J. Chem. EW. h t e 1992, 37, 288-274
200
Ackno-nl c
We are grateful to the members of the K value Screening Committee, especially to Dr. H. Kistenmacher, far encourage ment and advice, and we thank A. Chareton, A. Valh, P. Alall, C. Lafeuii, F. Fontaiba, J. Dest&ve, and D. Legret for their contribution.
IO 4
Rogbtry No. N2, 7727-37-9; COP, 124-38-9; H2S, 7783-064; propyC cyclohexane, 1878-92-8; toluene, 108-884 m-xylene, 108-36-3; mesh tylene, 108-67-8; propylbenzene, 103-65-1.
M8. Presswe v ~ ~ s uaxqmkbndlagem 8
forthe carbon dioxide
(1knproWlcyckhexam (2) system: (A)thls work at 313.1 K (X) this work at 393.2 K (V)this work at 472.8 K; ( 0 )from Kim et al. (12) at 315.4 K; ( 0 )from Kim et al. (72) at 392.7 K; (+) from Kim et al. (72)at 474.6 K.
uI
,
1 I
-
=a,
,
,
,
,
,
I
c
B 0,1
0, 01
Literature C i t d
(1) Richon, D.; Laugh,S.; Renon, H. J . Chem. Eng. Data 1001, 38, 104. (2) ReuMlltz, J. M.; Keeler, R. N. AI&€ J . 1011, 7, 399. (3) R m , H.; Leugler, S.; Schwarhentmber, J.; Richon, D. FknM phese E m . 1080, 51, 285. (4) Lqred, D.; Rlchon, D.; Renon, H. AI&€ J . 1081, 27, 203. (5) F-e, P.; Hom, J. F.; Laugkr, S.; Rem, H.; Rlchon, D.; Szwarc, H. AICW J . 1080, 28, 872. (6) Fontaka, F.; Rlchon, D.; Renon, H. Rev. Sd. Inabvn. 1084, 55,944. (7) Laugkr, S.; Rlchon, D. Rev. Sd. I n s t ” . 1006, 57, 469. (8) laugler, S.; Lqred, D.; Desth, J.; Rlchon, D.; R e m , H. QPA Research Report RR-SB Gas Procc~lson,Assodatlon: Tulsa, OK, 1982. (9) Laugh, S.; AkU, P.; Vak, A.; Chareton, A.; Fontalba, F.; Rlchon. D.; R m , H. QPA Remarch Report RR-75 Qes Procc~lson,Associetlon: Tulsa, OK, 1984. (10) Charston, A.; Vah, A.; LafeuU, C.; Laugh, S.; Rlchon, D.; Renon, H. (3PA Research Rem RR-101, Gas Processors Asso&tlon: Tulsa, OK, 1986. (11) T h ” a f m , J. Pny&o-chemhl Constants of h e olgenlc Compow&; Elmvler: Amterdam, 1965. (12) Klm, H.; Lln, H.M.; Chao, K. C. AIChESymp. Ser. 1085, 81 (244), 96.
1
10
100
Pressure, MPa
Flguv 4. Pattitkin coeffldent versus pressure diagram for the carbon dioxide (l)-n-propylcyclohexne (2) system: )(. thls work at 393.2 K (Indicated with error bars); (m)thls work at 472.8 K (IndICeted with error bars); (0) from Klm et al. (72) at 392.7 K; (X) from Kim et el. (72)at 474.8 K.
for review January 22, 1991. Revbed October 21, 1991. Accepted Ootober 21, 1991. We are grateful to the Gas Processors Assock-
~&ed
tbn for Rnanchl support.
Phase Compositions, Viscosities, and Densities for Aqueous Two-Phase Systems Composed of Polyethylene Glycol and Various Salts at 25 O C Steven M. Snyder, Kenneth D. Cole,’ and Davld C. Szlag National Institute of Standards and Technology, 325 Broadway, BouhAw, colored0 80303
Pha8e dlagram8 ol aqueous tw+phase gstem8 mnpo8ed of polyethylene glycol and varlour 8alt rdutlonr were meawred. The denrltle8 and vbcodth of th.w pha8e system8 were a b measured. Polyethylene glycol was uwd with three average moledar ma88e8 ol 1000, 3350, and 8000. The salts uwd were magne8ium d a t e , sodium wtfate, aodlum carbonate, sulfato, and potarrlum phoephate. Phase diagram data, as well a8 the den8itl.r and vl8co8ltles ol the pha8e8, were meamired at 25 ‘C.
Introduction UquM-llquM extractbn utillzing aqueous tw+phase systems (ATPs) has been used to separate and Putty bbbglcai products
from the complex mixtures In whlch they are produced ( 7,2). Data on the cOmpOSltiOn and properties of phase systems are necessary for the deslgn of ATPs extraction processes. Phesediagram data are also necessary for the development of models that predict phase partitioning (3-6). In this work, a comprehensive set of densities, viscoslties, and phase composltbns of ATPs composed of various polyethylene glycol (PEG) m a w s and salts were measured at 25 O C . A prevlous technique has been used to determine c om positions of polymer-polymer systems utilizing measurements of optlcei rotatbn and refractive index (7,2, 7, 8). However the PEG-salt systems have ilttle~or no optical activity. Potassium phosphate-PEG systems can be determlned by titration (7); however, this method does not apply to the other salts used. Therefore, the gravimetric method of determining phase composition described by Stewart and Todd (9) was used and
Thls article not subJect to U.S. Copyrlght. Pubilshed 1902 by the Amerlcan Chemical Society
Journal of Chemical and Engineering Data, Vol. 37, No. 2, 1992 200
s2z55;s
BX8SB
?????
8 88888 000000
00000
, ~
G
c C
m
L
c
c e
i
I 4
I
i
I-]
270 Jounel of Chemhxl and Enghwting Lkta, Vol. 37, No. 2, 1992
33333
88Ggg
? ???? 0 0 0 0 0
ucpunw
Journal of Chemical and Engineering Data, Vol. 37, No. 2, 1992 271 40:
a
1
B
0
-
3 30: a a
: -
s :
Y
0 0 0
20.
c
r(333;i
0
w
0
b
:
10:
40
0
10
Magnesium Sulfate (mass
20
%I
C
A
3 ul ul
2
Y
0 0
0 OJ
(3
W
n.
Magnesium S u l f a t e (mass %)
Flguro 1. Phase dlagrams of magnesium sulfate systems wlth PEG 1000 (a), PEG 3350 @), and PEG 8000 (c). The composltkns of these systems (A-M) are given in Table I.
compared wlth analysis via high-performance liquid chromatography (HPLC)-gel permeation chromatography.
M 8 h r l . k . Three lots of PEG were obtained commercially. The number-average molecular mass (M,) and the weightdvwage molecular mass (M,) were determined using gel permeatkn chrometography by A " polymer Standards Corp. (Mentor, OH). The columns used were an ultrahydrogel250 A and an ultrahydrogel 120 A In tandem. The mobile phase was water at 1.0 mL/mln at 30 OC. The salts used In the phase systems were reagent grade and anhydrous.
272 Jocxnal of Chemical and Engineering Data, Vol. 37,No. 2, 1992
b
a
rp m
m
E"
Y
0 0
0
aD
(I
W
n
POta88iUm P h o l p h a t o h a 8 8 %)
Flguro 2. Phase *gams of potassium phosphate, pH 8.0, systems with PEG 1000 (a) and PEG 8000 (b). The composltlons of these systems ( A d ) are ghren in Table 11.
a
I
503
b 50
40
ap Y)
m m 30
E
Y
0 v) c)
20
0
0
w
a 10 1
0
0
10 20 S o d i u m Carbonate (mass %)
C
A
3 m m
E
Y
0 0 0
m (I
W
n
Sodium C a r b o n a t e (mass %)
Flguro 3. Phase dlagrams of sodium carbonate systems with PEG 1000 (a). PEG 3350 (b), and PEG 8000 (c). The composlbns of thm phase systems (A-L) are ghren In Table 111.
Appwatw urd n. ATPs (45g) were constructed by we4ghhg stock polymer sdutkns and dry salts into a 5o-ml centrifuge tube. For the potassium phosphate systems, the Henderson-Hasseibach equation was used to determine the ratio of mono- and dibasic salts necessary to bring the pH to 8.0. ATPs were brought to 25 f 0.1 OC in a water bath. The systems wwe mixed for 2 min each with a vortexer, and them separated at 25 f 0.5 OC in a centrifuge at 5000g for 10 min, where 8 is the acceleration due to gravity. The phase composltlons of some of the systems were determined using a gravimetric method. Approximately 100 mg of phase was we@ted into a glass tube using an analytical balance rerrdhg f 0.1 mg. Two volumes of wam were added, and the soiutbns were shell frozen in a mlxture of dry ice and
acetone. A lyophilizer was used to sublimate the water under a vacuum of X13.3 Pa for 24 h, after which the tubes were again weighed. The sample tubes were placed on the surface of a hot plate at approximately 450 OC for 5 days. The PEG was oxidized and volatilized whlle the salt remained as a white ash. The tubes were repeatedly weighed until the mass was constant. The estimated reproducibility of the phase compositions was f0.4% wlw. The concentrations of salt and PEG in the upper and lower phases of some systems were measured using HPLC-gel permeation chromatography. The column used was a TSK G1000 PW (300 X 7.5 mm) with a mobile phase of 6.3 mM Na,HPO, and a flow rate of 0.7 mL1min. The PEG and salt were detected using a refractive index detector. Samples were
Jownel of Chemical and Enginewlng Data, Vol. 37, No. 2, 7992 273
Ammonium S u l f a t e (mass %)
Ammonium S u l f a t e (mass %)
FIguro 4. Phase diagrams of ammonium sulfate systems with PEG 1000 (a) and PEG 8000 (b). The compositions of them systems (A-H) are given In Table I V .
a
3
b
0
z
OD
el
:
Y
0 v)
m (3
O W
n
Sodium S u l f a t e ( m a s s %)
C
Sodium S u l f a t e ( m a s s %)
3
Sodium Sulfate (mass %)
Flgwo 5. Phase diagram of sodium sulfate systems with PEG 1000 (a), PEG 3350 (b), and PEG 8000 (c). The compositions of these p h s e systems (A-N) are given in Table V.
diluted with water 80 that the Megrated area of thelr peaks fell into the itnear range of standard curves ganerated using PEG 1000, PEG 8000, (NH&SO+ M@O,, and N8*SO& The phew compositions were reprducible withift &0.4% w/w. The dendtks were da#rmnsd d n g a vibretlng &tube densitometer with a t e m p t w e of p5 f 0.1 O C . The reproducibility of the densltbs of the phases wag estimated to be f0.8 kg/m3. The viswsbs wsrack&"d using a cone and plate viscometer wlth a temperature bath of 25 f 0.1 O C . The viscosities of the separated phases were reproducible within f0.0003 Pa-s.
The lot of PEG 1000 used in these experiments had an M, of 1075 and a M, of 1125 (M,/M, = 1.05). The PEG 3350 used had an M, of 3200 and an M, of 3400 (M,/M, = 1.06). The PEG 8000 used had an M, of 8070 and an M, of 9700 (MJM, = 1.20). The densities and vlscostties of the phase systems are compiled kr Tables I-V. The viscoslbies generaw increase wlth the increase of PEG molecular mass. The phase densities approach thet of water, due to the water content. I n each phase system, both properties tend to increase with
274
J. Chem. Eng. Data 1992, 37, 274-277
longer tie-line length, with exceptions occurring in the region close to the critical point. The phase-composition data from Tables I-V are plotted in phase diagrams in Figures 1-5, respectively. The tie lines are determined by connecting each corresponding set of total, bottom, and top phase points. The binodlai curve is drawn through the top and bottom phase points, and is estimated near the critical point on the basis of the bcations and trends of the top and bottom phase compositions and, in some cases, single phase points. The phase diagrams with (NH4),S04, MgSO,, or Na2S04 combined with PEG 1000 or PEG 8000 were determined using the HPLC method. The remainder of the diagrams were determined gravimetrically. Points from the HPLC data sets were repeated using the gravimetric method. While both methods had virtually the same reproducibility and resulted in values identical within their uncertainty, the gravimetric technique was simpler and less labor intensive.
Llterature Clted (1) Albetnson, P.-A. P a m of cell Pa&&s and A k m n o k & s , 3rd ed.;Wiiey-Interscience: New York, 1988. (2) Waiter, H.; Brooks, D. E.; F W , D. P8/t/th& h AqTWOPhsm Systems; Academk Press. Inc.: Orlando, FL. 1985. (3) Cabezas, H., Jr.; Kabirl-Bedr, M.; Szlag, D. C. BIoseperadbn 1990, 1 . 227. (4) Edmond, E.; Ogston. A. G. Blodwm. J . 1986, 109, 589. (5) King, R. S.; Blanch, H. W.; Rawnltz, J. M. AIJ . 1906, 31,1585. (8) KeMrcBedr, M. Ph.D. University of Arizona, Tucson. AZ, 1990. (7) Srlag, D. C.; Gulweno, K. A.; Snyder, S. M. In Downshem and Bkseperedbn; Hemel. J.4. P., Hunter, J. B., Skdar, S. K.. Eds.; ACS Symposium Serks 419 Amerlcan Chemlcai Socbty: Washington, DC. 1990; pp 71-88. (8) Diamond, A. D.;Hsu, J. 1.;Bktechnd. B&eng. 1989, 31. 1OOO. (9) Stewart, R.; Todd, P. In hnn F r o " in skprocessins I I ; Skdar, S., Todd. P., Bier, M., Eds.: ACS Symposium Series; Am arican Chemical Society: Washington, DC, in press.
Received for review July 26. 1991. Revised January 10, 1992. Accepted January 28, 1992.
Anion Exchange in Amberlite IRA-400 and Amberlite IRA-410 Ion Exchange Resins Modedo Lbpez, J o d Coca, and Hermlnlo Sadre' Department of Chemical Engineering, University of O v k b , 33071 O v k b , Spain
EqulHbrkm for tho binary exchange of anknr on Amberllte IRA400 and AmborHte IRA410 war moarwed. Standard metwere used to detormlne opwatlng characterktlcs of both rorlns. EquHlbrlum data were obtained by the batch method. The ftttlng of blnary lon exchange botherm oquatknr Is an Important aspect of data analyrk. The Lagmulr, FrwumHlch, Sllpr, and Kobk-Corrlgan hothem were t r a " e d to a llnear form and thdr adJmtabk paramotem estimated by Ilnear r.gr.rdon. Tho Langmulr kotherm k the moat ruttable for both corrolatlon of qqulHbrlum data and predktlon and lntorprotatlon of halhrorrph cwves. Introductlon The three factors that can affect the behavlor of an ion exchange column are equilibrium, kinetics, and mechenics . The degree of column efficiency depends primarily on equilibrium ( 1 , 2). Various methods have been used to obtain binary lon-exchange equiiibrlum data. The simplest is the batch method, proposed by Gregor and Bergman (3). The dimensionless equivalent &nic hcllon, x and y , for fluid and solid, respectively, are defined by
where C, = concentration of the ion species h the solution, Co = total concentration of the soknion phase, 0, = collcentratkn of the ion species in the solution phase, and 0 = total exchange capacity of the resin. To whom correspondence should be addressed.
0021-9568/92/1737-0274$03.00/0
This paper focuses on the study of the a n h equillkiun data of the AmberHte IRA-400 and IRA-410 resins fitted to linear transformetkns of different isotheml equatkns. The regesskn coefficients for each resin were determined.
Expwlmental Sectlon
The resin phase consisted of Amberlite IRA400 (type I) and Amberlite IRA410 (type 11) gel strong-base anbn exchange resins (supplled by Rohm and Haas Co.)in the X-form (X = C03H, Ci, OH, SO4). Solution phases were mixtves made up using sodium salts of both the Xanion and fluoride anion required to obtain a total concentration of 0.05 N. The resins were washed with distliled deionized water and regenerated or eluted with 4% sodium sulfate, except in eiutions of sulfate ion where 1 N sodium nitrate was used (I). The resins were condltbned by alternate conversbns to the hydroxide and c h h b forms and washed until no further chloride could be detected in the effluent. Part of this material was converted into the bicarbonate, chloride, hydroxide, and sulfate forms. To determine the weter content of th@ wet resin, resin sam pies were put in a Biichner funnel which was connected to a water vacuum pump to remove the interstltlai water. After weighing, the resin samples were drled in a desiccator over phosphorus pentoxide to constant weight. The resh, &ns& was determined in water by a standard-type picnometer. The to&/ w C & fOr W O W 8m9WltS (appr;oxknately 5 9) Of the ionic form of the resin wae d e t e " d by adding an excess of sodium salt of X-ankns, washing until no further X-anion could be detected in the effluent, rege"tlng the resin with exactly 1 L each of 4% sodium sulfate (or 1 N sodium nitrate), cdlecting this effluent in a vohme&lc flask, and finaHy det#mlnhg the X-anbn (5-8). Table I shows the physicochemical characteristics of the used resins. I n the batch equilibrium studies , weighed amounts (2-3 g) of moist resin were equilibrated with 100 mL of mixtures of 0 1992 American Chemical Society
