Journal of Colloid and Interface Science 251, 225229 (2002) doi:10.1006/jcis.2002.

8336

Effect of Acidic Surface Treatment of Red Mud on Mechanical Interfacial Properties of Epoxy/Red Mud Nanocomposites
Soo-Jin Park,,1 Dong-Il Seo, and Changwoon Nah
Advanced Materials Division, Korea Research Institute of Chemical Technology, P.O. Box 107, Yusong, Taejon 305-600, Korea; and Department of Polymer Science & Engineering, Chonbuk National University, Chonbuk 561-756, Korea Received January 23, 2002; accepted March 8, 2002; published online June 3, 2002

In this work, red mud (RM) was chemically modied by 0.1, 1, and 5 M H3 PO4 solution to prepare epoxy/RM nanocomposites. The effect of chemical treatment on a RM surface was studied in terms of pH, acidbase values, N2 /77 K gas adsorption, equilibrium spreading pressure (e ), and surface free energy. The mechanical interfacial properties of epoxy/RM nanocomposites was measured in the context of critical stress intensity factor (KIC ) and critical strain energy release rate (GIC ) measurements. From the experimental results, the acidic surface treatment led to a modication of the RM surface properties, such as the surface acid values, porosity, L specic surface area, e , and London dispersive component (S ) of surface free energy of the treated RM as an increase of the treatment concentration. In the fracture toughness (KIC and GIC ) measurements, the mechanical interfacial properties of epoxy/treated RM nanocomposites were intimately correlated with the improvement of interfacial adhesion between the RM surface and a matrix and L the increase of S of surface free energy of the RM due to the development of SBET or porosity of RM surfaces. C 2002 Elsevier Science (USA) Key Words: epoxy resin; red mud; nanocomposites; surface treatment; mechanical interfacial properties.

INTRODUCTION

Red mud is a waste obtained from the aluminum industry and has been accumulating at a rate of 30 million ton annually throughout the world. Under normal conditions when 1 ton of alumina is produced from bauxite, an equal amount of red mud is generated as a waste. The disposal of red mud is associated with space/real estate near industry, cost of disposal, and pollution, which are now crucial factors. Therefore, the exploitations of these by-products are a growing technological aspect of basic industries and the environment. A further aspect is their re-use as starting materials for other productions (1, 2). Red mud has been suggested as ller for polymer reinforcement or as a cheap adsorbent for removal of toxic metals or an acid by several researchers. Chand and Hashmi (3, 4) tried to improve the mechanical properties and abrasive wear properties

of polymer blend lled with red mud. Pradhan et al. (5) had reported that activated red mud as a good adsorbent was used for adsorption of phosphate or chromium. In addition, the mechanical and thermal properties of polymers are generally improved by the addition of inorganic additives. The challenges in this area of high-performance organic inorganic hybrid materials are to obtain signicant improvements in the interfacial adhesion between the polymer matrix and the reinforcing material since the organic matrix is relatively incompatible with the inorganic phase. Generally, a better interfacial bonding will impart better properties to a polymer composite such as high modulus, strength, and stiffness (6, 7). This reinforcement of polymer lled with inorganic material is largely dependent on the physical interfacial phenomenon between the ller and matrix. This can be determining the degree of adhesion by physical interaction, such as active functional groups, hydrogen bonding, Lewis acidbase interactions, surface energy, and crystallite faces of ller surface at the interface (8, 9). To increase physical interaction, various surface treatment techniques are applied, such as oxidation in acid solutions (10), dry oxidation in oxygen (11), anodic oxidation (12, 13), and plasma treatments (14). It is due to the fact that surface modication leads to development of surface functional activity on a ller surface, resulting in modication to achieve good interfacial adhesion between the reinforcement and the matrix. In this work, RM is chemically modied with phosphoric acid to investigate the effect of surface treatment of RM on mechanical interfacial properties of epoxy/RM nanocomposites. Changes in microstructure and surface energetics on a RM surface are also discussed with the results of mechanical interfacial properties of epoxy/RM nanocomposites.
EXPERIMENTAL

Materials and Sample Preparation The red mud used in this study was provided by Korea Chem. Co. (namely, RM) and the composition of RM is shown in Table 1. The average size and density of the RM used are about 80 nm and 2.4 g cm3 , respectively. The samples were
225
C

1 To whom correspondence should be addressed. Fax: +82-42-861-4151. E-mail: psjin@krict.re.kr.

0021-9797/02 $35.00
2002 Elsevier Science (USA) All rights reserved.

226

PARK, SEO, AND NAH

TABLE 1 Equivalent Mineral Oxide Composition of the Materials Studied

Contents Al2 O3 SiO2 Fe2 O3 TiO2 Na2 O CaO LOI (loss of ignition) Composition (%) 23.7 22.9 16.6 6.7 11.6 6.7 11.8

Mechanical Interfacial Properties The epoxy resin was mixed with the desired amount of RM to prepare curing samples of epoxy/RM nanocomposites. The RM was dispersed uniformly in epoxy monomer at 60 C for 2 h by sonicating and stirring. Then, curing agent was added into the epoxy/RM hybrid and mixed thoroughly. The mixtures were then degassed to remove bubbles before they were cast into a mold. The samples were cured with DDM for 2 h at 140 C and 1 h at 200 C. The mechanical interfacial properties of composites were conducted to measure properties which are sensitive to the degree of adhesion of the llermatrix interface. The critical stress intensity factor (K IC ), which is one of the fracture toughness parameters, can be characterized by a single-edge-notched (SEN) beam fracture toughness test (ASTM E399) in three-point bending exure. Notches were cut using a diamond wire saw, approximately half the depth of the specimen. A span-to-depth ratio of 4 : 1 and cross-head speed of 1 mm/min were used.
RESULTS AND DISCUSSION

chemically treated with 0.1, 1, and 5 M H3 PO4 solution for modication of the RM surface at room temperature for 12 h. In addition, prior to use of the following analysis, the RM was washed several times with distilled water and dried in a convection oven. Then, the RM, less than 53 m, using a sieve were applied for characterization. The epoxy resin that was used was the diglycidyl ether of bisphenol A (YD-128 supplied by Kukdo Chem. Co. of Korea). The epoxide equivalent weight of epoxy resin was 185190 g eq1 and the density was 1.16 g cm3 at 25 C. The 4,4 -diaminodiphenylmethane (DDM) was used as a curing agent for epoxy resin. The chemical structures of epoxy and DDM are shown in Fig. 1. Surface and Adsorption Properties The pH of the samples was measured with ASTM D3838. About 0.5 g of the as-received or treated RM was added to 20 ml of distilled water, the mixture was shaken for 12 h, and the pH of each sample was measured after ltration. The acid base values of the surface functional groups of the samples were determined by Boehms titration method (15). In the case of the acid value, about 1.0 g of the samples was added to 100 ml of 0.1 N NaOH solution and the mixture was shaken for 24 h. The solution was ltered through a membrane lter and titrated with 0.1 N HCl solution. Likewise, the base value was determined by converse titration. The specic surface area (SBET ) and porosity of the RM studied were investigated by N2 /77 K gas adsorption (16) using an automated adsorption apparatus (Micromeritics, ASAP 2010). The samples were outgassed at 573 K for 6 h to obtain a residual pressure of less than 103 Torr in high vacuum prior to adsorption measurements.

Surface and Adsorption Properties of Red Mud Table 2 shows the surface properties of the RM treated with differential concentration of H3 PO4 solution. It can be seen that the treatments largely inuence pH and acidbase values of the RM studied, resulting in increasing the Boehms acid value or decreasing the base value. The major constituent of RM is a metal alkoxide, such as Fe2 O3 , Al2 O3 , CaO, SiO2 , TiO2 , and Na2 O, and the surface of RM has a high pH due to the presence of sodium oxide (17), etc. The pH and base values of the RM surface are decreased and the acid value of the RM surface is increased by acidic surface treatment of RM in H3 PO4 solution due to the development of an acidic functional group or active hydrogen. In addition, it is well known that the physical properties of reinforcing agent, such as size, specic surface area, and surface functional groups play an important role in improving the thermal or mechanical properties of composites (18). In this work, the physical properties of a particle are investigated by the N2 gas adsorption method. Figure 2 shows N2 gas adsorption isotherms of the RM treated with a differential concentration in H3 PO4 solution. An adsorption isotherm of untreated RM shows typical TYPE II adsorption of an IUPAC standard (19) on particles which have macropores or nonpores, showing poor adsorption. The acidic surface

CH3 CH2 CH CH2 O O C CH3

OH O CH2 CH CH 2 O n

CH 3 C CH 3 O CH 2 CH CH 2 O

TABLE 2 pH and AcidBase Values of the RM Series Studied

pH Acid values (meq/g) 159 193 689 1393 Base values (meq/g) 628 567 180 30

Epoxy
NH 2

H 2N

CH 2

DDM

FIG. 1.

The chemical structures of epoxy resin and DDM.

As-received 0.1 M 1M 5M

11.6 8.9 5.1 3.9

ACIDIC SURFACE TREATMENT OF RED MUD

180
100

227

Vol. adsorbed (cm /g STP)

150 120 90 60 30 0 0.0

Vol. adsorbed (cm /g STP)

as-received 0.1M 1M 5M

80

as-received 0.1M 1M 5M

60

40

20

0.2

0.4

0.6

0.8

1.0

0 0.0

10

P/Po
FIG. 2. Dependence of N2 adsorption volume adsorbed at 77 K on the relative pressure of the RM studied.

Thickness ( )
FIG. 4. Dependence of N2 adsorption volume on the gas adsorbed thickness of the RM studied.

treatments lead to the development of micropores or mesopores on the RM surface and create localized adsorption sites on the RM surface, resulting in increasing adsorption volume of N2 gas at low relative pressure. Figure 3 shows the plot of P/P0 versus (P/P0 )/a[1 (P/P0 )], ranging from 0.05 to 0.25 in relative pressure, to investigate a change of microstructure of the RM surface applied by the BET method (16). The specic surface area (SBET ), value of BET constant (CBET ), and net heat of adsorption can be obtained from Fig. 3. It is well known that CBET provides information on adsorption, such as shape of adsorption isotherm, net heat of adsorption (NHA), and a comparison of the relative amount of adsorption. If CBET is high, a point of inection is clear at low relative pressure of the adsorption isotherm. The NHA is able to be obtained from the CBET as follows, E = E0 E L = RT ln(CBET )
0.035
as-received 0.1M 1M 5M

[1] [2]

where E 0 and E L refer to heat of adsorption and that of liquefaction at the monolayer of the RM surface, respectively. R and T refer to gas constant and temperature in Kelvin, respectively. Figure 4 shows a t-plot of thickness verse adsorption volume. Lippens and de Boer (20) suggested that micropore volume could be obtained by t-plots, which compares the adsorption isotherm of standard adsorbate to that of real samples. The slop of t-plot is related to an external surface of a solid, and the micropore volume of a solid is able to be obtained from a multiplying density conversion factor by the intercept of the t-plot. Table 3 shows values of SBET , CBET , NHA, R p , and pore volume obtained by the above-mentioned method. The SBET and total pore volume of untreated RM are about 36 m2 /g and 0.18 cm3 /g, and the pore structure of untreated RM has a meso porous structure of IUPAC standard between 20 and 500 A as shown in Table 3. The result of surface treatment with the differential solution concentration shows that both SBET and micropore volume of the treated RM are slightly developed. Especially, it is interesting to note that NHA is also developed. It is because micropores exhibit unusually high adsorption potentials due to the overlapping potential from the walls of the micropore (21).
TABLE 3 N2 Adsorption Data of the RM Series Studied
SBET a (m2 /g) CBET b 1158 1137 1260 2341 VT c (cm3 /g) 0.182 0.184 0.118 0.274 Vm d (cm3 /g) 0.0031 0.0030 0.0078 0.0341 RP e (A) 203 177 73 63 NHA f (kJ/g) 4.537 4.525 4.591 4.989

(P/P0)/ a[1-(P/P0)]

0.028

0.021

0.014

0.007

As-received 0.1 M 1M 5M
0.05 0.10 0.15 0.20 0.25 0.30
a

36 42 64 174

0.000

P/Po
FIG. 3. Plots of P/P0 and (P/P0 )/a[1 (P/P0 )] of the RM studied by the BET method.

Specic surface area from BET method by N2 adsorption at 77 K. BET constant. c Total pore volume. d Micropore volume obtained by t-plot method. e Average pore diameter from empirical equation of R = 4V /S p T BET . f Net heat of adsorption from C BET .
b

228
Fe2O3 + Al2O3 + Na2O + CaO
FIG. 5.

PARK, SEO, AND NAH

2H3PO4 2H3PO4 H3PO4 H3PO4

2FePO4 2AlPO4

+ 3H2O + 3H2O H2O H2O

Na2HPO4 + CaHPO4 +

Reaction mechanism of constituents on the RM surface.

Therefore, the acidic treatment has an effect on a change of structure of the RM surface, resulting in developing the porosity and SBET . It is considered that a change of structure of the RM surface is due to hydrolysis or leaching a metal salt out at the RM surface (22). Figure 5 shows the reaction mechanism of metal oxides at the RM surface when the RM surface is treated with H3 PO4 . Kasliwal and Sai (2) had reported that the RM surface is treated with HCl to investigate a method for enrichment of titanium dioxide in RM. The metal oxides are hydrolyzed and a metal salt is leached out at the RM surface, which is conrmed by a decrease of pH in treating solution in this work. Equilibrium Spreading Pressure and Surface Free Energy The equilibrium spreading pressure (e ) reects the reduction of surface tension of a solid. By measurement of e , the disperse component of the RM surface can be quantied. The calculation of e includes the specic surface area of the material and gives a physical quantity representative of the chemical surface state of the RM. When the nitrogen adsorbate is considered in the adsorption studies, e can be determined by using the integrated form of the Gibbs equation (23), e = RT N A a N 2 a0
1

where subscript S and superscript L represent respectively the solid-state and London dispersive component of surface free L energy and N2 = N2 = 10.5 mJ m2 (25) in this work. Table 4 shows the evolutions of e and SL of the treated RM. The acidic treatment leads to an increase of e with increases of treated concentration, which predicts interfacial interaction between the RM surface and the matrix. And the surface free energy is a physical phenomenon caused by intermolecular interactions at the interface. The London dispersive component of surface free energy ( SL ) is one of the main factors in the adsorption capacity, along with the specic surface area of the solid. An increase of SL of the treated RM is effective in improving the stability of dispersions of the RM, relating to the increase of specic surface area. Mechanical Interfacial Properties It is generally accepted that the mechanical interfacial properties of ller-reinforced composites depend strongly on the level of adhesion between the ller and the matrix. The degree of adhesion at the interface between RM and the matrix can be measured in terms of the critical stress intensity factor (K IC ). For the single-edge-notched (SEN) beam fracture toughness test, the value of K IC is calculated as follows (26), K IC = PL Y, bd 3/2 [6]

ad(ln P/P0 ),
0

[3]

where P is the rupture force, L the span between the supports, and Y the geometry factor described in ASTM E399, and b and d are the specimen width and thickness, respectively. In addition, in plane strain, the values for K IC and the critical strain energy release rate (G IC ) for the same material and testing conditions, in theory, are related as follows (27), G IC = (1 2 )K IC 2 , E [7]

where R is the ideal gas constant, T the absolute temperature, a the amount adsorbed at a relative pressure P/P0 , a0 the amount adsorbed for lling a monolayer, N A Avogadros number, and a N2 the cross-sectional area of the adsorbed nitrogen. By the BET method approach, it is then possible to determine the amount adsorbed for lling a monolayer, a0 , as follows, P/P0 (CBET 1)P/P0 1 + , = a(1 P/P0 ) CBET a0 CBET a0 [4]

where is Poissons ratio ( = 0.3 is evaluated in this work) and E is the tensile modulus at the same testing conditions as the fracture test. The K IC and G IC are described by the state of stress in the vicinity of the tip of a crack as functions of the specimen geometry, the crack geometry, and the applied load on the basis of linear elastic fracture mechanics. Likewise, the size, shape,
TABLE 4 L Spreading Pressure (e ) and London Dispersive Component ( S ) of Surface Free Energy of the RM Series Studied
e (mJ/m2 ) As-received 0.1 M 1M 5M 40 42.2 42.7 51.2
L S (mJ/m2 )

where CBET is a BET constant characteristic of the adsorbate adsorbent. In addition, it is of interest to determine the SL of the RM investigated in the adsorption studies at 77 K. The relationship between SL and e for a solid surface is indeed given by the following equation (18, 24),
L S =

e + 2 N2
L 4 N2

[5]

88.7 95.1 96.7 124

ACIDIC SURFACE TREATMENT OF RED MUD

229
CONCLUSIONS

FIG. 6. SEM microphotograph of the fracture surface of epoxy/RM nanocomposites as-received.

In this work, the RM, which is a by-product discharged from an Al2 O3 production factory with bauxite, is chemically treated with H3 PO4 solution to prepare epoxy/RM nanocomposites. And the effect of acidic surface treatment of the RM on mechanical interfacial properties of epoxy/RM nanocomposites is investigated. The treatments largely inuence pH and acidbase values of the RM studied, resulting from a development of an acidic functional group or active hydrogen on the RM surface. Moreover, the surface treatment plays an important role in a change of structure of the RM surface due to hydrolysis or leaching a metal salt out at the RM surface. In addition, the mechanical interfacial properties of composites are improved as the treatment concentration increases, resulting in increasing interfacial adhesion between the reinforcing agent and the matrix. Therefore, it can be conrmed that the value of K IC or G IC of epoxy/RM nanocomposites are closely correlated with the surface free energy of RM due to the development of SBET or porosity of RM surfaces.
REFERENCES
1. Griotheim, K., and Welch, B. J., Aluminium Smelter Technology. Verlag, Berlin, 1983. 2. Kasliwal, P., and Sai, P. S. T., Hydrometallurgy 53, 73 (1999). 3. Chand, N., and Hashmi, S. A. R., J. Sci. Ind. Res. 58, 795 (1999). 4. Chand, N., and Hashmi, S. A. R., Bull. Mater. Sci. 22, 801 (1999). 5. Pradhan, J., Das, S. N., and Thakur, R. S., J. Colloid Interface Sci. 217, 137 (1999). 6. Agag, T., Koga, T., and Takeichi, T., Polymer 42, 3399 (2001). 7. Jang, B. Z., Compos. Sci. Technol. 44, 333 (1992). 8. Park, S. J., and Kim, J. S., J. Colloid Interface Sci. 232, 311 (2000). 9. Park, S. J., and Cho, M. S., Carbon 38, 1053 (2000). 10. Donnet, J. B., and Bansal, R. C., Carbon Fibers, 2nd ed. Dekker, New York, 1990. 11. Yuan, L. T., Shyu, S. S., and Lai, J. Y., J. Appl. Polym. Sci. 42, 2525 (1991). 12. Ishikawa, Y., and Matsumoto, Y., Electrochim. Acta 46, 2819 (2001). 13. Park, S. J., and Kim, M. H., J. Mater. Sci. 35, 1901 (2000). 14. Dilsiz, N., Ebert, E., Weisweiler, W., and Akovali, G., J. Colloid Interface Sci. 170, 241 (1995). 15. Boehm, H. P., Adv. Catal. 16, 179 (1966). 16. Brunauer, S., Emmett, P. H., and Teller, E., J. Am. Chem. Soc. 60, 309 (1938). 17. Sglavo, V. M., Campostrini, R., Maurina, S., Carturan, G., Monagheddu, M., Budroni, G., and Cocco, G., J. Eur. Ceram. Soc. 20, 235 (2000). 18. Park, S. J., in Interfacial Forces and Fields: Theory and Applications (J. P. Hsu, Ed.), pp. 385492. Dekker, New York, 1999. 19. Gregg, S. J., and Sing, K. S. W., Adsorption, Surface Area and Porosity. Academic Press, London, 1982. 20. Lippens, B. C., and de Boer, J. H., J. Catal. 4, 319 (1965). 21. Lowell, S., and Shields, J. E., Powder Surface Area and Porosity, 3rd ed. Wiley, New York, 1991. 22. Park, S. J., Seo, D. I., Lee, J. R., and Kim, D. S., Polymer (Korea) 25, 587 (2001). 23. Park, S. J., and Kim, J. S., J. Colloid Interface Sci. 232, 311 (2000). 24. Hiemenz, P. C., and Rajagopalan, R., Principles of Colloid and Surface Chemistry, 3rd ed. Dekker, New York, 1997. 25. Fowkes, F. M., Ind. Eng. Chem. 56, 40 (1964). 26. Chen, M. C., Hourston, D. J., and Sun, W. B., Eur. Polym. J. 31, 199 (1995). 27. Pressly, T. G., Keskkula, H., and Paul, D. R., Polymer 42, 3043 (2001).

and surface properties of the reinforcing agent play an important role in improving the mechanical interfacial properties of composites. Figure 6 shows SEM microphotographs of the fracture surface of epoxy/RM nanocomposites to investigate the dispersivity and particle size of the RM studied into a matrix. The result shows that the RM particles, which are largely sized about 50100 nm, can be observed to be well-dispersed in a polymeric matrix. Table 5 shows the mechanical interfacial properties of epoxy/ RM nanocomposites lled with a differentially treated RM to investigate the relationship between the fracture toughness of the material and the interfacial adhesion in composites. The experimental results show that both K IC and G IC of epoxy/RM nanocomposites as the treatment concentration increases are improved. The acidic surface treatment leads to an increase of SL , relating to the development of SBET or porosity of RM, and improves an acidic functional group or active hydrogen at the RM surface as mentioned above. It is considered that the mechanical interfacial properties of epoxy/RM nanocomposites are increased due to the improvement of dispersivity of RM in a matrix and the increase of interaction potential between the RM surface and matrix.

TABLE 5 The K IC and G IC of the Epoxy/RM Nanocomposites Treated with Various Concentrations
K IC (MPa m1/2 ) As-received 0.1 M 1M 5M 3.15 3.47 3.65 4.26 G IC (kJ/m2 ) 5.64 5.71 6.15 6.97