Mineral membranes made of sintered clay: application to crossflow microfiltration

Mineral membranes made of sintered clay: application to crossflow microfiltration

DESALINATION ELSEVIER Desalination 146 (2002) 57-61 www.elsevier.com/locate/desal Mineral membranes made of sintered clay: application to crossflow...

368KB Sizes 1 Downloads 138 Views

DESALINATION ELSEVIER

Desalination

146 (2002) 57-61 www.elsevier.com/locate/desal

Mineral membranes made of sintered clay: application to crossflow microfiltration Jilali Bentama”“, Kamar Ouazzania, Philippe Schmitzb “Lahomtol-)>of the Processes and the Environment, High School of Technology, I?0. Box 2427, Fez, Morocco email: [email protected] “Institut de Mkanique des Fluides, UMR CNRSLNP-UPS 5502, Alike du Professeur Camille Soula, 31400 Toulouse, France, email: [email protected] Received 5 February 2002; accepted 19 March 2002

Abstract New mineral membranes made of sintered clay were performed and characterized in terms of porosity, hydraulic resistance, pore diameter and mechanical resistance. These membranes can be used as microfiltration membranes. The variations of the filtrate flux as a function of time are measured during crossflow microfiltration operations of dilute suspensions of bentonite and talc, for different transmembrane pressure values and mean flow rates. Four membranes of different porosities are tested. Then, we show that the clay membranes can be used in different applications such as water treatment and membrane bioreactor, according to the average pore diameter. Keywords: Membranes;

Crossflow microfiltration;

Water treatment;

1. Introduction Crossflow microfiltration processes can be considered as high performance separation techniques in many industrial applications such as water treatment, food and beverage manufacture and biotechnology. The membranes generally used in these processes can be classified into two groups according to the nature of the *Corresponding

bioreactor

material: organic or mineral. This paper concerns the characterization of new sintered clay microfiltration membranes. These membranes were made of clay from Fbs in Morocco. They take benefits of the ancestral knowledge of Moroccan handicraft and of recent progress in material chemistry. Four membranes of different porosity were elaborated. Preliminary filtration tests are made on model suspensions of bentonite and talc.

author.

Presented at the International July 7-12, 2002. 001 l-9164/02/$-

Membrane

Congress on Membranes

and Membrane

Processes

See front matter 0 2002 Elsevier Science B.V. Al1 rights reserved

PII: SO0 I I-9 I64(02)00479-4

(ICOM),

Toulouse, France,

J. Bentama et al. /Desalination

58

Then the filtration of a water rich in organic matter (dissolved and in suspension) exhibits promising results for the future use of sintered clay membranes in water treatment. Finally the coupling of fermentation and microfiltration operations are also studied using two different sintered clay membranes.

146 (2002) 57-61

3. Experimental

2. Characteristics of the sintered clay membranes This work is a part of a more complete study concerning clay material valorization [l]. The chemical composition of the clay used in this study (Table 1) was previously determined by spectrometry atomic absorption, after fusion in the borate of strontium [2]. It is essentially composed of silica, alumina and oxide of calcium (about 72%). Besides, the contents in oxide alkaline are weak and the presence of the iron oxide (4.92%) explains the rust colour of the clay. The characteristics of the sintered clay membranes used are given in Table 2. The different porosities are obtained by controlling the addition of a certain quantity of organic matter to the mixture before the sintering operation [3]. We chose four formulations: 5, 10, 15 and 20% agent of porosity represented respectively by C, C,,, C 15and C,,. The membranes are tubular; the inner and outer diameters are respectively equal to 15 mm and 20 mm, the length is equal to 155 mm.

set-up

The experimental set-up (Fig. 1) consists of a large fluid reservoir (4), a first volumetric pump (1) to ensure the circulation of the suspension, a second volumetric pump (2) for backflushing. Two pressure gauges (3) are used to measure the transmembrane pressure. The retentate is permanently recycled in the reservoir (4). The permeate is kept or used for backflushing. Two valves (6) allow us to control the flow rate and the pressure in the module in which the membrane is introduced (7). 4. Results and discussion Filtration operations are performed using the experimental set-up previously detailed. The operating conditions studied are: the nature of the _ * -i_

Fig. 1. Experimental set-up. Table 1 Chemical composition of clay

Oxide

SiOz

A1203

CaO

Fez03

MgO

Na,O

K,O

TiOz

MnO

SO,

%

47.05

10.34

14.53

4.92

3.91

1.40

1.73

0.55

0.05

0.57

Table 2 Characteristics of the membranes Membrane

CS

Cl0

Cl5

c20

Mechanical resistance, MPa Mean pore diameter, pm Porosity, % Hydraulic resistance, m/Pa.s

1.09 1.10 37 1.7 1o-9

0.95 1.31 43 2.0 1o-9

0.82 1.93 49 2.6 1O-9

0.73 3.70 53 3.8 1O-9

J. Bentama et al. /Desalination

suspension, the transmembrane and the flow rate.

59

146 (2002) 57-61

pressure (APtm)

";‘ i 1300 1 Z

4.1. Bentonite suspension Bentonite particles of 40 microns mean size are diluted in ultrapure water. The concentration of the suspension is 1 g/l. First, the effect of the transmembrane pressure on the filtrate flux magnitude is studied. As can be expected, the higher is the membrane porosity, the higher is the filtrate flux (Fig. 2). The increase of the hydraulic resistance of the cake as the transmembrane pressure increases due to the bentonite particle cake compression was already evidenced by several authors [4,5]. The favorable effect of the flow rate on the filtrate flux magnitude was also verified [6] (results not reported here) since the principle of the crossflow filtration is to take benefit of the existence of the wall shear stress to limit the cake formation.

2 'u 1000 1 E 5 700 -

400

I 0.5

0.7

0.9

1.1

1.3

1.5

Transmembrane pressure(bar)

Fig. 2. Effect of transmembrane pressure on filtrate flux: filtration of bentonite

1800 -

+C5

1500

c 2 _

I ~

-B-c10 1200

+c15 +c20

2 900 2

e 600 .z

I

300

4.2. Talc suspension The mean size of talc particles is about 3 microns. The concentration of the suspension is 1 g/l. The turbidity of the filtrate is measured after each filtration operation and compared to the turbidity of the feeding suspension. The results are reported in Table 3. The corresponding reduction of the filtrate flux with time is plotted in Fig. 3. It can be noticed that the shape of the curves for the membranes C, and C,, are classical and that the associated turbidity drop rate is total. This result is in agreement with the pore diameter of these two membranes, which is inferior to the mean particle size. On the contrary, a more rapid decrease of filtrate flux is observed for the

Table 3 Turbidity of the filtrate, filtration of talc Membrane

C5

Cl0

Cl5

Go

Turbidity, NTU Drop rate, o/o

0.95 99.59

0.95 99.59

10.32 95.51

61.07 73.45

1 0

25

50

75

100

125

150

Time(mn)

Fig. 3. Effect of membrane porosity on filtrate flux decline: filtration of talc.

membranes C,, and C,, for which the pore diameter is superior to the mean particle size. In this case, the fouling of some pores of the membrane can occur, reducing its hydraulic resistance. 4.3. Treatment of a water of surjface The attempts are made to observe the behaviour of the sintered clay membrane towards the water rich in organic matters (dissolved and in suspension). The water used in the study comes from a river in a rainy period. The characteristics of this water are presented in Table 4. For the three values of transmembrane pressure, the flux is moderate after 15 min of filtration (Table 5).

60

J. Bentama et al. /Desalination

surface water. The period of backflushing has been optimized in order to recover the initial hydraulic resistance.

Table 4 Characteristics of the surface water studied Parameters

146 (2002) 57-61

Values 7.8 1030 17.5 102.42 11,000

PH Conductivity yS.cm-’ Turbidity, NTU DCO, mg OJ’ FMAT UFC, ml-’

Table 5 Filtrate flux (l/h.m2) vs. Af’tm (bar): filtration of surface water ARm

C5

Go

Cl,

C 20

0.8 1 1.2

279 315 392

334 380 467

493 542 650

931 1020 1150

The filtrate flux increases with the transmembrane pressure. We notice that for the membrane C,,, the initial flow rate is easily recovered after backflushing [7]. This interesting phenomenon is more difficult to achieve for the membranes C, and C,,, and is practically absent for the membrane C,,,. Table 6 exhibits the quality of the filtrate flux obtained after 15 min of filtration, for a transmembrane pressure value of 1.2 bar. For this type of surface water, we recommend the membrane C,,. Indeed the debit is increased by alternating periods of wash and filtration. Successive sequences of filtration operation and periodic backflushing have been done using this

4.4. Coupling fermentation operations

and micofiltration

In this technique the cell culture is continuous, which allow to control the environment of cells and to obtain very high active concentrations by the recycling of the microorganisms. The biologic fluid mode1 used in different operations of microfiltration is a suspension of yeast in powder. The preparation of the suspension consists in hydrating the yeast in some physiological water (NaCl, 9 g.l-I). The concentration of yeast is 5 g.l-‘. The immediate measure of the concentration of yeast in the filtrate is made periodically by reading the optical density in 620 nm using a spectrophotometer. In Fig. 4, we present the evolution of filtrate flux with time for the membranes C,, and C,,. Concerning the membrane C15, a first drastic decline of the filtrate flux followed by a slow and progressive decline are observed. The final value is stabilised after four hours. This profile is the main characteristics of the phenomenon of formation of a superficial layer in the immediate neighbourhood of the membrane. However the filtration continues in the presence of this layer.

Table 6 Characteristics of the tiltrate: filtration of surface water Membrane

Turbidity, NTU

DCO, Mg OJ’

CS Cl0 C,S

0.8 1.5 2.1 3.2

30.8 31.7 35 42.2

c20

FMAT, UFC ml-’ 0 0 10 1300

0

50

100

150 Time

200

250

300

350

(min)

Fig. 4. Effect of membrane porosity on filtrate flux decline: filtration of yeast.

J. Bentama et al. /Desalination

61

146 (2002) 57-61

But this layer of high hydraulic resistance induces

a drastic reduction of the actual diameter membrane pores. The second decline of the flux observed in Fig. 4 is due to the external fouling of the superficial layer. The evolution of the filtrate flux with time for the membrane C,,, is characterized by an initial step less pronounced than that the one of the membrane C,,. After this step, the flux stabilizes quickly. It confirms the absence of a progressive supplementary superficial layer in the neighbourhood of the membrane. In Fig. 5, we have plotted the evolution of yeast concentration in the filtrate as a function of time of filtration. Using the membrane C,, the concentration of yeast in the filtrate is almost zero after sixty minutes in two steps (superficial layer and deposit of particles). A single decline of the yeast concentration in the filtrate is observed for the membrane C,,, and after a shorter time (30 min). Therefore, the membrane C ,. is recommended for the filtration of biologic fluid.

Fig. 5. Effect of membrane porosity on yeast concentration in the filtrate: filtration of yeast.

Acknowledgements

[51

The “Comite Mixte Interuniversitaire FrancoMarocain” is gratefully acknowledged for its financial support no 194F/SI/99.

161

References

[71

[I]

J. Bentama, A. Addaou et M. Rafiq, Physicochimie et etude du frittage d’une argile destinee a l’elaboration

O/ 0

40

80

120

160

Time (min)

VI 131

[41

de membranes, J. Chim. Phys., 5(95) (1998) IOOl1019. E. Jeanroy, Anal. 2( 10) (1974) 703-707. J. Bentama, P. Schmitz et A. Addaou, Elaboration et caracterisation de membrane a base d’argile. Recent Progres en Genie des Procedes, 15(81) (2001) 6976. H. Bauser, N. Stroh and E. Walitza, Interfacial effects with microfiltration membranes, J. Membr. Sci., 11 (1982) 321-332. A. Tawell and J. Landau, Mass transfer between solids sphere and oscillating fluids, J. Chem. Eng., 54 (1976) 532-539. R.H. Davis and J.D. Sherwood, Similary solution for steady state crossflow microfiltration, Chem. Eng. Sci., 45 (1990) 3204-3209. J. Bentama and P Schmitz, Elaboration and characterization of sintered clay membranes. Application to crossflow microfiltration, Word Filtration Congress 8, Brighton UK, 2000, pp. 435-438.