- Email: [email protected]
Transport of organic dyes through ether-type polyurethane membrane
Talanta 49 (1999) 757 – 771
Transport of organic dyes through ether-type polyurethane membrane Kathy Rzeszutek, Art Chow * Department of Chemistry, Uni6ersity of Manitoba, Winnipeg, MB, Canada R3T 2N2 Received 27 October 1998; received in revised form 2 February 1999; accepted 4 February 1999
Abstract Transport of various anthraquinone, acidic and basic dyes in aqueous solution through ether-type polyurethane membrane has been studied to better define the factors affecting the removal of organic compounds by the polyurethane membrane and to complement the previously proposed sorption mechanism. The effects of pH, salts, dye geometry and size, initial dye concentration, thickness of the membrane, and solution temperature on the rate of transport were investigated. Transport was found to be dependent upon the pH conditions of the starting and the receiving solutions. An increased rate of transport was observed with increased solution temperature and with the use of a thinner polyurethane membrane. The differences in the rates of transport can be attributed to the relative solubility of the organic dyes in the membrane and in solution, and to the strength and extent of intermolecular interactions with the polymer. Dye concentration, geometry and size, and the presence of salts in solution had no significant effect on the rate of transport. All of the studied dyes were found to exist as neutral species in the membrane. © 1999 Elsevier Science B.V. All rights reserved. Keywords: Organic dyes; Ether-type polyurethane membrane; Non-porous membrane transport
1. Introduction Due to more rigorous government regulations for the disposal of waste effluents and the increased environmental awareness of the public, the manufacturers and industrial consumers of organic compounds have been forced to strictly monitor their wastewater for the presence of organic chemicals. As a result, more focus has been placed on the development of advanced waste * Corresponding author. Fax: +1-204-4747608. E-mail address: [email protected] (A. Chow)
purification techniques for the removal of organic compounds from industrial effluents. Currently, the most extensively employed purification procedure is adsorption of organic pollutants on activated carbon [1]. Unfortunately, this process alone is neither very efficient nor economical [1]. Because adsorption takes place from an aqueous solution, the organic species which are polar and soluble in water are not efficiently removed by the relatively non-polar carbon; water-insoluble nonpolar compounds tend to form colloidal dispersions from which migration and adsorption on the carbon surface are also not very effective [1].
0039-9140/99/$ - see front matter © 1999 Elsevier Science B.V. All rights reserved. PII: S 0 0 3 9 - 9 1 4 0 ( 9 9 ) 0 0 0 7 1 - 5
758
K. Rzeszutek, A. Chow / Talanta 49 (1999) 757–771
Current industrial wastewater purification operations are a combination of physical, chemical and/or biological degradation procedures which are applied prior to exposure of the effluent to carbon. However, this entire process is quite expensive and time-consuming because of the many steps involved [1]. In the last few years, applications of membrane separations to organic pollution control have been extensively researched [2 – 14]. Membrane separation is a relatively new type of separation process which is predicted, ultimately, to replace a majority of the conventional separation systems. At present, there is little interest in using biological membranes for this purpose. On the other hand, the possibility of separations with various synthetic polymeric membranes is highly appealing because no conventional single step purification treatment offers the same potential and versatility as those of synthetic membranes. In membrane processes, the separation is achieved via a membrane which acts as a barrier between two homogeneous phases and as a sorbing medium. The separation is governed by both the chemical and the physical nature of the membrane material and it occurs because of differences in shape, size, chemical properties, or charge of the species to be separated [2 – 14]. Transport through the membrane can be achieved when a driving force such as a pressure or concentration difference, for example, is applied or created during adsorption. Membrane technology is already being used to some extent in many industrial areas such as food and beverage, metallurgy, pulp and paper, textile, pharmaceutical, automotive, dairy, biotechnological and chemical industries. The membrane processes have become particularly important in water treatment for industrial and domestic water where membrane technology can be applied in cleaning operations [2 – 14]. For instance, in textile plants membranes are used to treat dye house effluents to remove the dyestuff and allow the reuse of auxiliary chemicals for dyeing, or to concentrate the dyestuffs and auxiliaries and produce purified water [15]. One of the more important categories of membranes comprises the non-porous polymeric membranes which are most widely adopted in current
industrial membrane separations [2–15]. Polyurethane membrane is an example of a nonporous synthetic membrane. Ether-type polyurethane membrane is manufactured through a copolymerisation condensation reaction of polyols with polyisocyanates [16]. These two components aggregate into two distinct regions during the reaction which results in the formation of hard and soft domains. The hard domain is very rigid in structure due to the presence of aromatic rings; the soft domain, comprising the polyol segments, is more fluid and amorphous. Extensive intermolecular hydrogen bonding takes place between the individual crystalline segments and the soft segments. As a result, the ether-type polyurethane membrane is resistant to extreme pH and temperature conditions which makes it an attractive candidate for industrial membrane separation applications. Our earlier work on the mechanism of extraction of phenols and benzoic acids by the polyurethane membrane shows that this polymer is capable of sorbing organic compounds from solution [17,18]. Carbon adsorption of organic dyes from waste effluents is relatively ineffective for the reasons previously mentioned, and consequently effluents containing organic dyes are very difficult to treat in environmental systems [1]. We chose to investigate the transport of organic dyes for the purpose of better defining the factors affecting sorption of organic compounds by the polyurethane membrane in general, and to explore the possibility of practical industrial applications. This comprehensive study should help to promote the use of the polyurethane membrane for separations in various segments of industry.
2. Experimental
2.1. Instrumental analysis and reagents UV-visible spectra and absorbance readings were taken using a Hewlett-Packard Model 8452A diode-array spectrophotometer. Solution pH was measured with an Orion Expandable Ion Analyser EA940. Water was obtained from a Barnstead Nanopure II™ purification system fed with water
K. Rzeszutek, A. Chow / Talanta 49 (1999) 757–771
759
Table 1 Anthraquinone organic dyes
purified by reverse osmosis. Polyether-type polyurethane membranes PT6100S and PT6310S 0.025 and 0.051 mm thick, respectively were supplied by Deerfield Urethane Inc., South Deerfield, MA. The organic dyes were obtained from BASF Co., Ludwigshafen, Germany and from a biological stain kit, Chem-Supply Model No. BS-100, provided by Chem Service, Inc., West Chester, PA. The molecular structures and the Colour Index (C.I.) numbers of the dyes used for this study are listed in Tables 1 – 3. The purity of these dyes was checked by TLC using a 50:10:15 ratio of n-butanol:ethanol:water mobile phase composition. Because the compounds appeared to be in their pure form, they were used without further purification. All other chemicals were of a reagent grade.
2.2. Experimental apparatus and procedure Enough of each dye was weighed out to give a final solution concentration of about 1.9 ×10 − 5 M in 1.0 M HCl or in 1.0 M NaOH unless otherwise specified. The choice of this concentration was dictated by the solubility limit of the least soluble dye in 1.0 M HCl. Parameters such as salt, acid, or base concentrations were adjusted accordingly.
The dye classification shown Tables 1–3 was adopted from textile science and is based not only on the type of substituents on the dye molecule but also on the application conditions used in textile operations utilizing those particular dyes [19]. The apparatus used for membrane testing consisted of two separate glass ‘cells’ (a starting and a receiving cell) having capacities of 260 ml. The flanges of both cells were covered lightly with Dow Corning high vacuum silicone grease in order to obtain a better seal of the two cells. All cells had short sidearms which allowed access for solutions and sampling. Ether-type polyurethane membrane 0.025 mm thick (unless otherwise specified) was cut into squares and placed between the flanges of the two cells which were then clamped together. The membrane was not subjected to any cleaning prior to use and we assumed that it does not swell in aqueous solution due to its hydrophobic nature. The surface area of the membrane exposed to solutions was 23.89 0.2 cm2. The starting cell contained 250 ml of aqueous dye solution; the receiving cell contained 250 ml of 1.0 M HCl solution without the dye (unless otherwise specified). Teflon stirrers were placed inside each cell to keep the solutions stirred throughout the experiment. Sep-
760
K. Rzeszutek, A. Chow / Talanta 49 (1999) 757–771
Table 2 Acidic organic dyes
arate magnetic stirrers were used for each cell. After the sample and the receiving solutions were transferred into the cells, the sidearm openings of the cells were closed with glass stoppers and Parafilm™ to eliminate evaporation and pH variations. The entire apparatus was then wrapped in aluminum foil for the duration of the experiment to prevent any possible decomposition of the dye due to light. Each experiment was repeated at least twice to establish reproducibility. The UV-visible spectra (190 – 820 nm) of the original solutions were taken prior to the experiment (AI) and compared with the absorbances of
the starting and receiving solutions taken after a known period of time (AFS, AFR).% dye in starting cell and % dye in receiving cell are the concentrations of dye in the starting and the receiving solutions, respectively at a known time. The concentration of dye in the membrane at a specific time, reported as % dye in membrane, can be calculated. % Dye in starting cell=AFS/AI × 100%
(1)
% Dye in receiving cell= AFR/AI × 100%
(2)
% Dye in membrane= 100%− (% dye in starting cell+ % dye in receiving cell)
(3)
K. Rzeszutek, A. Chow / Talanta 49 (1999) 757–771
761
Table 3 Basic organic dyes
3. 1 Results and discussion
3.1. Effect of the solution pH on the transport of organic dyes through polyurethane membrane Our previous work on the mechanism of extraction of phenols [17] and benzoic acids [18] by the polyurethane membrane shows that two conditions must be met in order for extraction to occur. First, the organic compound must be transferred from solution onto the membrane surface. Second, more of this compound will be removed from solution after the initial sorption only if the species sorbed on the membrane surface can migrate deeper into the polymer. An increase in the degree of sorption is observed until the capacity of the membrane has been reached. Through the study of the extraction of phenols and benzoic acids, we have found that the polyurethane membrane can only hold un-
charged organic species. Based on what we know about the mechanism of extraction of phenols and benzoic acids, we hypothesized that it should be possible to increase the amount of the organic compound removed from the starting solution if the sorbed compound could be transported through the membrane into a receiving solution. To test this hypothesis, we investigated the transport of a series of organic dyes (Tables 1–3) which are representatives of dyes frequently used in the textile industry [19,20]. The solution conditions for these dyes were chosen such that majority of the dye molecules would exist as charged or as neutral species. Both 1.0 M HCl and 1.0 M NaOH solutions were used for the extraction and transport of the anthraquinone dyes Alizarin and Alizarin Red S, all of the acidic dyes, and the basic dye Wool Green S. Anthraquinone Disperse Blue 14 is insoluble in
K. Rzeszutek, A. Chow / Talanta 49 (1999) 757–771
762
Table 4 Qualitative analysis of transport of anthraquinone, acidic and basic organic dyesa Dye Anthraquinone Alizarin Alizarin Red S Disperse Blue 14 Acidic Solvent Yellow 2 Methyl Red Methyl Orange Alizarin Yellow GG Eriochrome Black T Martius Yellow Naphthol Yellow S Basic Victoria Blue R Wool Green S
Dye in starting solution
Dye in receiving solution
Dye in membrane
1.0 1.0 1.0 1.0 1.0
M M M M M
HCl: yellow NaOH: violet HCl: light yellow NaOH: purple HCl: pink
Yellow None Light yellow None Pink
Yellow None Light yellow None Blue
1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0
M M M M M M M M M M M M M
HCl: red NaOH: yellow HCl: red/pink NaOH: yellow HCl: red NaOH: yellow HCl: orange NaOH: yellow HCl: burgundy NaOH: yellow HCl: light yellow NaOH: orange HCl: light yellow
Red None Red/pink None None None Orange None Burgundy None Yellow None Light yellow
Red None Red/pink None None None Orange None Burgundy None Yellow None Light yellow
None None Blue None None
None None Blue None None
1.0 M NaOH dark yellow 1.0 M HCl: yellow Water: blue 1.0 M HCl: green 1.0 M NaOH: blue
a Conditions: 25.0 92.0°C, 250-ml aliquot of : 1.9×10−5 M aqueous dye solution in the starting cell, 250-ml aliquot of solvent in the receiving cell, 0.025 mm thick ether-type membrane, active surface area of 23.8 9 0.2 cm2.
basic solution, and therefore it was tested only in 1.0 M HCl. The basic dye, Victoria Blue R, is also insoluble in basic solution but it dissolves in water (pH :5.0) and in 1.0 M HCl. Table 4 shows the qualitative results for the extraction and transport of the selected dyes from the corresponding starting solutions. The receiving solution for each dye had a similar acidity or basicity to the starting solution in which the dye was dissolved. The color of each dye in the starting solution, the receiving solution, and the membrane corresponds to the type of species present (i.e. charged vs. uncharged). The anthraquinone and the acidic dyes (with the exception of Methyl Orange) were found to extract and transport out of acidic solution but not from basic solution. The acidic Methyl Orange and the basic dye Wool Green S showed no extraction or trans-
port from either solution used. Basic Victoria Blue R was extracted and transported out of water but not out of acidic solution. In 1.0 M HCl, the anthraquinone dyes that have acidic substituents are neutral, e.g. Alizarin and Alizarin Red S. During extraction and transport of these dyes from acidic solution, the membrane had the same color as the starting dye solutions. In 1.0 M NaOH, Alizarin and Alizarin Red S are negatively charged; no extraction or transport was observed from basic solutions of these dyes. Organic dyes such as disperse blue 14 having basic substituents are positively charged in 1.0 M HCl. During the transport of Disperse Blue 14, the starting and receiving solutions contained charged species (pink), but the membrane was blue in color. The blue color of Disperse Blue 14 is observed when
K. Rzeszutek, A. Chow / Talanta 49 (1999) 757–771
Fig. 1. Transport of Disperse Blue 14 through 0.025 mm thick ether-type polyurethane membrane. Conditions: 25.0 9 2.0°C, 250-ml aliquot of : 1.9× 10 − 5 M aqueous dye solution in 1.0 M HCl in the starting cell, 250-ml aliquot of 1.0 M HCl in the receiving cell, 0.025 mm thick ether-type membrane, active surface area of 23.8 9 0.2 cm2. Decrease in dye concentration in the starting cell, ; increase in dye concentration in the receiving cell, .
the dye molecules are uncharged (e.g. when the dye is dissolved in a non-polar solvent such as hexane). The data collected on transport of the anthraquinone dyes suggests that only a neutral form of these dyes is retained by the membrane. Clear visual evidence for this conclusion is the transport of Disperse Blue 14 during which we observe a color change when the dye is being transferred from solution into the polymer and vice versa. Similar conclusions were reached from extraction and transport of acid dyes. In 1.0 M HCl, a majority of acidic dye species will be neutral. Similarly to the anthraquinone dyes, the acidic dyes were extracted and transported from and into acidic solution but not from basic solution. Methyl Orange, however, was an exception. The lack of extraction of this dye can be accounted for by the high probability that the dye never actually exists as an extractable uncharged species i.e. the closest instance in which it resembles an uncharged species is when it forms a zwitterion. The
763
net charge on the zwitterion is zero, but the individual ends of this species are either positively or negatively charged. Consequently, no initial transfer onto the membrane surface can occur, and hence no transport into the receiving cell was observed. Similar reasoning can be used to account for the unextractability of basic Wool Green S. Like Methyl Orange, the probability of Wool Green S being uncharged is unlikely due to the presence of the basic substituents and the sulfonic acid group on the dye molecule. Under acidic conditions, the sulfonic acid group on the dye will be neutral, but amino groups (other than the ion-paired substituent which is not considered to be truly charged because the positive charge is closely associated with the negative ions) may be protonated. The opposite will be true in basic solution. Since no neutral species is formed, there is no extraction and therefore no transport. Basic Victoria Blue R, has only basic substituents in addition to the ion-paired quaternary ammonium group. The majority of Victoria Blue R molecules are neutral in water (again, the ion-paired group is not considered to be charged), and therefore this dye is extracted and transported out of water. No extraction or transport occurred out of 1.0 M HCl due to the presence of positively charged species. These results further support the hypothesis that the presence of a neutral species of an organic compound is a deciding factor which determines whether the initial transfer of an organic compound from solution onto the membrane surface will take place. After successful initial sorption, the transport through the membrane into a receiving solution will occur only if the receiving solution conditions are favorable for desorption of the organic species from the membrane. For example, the anthraquinone dyes are transported into acidic receiving solutions but not into water in which they are only sparingly soluble. In conclusion, the transport of organic dyes from one solution into another using the polyurethane membrane as the transporting medium is contingent upon both the initial sorption of the dye from solution onto the polymer surface and the chemical properties of the receiving solution.
K. Rzeszutek, A. Chow / Talanta 49 (1999) 757–771
764
3.2. Effect of dye geometry, size, and chemical properties on transport through polyurethane membrane As pointed out in Section 3.1, the first step in the overall transport process (Fig. 3) is the sorption of the dye onto the membrane surface. It is hypothesized that the efficiency with which the dye in solution is initially transferred onto the polymer surface will be dependent on the relative solubilities of the dye in the solvent and in the membrane which will take into account all the different possible specific and non-specific interactions [21]. After the initial sorption has taken place, the ability of the sorbed species to separate the individual segments of the polymer and migrate deeper into the bulk is largely driven by the strength and the extent of the intermolecular interactions that can take place between the species and the constituent groups on the hard and soft segments of the polymer [17,18]. Because all of the dyes used in this study are relatively non-polar with substituents capable of intermolecular interactions and aromatic rings which are also found in the matrix of the membrane, they should be soluble in this polymer. For the remainder of this discussion we will only be concerned with the neutral form of the dyes since
it is the only species that is retained by the membrane. Using this information on extraction and transport within the polymer matrix, we anticipated that the rate of transport of the dye from the starting into the receiving solution through the polymer will be affected by the morphology and the chemical characteristics of the dye molecule. To test this hypothesis, the dyes were dissolved in the solvent from which extraction and transport were previously observed (Table 4) to give a concentration of 1.9× 10 − 5 M. Ether-type 0.025 mm thick polyurethane membrane was used in all instances. The results reported in Table 5 show the time at which no further change was measured in the receiving and the starting cells, the percentage dye detected in each of the cells at that time, and the percentage dye calculated to be in the membrane. Unlike all other dyes that were tested, Eriochrome Black T and Naphthol Yellow S show significantly lower concentrations in the receiving cell than in the starting cell at the reported times. This suggests that the transport of the two dyes must be very slow, and therefore the change in concentration at the time of measurement may be too small to detect. We found that some dyes, for example Alizarin, Disperse Blue 14 and Solvent Yellow 2, are not well retained by the membrane but quite efficiently
Table 5 Effect of dye morphology and chemical properties on the rate of transporta Dye
Time (h)
% Dye in starting cell
% Dye in receiving cell
% Dye in membrane
Anthraquinone Alizarin Alizarin Red S Disperse Blue 14
163 929 147
4893 5093 479 1
42 94 45 93 46 9 1
10 93 5 93 791
Acidic Solvent Yellow 2 Methyl Red Alizarin Yellow GG Eriochrome Black T Martius Yellow Naphthol Yellow S
235 183 240 216 43 408
54 9 1 189 2 1493 3993 29 9 3 909 1
45 9 1 15 92 5 92 15 94 29 93 8 93
191 65 9 1 81 9 2 46 9 3 42 9 3 2 91
Basic Victoria Blue R
733
30 9 1
26 91
40 91
Conditions: 25.0 92.0°C, 250-ml aliquot of : 1.9×10−5 M aqueous dye solution in the starting cell in a solvent from which transport occurs (Table 4), 250-ml aliquot of solvent (identical to that in the starting cell) in the receiving cell, 0.025 mm thick ether-type membrane, active surface area of 23.8 90.2 cm2. a
K. Rzeszutek, A. Chow / Talanta 49 (1999) 757–771
transported. Other dyes such as Methyl Red, Alizarin Yellow GG, Eriochrome Black T, Martius Yellow, and Victoria Blue R are very well retained and transported through the polyurethane membrane. Two of the studied dyes, Alizarin Red S and Naphthol Yellow S, are very poorly extracted and transported by this polymer. Dyes which show very inefficient extraction and transport (Alizarin Red S and Naphthol Yellow S) are small, compact and both contain a single sulfonic acid substituent (Tables 1 and 2). Because of the small size of these dyes, the effect of individual substituents on the chemical characteristics of the entire molecule is more significant than it would be for a larger dye. Thus, in the case of Alizarin Red S and Naphthol Yellow S, the presence of the sulfonic acid group imparts more polarity to both compounds. The relatively polar character of these dyes hinders efficient initial sorption onto the membrane surface and results in the very slow transport into the receiving cell. The dyes which were effectively retained and transported by the membrane are generally non-polar in nature (Tables 2 and 3). For example Eriochrome Black T, in spite of the presence of the sulfonic acid group, remains quite non-polar because of a greater number of non-polar aromatic rings. Unlike Alizarin Red S and Naphthol Yellow S, this dye is sorbed and transported quite efficiently. Martius Yellow which has chemical structure identical to that of Naphthol Yellow S (Table 2) but lacks the sulfonic acid group is highly retained by the membrane and very quickly transported which again indicates that it is the resultant polarity of the dye molecule (and therefore the solubility) and not the polarity of the individual substituents that plays an important role in the transport process. Dyes which showed high extraction and efficient transport are quite diverse in geometry and size. Examples are large non-linear dyes such as Eriochrome Black T and Victoria Blue R, smaller linear Methyl Red and Alizarin Yellow GG, and even smaller Martius Yellow (Tables 2 and 3). We can therefore conclude that the morphology and size of the dye alone are not important factors controlling the transport process. The percent concentration of the linear dyes, Methyl Red and Alizarin Yellow GG, in the membrane is significantly higher than that of, for
765
Fig. 2. Transport of Disperse Blue 14 through 0.051 mm thick ether-type polyurethane membrane. Conditions: 25.0 9 2.0°C, 250-ml aliquot of :1.9 ×10 − 5 M aqueous dye solution in 1.0 M HCl in the starting cell, 250-ml aliquot of 1.0 M HCl in the receiving cell, 0.051 mm thick ether-type membrane, active surface area of 23.8 90.2 cm2. Decrease in dye concentration in the starting cell, ; increase in dye concentration in the receiving cell, .
example, Eriochrome Black T or Martius Yellow. This can be accounted for by the presence of carboxylic acid groups on both Methyl Red and Alizarin Yellow GG which are also found in the polymer matrix. The presence of the carboxylic acid groups increases the probability for more hydrogen bonding interactions between the dye and the polymer, and increases the solubility of the dye in the polymer matrix. These results support the hypothesis on the importance of intermolecular
Fig. 3. Pictorial representation of transport of an organic dye through a polyurethane membrane. SS is the starting solution, D is the neutral dye species, MS is the membrane surface on the starting solution side, MB is the membrane bulk, MS’ is the membrane surface on the receiving solution side, RS is the receiving solution.
766
K. Rzeszutek, A. Chow / Talanta 49 (1999) 757–771
interactions. Dyes such as Alizarin, Disperse Blue 14 and Solvent Yellow 2 (Tables 1 and 2) were not well retained by the membrane, but they were transported quite efficiently. These dyes are capable of forming hydrogen bonds and are relatively non-polar. The low concentration of these dyes in the membrane, in spite of the effective transport, must be a result of their relative solubility in the solvent and in the polymer. The strength of hydrogen bonding of the dyes with the solvent must be slightly higher than with the polymer, and therefore we observed preferential partitioning of these dyes into the aqueous phase. In conclusion, the rate of transport is not controlled by the geometry or size of the dye molecule but rather by the relative solubility of the dye in the polymer and in the solvent, and by the relative strength and extent of the intermolecular interactions with the polymer and the solvent. These findings are in agreement with the conclusions reached from our earlier study of the mechanism of extraction of phenols and benzoic acids by the polyurethane membrane [17,18].
3.3. Dependence of the rate of transport on thickness of the polyurethane membrane The relationship between the thickness of the polyurethane membrane and the rate of transport of the dye molecules through this polymer is exemplified by the transport of Disperse Blue 14. This dye having a concentration of 1.9× 10 − 5 M in 1.0 M HCl was transported into 1.0 M HCl using 0.025 and 0.051 mm thick ethertype polyurethane membranes. Figs. 1 and 2 show the time taken to achieve an equilibrium between the starting cell, the receiving cell and the membrane when the 0.025 and 0.051 mm thick membranes were used respectively. Equilibrium was achieved within about 195 h with the 0.025 mm thick membrane and within 219 h with the 0.051 mm thick membrane. At the observed equilibrium, the 0.025 mm thick membrane contained 6 92% of the starting concentration of the dye, and the 0.051 mm thick membrane contained 1991% of the initial
Table 6 Effect of the initial concentration of Disperse Blue 14 on the rate of transporta Initial concentration of Disperse Blue 14 1.9×10−6 M Equilibrium time (h) 146 % Dye in starting cell 27 9 4 % Dye in receiving 27 94 cell % Dye in membrane 46 9 4
1.9×10−5 M 146 4891 45 91 7 91
a Conditions: 25.0 92.0°C, 250-ml aliquot of aqueous dye solution of appropriate concentration in the starting cell in 1.0 M HCl, 250-ml aliquot of 1.0 M HCl in the receiving cell, 0.025 mm thick ether-type membrane, active surface area of 23.8 90.2 cm2.
dye solution concentration. The masses corresponding to the active surface (23.89 0.2 cm2) of 0.025 mm and 0.051 mm thick membranes prior to exposure to the dye were : 0.120 and 0.235 g, respectively. An attempt was made to determine the overall diffusion coefficient for the transport process. At a steady state, the diffusion coefficient can be evaluated by using a so called ‘time-lag’ method [22,23]. Initially, in the non-stationary state the amount of dye diffusing through the membrane is mathematically represented by Table 7 Effect of the initial concentration of Solvent Yellow 2 on the rate of transporta Initial concentration of Solvent Yellow 2 1.9×10−6 M Equilibrium time (h) 235 % Dye in starting cell 55 9 1 % Dye in receiving 43 9 1 cell % Dye in membrane 291
1.9×10−5 M 235 529 1 47 9 1 1 91
Conditions: 25.0 92.0°C, 250-ml aliquot of aqueous dye solution of appropriate concentration in the starting cell in 1.0 M HCl, 250-ml aliquot of 1.0 M HCl in the receiving cell, 0.025 mm thick ether-type membrane, active surface area of 23.8 90.2 cm2. a
K. Rzeszutek, A. Chow / Talanta 49 (1999) 757–771
767
Table 8 Successive removal of Disperse Blue 14 from aqueous solutiona Dye concentration in starting cell at equilibrium (M) 1.9×10−5
Initial conditionsb Initial transport First successive transport Second successive transport
Dye concentration in receiving cell Dye concentration in membrane at at equilibrium (M) equilibrium (M) 0
0
8.9×10−6 4.3×10−6
8.7×10−6 3.9×10−6
1.4×10−6 1.6×10−6
2.1×10−6
2.3×10−6
1.5×10−6
Conditions: 25.0 9 2.0°C, 250-ml aliquot of aqueous dye solution in the starting cell in 1.0 M HCl, 250-ml aliquot of 1.0 M HCl in the receiving cell, 0.025 mm thick ether-type membrane, active surface area of 23.8 9 0.2 cm2. b Note: initial conditions represent the initial concentration in the starting cell before any extraction or transport has occurred; the time required to reach equilibrium was : 147 h for the initial and successive transports. a
Q = FDxCA exp
Dt 1 2 ( −1)n − − % Dx 2 6 p 2n n2
n
(− D 2nn 2) x2
(4)
As time increases and the equilibrium is approached, the above equation can be simplified to Q=
DCAF Dx 2 t− Dx 6D
n
(5)
If the thickness of the membrane, Dx, and the time, Dt, taken to reach the steady state (‘timelag’) are known, the diffusion coefficient can be calculated using [22,23] D=
Dx 2 6Dt
(6)
The transport of disperse blue 14 through the 0.025 mm thick membrane required about 195 h for the dye in the membrane to reach a steady state (i.e. when no further change in dye concentration was observed in the starting and receiving solutions and in the membrane). The approximate diffusion coefficient calculated using Eq. (6) is 5.34×10 − 7 mm2 h − 1. The transport of Disperse Blue 14 through the 0.051 mm thick membrane required about 219 h to reach a steady state. The approximate value of the diffusion coefficient is 1.98×10 − 6 mm2 h − 1. As expected, the overall equilibrium for the transport of Disperse Blue 14 was achieved more quickly when the thinner membrane was used.
This can be accounted for by the shorter path that the dye had to diffuse through in order to reach the receiving cell solution. The more quickly the dye appears on the receiving cell side of the membrane, the more rapidly it can be desorbed into the receiving solution. The concentration of the dye in the membrane increases until the steady state is reached where no further sorption or transport are observed. The amount of dye in the thinner membrane was calculated to be approximately one third of that in the thicker polymer. The thicker membrane can retain a higher amount of the dye because it has more of the polymer material by weight that is available for interactions with the dye molecules. The calculated diffusion coefficients for the transport are broad approximations of the true values. The mathematical representation of the entire diffusion is beyond the scope of this paper and will be addressed in a later publication.
3.4. Dependence of the rate of transport on the initial dye solution concentration Disperse Blue 14 and Solvent Yellow 2 were used to study the effect of the initial concentration of the dye on the rate of transport through the polyurethane membrane. The choice of the initial dye concentrations in the sample cells was dictated by the solubility of the dyes in 1.0 M HCl and the detection limit of the UV-visible spectrophotometer. The starting concentrations used
K. Rzeszutek, A. Chow / Talanta 49 (1999) 757–771
768
for both Solvent Yellow 2 and Disperse Blue 14 were 1.9× 10 − 6 and 1.9× 10 − 5 M. Table 6 and Table 7 show the time taken to achieve equilibrium at the two concentrations for both dyes, and the corresponding percentage of each dye in the starting and the receiving cells and in the membrane. Regardless of the starting concentration used, equilibrium was achieved within :146 h for Disperse Blue 14 and 235 h for Solvent Yellow 2. When 1.9 ×10 − 6 M Disperse Blue 14 was extracted, a higher percentage of the dye was retained by the membrane and lower final concentrations were detected in the starting and the receiving solutions than in the extraction of 1.9× 10 − 5 M solution of this dye. This was not observed for the transport of the corresponding solutions of Solvent Yellow 2. In the case of Disperse Blue 14, although the membrane contained a higher proportion of the dye, the calculated molar concentration of the dye in the membrane was approximately the same as the molar concentration for the extraction of the more concentrated solution of Disperse Blue 14. This suggests that the membrane can remove the dye from solution until its capacity for the dye is reached. In the case of Solvent Yellow 2, the capacity of the membrane has probably been reached at 2% even at the 1.9× 10 − 6 M dye concentration. Therefore, a change in the percentage of this dye in the membrane was not observable as with the Disperse Blue 14. To determine whether the polyurethane membrane already containing a dye could be used to
transport more dye if a fresh receiving solution was provided, we investigated a stepwise removal of Disperse Blue 14. The starting concentration of Disperse Blue 14 was 1.9× 10 − 5 M. After equilibrium was achieved between the two cells and the membrane, the molar dye concentrations in both cells and in the membrane were calculated. Next, the receiving cell solution containing the dye was discarded and replaced with new 1.0 M HCl solution. The cells were again allowed to come to equilibrium, and the concentrations were calculated. This process was repeated until an instrumental detection limit for the dye was reached. The calculated molar concentrations of Disperse Blue 14 in the starting and receiving cells at equilibrium for each successive extraction are shown in Table 8. The concentration of Disperse Blue 14 in the membrane gradually increased during the initial extraction/transport until equilibrium. This membrane dye concentration remained constant for all of the subsequent successive extraction/transport experiments regardless of the time at which the measurements were taken and the solution dye concentration being transported. Furthermore, because the time needed to achieve equilibrium for all of the successive extractions was approximately the same as for the initial extraction/ transport, we can conclude that the transporting properties of the membrane have not been affected by the presence of the dye in the matrix. These results prove that the membrane will extract the dye until its capacity for the dye is reached (i.e. the solubility limit for the dye in the membrane is
Table 9 Effect of salt on the rate of transport of Disperse Blue 14 and Solvent Yellow 2a Equilibrium time (h)
% Dye in starting cell
% Dye in receiving cell
% Dye in membrane
Disperse Blue 14 No salt added 0.5 M LiCl
147 195
489 1 489 2
469 1 409 2
891 10 9 2
Solvent Yellow 2 No salt added 0.5 M LiCl
234 234
5291 589 2
469 1 419 2
291 192
Conditions: 25.0 9 2.0°C, 250-ml aliquot of : 1.9×10−5 M aqueous dye solution in the starting cell in 1.0 M HCl with salt concentration appropriately adjusted, 250-ml aliquot of 1.0 M HCl in the receiving cell, 0.025 mm thick ether-type membrane, active surface area of 23.89 0.2 cm2. a
K. Rzeszutek, A. Chow / Talanta 49 (1999) 757–771
769
Table 10 Effect of temperature on the rate of transport of Disperse Blue 14a Temperature (°C)
% Dye in starting cell at 24 h
% Dye in receiving cell at 24 h
% Dye in membrane at 24 h
4 23 37 66
8891 699 1 589 1 459 1
691 24 9 1 36 91 43 91
691 791 691 79 1
Conditions: temperature, 9 2.0°C, 250-ml aliquot of : 1.9×10−5 M aqueous dye solution in the starting cell in 1.0 M HCl, 250-ml aliquot of 1.0 M HCl in the receiving cell, 0.025 mm thick ether-type membrane, active surface area of 23.8 9 0.2 cm2. a
reached). In order for this membrane to continue removing the remaining dye in the starting solution, a flow of dye through the membrane into the receiving solution must occur.
3.5. Effect of salt on the rate of transport through polyurethane membrane Disperse Blue 14 and Solvent Yellow 2 were chosen as examples to illustrate the effect of salt on the rate of transport. The extraction efficiency and the time taken to achieve equilibrium for transport of both dyes (1.9× 10 − 5 M) from 1.0 M HCl solutions containing 0.5 M LiCl were compared with the results for solutions without the salt. 0.5 M and 1.0 M aqueous solutions of NaCl, KCl, RbCl, CaCl2, and Al2(SO4)3 in 1.0 M HCl and in water were also tested to determine whether any transport of the cations and/or anions was occurring. The percentages of the dyes in the starting and the receiving cells and in the membrane at equilibrium for the sample solutions with and without salt are presented in Table 9. None of the cations and anions were transported through the membrane from any of the starting salt solutions as confirmed by analysis using atomic absorption spectroscopy. The extraction efficiency of Disperse Blue 14 and Solvent Yellow 2 from starting solutions containing the salt and the rates of transport into the receiving cells are almost identical (within experimental error) to the corresponding solutions of these dyes without the salt. Furthermore, the presence of salt does not change the saturation amount of each of the dyes that can be retained by the polyurethane membrane. As discussed in Section 3.1, the majority of Disperse Blue 14 species in 1.0 M HCl are posi-
tively charged. These positively charged species are, however, at equilibrium with a small percentage of uncharged molecules of Disperse Blue 14. The opposite is true for the solution of Solvent Yellow 2 in 1.0 M HCl in which the predominant species is the uncharged dye molecule. When salt is added to solutions containing charged organic species, ion pair formation can take place [24–26] and the overall solution becomes more polar. Because of the increased solution polarity in the presence of salts, we would expect a ‘salting out’ effect where the non-polar uncharged species becomes more insoluble in the solvent, and therefore partitions more efficiently into the non-polar membrane which results in a more rapid extraction and transport of both dyes. Because no significant changes in extraction and transport rates have been observed, we concluded that addition of salt does not cause a sufficiently high change in the relative solubility of the dye species in solution and the membrane. These findings are in agreement with the conclusions reached for the earlier investigation of the extraction of phenols and benzoic acids by the polyurethane membrane [17,18].
3.6. Dependence of the rate of transport on solution temperature 1.9× 10 − 5 M solutions of Disperse Blue 14 in 1.0 M HCl were transported into a 1.0 M HCl solution using 0.025 mm polyurethane membrane at 4, 23, 37, and 66°C. The percentages of the dye in the starting and the receiving cells and in the membrane at 24 h for each of the temperatures are shown in Table 10. The percentage of the dye retained by the membrane is approximately the
770
K. Rzeszutek, A. Chow / Talanta 49 (1999) 757–771
same at all four temperatures. However, the rate of removal of the dye from the starting cell and the rate of transport into the receiving cell were found to increase with increased temperature. The above results do not correlate with the temperature data that we collected for extraction of phenols and benzoic acids by the ether-type polyurethane membrane [17,18]. For the extraction of phenols and benzoic acids, we found that as the temperature was increased, the extraction decreased because of the exothermic nature of the extraction process. In this study, however, in addition to extraction we have transport into a receiving solution. Unlike the simple extraction, the transport is probably controlled not only by thermodynamic factors but also by the concentration gradient created during the process. The fact that the membrane retains approximately the same amount of dye at all of the temperatures suggests that the solubility of the dye in the polymer remains relatively the same. The increased rates of removal and transport of the dye at higher temperatures suggest that the physical properties of the membrane matrix must be different at those temperatures. The most likely explanation is that due to the higher energy in the membrane segments at the increased temperature, some of the intermolecular interactions among the individual segments within the polymer may be weakened or broken; this results in the matrix being more fluid in nature and, consequently, more accessible to the dye species in solution. Therefore, at higher temperatures (e.g. 66°C), the dye species sorbed on the surface can migrate more freely through the bulk of the polymer and into the receiving solution which is evidenced by the significantly shorter time in which equilibrium is achieved (e.g. within 146 h at 23°C vs. 24 h at 66°C).
4. Conclusion The rate at which various organic dyes will transport through the ether-type polyurethane membrane is dependent on the solution pH, the overall polarity and capability of the dye molecule to engage in hydrogen bonding interactions with the polymer, the relative solubility of the dye
species in the polymer and in the solvent, and the temperature at which the transport is occurring. The ether-type polyurethane membrane can retain only neutral species of the organic dyes. In order to obtain an initial transport through the bulk of the membrane, the pH of the starting solvent must be favorable for the formation of a neutral species which is the only species soluble in the membrane. Furthermore, a flow of the species in the membrane into the receiving solution will be observed only if a driving force resulting from the differences in dye concentration in the starting and receiving solutions and the membrane (or a concentration gradient) is present. The rate of transport is dependent on chemical factors such as the efficiency with which the dye can partition into the membrane surface and then migrate through the membrane bulk, and on the physical experimental conditions such as the temperature. For example, species with comparable solubility in the polymer and in solution is transported from the starting cell into the receiving cell quite efficiently. On the other hand, if the solubility of the species is relatively higher in one of the media, the transport may be hindered due preferential partitioning. The partitioning of the dye between the solution and membrane phases and the rate of migration through the bulk of the polymer can be significantly altered by changing the temperature. The rate of transport of the dyes through the polyurethane membrane increased when higher temperatures were used. The increase in the transport rate at increased temperatures can be accounted for by change in the physical properties of the membrane matrix which becomes more fluid, and therefore more exposed and accessible to the dye species in solution. Overall, the results from this study provide a deeper insight into the mechanism of extraction and transport process of organic compounds by the polyurethane membrane and show that this membrane has considerable potential in commercial applications for removal of organic compounds from aqueous solutions.
Acknowledgements The authors would like to acknowledge the
K. Rzeszutek, A. Chow / Talanta 49 (1999) 757–771
financial support of the University of Manitoba and the Natural Sciences and Engineering Research Council of Canada.
[12] [13]
References [1] A. Reife, H.S. Freeman, Environmental Chemistry of Dyes and Pigments, Wiley-Interscience, New York, 1996, pp. 3 – 32. [2] R. Singh, CHEMTECH 28 (4) (1998) 33. [3] R. Singh, CHEMTECH 26 (6) (1996) 46. [4] K. Scott, R. Hughes, Industrial Membrane Separation Technology, Chapman and Hall, London, 1996. [5] M. Mulder, Basic Principles of Membrane Technology 2nd ed., Kluwer Academic Publishers, Dordrecht/Boston/ London, 1996. [6] J.A. Howell, A. Noworyta, Towards Hybrid Membrane and Biotechnology Solutions for Polish Environmental Problems, Wrocl*aw Technical University Press, Wrocl*aw, 1995. [7] K. Scott, Handbook of Industrial Membranes, 1st ed., Elsevier Advanced Technology, 1995. [8] R.D. Noble, S.A. Stern, Membrane Separation Technology: Principles and Applications, Elsevier Science, Amsterdam, 1995. [9] T. Matsuura, Synthetic Membranes and Membrane Separation Processes, CRC Press, Boca Raton, 1993. [10] P. Aptel, Membrane pressure driven processes in water treatment, in: J.G. Crespo, K.W. Broddeker (Eds.), Membrane Processes in Separation and Purification, NATO ASI Series, Series E: Applied Sciences, Kluwer Academic Publishers, 272 (1993) 263. [11] M. Mulder, The use of membrane separation processes in environmental problems, in: J.G. Crespo, K.W. Brod-
[14]
[15]
[16]
[17] [18] [19]
[20]
[21]
[22]
[23] [24] [25] [26]
771
deker, Membrane Processes in Separation and Purification, NATO ASI Series, Series E: Applied Sciences, Kluwer Academic Publishers, 272 (1993) 233. W.S. Winston Ho, K.K. Sirkar, Membrane Handbook, Van Norstrand Reinhold, New York, 1992. J.A. Howell, The membrane alternative: energy implications for industry, Report number 21, Elsevier, London and New York, 1990, pp. 2 – 8. L. Cecille, J.-C. Toussaint, Future industrial prospects of membrane processes, Elsevier, London and New York, 1989, pp. 3 – 7, 44, 55. L. Cecille, J.-C. Toussaint, Future Industrial Prospects of Membrane Processes, Elsevier, London and New York, 1989, pp. 190 – 195. G. Woods, The ICI Polyurethanes Book, ICI Polyurethanes and John Wiley & Sons Publication, Toronto, 1987, pp. 27 – 54. K. Rzeszutek, A. Chow, Talanta 46 (1998) 507. K. Rzeszutek, A. Chow, Talanta 47 (1998) 697. D.M. Nunn, The Dyeing of Synthetic-Polymer and Acetate Fibres, Dyers Company Publications Trust, Bradford, West Yorkshire, 1979, pp. 2 – 53. E.P.G. Gohl, L.D. Vilensky, Textile Science: an Explanation of Fibre Properties, 2nd ed., Longman Cheshire Pty Ltd., Melbourne, 1983, pp. 120 – 164. C.L. Bird, W.S. Boston, The Theory of Coloration of Textiles, The Dyers Company Publications Trust, Bradford, West Yorkshire, England, 1975, pp. 294 – 298. I.D. Rattee, M.M. Breuer, The Physical Chemistry of Dye Adsorption, Academic Press, London and New York, 1974, pp. 72 – 77. J. Crank, The Mathematics of Diffusion, Clarendon Press, Oxford, 1975, pp. 44 – 68. L. Shumack, A. Chow, Talanta 34 (1987) 957. P. Fong, A. Chow, Talanta 39 (1992) 497. R. Werbowesky, A. Chow, Talanta 43 (1996) 263.
. .
Transport of organic dyes through ether-type polyurethane membrane Kathy Rzeszutek, Art Chow * Department of Chemistry, Uni6ersity of Manitoba, Winnipeg, MB, Canada R3T 2N2 Received 27 October 1998; received in revised form 2 February 1999; accepted 4 February 1999
Abstract Transport of various anthraquinone, acidic and basic dyes in aqueous solution through ether-type polyurethane membrane has been studied to better define the factors affecting the removal of organic compounds by the polyurethane membrane and to complement the previously proposed sorption mechanism. The effects of pH, salts, dye geometry and size, initial dye concentration, thickness of the membrane, and solution temperature on the rate of transport were investigated. Transport was found to be dependent upon the pH conditions of the starting and the receiving solutions. An increased rate of transport was observed with increased solution temperature and with the use of a thinner polyurethane membrane. The differences in the rates of transport can be attributed to the relative solubility of the organic dyes in the membrane and in solution, and to the strength and extent of intermolecular interactions with the polymer. Dye concentration, geometry and size, and the presence of salts in solution had no significant effect on the rate of transport. All of the studied dyes were found to exist as neutral species in the membrane. © 1999 Elsevier Science B.V. All rights reserved. Keywords: Organic dyes; Ether-type polyurethane membrane; Non-porous membrane transport
1. Introduction Due to more rigorous government regulations for the disposal of waste effluents and the increased environmental awareness of the public, the manufacturers and industrial consumers of organic compounds have been forced to strictly monitor their wastewater for the presence of organic chemicals. As a result, more focus has been placed on the development of advanced waste * Corresponding author. Fax: +1-204-4747608. E-mail address: [email protected] (A. Chow)
purification techniques for the removal of organic compounds from industrial effluents. Currently, the most extensively employed purification procedure is adsorption of organic pollutants on activated carbon [1]. Unfortunately, this process alone is neither very efficient nor economical [1]. Because adsorption takes place from an aqueous solution, the organic species which are polar and soluble in water are not efficiently removed by the relatively non-polar carbon; water-insoluble nonpolar compounds tend to form colloidal dispersions from which migration and adsorption on the carbon surface are also not very effective [1].
0039-9140/99/$ - see front matter © 1999 Elsevier Science B.V. All rights reserved. PII: S 0 0 3 9 - 9 1 4 0 ( 9 9 ) 0 0 0 7 1 - 5
758
K. Rzeszutek, A. Chow / Talanta 49 (1999) 757–771
Current industrial wastewater purification operations are a combination of physical, chemical and/or biological degradation procedures which are applied prior to exposure of the effluent to carbon. However, this entire process is quite expensive and time-consuming because of the many steps involved [1]. In the last few years, applications of membrane separations to organic pollution control have been extensively researched [2 – 14]. Membrane separation is a relatively new type of separation process which is predicted, ultimately, to replace a majority of the conventional separation systems. At present, there is little interest in using biological membranes for this purpose. On the other hand, the possibility of separations with various synthetic polymeric membranes is highly appealing because no conventional single step purification treatment offers the same potential and versatility as those of synthetic membranes. In membrane processes, the separation is achieved via a membrane which acts as a barrier between two homogeneous phases and as a sorbing medium. The separation is governed by both the chemical and the physical nature of the membrane material and it occurs because of differences in shape, size, chemical properties, or charge of the species to be separated [2 – 14]. Transport through the membrane can be achieved when a driving force such as a pressure or concentration difference, for example, is applied or created during adsorption. Membrane technology is already being used to some extent in many industrial areas such as food and beverage, metallurgy, pulp and paper, textile, pharmaceutical, automotive, dairy, biotechnological and chemical industries. The membrane processes have become particularly important in water treatment for industrial and domestic water where membrane technology can be applied in cleaning operations [2 – 14]. For instance, in textile plants membranes are used to treat dye house effluents to remove the dyestuff and allow the reuse of auxiliary chemicals for dyeing, or to concentrate the dyestuffs and auxiliaries and produce purified water [15]. One of the more important categories of membranes comprises the non-porous polymeric membranes which are most widely adopted in current
industrial membrane separations [2–15]. Polyurethane membrane is an example of a nonporous synthetic membrane. Ether-type polyurethane membrane is manufactured through a copolymerisation condensation reaction of polyols with polyisocyanates [16]. These two components aggregate into two distinct regions during the reaction which results in the formation of hard and soft domains. The hard domain is very rigid in structure due to the presence of aromatic rings; the soft domain, comprising the polyol segments, is more fluid and amorphous. Extensive intermolecular hydrogen bonding takes place between the individual crystalline segments and the soft segments. As a result, the ether-type polyurethane membrane is resistant to extreme pH and temperature conditions which makes it an attractive candidate for industrial membrane separation applications. Our earlier work on the mechanism of extraction of phenols and benzoic acids by the polyurethane membrane shows that this polymer is capable of sorbing organic compounds from solution [17,18]. Carbon adsorption of organic dyes from waste effluents is relatively ineffective for the reasons previously mentioned, and consequently effluents containing organic dyes are very difficult to treat in environmental systems [1]. We chose to investigate the transport of organic dyes for the purpose of better defining the factors affecting sorption of organic compounds by the polyurethane membrane in general, and to explore the possibility of practical industrial applications. This comprehensive study should help to promote the use of the polyurethane membrane for separations in various segments of industry.
2. Experimental
2.1. Instrumental analysis and reagents UV-visible spectra and absorbance readings were taken using a Hewlett-Packard Model 8452A diode-array spectrophotometer. Solution pH was measured with an Orion Expandable Ion Analyser EA940. Water was obtained from a Barnstead Nanopure II™ purification system fed with water
K. Rzeszutek, A. Chow / Talanta 49 (1999) 757–771
759
Table 1 Anthraquinone organic dyes
purified by reverse osmosis. Polyether-type polyurethane membranes PT6100S and PT6310S 0.025 and 0.051 mm thick, respectively were supplied by Deerfield Urethane Inc., South Deerfield, MA. The organic dyes were obtained from BASF Co., Ludwigshafen, Germany and from a biological stain kit, Chem-Supply Model No. BS-100, provided by Chem Service, Inc., West Chester, PA. The molecular structures and the Colour Index (C.I.) numbers of the dyes used for this study are listed in Tables 1 – 3. The purity of these dyes was checked by TLC using a 50:10:15 ratio of n-butanol:ethanol:water mobile phase composition. Because the compounds appeared to be in their pure form, they were used without further purification. All other chemicals were of a reagent grade.
2.2. Experimental apparatus and procedure Enough of each dye was weighed out to give a final solution concentration of about 1.9 ×10 − 5 M in 1.0 M HCl or in 1.0 M NaOH unless otherwise specified. The choice of this concentration was dictated by the solubility limit of the least soluble dye in 1.0 M HCl. Parameters such as salt, acid, or base concentrations were adjusted accordingly.
The dye classification shown Tables 1–3 was adopted from textile science and is based not only on the type of substituents on the dye molecule but also on the application conditions used in textile operations utilizing those particular dyes [19]. The apparatus used for membrane testing consisted of two separate glass ‘cells’ (a starting and a receiving cell) having capacities of 260 ml. The flanges of both cells were covered lightly with Dow Corning high vacuum silicone grease in order to obtain a better seal of the two cells. All cells had short sidearms which allowed access for solutions and sampling. Ether-type polyurethane membrane 0.025 mm thick (unless otherwise specified) was cut into squares and placed between the flanges of the two cells which were then clamped together. The membrane was not subjected to any cleaning prior to use and we assumed that it does not swell in aqueous solution due to its hydrophobic nature. The surface area of the membrane exposed to solutions was 23.89 0.2 cm2. The starting cell contained 250 ml of aqueous dye solution; the receiving cell contained 250 ml of 1.0 M HCl solution without the dye (unless otherwise specified). Teflon stirrers were placed inside each cell to keep the solutions stirred throughout the experiment. Sep-
760
K. Rzeszutek, A. Chow / Talanta 49 (1999) 757–771
Table 2 Acidic organic dyes
arate magnetic stirrers were used for each cell. After the sample and the receiving solutions were transferred into the cells, the sidearm openings of the cells were closed with glass stoppers and Parafilm™ to eliminate evaporation and pH variations. The entire apparatus was then wrapped in aluminum foil for the duration of the experiment to prevent any possible decomposition of the dye due to light. Each experiment was repeated at least twice to establish reproducibility. The UV-visible spectra (190 – 820 nm) of the original solutions were taken prior to the experiment (AI) and compared with the absorbances of
the starting and receiving solutions taken after a known period of time (AFS, AFR).% dye in starting cell and % dye in receiving cell are the concentrations of dye in the starting and the receiving solutions, respectively at a known time. The concentration of dye in the membrane at a specific time, reported as % dye in membrane, can be calculated. % Dye in starting cell=AFS/AI × 100%
(1)
% Dye in receiving cell= AFR/AI × 100%
(2)
% Dye in membrane= 100%− (% dye in starting cell+ % dye in receiving cell)
(3)
K. Rzeszutek, A. Chow / Talanta 49 (1999) 757–771
761
Table 3 Basic organic dyes
3. 1 Results and discussion
3.1. Effect of the solution pH on the transport of organic dyes through polyurethane membrane Our previous work on the mechanism of extraction of phenols [17] and benzoic acids [18] by the polyurethane membrane shows that two conditions must be met in order for extraction to occur. First, the organic compound must be transferred from solution onto the membrane surface. Second, more of this compound will be removed from solution after the initial sorption only if the species sorbed on the membrane surface can migrate deeper into the polymer. An increase in the degree of sorption is observed until the capacity of the membrane has been reached. Through the study of the extraction of phenols and benzoic acids, we have found that the polyurethane membrane can only hold un-
charged organic species. Based on what we know about the mechanism of extraction of phenols and benzoic acids, we hypothesized that it should be possible to increase the amount of the organic compound removed from the starting solution if the sorbed compound could be transported through the membrane into a receiving solution. To test this hypothesis, we investigated the transport of a series of organic dyes (Tables 1–3) which are representatives of dyes frequently used in the textile industry [19,20]. The solution conditions for these dyes were chosen such that majority of the dye molecules would exist as charged or as neutral species. Both 1.0 M HCl and 1.0 M NaOH solutions were used for the extraction and transport of the anthraquinone dyes Alizarin and Alizarin Red S, all of the acidic dyes, and the basic dye Wool Green S. Anthraquinone Disperse Blue 14 is insoluble in
K. Rzeszutek, A. Chow / Talanta 49 (1999) 757–771
762
Table 4 Qualitative analysis of transport of anthraquinone, acidic and basic organic dyesa Dye Anthraquinone Alizarin Alizarin Red S Disperse Blue 14 Acidic Solvent Yellow 2 Methyl Red Methyl Orange Alizarin Yellow GG Eriochrome Black T Martius Yellow Naphthol Yellow S Basic Victoria Blue R Wool Green S
Dye in starting solution
Dye in receiving solution
Dye in membrane
1.0 1.0 1.0 1.0 1.0
M M M M M
HCl: yellow NaOH: violet HCl: light yellow NaOH: purple HCl: pink
Yellow None Light yellow None Pink
Yellow None Light yellow None Blue
1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0
M M M M M M M M M M M M M
HCl: red NaOH: yellow HCl: red/pink NaOH: yellow HCl: red NaOH: yellow HCl: orange NaOH: yellow HCl: burgundy NaOH: yellow HCl: light yellow NaOH: orange HCl: light yellow
Red None Red/pink None None None Orange None Burgundy None Yellow None Light yellow
Red None Red/pink None None None Orange None Burgundy None Yellow None Light yellow
None None Blue None None
None None Blue None None
1.0 M NaOH dark yellow 1.0 M HCl: yellow Water: blue 1.0 M HCl: green 1.0 M NaOH: blue
a Conditions: 25.0 92.0°C, 250-ml aliquot of : 1.9×10−5 M aqueous dye solution in the starting cell, 250-ml aliquot of solvent in the receiving cell, 0.025 mm thick ether-type membrane, active surface area of 23.8 9 0.2 cm2.
basic solution, and therefore it was tested only in 1.0 M HCl. The basic dye, Victoria Blue R, is also insoluble in basic solution but it dissolves in water (pH :5.0) and in 1.0 M HCl. Table 4 shows the qualitative results for the extraction and transport of the selected dyes from the corresponding starting solutions. The receiving solution for each dye had a similar acidity or basicity to the starting solution in which the dye was dissolved. The color of each dye in the starting solution, the receiving solution, and the membrane corresponds to the type of species present (i.e. charged vs. uncharged). The anthraquinone and the acidic dyes (with the exception of Methyl Orange) were found to extract and transport out of acidic solution but not from basic solution. The acidic Methyl Orange and the basic dye Wool Green S showed no extraction or trans-
port from either solution used. Basic Victoria Blue R was extracted and transported out of water but not out of acidic solution. In 1.0 M HCl, the anthraquinone dyes that have acidic substituents are neutral, e.g. Alizarin and Alizarin Red S. During extraction and transport of these dyes from acidic solution, the membrane had the same color as the starting dye solutions. In 1.0 M NaOH, Alizarin and Alizarin Red S are negatively charged; no extraction or transport was observed from basic solutions of these dyes. Organic dyes such as disperse blue 14 having basic substituents are positively charged in 1.0 M HCl. During the transport of Disperse Blue 14, the starting and receiving solutions contained charged species (pink), but the membrane was blue in color. The blue color of Disperse Blue 14 is observed when
K. Rzeszutek, A. Chow / Talanta 49 (1999) 757–771
Fig. 1. Transport of Disperse Blue 14 through 0.025 mm thick ether-type polyurethane membrane. Conditions: 25.0 9 2.0°C, 250-ml aliquot of : 1.9× 10 − 5 M aqueous dye solution in 1.0 M HCl in the starting cell, 250-ml aliquot of 1.0 M HCl in the receiving cell, 0.025 mm thick ether-type membrane, active surface area of 23.8 9 0.2 cm2. Decrease in dye concentration in the starting cell, ; increase in dye concentration in the receiving cell, .
the dye molecules are uncharged (e.g. when the dye is dissolved in a non-polar solvent such as hexane). The data collected on transport of the anthraquinone dyes suggests that only a neutral form of these dyes is retained by the membrane. Clear visual evidence for this conclusion is the transport of Disperse Blue 14 during which we observe a color change when the dye is being transferred from solution into the polymer and vice versa. Similar conclusions were reached from extraction and transport of acid dyes. In 1.0 M HCl, a majority of acidic dye species will be neutral. Similarly to the anthraquinone dyes, the acidic dyes were extracted and transported from and into acidic solution but not from basic solution. Methyl Orange, however, was an exception. The lack of extraction of this dye can be accounted for by the high probability that the dye never actually exists as an extractable uncharged species i.e. the closest instance in which it resembles an uncharged species is when it forms a zwitterion. The
763
net charge on the zwitterion is zero, but the individual ends of this species are either positively or negatively charged. Consequently, no initial transfer onto the membrane surface can occur, and hence no transport into the receiving cell was observed. Similar reasoning can be used to account for the unextractability of basic Wool Green S. Like Methyl Orange, the probability of Wool Green S being uncharged is unlikely due to the presence of the basic substituents and the sulfonic acid group on the dye molecule. Under acidic conditions, the sulfonic acid group on the dye will be neutral, but amino groups (other than the ion-paired substituent which is not considered to be truly charged because the positive charge is closely associated with the negative ions) may be protonated. The opposite will be true in basic solution. Since no neutral species is formed, there is no extraction and therefore no transport. Basic Victoria Blue R, has only basic substituents in addition to the ion-paired quaternary ammonium group. The majority of Victoria Blue R molecules are neutral in water (again, the ion-paired group is not considered to be charged), and therefore this dye is extracted and transported out of water. No extraction or transport occurred out of 1.0 M HCl due to the presence of positively charged species. These results further support the hypothesis that the presence of a neutral species of an organic compound is a deciding factor which determines whether the initial transfer of an organic compound from solution onto the membrane surface will take place. After successful initial sorption, the transport through the membrane into a receiving solution will occur only if the receiving solution conditions are favorable for desorption of the organic species from the membrane. For example, the anthraquinone dyes are transported into acidic receiving solutions but not into water in which they are only sparingly soluble. In conclusion, the transport of organic dyes from one solution into another using the polyurethane membrane as the transporting medium is contingent upon both the initial sorption of the dye from solution onto the polymer surface and the chemical properties of the receiving solution.
K. Rzeszutek, A. Chow / Talanta 49 (1999) 757–771
764
3.2. Effect of dye geometry, size, and chemical properties on transport through polyurethane membrane As pointed out in Section 3.1, the first step in the overall transport process (Fig. 3) is the sorption of the dye onto the membrane surface. It is hypothesized that the efficiency with which the dye in solution is initially transferred onto the polymer surface will be dependent on the relative solubilities of the dye in the solvent and in the membrane which will take into account all the different possible specific and non-specific interactions [21]. After the initial sorption has taken place, the ability of the sorbed species to separate the individual segments of the polymer and migrate deeper into the bulk is largely driven by the strength and the extent of the intermolecular interactions that can take place between the species and the constituent groups on the hard and soft segments of the polymer [17,18]. Because all of the dyes used in this study are relatively non-polar with substituents capable of intermolecular interactions and aromatic rings which are also found in the matrix of the membrane, they should be soluble in this polymer. For the remainder of this discussion we will only be concerned with the neutral form of the dyes since
it is the only species that is retained by the membrane. Using this information on extraction and transport within the polymer matrix, we anticipated that the rate of transport of the dye from the starting into the receiving solution through the polymer will be affected by the morphology and the chemical characteristics of the dye molecule. To test this hypothesis, the dyes were dissolved in the solvent from which extraction and transport were previously observed (Table 4) to give a concentration of 1.9× 10 − 5 M. Ether-type 0.025 mm thick polyurethane membrane was used in all instances. The results reported in Table 5 show the time at which no further change was measured in the receiving and the starting cells, the percentage dye detected in each of the cells at that time, and the percentage dye calculated to be in the membrane. Unlike all other dyes that were tested, Eriochrome Black T and Naphthol Yellow S show significantly lower concentrations in the receiving cell than in the starting cell at the reported times. This suggests that the transport of the two dyes must be very slow, and therefore the change in concentration at the time of measurement may be too small to detect. We found that some dyes, for example Alizarin, Disperse Blue 14 and Solvent Yellow 2, are not well retained by the membrane but quite efficiently
Table 5 Effect of dye morphology and chemical properties on the rate of transporta Dye
Time (h)
% Dye in starting cell
% Dye in receiving cell
% Dye in membrane
Anthraquinone Alizarin Alizarin Red S Disperse Blue 14
163 929 147
4893 5093 479 1
42 94 45 93 46 9 1
10 93 5 93 791
Acidic Solvent Yellow 2 Methyl Red Alizarin Yellow GG Eriochrome Black T Martius Yellow Naphthol Yellow S
235 183 240 216 43 408
54 9 1 189 2 1493 3993 29 9 3 909 1
45 9 1 15 92 5 92 15 94 29 93 8 93
191 65 9 1 81 9 2 46 9 3 42 9 3 2 91
Basic Victoria Blue R
733
30 9 1
26 91
40 91
Conditions: 25.0 92.0°C, 250-ml aliquot of : 1.9×10−5 M aqueous dye solution in the starting cell in a solvent from which transport occurs (Table 4), 250-ml aliquot of solvent (identical to that in the starting cell) in the receiving cell, 0.025 mm thick ether-type membrane, active surface area of 23.8 90.2 cm2. a
K. Rzeszutek, A. Chow / Talanta 49 (1999) 757–771
transported. Other dyes such as Methyl Red, Alizarin Yellow GG, Eriochrome Black T, Martius Yellow, and Victoria Blue R are very well retained and transported through the polyurethane membrane. Two of the studied dyes, Alizarin Red S and Naphthol Yellow S, are very poorly extracted and transported by this polymer. Dyes which show very inefficient extraction and transport (Alizarin Red S and Naphthol Yellow S) are small, compact and both contain a single sulfonic acid substituent (Tables 1 and 2). Because of the small size of these dyes, the effect of individual substituents on the chemical characteristics of the entire molecule is more significant than it would be for a larger dye. Thus, in the case of Alizarin Red S and Naphthol Yellow S, the presence of the sulfonic acid group imparts more polarity to both compounds. The relatively polar character of these dyes hinders efficient initial sorption onto the membrane surface and results in the very slow transport into the receiving cell. The dyes which were effectively retained and transported by the membrane are generally non-polar in nature (Tables 2 and 3). For example Eriochrome Black T, in spite of the presence of the sulfonic acid group, remains quite non-polar because of a greater number of non-polar aromatic rings. Unlike Alizarin Red S and Naphthol Yellow S, this dye is sorbed and transported quite efficiently. Martius Yellow which has chemical structure identical to that of Naphthol Yellow S (Table 2) but lacks the sulfonic acid group is highly retained by the membrane and very quickly transported which again indicates that it is the resultant polarity of the dye molecule (and therefore the solubility) and not the polarity of the individual substituents that plays an important role in the transport process. Dyes which showed high extraction and efficient transport are quite diverse in geometry and size. Examples are large non-linear dyes such as Eriochrome Black T and Victoria Blue R, smaller linear Methyl Red and Alizarin Yellow GG, and even smaller Martius Yellow (Tables 2 and 3). We can therefore conclude that the morphology and size of the dye alone are not important factors controlling the transport process. The percent concentration of the linear dyes, Methyl Red and Alizarin Yellow GG, in the membrane is significantly higher than that of, for
765
Fig. 2. Transport of Disperse Blue 14 through 0.051 mm thick ether-type polyurethane membrane. Conditions: 25.0 9 2.0°C, 250-ml aliquot of :1.9 ×10 − 5 M aqueous dye solution in 1.0 M HCl in the starting cell, 250-ml aliquot of 1.0 M HCl in the receiving cell, 0.051 mm thick ether-type membrane, active surface area of 23.8 90.2 cm2. Decrease in dye concentration in the starting cell, ; increase in dye concentration in the receiving cell, .
example, Eriochrome Black T or Martius Yellow. This can be accounted for by the presence of carboxylic acid groups on both Methyl Red and Alizarin Yellow GG which are also found in the polymer matrix. The presence of the carboxylic acid groups increases the probability for more hydrogen bonding interactions between the dye and the polymer, and increases the solubility of the dye in the polymer matrix. These results support the hypothesis on the importance of intermolecular
Fig. 3. Pictorial representation of transport of an organic dye through a polyurethane membrane. SS is the starting solution, D is the neutral dye species, MS is the membrane surface on the starting solution side, MB is the membrane bulk, MS’ is the membrane surface on the receiving solution side, RS is the receiving solution.
766
K. Rzeszutek, A. Chow / Talanta 49 (1999) 757–771
interactions. Dyes such as Alizarin, Disperse Blue 14 and Solvent Yellow 2 (Tables 1 and 2) were not well retained by the membrane, but they were transported quite efficiently. These dyes are capable of forming hydrogen bonds and are relatively non-polar. The low concentration of these dyes in the membrane, in spite of the effective transport, must be a result of their relative solubility in the solvent and in the polymer. The strength of hydrogen bonding of the dyes with the solvent must be slightly higher than with the polymer, and therefore we observed preferential partitioning of these dyes into the aqueous phase. In conclusion, the rate of transport is not controlled by the geometry or size of the dye molecule but rather by the relative solubility of the dye in the polymer and in the solvent, and by the relative strength and extent of the intermolecular interactions with the polymer and the solvent. These findings are in agreement with the conclusions reached from our earlier study of the mechanism of extraction of phenols and benzoic acids by the polyurethane membrane [17,18].
3.3. Dependence of the rate of transport on thickness of the polyurethane membrane The relationship between the thickness of the polyurethane membrane and the rate of transport of the dye molecules through this polymer is exemplified by the transport of Disperse Blue 14. This dye having a concentration of 1.9× 10 − 5 M in 1.0 M HCl was transported into 1.0 M HCl using 0.025 and 0.051 mm thick ethertype polyurethane membranes. Figs. 1 and 2 show the time taken to achieve an equilibrium between the starting cell, the receiving cell and the membrane when the 0.025 and 0.051 mm thick membranes were used respectively. Equilibrium was achieved within about 195 h with the 0.025 mm thick membrane and within 219 h with the 0.051 mm thick membrane. At the observed equilibrium, the 0.025 mm thick membrane contained 6 92% of the starting concentration of the dye, and the 0.051 mm thick membrane contained 1991% of the initial
Table 6 Effect of the initial concentration of Disperse Blue 14 on the rate of transporta Initial concentration of Disperse Blue 14 1.9×10−6 M Equilibrium time (h) 146 % Dye in starting cell 27 9 4 % Dye in receiving 27 94 cell % Dye in membrane 46 9 4
1.9×10−5 M 146 4891 45 91 7 91
a Conditions: 25.0 92.0°C, 250-ml aliquot of aqueous dye solution of appropriate concentration in the starting cell in 1.0 M HCl, 250-ml aliquot of 1.0 M HCl in the receiving cell, 0.025 mm thick ether-type membrane, active surface area of 23.8 90.2 cm2.
dye solution concentration. The masses corresponding to the active surface (23.89 0.2 cm2) of 0.025 mm and 0.051 mm thick membranes prior to exposure to the dye were : 0.120 and 0.235 g, respectively. An attempt was made to determine the overall diffusion coefficient for the transport process. At a steady state, the diffusion coefficient can be evaluated by using a so called ‘time-lag’ method [22,23]. Initially, in the non-stationary state the amount of dye diffusing through the membrane is mathematically represented by Table 7 Effect of the initial concentration of Solvent Yellow 2 on the rate of transporta Initial concentration of Solvent Yellow 2 1.9×10−6 M Equilibrium time (h) 235 % Dye in starting cell 55 9 1 % Dye in receiving 43 9 1 cell % Dye in membrane 291
1.9×10−5 M 235 529 1 47 9 1 1 91
Conditions: 25.0 92.0°C, 250-ml aliquot of aqueous dye solution of appropriate concentration in the starting cell in 1.0 M HCl, 250-ml aliquot of 1.0 M HCl in the receiving cell, 0.025 mm thick ether-type membrane, active surface area of 23.8 90.2 cm2. a
K. Rzeszutek, A. Chow / Talanta 49 (1999) 757–771
767
Table 8 Successive removal of Disperse Blue 14 from aqueous solutiona Dye concentration in starting cell at equilibrium (M) 1.9×10−5
Initial conditionsb Initial transport First successive transport Second successive transport
Dye concentration in receiving cell Dye concentration in membrane at at equilibrium (M) equilibrium (M) 0
0
8.9×10−6 4.3×10−6
8.7×10−6 3.9×10−6
1.4×10−6 1.6×10−6
2.1×10−6
2.3×10−6
1.5×10−6
Conditions: 25.0 9 2.0°C, 250-ml aliquot of aqueous dye solution in the starting cell in 1.0 M HCl, 250-ml aliquot of 1.0 M HCl in the receiving cell, 0.025 mm thick ether-type membrane, active surface area of 23.8 9 0.2 cm2. b Note: initial conditions represent the initial concentration in the starting cell before any extraction or transport has occurred; the time required to reach equilibrium was : 147 h for the initial and successive transports. a
Q = FDxCA exp
Dt 1 2 ( −1)n − − % Dx 2 6 p 2n n2
n
(− D 2nn 2) x2
(4)
As time increases and the equilibrium is approached, the above equation can be simplified to Q=
DCAF Dx 2 t− Dx 6D
n
(5)
If the thickness of the membrane, Dx, and the time, Dt, taken to reach the steady state (‘timelag’) are known, the diffusion coefficient can be calculated using [22,23] D=
Dx 2 6Dt
(6)
The transport of disperse blue 14 through the 0.025 mm thick membrane required about 195 h for the dye in the membrane to reach a steady state (i.e. when no further change in dye concentration was observed in the starting and receiving solutions and in the membrane). The approximate diffusion coefficient calculated using Eq. (6) is 5.34×10 − 7 mm2 h − 1. The transport of Disperse Blue 14 through the 0.051 mm thick membrane required about 219 h to reach a steady state. The approximate value of the diffusion coefficient is 1.98×10 − 6 mm2 h − 1. As expected, the overall equilibrium for the transport of Disperse Blue 14 was achieved more quickly when the thinner membrane was used.
This can be accounted for by the shorter path that the dye had to diffuse through in order to reach the receiving cell solution. The more quickly the dye appears on the receiving cell side of the membrane, the more rapidly it can be desorbed into the receiving solution. The concentration of the dye in the membrane increases until the steady state is reached where no further sorption or transport are observed. The amount of dye in the thinner membrane was calculated to be approximately one third of that in the thicker polymer. The thicker membrane can retain a higher amount of the dye because it has more of the polymer material by weight that is available for interactions with the dye molecules. The calculated diffusion coefficients for the transport are broad approximations of the true values. The mathematical representation of the entire diffusion is beyond the scope of this paper and will be addressed in a later publication.
3.4. Dependence of the rate of transport on the initial dye solution concentration Disperse Blue 14 and Solvent Yellow 2 were used to study the effect of the initial concentration of the dye on the rate of transport through the polyurethane membrane. The choice of the initial dye concentrations in the sample cells was dictated by the solubility of the dyes in 1.0 M HCl and the detection limit of the UV-visible spectrophotometer. The starting concentrations used
K. Rzeszutek, A. Chow / Talanta 49 (1999) 757–771
768
for both Solvent Yellow 2 and Disperse Blue 14 were 1.9× 10 − 6 and 1.9× 10 − 5 M. Table 6 and Table 7 show the time taken to achieve equilibrium at the two concentrations for both dyes, and the corresponding percentage of each dye in the starting and the receiving cells and in the membrane. Regardless of the starting concentration used, equilibrium was achieved within :146 h for Disperse Blue 14 and 235 h for Solvent Yellow 2. When 1.9 ×10 − 6 M Disperse Blue 14 was extracted, a higher percentage of the dye was retained by the membrane and lower final concentrations were detected in the starting and the receiving solutions than in the extraction of 1.9× 10 − 5 M solution of this dye. This was not observed for the transport of the corresponding solutions of Solvent Yellow 2. In the case of Disperse Blue 14, although the membrane contained a higher proportion of the dye, the calculated molar concentration of the dye in the membrane was approximately the same as the molar concentration for the extraction of the more concentrated solution of Disperse Blue 14. This suggests that the membrane can remove the dye from solution until its capacity for the dye is reached. In the case of Solvent Yellow 2, the capacity of the membrane has probably been reached at 2% even at the 1.9× 10 − 6 M dye concentration. Therefore, a change in the percentage of this dye in the membrane was not observable as with the Disperse Blue 14. To determine whether the polyurethane membrane already containing a dye could be used to
transport more dye if a fresh receiving solution was provided, we investigated a stepwise removal of Disperse Blue 14. The starting concentration of Disperse Blue 14 was 1.9× 10 − 5 M. After equilibrium was achieved between the two cells and the membrane, the molar dye concentrations in both cells and in the membrane were calculated. Next, the receiving cell solution containing the dye was discarded and replaced with new 1.0 M HCl solution. The cells were again allowed to come to equilibrium, and the concentrations were calculated. This process was repeated until an instrumental detection limit for the dye was reached. The calculated molar concentrations of Disperse Blue 14 in the starting and receiving cells at equilibrium for each successive extraction are shown in Table 8. The concentration of Disperse Blue 14 in the membrane gradually increased during the initial extraction/transport until equilibrium. This membrane dye concentration remained constant for all of the subsequent successive extraction/transport experiments regardless of the time at which the measurements were taken and the solution dye concentration being transported. Furthermore, because the time needed to achieve equilibrium for all of the successive extractions was approximately the same as for the initial extraction/ transport, we can conclude that the transporting properties of the membrane have not been affected by the presence of the dye in the matrix. These results prove that the membrane will extract the dye until its capacity for the dye is reached (i.e. the solubility limit for the dye in the membrane is
Table 9 Effect of salt on the rate of transport of Disperse Blue 14 and Solvent Yellow 2a Equilibrium time (h)
% Dye in starting cell
% Dye in receiving cell
% Dye in membrane
Disperse Blue 14 No salt added 0.5 M LiCl
147 195
489 1 489 2
469 1 409 2
891 10 9 2
Solvent Yellow 2 No salt added 0.5 M LiCl
234 234
5291 589 2
469 1 419 2
291 192
Conditions: 25.0 9 2.0°C, 250-ml aliquot of : 1.9×10−5 M aqueous dye solution in the starting cell in 1.0 M HCl with salt concentration appropriately adjusted, 250-ml aliquot of 1.0 M HCl in the receiving cell, 0.025 mm thick ether-type membrane, active surface area of 23.89 0.2 cm2. a
K. Rzeszutek, A. Chow / Talanta 49 (1999) 757–771
769
Table 10 Effect of temperature on the rate of transport of Disperse Blue 14a Temperature (°C)
% Dye in starting cell at 24 h
% Dye in receiving cell at 24 h
% Dye in membrane at 24 h
4 23 37 66
8891 699 1 589 1 459 1
691 24 9 1 36 91 43 91
691 791 691 79 1
Conditions: temperature, 9 2.0°C, 250-ml aliquot of : 1.9×10−5 M aqueous dye solution in the starting cell in 1.0 M HCl, 250-ml aliquot of 1.0 M HCl in the receiving cell, 0.025 mm thick ether-type membrane, active surface area of 23.8 9 0.2 cm2. a
reached). In order for this membrane to continue removing the remaining dye in the starting solution, a flow of dye through the membrane into the receiving solution must occur.
3.5. Effect of salt on the rate of transport through polyurethane membrane Disperse Blue 14 and Solvent Yellow 2 were chosen as examples to illustrate the effect of salt on the rate of transport. The extraction efficiency and the time taken to achieve equilibrium for transport of both dyes (1.9× 10 − 5 M) from 1.0 M HCl solutions containing 0.5 M LiCl were compared with the results for solutions without the salt. 0.5 M and 1.0 M aqueous solutions of NaCl, KCl, RbCl, CaCl2, and Al2(SO4)3 in 1.0 M HCl and in water were also tested to determine whether any transport of the cations and/or anions was occurring. The percentages of the dyes in the starting and the receiving cells and in the membrane at equilibrium for the sample solutions with and without salt are presented in Table 9. None of the cations and anions were transported through the membrane from any of the starting salt solutions as confirmed by analysis using atomic absorption spectroscopy. The extraction efficiency of Disperse Blue 14 and Solvent Yellow 2 from starting solutions containing the salt and the rates of transport into the receiving cells are almost identical (within experimental error) to the corresponding solutions of these dyes without the salt. Furthermore, the presence of salt does not change the saturation amount of each of the dyes that can be retained by the polyurethane membrane. As discussed in Section 3.1, the majority of Disperse Blue 14 species in 1.0 M HCl are posi-
tively charged. These positively charged species are, however, at equilibrium with a small percentage of uncharged molecules of Disperse Blue 14. The opposite is true for the solution of Solvent Yellow 2 in 1.0 M HCl in which the predominant species is the uncharged dye molecule. When salt is added to solutions containing charged organic species, ion pair formation can take place [24–26] and the overall solution becomes more polar. Because of the increased solution polarity in the presence of salts, we would expect a ‘salting out’ effect where the non-polar uncharged species becomes more insoluble in the solvent, and therefore partitions more efficiently into the non-polar membrane which results in a more rapid extraction and transport of both dyes. Because no significant changes in extraction and transport rates have been observed, we concluded that addition of salt does not cause a sufficiently high change in the relative solubility of the dye species in solution and the membrane. These findings are in agreement with the conclusions reached for the earlier investigation of the extraction of phenols and benzoic acids by the polyurethane membrane [17,18].
3.6. Dependence of the rate of transport on solution temperature 1.9× 10 − 5 M solutions of Disperse Blue 14 in 1.0 M HCl were transported into a 1.0 M HCl solution using 0.025 mm polyurethane membrane at 4, 23, 37, and 66°C. The percentages of the dye in the starting and the receiving cells and in the membrane at 24 h for each of the temperatures are shown in Table 10. The percentage of the dye retained by the membrane is approximately the
770
K. Rzeszutek, A. Chow / Talanta 49 (1999) 757–771
same at all four temperatures. However, the rate of removal of the dye from the starting cell and the rate of transport into the receiving cell were found to increase with increased temperature. The above results do not correlate with the temperature data that we collected for extraction of phenols and benzoic acids by the ether-type polyurethane membrane [17,18]. For the extraction of phenols and benzoic acids, we found that as the temperature was increased, the extraction decreased because of the exothermic nature of the extraction process. In this study, however, in addition to extraction we have transport into a receiving solution. Unlike the simple extraction, the transport is probably controlled not only by thermodynamic factors but also by the concentration gradient created during the process. The fact that the membrane retains approximately the same amount of dye at all of the temperatures suggests that the solubility of the dye in the polymer remains relatively the same. The increased rates of removal and transport of the dye at higher temperatures suggest that the physical properties of the membrane matrix must be different at those temperatures. The most likely explanation is that due to the higher energy in the membrane segments at the increased temperature, some of the intermolecular interactions among the individual segments within the polymer may be weakened or broken; this results in the matrix being more fluid in nature and, consequently, more accessible to the dye species in solution. Therefore, at higher temperatures (e.g. 66°C), the dye species sorbed on the surface can migrate more freely through the bulk of the polymer and into the receiving solution which is evidenced by the significantly shorter time in which equilibrium is achieved (e.g. within 146 h at 23°C vs. 24 h at 66°C).
4. Conclusion The rate at which various organic dyes will transport through the ether-type polyurethane membrane is dependent on the solution pH, the overall polarity and capability of the dye molecule to engage in hydrogen bonding interactions with the polymer, the relative solubility of the dye
species in the polymer and in the solvent, and the temperature at which the transport is occurring. The ether-type polyurethane membrane can retain only neutral species of the organic dyes. In order to obtain an initial transport through the bulk of the membrane, the pH of the starting solvent must be favorable for the formation of a neutral species which is the only species soluble in the membrane. Furthermore, a flow of the species in the membrane into the receiving solution will be observed only if a driving force resulting from the differences in dye concentration in the starting and receiving solutions and the membrane (or a concentration gradient) is present. The rate of transport is dependent on chemical factors such as the efficiency with which the dye can partition into the membrane surface and then migrate through the membrane bulk, and on the physical experimental conditions such as the temperature. For example, species with comparable solubility in the polymer and in solution is transported from the starting cell into the receiving cell quite efficiently. On the other hand, if the solubility of the species is relatively higher in one of the media, the transport may be hindered due preferential partitioning. The partitioning of the dye between the solution and membrane phases and the rate of migration through the bulk of the polymer can be significantly altered by changing the temperature. The rate of transport of the dyes through the polyurethane membrane increased when higher temperatures were used. The increase in the transport rate at increased temperatures can be accounted for by change in the physical properties of the membrane matrix which becomes more fluid, and therefore more exposed and accessible to the dye species in solution. Overall, the results from this study provide a deeper insight into the mechanism of extraction and transport process of organic compounds by the polyurethane membrane and show that this membrane has considerable potential in commercial applications for removal of organic compounds from aqueous solutions.
Acknowledgements The authors would like to acknowledge the
K. Rzeszutek, A. Chow / Talanta 49 (1999) 757–771
financial support of the University of Manitoba and the Natural Sciences and Engineering Research Council of Canada.
[12] [13]
References [1] A. Reife, H.S. Freeman, Environmental Chemistry of Dyes and Pigments, Wiley-Interscience, New York, 1996, pp. 3 – 32. [2] R. Singh, CHEMTECH 28 (4) (1998) 33. [3] R. Singh, CHEMTECH 26 (6) (1996) 46. [4] K. Scott, R. Hughes, Industrial Membrane Separation Technology, Chapman and Hall, London, 1996. [5] M. Mulder, Basic Principles of Membrane Technology 2nd ed., Kluwer Academic Publishers, Dordrecht/Boston/ London, 1996. [6] J.A. Howell, A. Noworyta, Towards Hybrid Membrane and Biotechnology Solutions for Polish Environmental Problems, Wrocl*aw Technical University Press, Wrocl*aw, 1995. [7] K. Scott, Handbook of Industrial Membranes, 1st ed., Elsevier Advanced Technology, 1995. [8] R.D. Noble, S.A. Stern, Membrane Separation Technology: Principles and Applications, Elsevier Science, Amsterdam, 1995. [9] T. Matsuura, Synthetic Membranes and Membrane Separation Processes, CRC Press, Boca Raton, 1993. [10] P. Aptel, Membrane pressure driven processes in water treatment, in: J.G. Crespo, K.W. Broddeker (Eds.), Membrane Processes in Separation and Purification, NATO ASI Series, Series E: Applied Sciences, Kluwer Academic Publishers, 272 (1993) 263. [11] M. Mulder, The use of membrane separation processes in environmental problems, in: J.G. Crespo, K.W. Brod-
[14]
[15]
[16]
[17] [18] [19]
[20]
[21]
[22]
[23] [24] [25] [26]
771
deker, Membrane Processes in Separation and Purification, NATO ASI Series, Series E: Applied Sciences, Kluwer Academic Publishers, 272 (1993) 233. W.S. Winston Ho, K.K. Sirkar, Membrane Handbook, Van Norstrand Reinhold, New York, 1992. J.A. Howell, The membrane alternative: energy implications for industry, Report number 21, Elsevier, London and New York, 1990, pp. 2 – 8. L. Cecille, J.-C. Toussaint, Future industrial prospects of membrane processes, Elsevier, London and New York, 1989, pp. 3 – 7, 44, 55. L. Cecille, J.-C. Toussaint, Future Industrial Prospects of Membrane Processes, Elsevier, London and New York, 1989, pp. 190 – 195. G. Woods, The ICI Polyurethanes Book, ICI Polyurethanes and John Wiley & Sons Publication, Toronto, 1987, pp. 27 – 54. K. Rzeszutek, A. Chow, Talanta 46 (1998) 507. K. Rzeszutek, A. Chow, Talanta 47 (1998) 697. D.M. Nunn, The Dyeing of Synthetic-Polymer and Acetate Fibres, Dyers Company Publications Trust, Bradford, West Yorkshire, 1979, pp. 2 – 53. E.P.G. Gohl, L.D. Vilensky, Textile Science: an Explanation of Fibre Properties, 2nd ed., Longman Cheshire Pty Ltd., Melbourne, 1983, pp. 120 – 164. C.L. Bird, W.S. Boston, The Theory of Coloration of Textiles, The Dyers Company Publications Trust, Bradford, West Yorkshire, England, 1975, pp. 294 – 298. I.D. Rattee, M.M. Breuer, The Physical Chemistry of Dye Adsorption, Academic Press, London and New York, 1974, pp. 72 – 77. J. Crank, The Mathematics of Diffusion, Clarendon Press, Oxford, 1975, pp. 44 – 68. L. Shumack, A. Chow, Talanta 34 (1987) 957. P. Fong, A. Chow, Talanta 39 (1992) 497. R. Werbowesky, A. Chow, Talanta 43 (1996) 263.
. .