1300
Ind. Eng. Chem. Res. 1998, 37, 1300-1309
Adsorption of Organic Acids on Polyaminated Highly Porous Chitosan: Equilibria Wataru Takatsuji*,† Industrial Technology Center of Wakayama Prefecture, 60 Ogura, Wakayama 649-6261, Japan
Hiroyuki Yoshida Department of Chemical Engineering, Osaka Prefecture University, 1-1 Gakuen-Cho, Sakai 599-8531, Japan
The adsorption of organic acids on a new weakly basic ion exchanger, highly porous polyaminated chitosan (Chitopearl CCS), which has the primary amino group of chitosan and the primary, secondary, and tertiary amino groups of poly(ethylene imine), appeared feasible technically. Three different organic acids, acetic acid (R′-COOH), malic acid (R′′-(COOH)2) and citric acid (R′′′-(COOH)3) were used in this experimental study. These organic acids were adsorbed on the resin by an acid/base neutralization reaction. The adsorption isotherms were independent of the initial concentrations of organic acids. The theoretical equations for the adsorption isotherms were derived by considering the dissociation of organic acids in the solution and the adsorption on each functional group and by applying the mass action law. They correlated the experimental adsorption isotherms and titration curves well. Chitopearl CCS was observed to be a feasible medium for the adsorption of organic acids. Especially in high pH region (low concentration region), Chitopearl CCS could adsorb more organic acids than a commercial ion exchanger, DIAION WA30. 1. Introduction Organic acids have been used in a number of chemical industries. Recently, attention has been focused on the main components of biodegradable plastics. Almost all organic acids have been produced by decomposition of organics using a microorganism or fermentation. Waste sludge may become important organics that produce organic acids, and this may give one of the solutions to solve the big problem of waste sludge. After producing organic acids by the above methods, separation and purification are important from the standpoint of cost and quality of products. Although public information is limited, organic acids have been separated by means of esterification and calcium precipitation in almost all commercial plants. These methods do not offer highly purified organic acids. In addition, they consume a lot of energy and also the capital costs are high. Recently, new purification technologies, such as electrodialysis and adsorption, have been investigated to overcome the above problems. Weakly basic resins are commonly used for removal and/or recovery of acids from industrial aqueous streams. Helfferich (1965) proposed a model that HCl was adsorbed on weakly basic resins by ion exchange accompanied by an acid/base neutralization reaction. After his review, Adams et al. (1969), Warner and Kennedy (1970), Hubner and Kadlec (1978), Rao and Gupta (1982a,b), Helfferich and Hwang (1985) and Bhandari et al. (1992, 1993) presented papers dealing with the adsorption of acids on weakly basic resins. However, few studies have been reported on the adsorption of weak acids on weakly basic resins. Takatsuji and Yoshida (1994) reported that DIAION WA30 (Mit† Telephone: INT+81-734-77-1271. Fax: INT+81-734-772880.
subishi Chemical Co., Japan, a comercial weakly basic resin) and Chitopearl CCS (Fuji Spinning Co., Japan, polyaminated highly porous chitosan bead) were excellent adsorbents for organic acids from wine, which contained ethanol, glucose, and various organic acids. They also presented the theoretical equations for the adsorption isotherms of the organic acids on DIAION WA30, which has a tertiary amino group as a fixed functional group (Takatsuji and Yoshida, 1997). As standard weakly basic ion exchangers such as Dowex WGR, DIAION WA10 and Russian AN have two to three different fixed amino groups in the resin phase, a more general approach is necessary for understanding the adsorption of organic acids on weakly basic resins. In the present work, we have investigated the adsorption isotherms of organic acids on Chitopearl CCS. It is a new weakly basic ion exchanger that was fabricated by introducing poly(ethylene imine) into the macropore of highly porous cross-linked chitosan. Chitosan is produced by the deacetylation of chitin, which is a natural biopolymer extracted from the shell of arthropods such as lobsters, shrimp, and crabs. Since such arthropods are abundantly available, chitosan may be produced from them very cheaply, and as chitosan is harmless to human, it may be utilized for ion exchangers and adsorbents in the food and pharmaceutical industries. Chitopearl CCS may have at least four different fixed groups, the primary amino group of chitosan and the primary, secondary, and tertiary amino groups of poly(ethylene imine). The derivation of the theoretical equations on the adsorption of organic acids is more complicated than DIAION WA30. Yoshida et al. (1994) derived theoretical equations for adsorption of HCl on the polyaminated highly porous chitosan bead. In this work, to make clear the effect of the number of carboxylic group of organic acid on the adsorption
S0888-5885(97)00567-8 CCC: $15.00 © 1998 American Chemical Society Published on Web 02/18/1998
Ind. Eng. Chem. Res., Vol. 37, No. 4, 1998 1301 Table 1. Experimental Physical Properties of Chitopearl CCS and DIAION WA30 functional group
diameter (cm)
apparent density (kg/m3)
porosity
R′, -NHR; R′′, -NHCH2C(OH)HCH2-PEIc
0.06254
1136
0.831
0.06505
1131
0.486
resin chitosan derivative (spherical particles) Chitopearl CCSa styrene DIAION WA30b
–CH–CH2– CH–
–CH– a
CH2N(CH3)2
b
Fuji Spinning Co., Ltd. Mitsubishi Chemical Co., Ltd. c PEI: poly(ethylene imine) of which the molecular weight is 10000. Structure of Chitosan resin:
isotherm, acetic acid (R′-COOH), malic acid (R′′(COOH)2), and citric acid (R′′′-(COOH)3) were used in the experimental study. Assuming that these organic acids are adsorbed by the neutralization reaction with each functional group on the weakly basic resin and applying the mass action law, we derived theoretical equations for adsorption isotherms. The experimental adsorption isotherms and titration curves are compared with those theoretical equations. Further, the results obtained in the present work are compared with the results for adsorption of organic acids on DIAION WA30, which has one kind of functional group. 2. Experimental Section Before measuring the adsorption isotherms, about 10 cm3 of the ion exchange particles were placed in a column, and 0.5 kmol/m3 of NaOH aqueous solution was allowed to flow through the bed at a flow rate of 7 × 10-5 m3/h for 3 days. Thereafter the bed was washed with distilled and deionized water thoroughly. Finally, the resin particles were kept in pure water. Adsorption isotherms were measured by the batch method. Before the resin particles were weighed, the water around the particles was removed using a centrifugal filter (Sanyou Rikagaku-kiki Seisakusho) rotated at 5000 rpm for 3 min. Thereafter, they were put in contact with the HCl solution or each organic acid solution and well mixed. The amounts of HCl, acetic acid, malic acid, and citric acid adsorbed on the resin were measured after 5, 7, and 11 days. Since there was no difference between the results for 5 and 7 days with respect to HCl, the resin particles and HCl solution were put in contact for 5 days. Since the equilibrium for acetic acid, malic acid and citric acid were fully reached after 7 days, the resin particles and each organic acid solution were put in contact for 7 days. The pH of the solution was analyzed with a Horiba pH meter F-7AD. The concentration of organic acid was analyzed with a Shimadzu high performance liquid chromatography organic acid analysis system. The adsorbed phase concentration was calculated according to eq 1, where
q)
(C0 - C)FV W
(1)
C0 and C are the initial concentration and the equilib-
Table 2. Experimental Coefficients of HCl and Acetic Acid (AH)a i)C
i ) P1
i ) P2
i ) P3
tot.
Adsorption of HCl Chitopearl CCS Ki,HCl (m3/kmol) 8.2 × 105 76 Qi (kmol/m3) 0.598 0.614 DIAION WA30 Ki,HCl (m3/kmol) Qi (kmol/m3)
1.4 × 103 2.9 × 104 0.930 0.692 2.83 3.8 × 104 2.80 2.80
Adsorption of AH Chitopearl CCS Ki,A (m3/kmol) Qi,A (kmol/m3) DIAION WA30 Ki,A (m3/kmol) Qi,A (kmol/m3) a
5.5 × 105 19 0.630 0.463
34 0.868
2.8 × 103 0.690 2.65 1.5 × 102 2.94 2.94
AH: CH3COOH.
rium concentration in the liquid phase (kmol/m3), respectively. q denotes the adsorbent-phase concentration (kmol/m3 of wet resin). V and W are the volume of the solution (m3) and the weight of the wet resin particles (kg), respectively. F is the apparent density (kg of wet resin/m3 of wet resin). All experiments were carried out at 293 K. 3. Ion Exchanger The ion exchanger used in this experimental study was a new weakly basic resin, Chitopearl CCS (Fuji Spinning Co.). Chitopearl CCS was fabricated by introducing poly(ethylene imine) (hereafter called PEI) with a molecular weight of 10 000 into the cross-linked chitosan, and its functional groups consisted of four different amino groups, the primary amino group of chitosan and the primary, secondary and tertiary amino groups of PEI. The experimental physical properties of Chitopearl CCS are compared with those of DIAION WA30 (Mitsubishi Chemical Co.), which is a commerial weakly basic resin, in Table 1. The concentration of each amino group in the resin was determined by measuring the equilibrium isotherm for adsorption of HCl. The experimental equilibrium coefficients K and saturation capacities Q were determined according to the procedure presented by Yoshida et al. (1994). They are listed in Table 2.
1302 Ind. Eng. Chem. Res., Vol. 37, No. 4, 1998
4. Results and Discussion 4.1. Adsorption of Acetic Acid. Figure 1a shows the experimental adsorption isotherm for the Chitopearl CCS-acetic acid system. The adsorption isotherm is not affected by the initial liquid-phase concentration of acetic acid CA0. Since acetic acid molecule (AH) has one carboxylic group, it may be adsorbed on each functional group of the resin by the following reactions: Ka
AH y\z H+ + AKC,A
RC-NH2 + AH y\z RC-NH3+AKP3,A
RP3-N + AH y\z RP3-NH+AKP2,A
RP2-NH + AH y\z RP2-NH2+AKP1,A
RP1-NH2 + AH y\z RP1-NH3+A-
(2) (3) (4) (5) (6)
The elementary reactions are shown in Appendix 1. Applying the mass action law to eqs 2-6, eqs 7-10 are derived, where qC,A, qP3,A, qP2,A, and qP1,A denote the
qC,A )
qP3,A )
qP2,A )
qP1,A )
KC,AQC,A[AH] 1 + KC,A[AH] KP3,AQP3,A[AH] 1 + KP3,A[AH] KP2,AQP2,A[AH] 1 + KP2,A[AH] KP1,AQP1,A[AH] 1 + KP1,A[AH]
(7)
(8)
(9)
(10)
equilibrium concentrations of acetic acid adsorbed on RC-NH2, RP3-N, RP2-NH, and RP1-NH2, respectively (kmol/m3 wet resin). QC,A, QP3,A, QP2,A, and QP1,A show the saturation capacities on RC-NH2, RP3-N, RP2-NH, and RP1-NH2, respectively (kmol/m3 wet resin). [AH] and [A-] represent the equilibrium concentrations of CH3COOH and CH3COO- in the liquid phase (kmol/ m3), respectively. [AH] can be calculated using the Henderson-Hasselbalch equation (Glasstone and Lewis, 1960). The equilibrium concentration of acetic acid in the liquid phase is given by eq 11, where pKa is 4.56 at -
CA ) [AH] + [A ] )
10-pH(1 + 10pH-pKa) 10pH-pKa
(11)
293 K (Martell and Smith, 1974, 1975, 1977). The total amount of acetic acid adsorbed on Chitopearl CCS is given by eq 12.
qA ) qC,A + qP3,A + qP2,A + qP1,A
Figure 1. Experimental equilibrium isotherms for adsorption of acetic, malic, and citric acids on Chitopearl CCS. Key: (O) C0 ) 0.6 kmol/m3; (4) C0 ) 0.52 kmol/m3; (0) C0 ) 0.5 kmol/m3; (3) C0 ) 0.42 kmol/m3; (]) C0 ) 0.4 kmol/m3; (y) C0 ) 0.3 kmol/m3; (5) C0 ) 0.26 kmol/m3; (!) C0 ) 0.25 kmol/m3; (8) C0 ) 0.2 kmol/m3; (*) C0 ) 0.12 kmol/m3; (Y) C0 ) 0.1 kmol/m3; (6) C0 ) 0.05 kmol/ m3; (@) C0 ) 0.03 kmol/m3; (7) C0 ) 0.02 kmol/m3, (&) C0 ) 0.01 kmol/m3; (b) C0 ) 0.005 kmol/m3; (2) C0 ) 0.002 kmol/m3; (9) C0 ) 0.001 kmol/m3.
(12)
The values of K and Q in eqs 7-10 were determined by the same procedure as the adsorption for HCl shown by Yoshida et al. (1994). Since RC-NH2 shows the strongest basicity, only eq 3 occurs in the first step. The
Figure 2. Plots of data for pH > 4.8 for adsorption of acetic acid on Chitopearl CCS based on eq 13.
adsorption isotherm is expressed by eq 7, and it is transformed to eq 13.
[AH] ) -
[AH] 1 + QC,A KC,A qA
(13)
Figure 2 shows the plots of the data for pH > 4.8 based on eq 13. The straight line was determined using the least-squares method in pH > 4.8. The correlation coefficient was 0.969. KC,A and QC,A were determined from the intercept and slope of the straight line, respectively. They are listed in Table 2. Next, acetic acid is adsorbed on RC-NH2 and RP3N, which shows the second strongest basicity, simultaneously in the second step. Therefore, eqs 3 and 4 occur simultaneously, and eq 14 is valid. Equation 14 is transformed to eq 15 using eq 7, in which the values of
Ind. Eng. Chem. Res., Vol. 37, No. 4, 1998 1303
Figure 3. Plots of data for 3.5 < pH < 4.8 for adsorption of acetic acid on Chitopearl CCS based on eq 15.
Figure 5. Plots of data for 2.3 < pH < 4.8 for adsorption of acetic acid on Chitopearl CCS based on eq 18.
Figure 4. Plots of data for 2.7 < pH < 4.8 for adsorption of acetic acid on Chitopearl CCS based on eq 17.
KC,A and QC,A were determined above (Table 2). Figure
qA )
KC,AQC,A[AH] 1 + KC,A[AH]
[AH] ) -
+
KP3,AQP3,A[AH] 1 + KP3,A[AH]
[AH] 1 + QP3,A KP3,A qA - qC,A
(14)
step), eq 18 is derived, where KP1,A and QP1,A are
(15)
3 shows the plots of the data for 3.5 < pH < 4.8 based on eq 15. The data were correlated by a straight line. The correlation coefficient was 0.973. KP3,A and QP3,A determined from the straight line are listed in Table 2. Similarly, assuming that eqs 3-5 occur simultaneously in the third step, eqs 16 and 17 are obtained,
qA )
KC,AQC,A[AH] 1 + KC,A[AH]
+
KP3,AQP3,A[AH]
+ 1 + KP3,A[AH] KP2,AQP2,A[AH] 1 + KP2,A[AH]
[AH] ) -
Figure 6. Titration curve for adsorption of acetic acid on Chitopearl CCS and DIAION WA30. Key: (sbs) Chitopearl CCS; (--O--) DIAION WA30.
(16)
[AH] 1 + QP2,A (17) KP2,A qA - (qC,A + qP3,A)
where the values of KC,A, QC,A, KP3,A, and QP3,A are known (Table 2). Figure 4 shows the plots of the data for 2.7 < pH < 4.8 based on eq 17. The data were correlated by a straight line. The correlation coefficient was 0.865. KP2,A and QP2,A obtained from the straight line are listed in Table 2. Finally, assuming that all reactions of eqs 3-6 occur simultaneously in the whole pH region (in the fourth
[AH] ) -
[AH] 1 + QP1,A KP1,A qA - (qC,A + qP3,A + qP2,A) (18)
unknown and other values of K and Q are known (Table 2). Figure 4 shows the plots of the data for 2.3 < pH < 4.8 based on eq 18. The data are correlated by a straight line for pH > 2.3. The correlation coefficient was 0.912. KP1,A and QP1,A determined from the straight line are listed in Table 2. Figure 6 shows the titration curves for adsorption of acetic acid on chitopearl CCS and DIAION WA30. The solid lines in Figures 1a and 6 show the theoretical ones calculated from eqs 7-12. They correlate the data reasonably well. As can be seen from Table 2 the saturation capacities of each amino group for adsorption of acetic acid coincide nearly with those for adsorption of HCl. The dashed line in Figure 6 shows the theoretical one for the adsorption of the acetic acid on DIAION WA30 calculated from eq 8 with K and Q listed in Table 2 (Takatsuji and Yoshida, 1997). The two theoretical lines in Figure 6 intersect at about pH 3.5. For pH < 3.5 DIAION WA30 was able to adsorb the more acetic acid than Chitopearl CCS, but for pH > 3.5 the amount of acetic acid adsorbed on Chitopearl CCS surpassed DIAION WA30. 4.2. Adsorption of Malic Acid. Figure 1b shows the experimental equilibrium isotherm for adsorption of malic acid on Chitopearl CCS. The adsorption isotherm is not affected by the initial liquid-phase
1304 Ind. Eng. Chem. Res., Vol. 37, No. 4, 1998
concentration of malic acid, CM0. Since malic acid (MH2) has two carboxylic groups, it may be adsorbed on the resin by the following reactions: Kma1
MH2 y\z H+ + MHKma2
MH- y\z H+ + M2KC,M1
RC-NH2 + MH2 y\z RC-NH3+MHKC,M2
2RC-NH2 + MH2 y\z
(RC-NH3+)2M2-
KP3,M1
RP3-N + MH2 y\z RP3-NH+MHKP3,M2
2RP3-N + MH2 y\z (RP3-NH+)2M2KP2,M1
RP2-NH + MH2 y\z RP2-NH2+MH-
(19) (20) (21)
qM ) qC,M + qP3,M + qP2,M + qP1,M (22) (23) (24) (25)
KP2,M2
KP1,M1
(27)
KP1,M2
2RP1-NH2 + MH2 y\z (RP1-NH3+)2M2- (28) The elementary reactions are shown in Appendix 1. Applying the mass action law to eqs 19-28, and using the procedure of the previous paper (Takatsuji and Yoshida, 1997), eqs 29-32 were obtained. Eqs 21 and 22:
qC,M ) QC,M/2 + (1 - KC,M1[MH2]){(1 + KC,M1[MH2]) -
x(1 + KC,M1[MH2])2 + 8KC,M2[MH2]QC,M}/
8KC,M2[MH2] (29)
Eqs 23 and 24:
CM ) [MH2] + [MH-] + [M2-] ) 10pH-pKma1 + 2 × 10pH-pKma1 × 10pH-pKma2
8KP3,M2[MH2] (30)
Eqs 25 and 26:
qP2,M ) QP2,M/2 + (1 - KP2,M1[MH2]){(1 + KP2,M1[MH2]) -
x(1 + KP2,M1[MH2])2 + 8KP2,M2[MH2]QP2,M}/
8KP2,M2[MH2] (31)
Eqs 27 and 28:
qP1,M ) QP1,M/2 + (1 - KP1,M1[MH2]){(1 + KP1,M1[MH2]) -
x(1 + KP1,M1[MH2])2 + 8KP1,M2[MH2]QP1,M}/
8KP1,M2[MH2] (32)
(34)
and 4.71 at 293 K, respectively (Martell and Smith, 1974, 1975, 1977). Resin has multiple sites of amino groups. There is a possibility for the carboxylic groups of malic acid to adsorb on different types of amino groups of the resin. However since the basicities of different types of amino groups are different, two carboxylic groups of malic acid may adsorb on the same type of amino group at a given pH value. In addition the distance between carboxylic groups in one molecule of malic acid is long enough to associate with the two amino groups in chitosan and with the two amino groups of the same type in PEI. Therefore the values of K in eqs 29-32 were determined according to the procedures presented by Yoshida et al. (1994), and Takatsuji and Yoshida (1997). We also assumed that QC,M ) QC, QP3,M ) QP3, QP2,M ) QP2, and QP1,M ) QP1. When the concentration of malic acid in the liquid phase is very low, eqs 21 and 22 occur since RC-NH2 shows the strongest basicity. The adsorption isotherm for malic acid in the first step is expressed by eq 29. When qC,M ) QC/2 is substituted in eq 29, eq 35 is
KC,M1 )
qP3,M ) QP3,M/2 + (1 - KP3,M1[MH2]){(1 + KP3,M1[MH2]) -
x(1 + KP3,M1[MH2])2 + 8KP3,M2[MH2]QP3,M}/
(33)
calculated by using the Henderson-Hasselbalch equation (Glasstone and Lewis, 1960). The equilibrium concentration of malic acid in the liquid phase CM (kmol/ m3) is given by eq 34, where pKma1 and pKma2 are 3.24
10-pH(1 + 10pH-pKma1 + 10pH-pKma1 × 10pH-pKma2)
2RP2-NH + MH2 y\z (RP2-NH2+)2M2- (26) RP1-NH2 + MH2 y\z RP1-NH3+MH-
Here [MH2], [MH-], and [M2-] represent the equilibrium concentrations of C4H6O5, C4H5O5-, and C4H4O52- in the liquid phase (kmol/m3), respectively. qP3,M, qP3,M, qP2,M, and qP1,M denote the equilibrium amounts of malic acid adsorbed on RC-NH2, RP3-N, RP2-NH, and RP1-NH2, respectively (kmol/m3 wet resin). QC,M, QP3,M, QP2,M, and QP1,M show the saturation capacities on RC-NH2, RP3N, RP2-NH, and RP1-NH2, respectively (kmol/m3 wet resin). The total amount of malic acid adsorbed on Chitopearl CCS is given by eq 33. [MH2] can be
[
]
1 [MH2]
) QC/2
(35)
qM
obtained. The value of KC,M1 can be determined from the value of [MH2] at qM ) QC/2 on the experimental adsorption isotherm. The value of KC,M2 is determined from eq 36, to which eq 29 is transformed. QC + 2(qM
QC +
(
)
QC (1 + KC,M1[MH2]) 2 ) 1 - KC,M1[MH2]
2 qM -
(
)
QC 2 [MH2] 2 (36) KC,M2 (1 - KC,M1[MH2])2 8 qM -
- QC/2)(1 + KC,M1[MH2])/(1 - KC,M1[MH2]) was plotted against 8(qM - QC/2)2[MH2]/(1 - KC,M1[MH2])2 using the data in pH > 6.3. The plots were correlated well by a straight line. The correlation coefficient was 0.936. The value of KC,M2 was determined from the slope of the straight line. KC,M1 and KC,M2 are listed in Table 3.
Ind. Eng. Chem. Res., Vol. 37, No. 4, 1998 1305 Table 3. Experimental Coefficients of Malic Acid (MH2) and Citric Acid (CH3)a i)C
i ) P1
i ) P2
i ) P3
25 350
8.1 × 104 3.9 × 108
Adsorption of MH2 Chitopearl CCS Ki,M1 (m3/kmol) Ki,M2 (m3/kmol)2 DIAION WA30 Ki,M1 (m3/kmol) Ki,M2 (m3/kmol)2
5.6 × 1010 8.4 × 1010
3.6 11
3.5 × 104 2.1 × 107 Adsorption of CH3
Chitopearl CCS Ki,C1 (m3/kmol) Ki,C2 (m3/kmol)2 Ki,C3 (m3/kmol)3 DIAION WA30 Ki,C1 (m3/kmol) Ki,C2 (m3/kmol)2 Ki,C3 (m3/kmol)3
1.0 × 1011 8.0 × 1015 1.0 × 1017
6.8 1.5 0.015
41 54 3.3
qM ) qC,M + QP3/2 + (1 - KP3,M1[MH2]){(1 + KP3,M1[MH2]) -
x(1 + KP3,M1[MH2])2 + 8KP3,M2[MH2]QP3}/
8KP3,M2[MH2] (37)
When qM - qC,M ) QP3/2 is substituted in eq 37, eq 38 is obtained. The value of KP3,M1 can be determined
[
]
1 [MH2]
qM-qC,M
) QP3/2
(
(
)
QP3 + 2(qM - qC,M - QP3/2)(1 + KP3,M1[MH2])/(1 KP3,M1[MH2]) was plotted against 8(qM - qC,M - QP3/ 2)2[MH2]/(1 - KP3,M1[MH2])2 using the data for 6.3 > pH > 5.4 based on eq 39. The plots were correlated well by a straight line for 6.3 > pH > 5.4. The correlation coefficient was 0.921. KP3,M1 and KP3,M2 are listed in Table 3. In the third step, eqs 21-26 occur simultaneously and eq 40 is obtained. When qM - qC,M - qP3,M ) QP2/2 is
qM ) qC,M + qP3,M + QP2/2 + (1 - KP2,M1[MH2]){(1 + KP2,M1[MH2]) -
x(1 + KP2,M1[MH2])2 + 8KP2,M2[MH2]QP2}/
8KP2,M2[MH2] (40)
substituted into eq 40, eq 41 is obtained. The value of
(41)
)
QP2 (1 + KP2,M1[MH2]) 2 ) 1 - KP2,M1[MH2] QP2 2 8 qM - qC,M - qP3,M [MH2] 2 (42) KP2,M2 (1 - KP2,M1[MH2])2
2 qM - qC,M - qP3,M -
(
)
eq 40 is transformed. QP2 + 2(qM - qC,M - qP3,M - QP2/ 2)(1 + KP2,M1[MH2])/(1 - KP2,M1[MH2]) was plotted against 8(qM - qC,M - qP3,M - QP2/2)2[MH2]/(1 KP2,M1[MH2])2 using the data for 6.3 > pH > 2.6 based on eq 42. The plots were correlated by a straight line. The correlation coefficient was 0.847. KP2,M1 and KP2,M2 are listed in Table 3. Finally, eqs 21-28 occur simultaneously in the fourth step, and eq 43 is obtained. The value of KP1,M1 can be
qM ) qC,M + qP3,M + qP2,M + QP1/2 + (1 - KP1,M1[MH2]){(1 + KP1,M1[MH2]) -
x(1 + KP1,M1[MH2])2 + 8KP1,M2[MH2]QP1}/
8KP1,M2[MH2] (43)
determined from the value of [MH2] at qM - qC,M - qP3,M - qP2,M ) QP1/2 on the experimental adsorption isotherm as shown by eq 44. The value of KP1,M2 is determined
)
QP3 2 qM - qC,M (1 + KP3,M1[MH2]) 2 QP3 + ) 1 - KP3,M1[MH2] QP3 2 8 qM - qC,M [MH2] 2 (39) KP3,M2 (1 - KP3,M1[MH2])2
) QP2/2
qM-qC,M-qP3,M
(38)
from the value of [MH2] at qM - qC,M ) QP3/2 on the experimental adsorption isotherm. The value of KP3,M2 is determined from eq 39, to which eq 37 is transformed.
(
]
1 [MH2]
QP2 +
1.0 × 105 8.0 × 108 2.0 × 1010
Next, malic acid is adsorbed on RC-NH2 and RP3-N simultaneously in the second step. Therefore, eqs 2124 occur simultaneously and eq 37 is valid.
[
KP2,M1 can be determined from the value of [MH2] at qM - qC,M - qP3,M ) QP2/2 on the experimental adsorption isotherm. The value of KP2,M2 is determined from eq 42 to which
1.4 × 105 6.0 × 108 1.0 × 1011
a Key: MH , C(CH COOH)H(OH)COOH; CH , C(CH COOH) 2 2 3 2 2 (OH)COOH.
KP3,M1 )
KP2,M1 )
KP1,M1 )
[
1 [MH2]
]
qM-qC,M-qP3,M-qP2,M
) QP1/2
(44)
from eq 45 to which eq 43 is transformed. QP1 + 2(qM QP1 +
(
)
QP1 (1 + KP1,M1[MH2]) 2 ) 1 - KP1,M1[MH2]
2 qM - qC,M - qP3,M - qP2,M -
(
8 qM - qC,M - qP3,M - qP2,M KP1,M2
(1 - KP1,M1[MH2])
2
)
QP1 2 [MH2] 2
(45)
- qC,M - qP3,M - qP2,M - QP1/2)(1 + KP1,M1[MH2])/(1 KP1,M1[MH2]) was plotted against 8(qM - qC,M - qP3,M qP2,M - QP1/2)2[MH2]/(1 - KP1,M1[MH2])2 using the data for 6.3 > pH > 1.8 based on eq 45. The plots were correlated by a straight line. The correlation coefficient was 0.770. KP1,M1 and KP1,M2 are listed in Table 3. Figure 7 shows the titration curves for adsorption of malic acid on Chitopearl CCS and DIAION WA30. The solid lines in Figures 1b and 7 show the theoretical ones for Chitopearl CCS calculated from eqs 29-34 using the
1306 Ind. Eng. Chem. Res., Vol. 37, No. 4, 1998
values of K given in Table 3. They correlate the data reasonably well. The saturation capacity for adsorption of malic acid is total concentration of amino groups fixed in the resin. The dashed line in Figure 7 shows the theoretical one for DIAION WA30 calculated from eq 30 with K listed in Table 3 (Takatsuji and Yoshida, 1997). For pH > 5 Chitopearl CCS could adsorb more malic acid than DIAION WA30, because only RC-NH2 could react with malic acid for pH > 6. For pH < 5 DIAION WA30 with a greater concentration of RP3-N, which shows the second strongest basicity, adsorbed more malic acid than Chitopearl CCS. 4.3. Adsorption of Citric Acid. Figure 1c shows the experimental equilibrium isotherm for adsorption of citric acid on Chitopearl CCS. The adsorption isotherm is not affected by the initial liquid-phase concentration of citric acid CC0. Since citric acid (CH3) has three carboxylic groups, it may be adsorbed on the resin by the following reactions: Kca1
CH3 y\z H+ + CH2Kca2
CH2- y\z H+ + CH22-
CH
Kca3
+
3-
y\z H + C KC,C1
RC-NH2 + CH3 y\z RC-NH3+CH2-
Figure 7. Titration curve for adsorption of malic acid on Chitopearl CCS and DIAION WA30. Key: (sbs) Chitopearl CCS; (--O--) DIAION WA30.
Qi,C + 3 2Ki,C22[CH3] 2Ki,C2 1 - 2Ki,C1[CH3] + + 27Ki,C3 9Ki,C3
qi,C )
(46)
(
(47)
(
)
)
4Ki,C2 2 1 A + 3 27Ki,C1Ki,C3 3Ki,C1[CH3] i Ki,C2 A2i (i ) C, P3, P2, P1) (61) 2 3Ki,C1 [CH3]
(48) (49) where
KC,C2
2RC-NH2 + CH3 y\z (RC-NH3+)2CH2- (50) KC,C3
3RC-NH2 + CH3 y\z
(RC-NH3+)3C3-
KP3,C1
RP3-N + CH3 y\z RP3-NH+CH2KP3,C2
2RP3-N + CH3 y\z (RP3-NH+)2CH2KP3,C3
3RP3-N + CH3 y\z (RP3-NH+)3C3KP2,C1
RP2-NH + CH3 y\z RP2-NH2+CH2-
(
Ai ) -
(51)
(53)
KP1,C1
RP1-NH2 + CH3 y\z RP1-NH3+CH2KP1,C2
2RP1-NH2 + CH3 y\z (RP1-NH3+)2CH2-
+ -
Ri )
)
bi - R 2 x i θi 3
(57) (58) (59)
KP1,C3
3RP1-NH2 + CH3 y\z (RP1-NH3+)3C3- (60) The elementary reactions are shown in Appendix 1. Applying the mass action law to eqs 46-60, and using the procedure of the previous paper (Takatsuji and Yoshida, 1997), eq 61 was obtained
bi )
(
32Ki,C3
Ki,C1 -
33Ki,C32
(
32Ki,C3
33Ki,C3
(Ri < 0)
)
22Ki,C22
2Ki,C13Ki,C2 23Ki,C22
(62) (63)
(64)
(65)
2x-a3i
(55) Ki,C12
(Ri > 0)
bi
cos θi ) -
ai )
1/3
b2i + a3i 4
(54)
KP2,C2
KP2,C3
1/3
Ai ) 2x-ai cos
(52)
2RP2-NH + CH3 y\z (RP2-NH2+)2CH2- (56) 3RP2-NH + CH3 y\z (RP2-NH2+)3C3-
) (
bi + R 2 x i
[CH3]2 +
)
Ki,C12
[CH3] 32Ki,C3 (66)
- Ki,C1 [CH3]3 -
(
)
Ki,C13 2Ki,C2 + Qi,C [CH3]2 (67) 3Ki,C3 32K i,C3
where [CH3], [CH2-], [CH2-], and [C3-] denote the equilibrium concentrations of C6H8O7, C6H7O7-, C6H6O72-, and C6H5O73- in the liquid phase (kmol/m3), respectively. qC,C, qP3,C, qP2,C, and qP1,C are the equilibrium amounts of citric acid adsorbed on RC-NH2, RP3N, RP2-NH, and RP1-NH2, respectively (kmol/m3 wet
Ind. Eng. Chem. Res., Vol. 37, No. 4, 1998 1307
adsorb more citric acid than DIAION WA30 for pH > 5. This is because only RC-NH2 on Chitopearl CCS, which shows the strongest basicity, could react with citric acid in the low concentration for pH > 6, but since DIAION WA30 has a higher concentration of RP3-N than Chitopearl CCS, it could adsorb more citric acid for pH < 5. 5. Conclusion
Figure 8. Titration curve for adsorption of ciric acid on Chitopearl CCS and DIAION WA30. Key: (sbs) Chitopearl CCS; (--O--) DIAION WA30.
resin). QC,C, QP3,C, QP2,C, and QP1,C show the saturation capacities on RC-NH2, RP3-N, RP2-NH, and RP1-NH2, respectively (kmol/m3 wet resin). The total amount of citric acid adsorbed on Chitopearl CCS is given by eq 68. [CH3] can be calculated by using the Henderson-
qC ) qC,C + qP3,C + qP2,C + qP1,C
(68)
Hasselbalch equation (Glasstone and Lewis, 1960). The equilibrium concentration of citric acid in the liquid phase CC (kmol/m3) is given as
CC ) [CH3] + [CH2-] + [CH2-] + [C3-] ) 10-pH(1 + 10pH-pKca1 + 10pH-pKca1 × 10pH-pKca2 + 10pH-pKca1 × 10pH-pKca2 × 10pH-pKca3)/[10pH-pKca1 + 2 × 10pH-pKca1 × 10pH-pKca1 + 3 × 10pH-pKca1 × 10pH-pKca2 × 10pH-pKca3] (69) where pKca1, pKca2, and pKca3 are 2.87, 4.35, and 5.69 at 293 K, respectively (Martell and Smith, 1974, 1975, 1977). Since the distances among three carboxylic groups in one molecule of citric acid are long enough to associate with the three amino groups in chitosan and with the three amino groups of the same type in PEI and since the basicities of different types of amino groups are different, three carboxylic groups of citric acid may adsorb on the same type of amino group at a given pH value. We also assumed that QC,C ) QC, QP3,C ) QP3, QP2,C ) QP2, and QP1,C ) QP1, and the values of K in eq 61 were evaluated by fitting qC with experimental values in turn from high pH to low pH. They are listed in Table 3. Figure 8 shows the titration curves for adsorption of citric acid on Chitopearl CCS and DIAION WA30. The solid lines in Figures 1c and 8 show the theoretical ones for Chitopearl CCS calculated from eqs 61-69, and they correlate the data reasonably well. The saturation capacity for adsorption of citric acid is also total concentration of amino groups fixed in the resin. The dashed line in Figure 8 shows the theoretical one for DIAION WA30 (Takatsuji and Yoshida, 1997). Similar to the adsorption of malic acid, Chitopearl CCS could
The adsorption of organic acids (acetic acid, malic acid, and citric acid) on Chitopearl CCS, which is polyaminated highly porous chitosan, appeared feasible technically. 1. Chitopearl CCS could adsorb organic acids as well as DIAION WA30 which was a commerial weakly basic resin. In high pH region (low concentration region), Chitopearl CCS could adsorb more organic acids than DIAION WA30. 2. The adsorption isotherms were not affected by the initial concentration of organic acids. 3. The theoretical equations for the adsorption isotherms were derived by assumig that organic acids were adsorbed by an acid/base neutralization reaction on each fixed functional group (four different amino groups) and by considering the dissociation of organic acids. In an organic acid with one carboxylic group in one molecule, such as acetic acid, the adsorption isotherm was expressed by eqs 7-12. In an organic acid with two carboxylic groups in one molecule, such as malic acid, the adsorption isotherm was given by eqs 29-34. In an organic acid with three carboxylic groups in one molecule, such as citric acid, the adsorption isotherm was expressed by eqs 61-69. The equilibrium coefficients for adsorption on each fixed functional group were determined from the experimental titration curve in turn from high pH to low pH. The theoretical lines agreed resonably well with the experimental adsorption isotherms and titration curves. Nomenclature Cj ) equilibrium concentration of organic acid (j ) A; acetic acid, M; malic acid, C; citric acid) in liquid phase, kmol/ m3 Cj0 ) initial concentration of organic acid (j ) A; acetic acid, M; malic acid, C; citric acid) in liquid phase, kmol/m3 Ki,HCl ) equilibrium constant for adsorption of HCl on the functional group (i ) C; the primary amino group of chitosan, P1; the primary amino group of PEI, P2; the secondary amino group of PEI, P3; the tertiary amino group of PEI) fixed in adsorbent phase, m3/kmol Ka ) equilibrium constant (eq 2), kmol/m3 KC,A ) equilibrium constant (eq 3), m3/kmol KP1,A ) equilibrium constant (eq 6), m3/kmol KP2,A ) equilibrium constant (eq 5), m3/kmol KP3,A ) equilibrium constant (eq 4), m3/kmol Kma1 ) equilibrium constant (eq 19), kmol/m3 Kma2 ) equilibrium constant (eq 20), kmol/m3 KC,M1 ) equilibrium constant (eq 21), m3/kmol KP1,M1 ) equilibrium constant (eq 27), m3/kmol KP2,M1 ) equilibrium constant (eq 25), m3/kmol KP3,M1 ) equilibrium constant (eq 23), m3/kmol KC,M2 ) equilibrium constant (eq 22), (m3/kmol)2
1308 Ind. Eng. Chem. Res., Vol. 37, No. 4, 1998 KP1,M2 ) equilibrium constant (eq 28), (m3/kmol)2 KP2,M2 ) equilibrium constant (eq 26), (m3/kmol)2 KP3,M2 ) equilibrium constant (eq 24), (m3/kmol)2 Kca1 ) equilibrium constant (eq 46), kmol/m3 Kca2 ) equilibrium constant (eq 47), kmol/m3 Kca3 ) equilibrium constant (eq 48), kmol/m3 KC,C1 ) equilibrium constant (eq 49), m3/kmol KP1,C1 ) equilibrium constant (eq 58), m3/kmol KP2,C1 ) equilibrium constant (eq 55), m3/kmol KP3,C1 ) equilibrium constant (eq 52), m3/kmol KC,C2 ) equilibrium constant (eq 50), (m3/kmol)2 KP1,C2 ) equilibrium constant (eq 59), (m3/kmol)2 KP2,C2 ) equilibrium constant(eq 56), (m3/kmol)2 KP3,C2 ) equilibrium constant (eq 53), (m3/kmol)2 KC,C3 ) equilibrium constant (eq 51), (m3/kmol)3 KP1,C3 ) equilibrium constant (eq 60), (m3/kmol)3 KP2,C3 ) equilibrium constant (eq 57), (m3/kmol)3 KP3,C3 ) equilibrium constant (eq 54), (m3/kmol)3 Qi ) saturation capacity for adsorption of HCl on the functional group (i ) C; the primary amino group of chitosan, P1; the primary amino group of PEI, P2; the secondary amino group of PEI, P3; the tertiary amino group of PEI) fixed in adsorbent-phase, kmol/m3 Qi,j ) saturation capacity for adsorption of organic acid (j ) A; acetic acid, M; malic acid, C; citric acid) on the functional group (i ) C; the primary amino group of chitosan, P1; the primary amino group of PEI, P2; the secondary amino group of PEI, P3; the tertiary amino group of PEI) fixed in adsorbent phase, kmol/m3 qj ) equilibrium concentration of organic acid (j ) A; acetic acid, M; malic acid, C; citric acid) in adsorbent phase, kmol/m3 qi,j ) equilibrium concentration of organic acid (j ) A; acetic acid, M; malic acid, C; citric acid) on the functional group (i ) C; the primary amino group of chitosan, P1; the primary amino group of PEI, P2; the secondary amino group of PEI, P3; the tertiary amino group of PEI) fixed in adsorbent phase, kmol/m3 V ) volume of solution, m3 W ) weight of wet resin particle, kg F ) apparent density, kg of wet resin/m3 of wet resin
in the same manner as described above, respectively. MH2 Kma1
MH2 y\z H+ + MHKma2
MH- y\z H+ + M2K1
RC-NH2 + H+ y\z RC-NH3+ K3
RC-NH3+ + MH- y\z RC-NH3+MH-
Ka
AH y\z H+ + AK1
RC-NH2 + H+ y\z RC-NH3+ K2
RC-NH3+ + A- y\z RC-NH3+A-
KC,A
(A-3)
Equations 21 and 22 are obtained by summing eqs 19, 20, A-1, A-3, and A-4. KC,M1
RC-NH2 + MH2 y\z RC-NH3+MHKC,M2
2RC-NH2 + MH2 y\z (RC-NH3+)2M2-
(21) (22)
RP3-N, RP2-NH, and RP1-NH2 react with malic acid in the same manner as described above, respectively. CH3 Kca1
CH3 y\z H+ + CH2Kca2
CH2- y\z H+ + CH2Kca3
CH2- y\z H+ + C3K1
RC-NH2 + H+ y\z RC-NH3+ K5
RC-NH3+ + CH2- y\z RC-NH3+CH2K6
2RC-NH3+ + CH2- y\z (RC-NH3+)2CH2-
(46) (47) (48) (A-1) (A-5) (A-6) (A-7)
Equations 49, 50, and 51 are obtained by summing eqs 46, 47, 48, A-1, A-5, A-6, and A-7. KC,C1
RC-NH2 + CH3 y\z RC-NH3+CH2KC,C2
(49)
(2)
2RC-NH2 + CH3 y\z (RC-NH3+)2CH2
(A-1)
3RC-NH2 + CH3 y\z (RC-NH3+)3C3-
(A-2)
RP3-N, RP2-NH, and RP1-NH2 react with citric acid in the same manner as described above, respectively.
Equation 3 is obtained by summing eqs 2, A-1, and A-2.
RC-NH2 + AH y\z RC-NH3+A-
(A-1)
K4
3RC-NH3+ + C3- y\z (RC-NH3+)3C3-
The neutralization reactions between the amino groups fixed in the resin phase and organic acids in the liquid phase are given as follows: AH
(20)
2RC-NH3+ + M2- y\z (RC-NH3+)2M2- (A-4)
K7
Appendix 1
(19)
(3)
RP3-N, RP2-NH, and RP1-NH2 react with acetic acid
KC,C3
(50) (51)
Literature Cited Adams, G.; Jones, P. M.; Millar, J. R. Kinetics of Acid Uptake by Weak-base Anion Exchangers. J. Chem. Soc. A 1969, 2543. Bhandari, V. M.; Juvekar, V. A.; Patwardhan, S. R. Sorption Studies on Ion Exchange Resins. 1. Sorption of Strong Acids on Weak Base Resins. Ind. Eng. Chem. Res. 1992a, 31, 1060.
Ind. Eng. Chem. Res., Vol. 37, No. 4, 1998 1309 Bhandari, V. M.; Juvekar, V. A.; Patwardhan, S. R. Sorption Studies on Ion Exchange Resins. 2. Sorption of Weak Acids on Weak Base Resins. Ind. Eng. Chem. Res. 1992b, 31, 1073. Bhandari, V. M.; Juvekar, V. A.; Patwardhan, S. R. Sorption Sorption of Dibasic Acids on Weak Base Resins. Ind. Eng. Chem. Res. 1993, 32, 200. Glasstone, S.; Lewis, D. Elements of Physical Chemistry; D. Van Nostrand Company, Inc.: New York, 1960. Helfferich, F. Ion-Exchange Kinetics. V. Ion Exchange Accompanied by Reactions. J. Phys. Chem. 1965, 69, 1178. Helfferich, F. G.; Hwang, Y. L. Kinetics of Acid Uptake by Weak Base Anion Exchangers: Mechanism of Proton Transfer. AIChE Symp. Ser. 1985, 81, 17. Hubner, P.; Kadlec, V. Kinetics Behavior of Weak Base Anion Exchangers. AIChE J. 1978, 24, 149. Martell, A. E.; Smith, R. M. Critical Stability Constants.; Plenum Press: New York and London, 1974, 1975, and 1977. Rao, M. G.; Gupta, A. K. Ion Exchange Processes Accompanied by Ionic Reaction. Chem. Eng. J. 1982a, 24, 181. Rao, M. G.; Gupta, A. K. Kinetics of Ion Exchage in Weak Base Anion Exchange Resins. AIChE Symp. Ser 1982b, 78, 96.
Takatsuji, W.; Yoshida, H. Removal of Organic Acids from Wine by Adsorption on Weakly Basic Ion Exchangers: Equilibria for Single and Binary Systems. Sep. Sci. Technol. 1994, 29, 1473. Takatsuji, W.; Yoshida, H. Adsorption of Organic Acids on Weakly Basic Ion Exchanger: Equilibria, J. Chem. Eng. Jpn. 1997, 30, 396. Warner, R. E.; Kennedy, A. M. Kinetics of Neutralization of Weak Electrolyte Ion-Exchange Resins. J. Macromol. Sci. 1970, A4, 1125. Yoshida, H.; Kishimoto, N.; Kataoka, T. Adsorption of Strong Acid on Polyaminated Highly Porous Chitosan: Equilibria, Ind. Eng. Chem. Res. 1994, 33, 854.
Received for review August 19, 1997 Revised manuscript received January 9, 1998 Accepted January 10, 1998 IE970567X