compounds. Cowan and Stoddard (1964) stated that glasses could not be used in thermionic converter seals because of alkali metal (cesium) corrosion. Alkaline Earth Metals

77

 Jaeger and Krasemann (1952) observed no reaction of calcium, barium, or strontium to their boiling points (about 1150 OC). This is likely, since BaO and SRO can be reduced to the metals (thermite reaction) by metallic aluminum at 1100 OC in a vacuum (Gvelisiani and Pazukhin). Magnesium also shows no attract to its boiling point (1100 OC). Aluminum Aluminum reacts with alumina at 1100 OC to form Al2O, and at 1600 OC to form AlO (Hoch and Johnston). Antimony, Arsenic Jaeger and Krasemann claimed no attract of sintered alumina. Beryllium Beryllium (melting point 1500 OC) shows no attract of sintered alumina. A slight darkening of the Alumina, caused by formation an interfacial layer of chrysoberyl, BeO Al2O3 (melting point 1870 OC), and Be 3Al2O3 (melting point 1910 OC) are identified compounds (Lang, Fillmore, and Maxwell, 1952; Galakhov, 1957). Bishmut The corrosion resistance is good to 1400 OC (Reed). The reaction product is Bi2O32Al2O3 (Levin and Roth, 1964). Carbon Carbon reduces alumina to normal carbide Al4C3 (Prescott and Hinckle) at above 1700 OC (Jaeger and Kresemann), initiating at as low as 1310 OC (Komarek et al.),

78

but requiring about 2400 OC for complete conversion (Kohlmeyer and Lundquist). Two oxicarbides also exist at about 2030 OC, Al4O4C (Foster, Long and Hunter; Cox and Pidgeon). The presence of Fe2O3, SiO2, TiO2, and V2O5 as impurities in sintered alumina might induce deterioration at lower temperatures (Stroup), as low as 1380 OC (Kroll and Schlechten). The reduction to metallic aluminum occurs at about 2000 OC (Miller, Foster, and Baker). Graphite does not wet molten alumina, but severely pitted in contact with it in water-free, inert atmosphere or vacuum (Barlett and Hall). Cerium Ceric oxide, CeO, forms no compounds with alumina (Wartenberg and Eckhardt). Cerous oxide, Ce2O3, forms Ce2O3 11 Al2O3 and Ce2O3 Al2O3, both of which decompose in air to form Al2O3 and CeO2 at temperatures above 800 OC (Leonov and Keler). Chlorine Chlorine does not attack, except in the presence of carbon (Singer and Thurnauer). Chromium Chromium wets alumina at 1650 OC in a reducing atmosphere (Blackburn, Shelvin and Lowers). Cobalt Cobalt neither wets nor reacts with sintered alumina to above its melting point (1480 OC) in a reducing atmosphere (Sieverts and Moritz; Tumanov et. al.). Copper

79

Copper reacts with the transition aluminas at 800 OC in air to form CuAl2O4, which is stable to 1000 OC. It then converts to CuAlO2, which is stable in air to about 1260 OC (Hahn and de Lorent, 1955; Misra and Chaklader, 1963). Fluorine Winzer claimed attack of sintered alumina at 1700 OC by dry fluorine. Gallium Sintered alumina is inert to gallium to 1000 OC (Kelman et al.). Wartenberg and Reusch (1932) found solid solutions above 810 OC with Ga2O3. Hydrogen Jaeger and Krasemann observed no reduction of alpha alumina in hydrogen up to the melting point, only a surface darkening L. J. Trostel, Jr. (1965) noted that hydrogen attacks alumina refractories below 1600 OC, however in the presence of water vapor. Iron Iron can be melted in sintered alumina under reducing conditions, but wets about 1600 OC (Blckburn et. al.). The spinel FeO Al2O3, dissolves to about 6% alumina at 1750 OC (McIntosh, Rait, and Hay; Fischer and Hoffman). Muan and Gee (1956), and Muan (1958) found limited solubility for Fe2O3 in corundum. Richards and White; Atlas and Sumida; and Turnoc and Lindsley investigated the spinel reactions; Fe2O3 Al2O3 has a structure similar to kappa alumina, and requires above 1320 OC prepare. Lead

80

No reaction occurs with alumina at the melting point of lead (327 OC). In lead- bismuth eutectic alloy (44.5% Pb), Gangler found 0.000 mils/year loss at 1090
OC. Lead aluminate is unstable above 970 OC (Geller and Bunting).

Lithium (See alkali metals) Manganese

Manganese does not attack sintered alumina to above the melting point (1260
OC) in a reducing atmosphere. Although it is more active than iron, cobalt or

nickel it can be distilled in sintered alumina to give a spectroscopicially pure product (Sieverts and Moritz). Mercury Kelmann et al., and Hahn, Frank, et. al. found no reaction with alumina at 300 OC. Molybdenum Alumina is not reduced by molybdenum even above the melting point of alumina. Discoloration may occur at 2100 OC in a dry, inert atmosphere (He). Nickel No attack occurs in dry inert atmosphere Nickel can be melted in sintered alumina in hydrogen atmosphere (Economos and Kingery). Wetting occurs at 1800 OC. Niobium

81

Mass spectrometric and thermogravimetric analysis at 1800 to 2200 OC indicates the principal reaction is !!! !! + 3 !" 2 !" ! + 3!"# ! . Secondary reactions under neutral conditions are: !!! !! + !" !!! ! ! + !"!! ! !!! !! + 2 !" + 1 2 !! !!! ! ! + !"!! + !"# (!) (Grosmann, 1966). Nitrogen Nitrogen does not attack (Jaeger and Krasemann). Palladium, Platinum Both metals can be handled in the molten conditions in sintered alumina. (Jaeger and Krasemann) Phosphorus No attack was observed in moderate temperatures (Jaeger). Silver No compounds can be prepared with silver and alumina (Hahn, Frank et. al.). Sulfur, Selenium, Tellurium These elements do not attack alumina. Tantalum Tantalum reacts only slightly with molten alumina in water free inert atmosphere, H2, N2, CO, or in vacuum (Bartlett and Hall). Alloy 90 Ta-10W shows 82

slight reaction. Tantalum carbide and 4 TaC ZrC react only slightly (Bartlett and Hall). Tin No reaction occurred in molten tin at 1000 OC. Titanium, Zirconium No reaction occurred in an inert atmosphere below 1800 OC, at which temperature black discoloration of the grains and corrosion occurred (Economs and Kingery). Titanium nitride wets molten alumina with negligible corrosion in water free inert atmosphere (Barlett and Hall). Titanium + Aluminum-Hardened Nickel-Base Alloy Decker, Rowe and Freeman found that trance amounts of zirconium or boron picked up from zirconia or magnesia crucibled reduced cracking of the hardened nickel-base alloys during hot-working and increased their rupture strength and ductility. It was desirable to compensate for this effect when making heats in sintered alumina crucibles. Tungusten Jager and Kresemann observed no reaction between tungusten and alumina. Wallace et. al. (1961) investigated the reactions in a Knudsen cell-oven operable at 2500 OC in conjunction with a Nier-type mass spectrometer. Jacodine (1961) observed a growth of nodules or hillocks of alumina in the investigation of heater-cathode breakdown of alumina-coated heaters operated at 1200 OC and 180 volts dc. Tungusten wets molten alumina with negligible corrosion in water- free inert atmosphere (Barlett and Hall).

83

Uranium Jaeger and Krasemann found no reaction between uranium and sintered alumina to 1200 OC. Dykstra (1960) found no solid solution between Al2O3 and uranium oxide. Vanadium Burdese obtained reaction between V2O5 and gamma alumina at 500 OC to form Al2O3 V2O5. Zirconium Boride (ZrB2) and Zirconium Carbide (ZrC) Molten alumina wets and reacts moderately in water free, inert atmosphere (Barlett and Hall).

84

 REFERENCES Walter H. Gitzen; Alumina as a ceramic material; The American Ceramic Society Ohio; 1970 Ayral Andre; Colloidal Alumina: Definitions, Properties, Hydrosol Making and Application for Ceramic Memranes ; CNRS-ENSCM- UM2; France Bhatnagar A., Kumar E., Sillanp M., Nitrate removal from water by nano- alumina: Characterization and sorption studies, Chemical Engineering Journal, Volume 163 (2010) 317-323 Diniz da Costa J., Reed G., Thambimuthu K., High Temperature Gas Separation Membranes in Coal Gasification, Energy Procedia. 1 (2009) 295-302 Diniz da Costa J.C., Reed G.P., Thambimuthu K., High Temperature Gas Separation Membranes in Coal Gasification, Energy Procedia, Volume 1, Issue 1 (2009) 295- 302 Favre E., Carbon dioxide recovery from post-combustion processes : Can gas permeation membranes compete with absorption, Journal of Membrane Science 294 (2007) 5059 Favre, E., Carbon dioxide recovery from post-combustion processes: can gas permeation membranes compete with absorption, Journal of Membrane Science 294 (2007) 5059 Feng X., Pan C., Ivory J., Ghosh D., Integrated membrane/adsorption process for gas separation, Chemical Engineering Science, Vol. 53 (1998) 168 198 Franz J., Scherer V., An evaluation of CO2 and H2 selective polymeric membranes for CO2 separation in IGCC processes, Journal of Membrane Science (2010) 173- 183 Franz J., Scherer V., Impact of ceramic membranes for CO2 separation on IGCC power plant performance, Energy Procedia 4 (2011) 645652 Gu Y., Hacarlioglu P., Oyama S.T., Hydrothermally stable silicaalumina composite, membranes for hydrogen separation, Journal of Membrane Science 310 (2008) 2837 Hofmann Ph., Panopoulos K.D., Fryda L.E., Schweigerc A., Ouweltjes J.P, Karl J., Integrating biomass gasication with solid oxide fuel cells: Effect of real product gas tars, uctuations and particulates on Ni-GDC anode, International journal of hydrogen energy 33 (2008) 2834 2844 Hsieh H., Bhave R., Fleming H., Microporous alumina membranes, Journal of 85

Membrane Science, 39 (1988) 221-241 Hyun S., Jo S., Kang B., Surface modification of y-alumina membranes by silane coupling for CO2 separation, Journal of Membrane Science 120 (1996) 197-206 Hyun S., Jo S., Kang B., Surface modification of -alumina membranes by silane coupling for CO2 separation, Journal of Membrane Science, Volume 120, Issue 2 (1996) 197-206 Huang P., Xu N., Shi J., Lin Y.S., Recovery of volatile organic solvent compounds from air by ceramic membranes, Ind. Chemical Engineering, Res. 36 (1997) 3815 Idrees M., Yan L., Hamdani S., Synthesis and characterization of ceramic (alumina) membrane sheet, Proceedings of The Eleventh Symposium of Malaysia Chemical Engineers (1995) Javaid A., Ford D.M., Solubility-based gas separation with oligomer- modied inorganic membranes: Part II. Mixed gas permeation of 5 nm alumina membranes modied with octadecyltrichlorosilane, Journal of Membrane Science (2003) 157-163 Javaid A., Hughey M.P., Varutbangkul V., Ford D.M., Solubility-based gas separation with oligomer-modied inorganic membranes, Journal of Membrane Science (2001) 141-155 Javaid A., Membranes for solubility-based gas separation applications Review, Chemical Engineering Journal (2005) 219226 Jiraratananon R., Uttapap D., Tangamornsuksun C., Self-forming dynamic membrane for ultrafiltration of pineapple juice, Journal of Membrane Science 129 (1997) 135-143 Jones C., Fidalgo M., Wiesner M., Barron A., Alumina ultraltration membranes derived from carboxylatealumoxane nanoparticles, Journal of Membrane Science 193 (2001) 175184 Julbe A., Farrusseng D., Guizard C., Porous ceramic membranes for catalytic reactors- overview and new ideas, Journal of Membrane Science 181 (2001) 3-20 Kaldis S., Energy and capital cost analysis of CO2 capture in coal IGCC processes via gas separation membranes, Fuel Processing Technology (2004) 337-346 Koros W.J., Evolving Beyond the Thermal Age of Separation Processes: Membranes Can Lead the Way, www.interscience.wiley.com (2011) Koros W.J., Fleming G.K., Membrane-based gas separation, Journal of Membrane Science (1993) 83-89 Koutsonikolas D.E., Kaldis S.P., Sklari S.D., Pantoleontos G., Zaspalis V.T, Sakellaropoulos G.P., Preparation of highly selective silica membranes on defect- free c-Al2O3 membranes using a low temperature CVI technique, Microporous and Mesoporous Materials 132 (2010) 276281 86

Krishnan G., Simulation of a Process to Capture CO2 From IGCC Syngas Using a High Temperature PBI, Membrane, Energy Procedia (2009) 4079-4088 Kuzniatsova T., Mottern M.L., Shqau K., Yu D., Verweij H., Micro-structural optimization of supported alumina membranes, Journal of Membrane Science 316 (2008) 8088 Lafarga D., Santamaria J., Menndez M., Methane oxidative coupling using porous ceramic membrane reactorsI. reactor development, Chemical Engineering Science, Volume 49, Issue 12 (1994) 2005-2013 Laitinen N., Development of a ceramic membrane filtration equipment and its applicability for different wastewaters, Lappeenranta University of Technology (2002) Laitinen N., Kulovaarab M., Levnenc E., Luonsib A., Teilleriab N., Nystrma M., Ultrafiltration of stone cutting mine wastewater with ceramic membranes a case study, Desalination 149 (2002) 121125 Laltmen N., Luonsi A. , Levanen E., Mantyla T., Modified and unmodified alumina membranes in ultrafiltration of board mill wastewater fractions, Desalination 115 (1998) 63-70 Li W., Cao C.-Y., Wu L.-Y., Ge M.-Fa., Superb uoride and arsenic removal performance of highly ordered mesoporous alumina, Journal of Hazardous Materials 198 (2011) 143 150 Li W., Cao C.-Y., Wu L.-Y., Ge M.-F., Song W.-G., Superb fluoride and arsenic removal performance of highly ordered mesoporous aluminas, Journal of Hazardous Materials, Volume 198 (2011) 143-150 Metz B., Davidson, O., Coninck H., Loos M., Meyer, L., Stockage du dioxide de carbone, Rapport spcial du GIEC (IPCC) rsum lintention des dcideurs et rsum technique, International Panel on Climate Change (2005) Nicolas C.-H., Sublet J., Schuurman Y., Role of adsorption and diffusion pathwa ys on the CO 2/N 2 separation performance of nanocomposite (B)-MF I-alumina membranes, Chemical Engineering Science 66 (2011) 6057 6068 Pacheco S., Tapia J., Medina M., Rodriguez R., Cadmium ions adsorption in simulated wastewater using structured aluminasilica nanoparticles, Journal of Non-Crystalline Solids 352 (2006) 54755481 Pagana A.E., Sklari S.D., Kikkinides E.S., Zaspalis V.T., Microporous ceramic membrane technology for the removal of arsenic and chromium ions from contaminated water, Microporous and mesoporous materials V.110 (2008) 150- 156 Pruneanu S., Mihailescu G., Neamtu S., Olenic L., Biro L. P., Applications of nanoporous alumina membranes, IEEE (2001) 59-62

87

Rabiller-Baudry M., Chaufer B., Lucas D., Michel F., Ultrafiltration of mixed protein solutions of lysozyme and lactoferrin: role of modified inorganic membranes and ionic strength on the selectivity, Journal of Membrane Science, Volume 184 (2001) 137 -145 Ciora R., Liu P., Ceramic Membranes for Environmental Related Applications, Media and Process Technology Inc. Fluid/Particle Separation Journal, Vol. 15 (2003) 45-51 Sandeep K. Dalvie and Ruth E Baltus, Transport studies with porous alumina membranes, Journal of Membrane Scrence (1992) 241-254 Sandeep K.D., Ruth B., Transport studies with porous alumina membranes, Journal of Membrane Science, 71 (1992) 241-255 Sen M., Manna A., Parimal P., Removal of arsenic from contaminated groundwater by membrane-integrated hybrid treatment system, Journal of Membrane Science, Volume 354, Issues 12 (2010) 108-113 Xomeritakis, G., Tsai, C.Y., Jiang, Y.B., Brinker, C.J., Tubular ceramic- supported solgel silica-based membranes for ue gas carbon dioxide capture and sequestration, Journal of membrane Science 341 (2009) 3036 A.F .M . Leenaars and A .J . Burggraaf, The formation of supported membranes, J . Colloid Interface Sci, 105 (1985) 27. K. Vasanthakumar, S. C. Vanithakumari, and U. Kamachi Mudali Fabrication of Nanoporous Alumina Membranes by Single Step Anodization and Their Microscopic Characterization Corrosion Science and Technology Group, Indira Gandhi Centre for Atomic Research, Kalpakkam 603102, India H.P . Hsieh*, R.R. Bhave And H .L. Fleming; Microporous Alumina Membranes Aluminum Company Of America, Alcoa Laboratories, Alcoa Center, PA 15069 (U.S.A .), 1988 R.W. Baker, Membrane Separation systems: Recent Developments and Future Directions, Noyes Data Corp., Park Ridge, NJ, 1991. Inorganic Membranes: Markets, Technologies, Players, Business Communications Co. Inc., Norwalk, CT, Marketing Study GB-112N 1997 Membrane Sicence And Technology Series,3; Inorganic Membranes For Separation And Reaction; H. P. HSIEH, Elseiver, 1996 A. Kirchner , K.J.D. MacKenzie, I.W.M. Brown, T. Kemmitt, M.E. Bowden MacDiarmid; Structural characterization of heat-treated anodic alumina membranes prepared using a simplified fabrication process Institute for Advanced Materials and Nanotechnology

88

G.G. Avci, A. Misirli, V. Gunay, Processing and characterization of microfiltration supports prepared from alumina powders, Ceram. Int. 22 (1996) 2326. Christopher D. Jones a, Maria Fidalgo b, Mark R. Wiesner b, Andrew R. Barron ; Alumina ultrafiltration membranes derived from carboxylatealumoxane nanoparticles XiaoRong Huang*, GenLiang Meng, ZhongTao Huang, JianMing Geng ; Preparation of unsupported alumina membrane by sol-gel techniques Jiansheng Li, Xiaoyang Wang, Lianjun Wang , Yanxia Hao, Youla Huang, Ye Zhang, Xiuyun Sun, Xiaodong Liu Nanjing; Preparation of alumina membrane from aluminium chloride University of Science and Technology, School of Chemical Engineering, Weon-Sik Chae, Sung-Jae Im, Jin-Kyu Lee, and Yong-Rok Kim; Novel Fabrication of Nanoporous Alumina Membrane Microtubes: 2-Dimensional Nanoporous Arrays on Every Facets of Microtubes; December 15, 2004 K.L. Yeung, J.M. Sebastian, A. Varma; Mesoporous alumina membranes: Synthesis, characterization, thermal stability and nonuniform distribution of catalyst; USA 1996 Rocio Redon, A. Vazques-Olmos, M.E Mata Zamora, A. Ordonez- Medrano,; Contact Angle Studies on Anodic Porous Alumina; Mexico 2005

89