- Email: [email protected]
Progressive wetting of supported liquid membranes by aqueous solutions
Journal of Membrane Science, 42 (1989) 183-188 Elsevier Science Publishers B.V., Amsterdam - Printed in The Netherlands
183
Short Communication
PROGRESSIVE WETTING OF SUPPORTED LIQUID MEMBRANES BY AQUEOUS SOLUTIONS
HIROSHI TAKEUCHI* and MAKOTO NAKANO Department of Chemical Engineering, Nagoya University, Nagoya, 464-01 (Japan) (Received December 21,1987; accepted in revised form May 25,1988)
In earlier work [ 11, studies were made of the lifetime of supported liquid membranes ( SLMs) with use of various polymeric solid supports and organic solvents. These showed that more stable SLMs can be obtained by use of membrane solvents having higher interfacial tension, and of more hydrophobic solids with smaller pore size. In SLM operations, however, the membrane solution contains a carrier such as an extraction reagent with surface-active character, and a pressure difference exists between the two sides of the membrane. More recently, Kim and Harriott [2] discussed the minimum entry pressure for water into a hydrophobic membrane (PTFE) with and without impregnation by an organic liquid in terms of an effective contact angle, &, giving anomalously low entry pressure for the organic solvent systems. In other words, an SLM allows water to be imbibed below a critical pressure difference defined by (2y/a)cos1&. The progressive wetting by an aqueous solution of a polymeric solid support impregnated with an organic solution thus presents an important problem in SLM operations. In this paper consideration is given to a submerged aqueous drop on a horizontal porous sheet in an organic solution, as shown in Fig. 1. The work of immersional wetting may be defined in a way similar to that for drops on a non-porous plate, (1)
Wi=Yeo-YYsw=YobwC0s8
With a contact angle 8> 90” and Wi> 0, the wetting becomes immersion; therefore, the liquid-liquid interfacial tension yawhas no influence on the wettability due to the difference between y(+,)_, and y(8/0)_-w, which can only be expressed by y,,,cos& In the present communication, the results from the most *To whom correspondence should be addressed.
0376-7388/89/$03.50
0 1989 Elsevier Science Publishers B.V.
184
Organic
jb,
liq.
T-
Fig. 1. Schematic organic solution.
representation
of aqueous drop on a horizontal
porous sheet, submerged
recent investigation of contact angle behavior will be reported, of evaluating the progressive wettability of SLM.
in an
with the object
Experimental Contact angle measurements
Since small drops, when put on a flat surface, take the form of a partially flattened sphere, it is possible to determine the contact angle, 8, from the drop volume, V, and the diameter of the bases of the drop, do, by using the following equation [ 3 ] : tan (0/Z) = 2h/d,
(2)
where h is the height of the drop. A known volume (2 or 8 ~1) of aqueous solution was delivered from a microsyringe on to a polymeric sheet in contact with a sintered glass plate placed horizontally within a glass spectrophotometric cell filled with an organic solution. The height of the drop was measured with a travelling microscope as a function of elapsed time. In the present study, three commercially available microporous sheets were used, as in the previous work [l] : Fluoropore 100 (FP-lOO), Duragard 2500, and Nuclepore membranes. The organic liquids used were n-heptane, toluene, kerosene, and their solutions of commercial extraction reagents. Extraction systems comprised Ni (NO,) 2- [ di- (2-ethylhexyl )phosphoric acid (DBEHPA or HDEHP)/n-heptane], Co(NO,),-[di-(2-ethylhexyl)phosphonic acid mono (2-ethylhexyl) ester (PC-88A or EHPNA)/n-heptane] and K&r,07[ tri-n-octylamine (TOA) /toluene] . Interfacial tension for the various extraction systems was measured under conditions without mass transfer by the drop volume method. A water drop was formed within the organic phase by means of a glass nozzle, whereby the two liquids were mutually saturated. The feed rate was kept at 0.0219 cm”/sec
185
by a microfeeder, and the drop volume was determined by measuring the time of drop formation. Results and discussion Figure 2 presents typical results for ldW ( = y) for the extraction systems, indicating that both the organophosphorous extractants are surface-active and the corresponding metal complexes have even more active properties. The values of y for the organic solutions decrease with an increase in the metal-loading ratio; a notable lowering in y is found for the Ni- (DBEHPA/n-heptane) system. This suggests that capillary force on the upstream side in SLM operations becomes smaller than that on the downstream side; the organic phase-water boundary is thus liable to become drawn into the SLM on the upstream side. Values of the contact angle of an aqueous drop for organic solvent-impregnated FP-100 and Duragard 2500 membranes are given in Table 1; both toluene and kerosene systems show a slight decrease in contact angle with elapsed time, indicating low progressive wettability. Similar behavior was also observed during the extraction of metal ions, with no significant difference for the three extraction systems used (see Table 1) . However, in actual SLM operations for the Ni- (DBEHPA/n-heptane ) and Cr- (TOA/toluene ) systems, the lifetimes of the SLM were comparatively short [ 41. Further measurements of the contact angle were conducted under conditions of back-extraction, as follows: (A) (DBEHPA/n-heptane ) loaded with Ni ( N03)2-0.1 N HNO,; concentrations of DBEHPA and Ni in the organic phase, [ (RH),] =0.04 M and [ Ni] = 0.04 M, respectively. (B) (C)
(PC-88A/n-heptane) loaded with Co(NO,),-0.1 0.04 M and [Co] = 7.3 mM.
N HNO,; [ (RH),]
=
(TOA/toluene) loaded with K,Cr,O,-0.1 N NaOH; [TOA] =0.0277 M and [Cr(VI)] =0.55 mM. Typical values of 13obtained for the three different membranes are plotted in Fig. 3 as a function of elapsed time. Although system A exhibited pronounced turbulence due to the Marangoni effect, both FP-100 and Duragard 2500 membranes showed similar behavior with time, an unusual variation in 6’ being observed after a contact time of 20 to 25 minutes for Nuclepore membranes the contact angle decreased rapidly and attained a constant value after about 5 minutes. System B was stable, and did not exhibit interfacial turbulence; however, the 6’ value for the Nuclepore membrane became smaller than 90”. Such an “SLM” thus appeared to be completely wetted by acid solutions. For system C it was not easy to form a single aqueous drop in the organic phase; furthermore, the sessile drop formed on FP-100 was flattened even with a volume as small as 2 ~1. In a preliminary test, the contact angle was deter-
186
0::
0
*
2x10-3
I
I
0’ [RI, ( kmol/m3)
0 Co-(EHPNA/n-hep) n NI -(EHPNA/n-hep) 0 Nb(HDEHP/n-hep) A CrW(TOA/tol 1
10-l
, water / water / water , water
Fig. 2. Results of interfacial tension measurements. TABLE 1 Values of contact angle of aqueous drop on membranes in organic solution under conditions with and without metal extraction Porous sheet
FP-100 Duragard 2500 Duragard 2500
FP-100 FP-100 FP-100
Liquid-liquid system n-heptane-water kerosene-water toluene-water
system A’ system B’ system C’
Contact angle (degrees) 0
2
4
20 hr
117 114 119
117 112 118
117 111 117
117 104 111
0
20
41
65
90 hr
104 110 112
105 103 106
100 106 109
99 105 106
96 105 95
System A’: (DZEHPA/n-heptane)-0.01 M Ni(N03)2 (pH 4). System B’: (PC-MA/n-heptane)-0.01 M Ni(NO,), (pH 4). System C’: (TOA/toluene)-100 ppm Cr(V1) (pH 1.6).
mined from the diameter of a yellow spot, do, deposited on the which was removed from the cell after 30 min; its value was in agreement with that obtained from the drop height. Thus, although method was used with system C, for higher concentrations of Cr
membrane, reasonable the present (VI) in the
187
m------_-_--_--_------__, ~~1=22&?+07N I
I
0
20
I
40 Time Cminl
NaOH I
I.
60
”
I
2ohr
Fig. 3. Variation of contact angle as function of elapsed time.
10'
10'
Time elapsed
(min)
ld
lo*
Fig. 4. Results of wetting studies.
organic phase no values could be obtained, because of the rapid change in 8 and also the deep colour of the organic phase. Figure 3 suggests that system C is highly wettable by alkaline solutions, and therefore the SLM based on this system appears to be the most unstable. As a means of inferring the time required for complete wetting of SLMs by aqueous solution, the values of y co4 for the three different systems are plotted in Fig. 4 against the elapsed time; the broken lines represent the respective values of y corresponding to COS& 1. The point of intersection of the extrapolated line (dashed-and-dotted line) through the data points with the broken line may thus be defined as a “critical wetting time”, t,, which can be used as
188
an index of the progressive wettability and, therefore, of the effective lifetime of an SLM. The critical wetting time for system A decreases in the following sequence (as can be seen from Fig. 4): Duragard 2500, FP-100, Nuclepore membranes; system B gives a large value even with Nuclepore. The result for system C suggests that, even with FP-100, the SLM is so unstable that it is replaced by the aqueous solution within 120 to 200 minutes. The value of t, as defined above corresponds to the time required for complete displacement of an organic liquid from a polymer support by an aqueous solution under static conditions. However, in practical SLM separations, membrane breakdown may quite probably be caused in a shorter time by the pressure difference between the two sides of the SLM. It can be concluded that the relative lifetime based on the progressive wettability of an actual SLM can be evaluated in terms of critical wetting time from a plot of ycos0 versus t.
References 1 2 3 4
H. Takeuchi, K. Takahashi and W. Goto, Some observations on the stability of supported liquid membranes, J. Membrane Sci., 34 (1987) 19. B.S. Kim and P. Harriott, Critical entry pressure for liquids in hydrophobic membrane, J. Colloid Interface Sci., 115 (1987) 1. J.J. Bikerman, Surface Chemistry, Theory and Applications, 2nd edn., Academic Press, New York, NY, 1958, p. 343. M. Nakano, K. Takahashi and H. Takeuchi, A method for continuous operation of supported liquid membrane,
J. Chem. Eng. Jpn., 20 (1987)
326.
183
Short Communication
PROGRESSIVE WETTING OF SUPPORTED LIQUID MEMBRANES BY AQUEOUS SOLUTIONS
HIROSHI TAKEUCHI* and MAKOTO NAKANO Department of Chemical Engineering, Nagoya University, Nagoya, 464-01 (Japan) (Received December 21,1987; accepted in revised form May 25,1988)
In earlier work [ 11, studies were made of the lifetime of supported liquid membranes ( SLMs) with use of various polymeric solid supports and organic solvents. These showed that more stable SLMs can be obtained by use of membrane solvents having higher interfacial tension, and of more hydrophobic solids with smaller pore size. In SLM operations, however, the membrane solution contains a carrier such as an extraction reagent with surface-active character, and a pressure difference exists between the two sides of the membrane. More recently, Kim and Harriott [2] discussed the minimum entry pressure for water into a hydrophobic membrane (PTFE) with and without impregnation by an organic liquid in terms of an effective contact angle, &, giving anomalously low entry pressure for the organic solvent systems. In other words, an SLM allows water to be imbibed below a critical pressure difference defined by (2y/a)cos1&. The progressive wetting by an aqueous solution of a polymeric solid support impregnated with an organic solution thus presents an important problem in SLM operations. In this paper consideration is given to a submerged aqueous drop on a horizontal porous sheet in an organic solution, as shown in Fig. 1. The work of immersional wetting may be defined in a way similar to that for drops on a non-porous plate, (1)
Wi=Yeo-YYsw=YobwC0s8
With a contact angle 8> 90” and Wi> 0, the wetting becomes immersion; therefore, the liquid-liquid interfacial tension yawhas no influence on the wettability due to the difference between y(+,)_, and y(8/0)_-w, which can only be expressed by y,,,cos& In the present communication, the results from the most *To whom correspondence should be addressed.
0376-7388/89/$03.50
0 1989 Elsevier Science Publishers B.V.
184
Organic
jb,
liq.
T-
Fig. 1. Schematic organic solution.
representation
of aqueous drop on a horizontal
porous sheet, submerged
recent investigation of contact angle behavior will be reported, of evaluating the progressive wettability of SLM.
in an
with the object
Experimental Contact angle measurements
Since small drops, when put on a flat surface, take the form of a partially flattened sphere, it is possible to determine the contact angle, 8, from the drop volume, V, and the diameter of the bases of the drop, do, by using the following equation [ 3 ] : tan (0/Z) = 2h/d,
(2)
where h is the height of the drop. A known volume (2 or 8 ~1) of aqueous solution was delivered from a microsyringe on to a polymeric sheet in contact with a sintered glass plate placed horizontally within a glass spectrophotometric cell filled with an organic solution. The height of the drop was measured with a travelling microscope as a function of elapsed time. In the present study, three commercially available microporous sheets were used, as in the previous work [l] : Fluoropore 100 (FP-lOO), Duragard 2500, and Nuclepore membranes. The organic liquids used were n-heptane, toluene, kerosene, and their solutions of commercial extraction reagents. Extraction systems comprised Ni (NO,) 2- [ di- (2-ethylhexyl )phosphoric acid (DBEHPA or HDEHP)/n-heptane], Co(NO,),-[di-(2-ethylhexyl)phosphonic acid mono (2-ethylhexyl) ester (PC-88A or EHPNA)/n-heptane] and K&r,07[ tri-n-octylamine (TOA) /toluene] . Interfacial tension for the various extraction systems was measured under conditions without mass transfer by the drop volume method. A water drop was formed within the organic phase by means of a glass nozzle, whereby the two liquids were mutually saturated. The feed rate was kept at 0.0219 cm”/sec
185
by a microfeeder, and the drop volume was determined by measuring the time of drop formation. Results and discussion Figure 2 presents typical results for ldW ( = y) for the extraction systems, indicating that both the organophosphorous extractants are surface-active and the corresponding metal complexes have even more active properties. The values of y for the organic solutions decrease with an increase in the metal-loading ratio; a notable lowering in y is found for the Ni- (DBEHPA/n-heptane) system. This suggests that capillary force on the upstream side in SLM operations becomes smaller than that on the downstream side; the organic phase-water boundary is thus liable to become drawn into the SLM on the upstream side. Values of the contact angle of an aqueous drop for organic solvent-impregnated FP-100 and Duragard 2500 membranes are given in Table 1; both toluene and kerosene systems show a slight decrease in contact angle with elapsed time, indicating low progressive wettability. Similar behavior was also observed during the extraction of metal ions, with no significant difference for the three extraction systems used (see Table 1) . However, in actual SLM operations for the Ni- (DBEHPA/n-heptane ) and Cr- (TOA/toluene ) systems, the lifetimes of the SLM were comparatively short [ 41. Further measurements of the contact angle were conducted under conditions of back-extraction, as follows: (A) (DBEHPA/n-heptane ) loaded with Ni ( N03)2-0.1 N HNO,; concentrations of DBEHPA and Ni in the organic phase, [ (RH),] =0.04 M and [ Ni] = 0.04 M, respectively. (B) (C)
(PC-88A/n-heptane) loaded with Co(NO,),-0.1 0.04 M and [Co] = 7.3 mM.
N HNO,; [ (RH),]
=
(TOA/toluene) loaded with K,Cr,O,-0.1 N NaOH; [TOA] =0.0277 M and [Cr(VI)] =0.55 mM. Typical values of 13obtained for the three different membranes are plotted in Fig. 3 as a function of elapsed time. Although system A exhibited pronounced turbulence due to the Marangoni effect, both FP-100 and Duragard 2500 membranes showed similar behavior with time, an unusual variation in 6’ being observed after a contact time of 20 to 25 minutes for Nuclepore membranes the contact angle decreased rapidly and attained a constant value after about 5 minutes. System B was stable, and did not exhibit interfacial turbulence; however, the 6’ value for the Nuclepore membrane became smaller than 90”. Such an “SLM” thus appeared to be completely wetted by acid solutions. For system C it was not easy to form a single aqueous drop in the organic phase; furthermore, the sessile drop formed on FP-100 was flattened even with a volume as small as 2 ~1. In a preliminary test, the contact angle was deter-
186
0::
0
*
2x10-3
I
I
0’ [RI, ( kmol/m3)
0 Co-(EHPNA/n-hep) n NI -(EHPNA/n-hep) 0 Nb(HDEHP/n-hep) A CrW(TOA/tol 1
10-l
, water / water / water , water
Fig. 2. Results of interfacial tension measurements. TABLE 1 Values of contact angle of aqueous drop on membranes in organic solution under conditions with and without metal extraction Porous sheet
FP-100 Duragard 2500 Duragard 2500
FP-100 FP-100 FP-100
Liquid-liquid system n-heptane-water kerosene-water toluene-water
system A’ system B’ system C’
Contact angle (degrees) 0
2
4
20 hr
117 114 119
117 112 118
117 111 117
117 104 111
0
20
41
65
90 hr
104 110 112
105 103 106
100 106 109
99 105 106
96 105 95
System A’: (DZEHPA/n-heptane)-0.01 M Ni(N03)2 (pH 4). System B’: (PC-MA/n-heptane)-0.01 M Ni(NO,), (pH 4). System C’: (TOA/toluene)-100 ppm Cr(V1) (pH 1.6).
mined from the diameter of a yellow spot, do, deposited on the which was removed from the cell after 30 min; its value was in agreement with that obtained from the drop height. Thus, although method was used with system C, for higher concentrations of Cr
membrane, reasonable the present (VI) in the
187
m------_-_--_--_------__, ~~1=22&?+07N I
I
0
20
I
40 Time Cminl
NaOH I
I.
60
”
I
2ohr
Fig. 3. Variation of contact angle as function of elapsed time.
10'
10'
Time elapsed
(min)
ld
lo*
Fig. 4. Results of wetting studies.
organic phase no values could be obtained, because of the rapid change in 8 and also the deep colour of the organic phase. Figure 3 suggests that system C is highly wettable by alkaline solutions, and therefore the SLM based on this system appears to be the most unstable. As a means of inferring the time required for complete wetting of SLMs by aqueous solution, the values of y co4 for the three different systems are plotted in Fig. 4 against the elapsed time; the broken lines represent the respective values of y corresponding to COS& 1. The point of intersection of the extrapolated line (dashed-and-dotted line) through the data points with the broken line may thus be defined as a “critical wetting time”, t,, which can be used as
188
an index of the progressive wettability and, therefore, of the effective lifetime of an SLM. The critical wetting time for system A decreases in the following sequence (as can be seen from Fig. 4): Duragard 2500, FP-100, Nuclepore membranes; system B gives a large value even with Nuclepore. The result for system C suggests that, even with FP-100, the SLM is so unstable that it is replaced by the aqueous solution within 120 to 200 minutes. The value of t, as defined above corresponds to the time required for complete displacement of an organic liquid from a polymer support by an aqueous solution under static conditions. However, in practical SLM separations, membrane breakdown may quite probably be caused in a shorter time by the pressure difference between the two sides of the SLM. It can be concluded that the relative lifetime based on the progressive wettability of an actual SLM can be evaluated in terms of critical wetting time from a plot of ycos0 versus t.
References 1 2 3 4
H. Takeuchi, K. Takahashi and W. Goto, Some observations on the stability of supported liquid membranes, J. Membrane Sci., 34 (1987) 19. B.S. Kim and P. Harriott, Critical entry pressure for liquids in hydrophobic membrane, J. Colloid Interface Sci., 115 (1987) 1. J.J. Bikerman, Surface Chemistry, Theory and Applications, 2nd edn., Academic Press, New York, NY, 1958, p. 343. M. Nakano, K. Takahashi and H. Takeuchi, A method for continuous operation of supported liquid membrane,
J. Chem. Eng. Jpn., 20 (1987)
326.