Ca e Mg da agua do mar remoçao

8/4/2019 Ca e Mg da agua do mar remoao

1/16

REACTIVE&

FUNCTI NALsOLY ERSELSEVIER Reactive & Functional Polymers 28 (1996) 111-126

Separation and concentration of calcium and magnesium from seawater by carboxylic resins with temperature-induced selectiqityDm itri Muraviev *, Joan Noguerol, Ma nuel Valiente *

Departam ent de Quimica, Q &mica Analit ica, Univ ersitat Aut bnoma de Barcelona, E-08193 Bellaterra, Barcelona, SpainReceived 20 June 1994; revised version accepted 16 February 1995

Abstrac t

Processes of concentration and separation of calcium and magnesium from artificial and natural sea water bycarboxylic ion-exchange resins of acrylic and methacrylic types at different temperatures have been inve$tigated.Th e values of equilibrium separation factor o! for Ca*+-Na+, Mg2+-N a+ and Ca*+-Mg *+ exch anges in ternarysystems have been determined in the temperature range of 10C to 80C. A significant increase of (Y vblues atelevated temperatures has been observed in the first two cases while for Ca+-Mg*+ exchange less redarkabletemperature dependence of a! can be distinguished. This effect has been show n to allow a selective thermosfrippingof Ca*+ and Mg2 + from the resins equilibrated at 80C with sea water in applying cool sea water at lo&. Thethermostripping leads to a selective desorption of both Ca*+ and Mg*+ while Na+ ions remain sorbed, resdlting inthe increase of Ca2+ and Mg*+ concentration in the eluate up to 50% (in comparison with the initial sea waler) anda decrease of 10% for Na+ concentration. These results may be considered as unique in polythermal concehtrationin comp arison with, e.g. conventional evaporation technique. The results of consecutive sorption-therm ostrippingcycles have shown the possibility to concentrate calcium and magnesium from natural sea water m ore thain threetimes by applying reagentless (and waste less as a result) ion-exchan ge technique. The results of frontal sedarationof Ca*+ and Mg*+ on acrylic resin in Na+-fo rm from natural sea wate r and thermostrip ping solutions obtainedare also presented. The novel approach for forecasting temperature dependences of the resin selectivity has beenproposed. The approach is based on a thermodynamic interpretation of the results obtained that allows to ;predictthe temperature dependences of both a! (for binary M g *+-Na+ exchange) and the apparent equilibrium constant ofternary Naf-Ca *+-Mg*+ exchange.Keywords: Ion exchan ge; Calcium; Magnesium ; Separation; Dual-tem perature concentration; Carboxy lic resins,thermodynamics; Carboxylic resins, sea water treatment

1. IntroductionAt present, around 25% of overall world pro-

duction of magn esium is yielded from hydromin-era1 resources (sea water, underground brinesand bitterns of some salty lakes) [l]. Traditional* Corresponding authors.

methods for producing magne sium by proc ssinghydromineral sources, despite their profit a ility,do not satisfy growing ecological standards [2].Consequently, new alternative ecologicall$ cleantechnologies, based on e.g. ion-exchange sepa-ration methods, have to be developed. Sbveralion-exchange methods for separation of dagne-sium and calcium have been discussed iin the

1381~5 148/96/$15.00 0 1996 Elsevier Science B.V. All rights reserved.SSDI 1381-5148(95)00046-l


2/16

1 1 2 D. Muraviev et al. IReactive & Functional Polymers 28 (1996) 111-126

literature. Some of them are based on applyingcomplexing agents for enhancing the separation(see, e.g. [3,4]). Barba et al. [5] have appliedsupersulfonated resins (containing more than 1sulfogroup per benzene ring of PS-DVB ma-trix) for the same purpose. The methods men-tioned may be called convent ional ion-exchangesepara t ion t echniques which are based as a ruleon the multistage procedures, involving sorption,desorption (elution) and regeneration steps. Tocarry out the last two stages one needs the ex-penditure of auxiliary chemicals that results inproduction of wastes.

Another group of separation methods isbased on the exploitation of the variation of resinaffinity towards target ions with temperaturefor governing the separation process. Paramet-ric pumping technique (see, e.g. [6-ll]), dual-temperature ion-exchange processes [12-141 andthermal ion-exchange fractionation [15,16] canbe refered to the ion-exchange separation meth-ods of this type. One feature alone makes theseseparation techniques most attractive. Actually,all of these separation methods allow to excludethe resin regeneration step (completely or par-tially), which is known to be the main source ofwastes in ion-exchange technology. In this sense,ion-exchange processes based on the above men-tioned techniques are ecologically clean andpractically wasteless. Nevertheless, the practi-cal interest towards these separation methodsis still very limited due to the lack of informa-tion about ion-exchange systems with propertiesfor designing reagentless and wasteless (as a re-sult) technologies based on the dual-temperatureion-exchange fractionations. One type of con-ventional ion exchangers has been used in themajority of studies carried out: sulfonic resins(see, e.g. [15,16]). On the other hand, Klein et al.[17] and later Ivanov et al. [18-211 have reportedstrong temperature dependence of ion-exchangeequilibrium in systems including carboxylic resinsand the binary model alkali/alkali-earth metal-ions mixtures. Carboxylic ion exchangers havebeen applied for the recovery of magnesiumfrom sea water after preliminary decalcination

on zeolites (see [l, and refs. 35-37 therein]),but no data on dual-temperature ion-exchangeconcentration and separation of Ca2+ and Mg2+from sea water using this type of ion exchangerscan be found in the literature. The same con-cerns with the equilibrium data in ternary ionicsystems involving carboxylic resins and sea wa-ter metal ions. For that reasons the followingproblems are studied in this paper:

(1) ion-exchange equilibrium on commer-cially available carbo@ic resins of acrylic andmethacrylic types in ternary ionic systems, in-cluding the macrocomponents of sea waterMg 3+ Ca2+ and Na+, at different temperatures;(2) a novel approach for predicting tempera-ture dependences of resin selectivities from theenthalpies of binary exchange on the same resins.

(3) kinetics of thermodesorption of the stud-ied metal-ions with natural sea water at lowtemperature from the resin pre-equilibrated withsea water at high temperature.

(4) prediction and experimental determina-tion of concentration factors for Ca2+, Mg2+ andNa+ in consecutive thermo-sorption-strippingcycles.

(5) frontal separation of the studied metal-ions by the carboxylic resin, and

(6) conditions for the wasteless conversion(regeneration) of the resin.

2. Experimental2.1. Ma t er i a l s a n d a n a l y t i ca l m e t h o d s

Samples of the natural Mediterranean sea wa-ter used in the present study were obtained fromthe area near Tarragona (Spain). Natural sea wa-ter was boiled and then filtered for removal oforganic matter before the use in ion-exchangeequilibrium studies. Artificial sea water sampleswere prepared from NaCl, MgC12, CaC12 andNa2SOs (Probus, Spain) of p.a. quality used asreceived. The composition of artificial sea wa-ter used was as follows; ionic species/C (g-equiv/1); Cl-/0.48; Naf/0.40; Mg2+/0.11; Ca2+/0.02;so:-IO.05


3/16

D. Muraviev et al. /React iv e & Functi onal Poly mers 28 (1996) III-126 113

Table 1Properties of Lew atit carboxilyc ion exchange resinsPropertyFuncional groupCapacityAverage pore diameterSurface area (BET)

R 249-Kmethacrylic acid1.3 molil270 A200 m/g

R 250-Kacrylic acid2.9 molil260 A100 m2/g

Macropo rous carboxylic ion exchangers Lew -atit R 249-K an d Lewatit R 250-K were receivedfrom Bayer Hispania Industrial, S.A. The mainproperties of the resins are shown in Table 1[22].

The concentrations of metal ions were de-termined by atomic emission spectroscopy usingICP-AES technique with ARL M odel 3410 spec-trometer (Fisons, U.S.A.) provided with mini-torch. Determination of H+ and OH- ions werecarried ou t by potentiome tric titration using aCrison pH-meter provided with a combined glasselectrode.2.2. M e t h o d s2.2.1. Ion- exchange equi l ibrium

The ion-exchange equilibrium was studied un-der dynamic conditions in thermostatic glasscolumns (inn. diam. = 1.4 cm). The columnswere loaded with a certain amount of the ionexchanger which remained constant during thegiven series of exper iments. The total capacity ofeach resin bed was determined after convertingthe ion exchangers into Na+-form with 0.5 MNaOH solution followed by rinsing w ith 0.5 MNaC l solution and passing a known volume of0.5 M HCl through the bed. Hydrochloric acidsolution was collected into a volumetric flaskand its concentration was then determined. Theion-exchange capacity of the resin bed (Q) wascalculated as the amount of acid retained bythe resin as follows: Q = V(C, - C i), whereV is the volume of HCl so lution passed throughthe bed, dm3 and C, and Ci are the initial and1Rinsing with sodium chloride solution w as used to suppressthe hydrolysis of the resin in Na+-form which takes placewhen the rinsing is made just with water.

the final concentration of the acid, mol/dm3,respectively. Then, the resins were convertedback into Na+-form by rinsing with NaC l so-lution and equilibrated at certain temperature.Initial solution (artificial or natural sea water)was passed through the columns at constant flowrate (3 cm3/min) up to achieving the equilib-rium. The eluate was collected in portions ofknown volume w here concentrations off Na+ ,Ca2+ and Mg 2+ were determined. Achieving toion-exchange equilibrium in the systems understudy was controlled by the periodical compari-son of metals concentration in the solution leav-ing the column with that of the feed solution.When the solution from the column outlet hadthe concentration of Na+ , Mg2+ and Ca2 closeto the feed solution the flow was stopped andthen resumed after certain period of time. Theequality of the feed concentration with that ofthe solution collected after the break was consid-ered as a criterion of achieving the equilibriumin the system . After equilibration, the solutionphase w as separated from the resin by evacuatingwith a water pump followed by the stripping with0.5 M HCl solution. The an alysis of Na+ , Mg 2+and Ca2+ in the corresponding eluate w as car-ried out. Then the resin was converted back intoNa+-form and prepared for the next run. The re-sults of the stripping solution analysis were usedto determine the separation factors c$& e$$ andc~$~by the use of the following expression:

Me 2 YMe2 XMe,(31Me~= r xMel XMe,where Y and X are the equivalent fractions ofions under separation in the resin and solutionphases, respectively; indices 1 and 2 are chosenso thata > 1.

The values of Na+, Ca2+ and Mg2+ concen-tration in solution samp les collected during theequilibration of the resin were used to obtain therespective breakthrough curves (see Fig. 3) and tocalculate c& values using the technique based onthat proposed by Spedding et al. [23] for determi-nation of sepa ration factors for nitrogen isotopes.The technique allows to calculate Q values with-


4/16

1 1 4 D. Muraviev et al. IReact iv e & Functi onal Poly mers 28 (1996) 111-126

out the direct analysis of the resin phase compo-sition [24,25] in applying the following equation:

& ViCi(Xo - Xi)1+&X0 = ci=l QJfo(l - xo)where E = a! - 1; X0 and Xi are the equivalentfractions of the better sorbed ion in the initialmixture and in the i eluate fraction, respectively(Ca2+ in our case); Vi is the volume of eluateportion i (cm3); Ci is the total concentration ofseparated ions mixture (meq/cm3); Q is the fullion-exchange capacity of the resin bed (meq); jis the number of eluate fractions where Xi # X0.

Values of CXM,etermined by the two inde-pendent methods (see Eqs. 1 and 2) were char-acterized by the uncertainties less than f5% forthe first and around f7% for the second.2.2.2. Thermostripping and thermosorption

Experiments on thermostripping and ther-mosorption of Ca2+ and Mg2+ were carried outusing the following technique: after equilibrationof the resin with sea water at certain tempera-ture the excess of the equilibrium solution wasremoved from the column so that its level coin-cided with that of the resin bed. Then the cool-ing (in case of thermostripping) or heating (forthermosorption) of the column was started. Afterreaching the appropriate temperature the sea wa-ter was passed through the column and collectedin portions followed by the analysis of Ca2+,Mg2+ and Na+ content. In the series of exper-iments of consecutive sorption-thermostrippingcycles, the composition of the initial solution usedin each cycle corresponded to that of the sampleof thermostripping solution obtained during theprevious cycle which contained the highest con-centration of Ca2 and Mg2+.2.2.3. Kinetics of thermosorption

Kinetic experiments on thermodesorption ofCa2+ and Mg2+ were carried out in thermostaticcolumns applying the shallow bed technique [26].The narrow granulometric fraction of Lewatit R250-K was obtained by dry sieving of air-dry

samples of the resin using 0.42 mm mesh, so,only those resin beads stuck in the holes of thesieve were collected. The average diameter ofthe beads was determined by microscopic tech-nique and appeared to be 0.053 f 0.002 cm. Theheight of the resin bed used was always constantand -0.5 cm. The resin was pre-equilibrated withthe natural sea water at 80C and at flow rate of3 cm3/min, then, the solution was removed fromthe column and the column was cooled at 10C.Cold sea water (at 1OC)was passed through theresin bed at high flow rate (43 cm3/min)2 andcollected in portions where Ca2+ and Mg2+ con-centrations were determined. The degree of theresin bed conversion, F, was calculated from theresults of the analysis as follows:

ic i(Ci - Co>F= i

Q Me (3)where Vi is the volume of eluate sample number7; Ci and Co are the concentrations of Ca2+ orMg2+ in the i eluate sample and in the initialsea water, respectively, expressed in mequiv/cm3;QM~ is the capacity of the resin bed towardsthe ionic species under consideration, and j isthe number of eluate sample where Ci - Co 5AC, AC being the absolute uncertainty on thedetermination of the concentration.3. Results and discussion3. I. Ion-exchange equilibrium3.1.1. Artificial sea water

Temperature dependences of a, : and a,,determined in the experiments with artificial seawater on Lewatit R 249-K and Lewatit R 250-K resins by direct method3 (using Eq. 1) are2 This value of the so lution flow rate has been shown to fitthe conditions of the shallow bed technique.3 Since this method of o determination is based on the resultsof the direct analysis of the phases com position here a ndfurther in the text it is called direct method . The secondmethod used (see Eq. 2) will be mentioned as indirectmethod for 01determination.


5/16

D. Murav iev et al. IReact iv e & Functi onal PolJtners 28 (1996) III-126 11.5200

15 0

a 100

50

I I I20 40 60 80

Temperature (C)Fig. 1. Tem perature dependencies of equilibrium separationfactors for Ca2+-N a+ and Mg*+-Na+ exchanges from artifi-cial sea water on Lewatit R 249-K (white circles) and LewatitR 250-K (black circles).

shown in Fig. 1. As seen in Fig. 1 , for bothacrylic and methacry lic resins , strong positiveinfluence of tempe rature on their selectivity isobserved, i.e. separation factors for Ca*+-N afand Mg* +-Naf exchanges increase significantlyat elevated temperature. The absolute a valuesare lower for the acrylic ion exchan ger than thosedeterm ined for the methacry lic one. This resultscan be attributed to the significan t d ifference intheir capac ity (see Table 1). The selectivity ofion exchangers are known to decrease with theincrease of their capacity [27,28].

The temperature dependences of c&i and a,,are a lso different (in the first case it is muc hstronger than in the second). In this case, thedifference has to be connected with the respec-tive enthalpies (AH) of Ca*+ -Na+ and Mg *+-Na+ exchange reactions. Despite the separationfactor is not a thermodynamically meaningful pa-rameter, it can be associated in certain cases withthe thermodynamic equilibrium constant, K, fion-exchange reaction which can be written asfollows [29-311:

s11ogK = log/c(Y) dY (4)0where Y has the same meaning as in Eq. 1 and Kis the dimensionless equilibrium constant [31]defined for, e.g. Na+-C a*+ binary exchange by:

(5)where NN ~ and N ca are the normalities of Na+and Ca*+ in the equilibrium solution (seal water)and ma and yea are the activity coeffic$nts ofNa+ and Ca*+ in sea water, respectively. Theactivity coefficient ratio for sea water is c$ns tantand can be included in K. Equation (5) ban berewritten for the equiva lent fraction concen tra-tion scale as follows:

where Y and X have the same meaning asi aboveand N ,,, is the total n ormality of metal iions insea water.

Com parison of Eqs. 6 and 1 show s thpt a$and KC a,Na re connected with each other iby thefollowing expression:

CaaN a = KCa,Na

The term in brackets is practically coastant,i.e. XC , = con st; IV ,,, = const and Y$ haspractically a constant value a t least in thq tem-perature range from 20 to 80C (see T4ble 3below). Since Y$! remains also practically con-stant in the same temperature interval, Eg. 7 isapplicable for connecting czNag and KQ-NL an dcan be rewritten as follows:

MeaN a = KMe,Na X a (8 )where Me is Ca2+ or Mg2+ and a =Y$/X$;N;i = const. This allows for atttiibut-ing the temperature dependences of a,: anti c$:to the respective AH values determined fir theexchange of the same ion couples on carbion exchanger KB-4 in binary systems [l


6/16

11 6 D. M uraviev et al. IReactive & Functional Polymers 28 (19 96) Ill-12 6Thus, Timofeevskaya et al. determined differ-ential enthalpies in applying ion-exchange tech-nique which appeared to be 8 and 5 kJ/equivfor Ca2+-Na+ and Mg2+-Na+ exchange, re-spectively, for concentrated solutions [18]. Thesevalues are close to those determined calorimet-rically for the exchange in diluted solutions i.e.8.4 and 7.1 kJ/equiv for the respective ion cou-ples [32]. Smaller AH value for Mg2+-Na+ ex-change in comparison with that of Ca2+-Na+ion-exchange reaction will result in a weakertemperature dependence of a$: than in the caseof a,;*

The above qualitative interpretations of a! vsT dependences can be confirmed by the quanti-tative predictions of these dependences for dif-ferent temperature intervals, which are based onthe following approach: the standard enthalpy ofion-exchange reaction corresponding to a com-plete exchange of one ion by another can bedetermined as follows:

s1

AH0 = AHapdY (9)0

where Y is the same as in equation (l), and AHapis the apparent enthalpy, depending on Y.

AHap can be determined from the tempera-ture dependences of Q! hrough the use of thevan? Hoff type equation [33-351:

6lnKAHaP = - 6(1/T)[ 1lwhere Yl corresponds to the equivalent fractionof the higher sorbed component in the resinphase, equilibrated with solution of the givencomposition (sea water in our case). Equation 10can be also applied for the quantitative predic-tion of CI! s T dependences. For this purpose Eq.10 must be integrated (assuming AHap indepen-dent of T) and rewritten as follows:yl ln !2 = - Hap

KI Rsubstituting K by a! from equation (8) gives:yl ln !2 = - Ha,Ql R

(11)

(12)

Since differentiation in Eq. 10 is based on theassumption YI = const, the same condition mustbe fulfilled for the validity of Eqs. 11 and 12. Asseen from Table 3 (see below), Yc, and YM~ avepractically constant values for both resins stud-ied in the temperature range between 20C and80C. An estimation of AHapfor the ternary sys-tem under study, assuming the independence ofthe respective binary exchanges, can be done byapplying Eq. 10 and assuming the usual Arrhe-nius dependences of (;II s l/T. Such estimationresults on the following values: for Na-Mg2+exchange on Lewatit R 250-K, AH,, = 6.7 kJ/equiv and 7.4 kJ/equiv for Lewatit R 249-K ForNa-Ca2+ exchange AH,, = 4.1 kJ/equiv for thefirst and 4.6 kJ/equiv for the second resin.As follows from the above values of A HapforNa+-Mg2+ exchange, it correlates very well withAHHaP 7.1 kJ/equiv reported by Samchemko[32] for the same binary exchange. This testifiesto the practical independence of AH,, of theresin composition (see Eq. 10) and allows forsubstituting AHapby AH in this case.

Hence, quantitative predictions of ln(a2/al)for Na-Mg2+ exchange in different T2 - Tl in-tervals, can be accomplished using Eq. 12 withA Hap = AH0 = 7.1 kJ/equiv determined inde-pendently (calorimetrically). The results of thesecalculations are presented in Table 2.

As seen from the results given in Table 2, thecalculated ln(a;?/at) values demonstrate a satis-factory fit with those determined experimentallyin the ternary system.

However, for Na+-Ca2+ binary exchange inthe ternary system, the results of the quanti-tative predictions of ln(a2/al) vs (T2 - TI) in-tervals give remarkable deviations of the exper-imental values from the computed ones. Thisdisagreement can be attributed by a strong de-pendence of AH,, vs Y that follows from theremarkable difference between estimated AHapvalues 4.1 and 4.6 kJ/equiv mentioned before,and AH = 8 and 8.4 kJ/equiv reported in theliterature [18,32].

Consider now the overall ion-exchange pro-cess in the system under study, which can be


7/16

D. Murav iev et al. I React iv e & Functi onal Poly mers 28 (I 996) I I I- 126 117

Table 2Temperature dependences of oNag for carboxylic resins, calculated and experimentally determined from artificial sea waterTz-TI Lewatit R 250-K Lewatit R 249-K In(o2/cyl) calculated for

a1 ff 2 Waz/ad exe . 01 ff2 Wadad exe. AH = 7.1 kJ/mol Prom Eq. 7353-333 28.12 36.21 0.15 39.53 47.09 0.11 0.14353-313 22.85 36.21 0.27 28.76 47.09 0.30 0.31353-293 16.00 36.21 0.47 19.91 47.09 0.51 0.50333-313 22.85 28.12 0.12 28.76 39.53 0.19 0.16333-293 16.00 28.12 0.32 19.91 39.53 0.40 0.35313-293 16.00 22.85 0.21 19.91 28.76 0.21 0.19

Table 3Composition of resin phases and apparent equilibrium constants of ternary Na +-M g +-Ca2+ exchange determined from artificialsea water at different temperaturesT $128 329 331 333 335 3

Lewatit R 250-KYea YMS0.33 0.500.31 0.560.32 0.590.34 0.580.34 0.60

YN a k0.17 9.540.13 12.890.09 18.230.08 22.270.06 29.09

Lewatit R 249-KYea YM S0.31 0.470.33 0.570.34 0.590.35 0.590.36 0.59

YN a

0.220.100.070.060.05

k _7.35

16.3123.0431.0237.89

described by the following equation:4 R- COONa + Ca*+ + Mg2+ +

2R-COOCa + 2R-COOMg + 4Na+ (13)The apparent constant, k , of the above tri-ionic-equilibrium can be written as follows [36-381 :

(14)where X and Y have the same meaning as in Eq. 1.

The values of k determined by using artificialsea water at different temp eratures are collectedin Table 3.

Equation 13 can be represented as a super-position of the two equations describing the bi-nary Na+-Ca*+ and Na-Mg2+ exchanges. Con-sequently, the standard enthalpy (A HP,,) of ion-exchange reaction 13 is equal to the sum of AHvalues for the exchange of the respective ion cou-ples. Following this, an estimation of ln(kz/ kl) fordifferent temperature intervals using Eq. 11 andassuming yt x landAH,, % AH& can be ob-tained. The values of A HP,,can be taken from the

Table 4Calculate d and experimentally determined In(kllkz) forNaZ+-CaZ+-Mg2+ equilibrium on carboxylic resins a t dif-ferent temperatures (see text)T?-T, Calculated Experimental

(1 ) (2 ) Lewatit LewatitR 250-K R 249-K353-333 0.27 0.32 0.27 0.20353-313 0.57 0.68 0.47 0.50353-293 0.91 1.08 0.81 0.84353-283 a 1.10 1.31 1.12 1.64333-313 0.30 0.36 0.20 0.30333-293 0.64 0.76 0.55 0.64333-283 a 0.83 0.99 0.85 1.44313-293 0.34 0.41 0.35 0.343 13-283 a 0.53 0.63 0.65 1.14293-283 a 0.19 0.22 0.30 0.20(1) AHtot = 13 kJ/mol; (2) AHtot = 15.5 kJ/mol.a The data refered to the temperature intervals involving 283K have been included to illustrate that the deviations of YQ ,and YM ~ rom the constant values (see Ta ble 4) resuh in thestronger disagreement of experimental and calculated results(see text).

above references [18,32], i.e. (1) AH;,, = 13 kJ/equiv and (2) A HE,, = 15.5 kJ/equiv. The resultsof these c alculations are presented in Table1 4.


8/16

118 D. Muraviev et al. I Reactive & Functional Polym ers 28 (1996) Ill-126

Comparison of the results shown in Table 4testifies to the correction of the approach pro-posed and to the validity of the assumptionsmade. As seen from Table 4 the values ofln(k,/kz) calculated from AH:,, = 13 kJ/equivare in a better agreement with those determinedexperimentally. This indicates that AH0 valuesdetermined for the binary Na+-Ca2+ and Na+-Mg2+ exchanges in concentrated solutions [18]are most appropriate for the description of thetemperature dependence of ion-exchange equi-librium in the ternary system under study.

With respect to Ca2+-Mg2+ exchange, thetemperature dependences of cogs for Lewatit R250-K and R 249-K resins are presented in Fig. 2where aM,is obtained by direct and indirectmethods. The problem of elucidating the tem-perature influence on Ca2+-Mg2+ exchange iscomplicated because of the weak temperatureeffect obtained (much weaker than in the pre-vious two cases discussed) which was expectedfrom the corresponding value of AH = 3 kJ/equiv [18]. As seen from the data shown inFig. 2, o$$ values determined by indirect methoddemonstrate a slight increase with the increase oftemperature for both ion-exchange resins stud-

4-

Caa m 3

2-

10 20 40 60 80Temperature (c)

ied while the direct method does not allow todistinguish any clear influence of temperatureon (Y.This disagreement of a values determinedby different methods may be associated withone feature of the indirect method which makesit advantageous in comparison with the directone. Determination of cr by indirect methodis carried out w i t h o u t s ep a r a t i o n o f p h a s e s afterequilibration, which is known to be the sourceof additional and hardly measurable experimen-tal errors [24]. This is particularly true for theion-exchange equilibrium on carboxylic resins atdifferent temperatures, since equilibrium, in thiscase, may be easily shifted due to either hydrol-ysis of ion exchanger (in applying rinsing tech-nique) or indefinite change of the temperatureinside the column (evacuating solution with wa-ter pump is accompanied by filtration of cool airthrough the resin bed). For a better elucidationof the temperature dependence on the separa-tion of magnesium and calcium, some additionalinformation has been applied. This informationcan be extracted from concentration-volume his-tories (breakthrough curves) obtained in eachexperiment carried out for phases equilibration,see Fig. 3. As seen in Fig. 3, the sorption front of

5

4-

caa 3-Mg

2-

1 (0

I I I I20 40 60 80

Temperature (S)Fig. 2. Temp erature dependencies of equilibrium separation factors for Ca2+ -Mg* + exchang e determined by direct (squares) andindirect (circles) methods from artificial sea water by Lewa tit R 250-K (a) and Lew atit R 249-K (b) resins.


9/16

D. Mum viev et al. I Reactive & Functional Polymers 28 (1996) Ill-I26 1191 . 5

1 . 0

C/CO

0.5

0.0 , 1 / ,0 500 1000 1 5 0 0 2000 2500

Volume (ml)Fig. 3. Concentration-volume history of frontal separation oflVlg=+ and Ca2+ by Lewatit R 250-K resin in Na+-form at80C. Q = 57 meq.

Ca2 + can be chosen as one of the parameters forcharacterization of M g2+ and Ca2+ separation atdifferent temperatures. Since the experimentalcond itions influencing the length of this front,such as solution concentration and flow rate,were kept constant in all the experiments, anyalteration of the front p osition can be evidentlyattributed to the change of the temperature ofthe system. This is clearly observed from the datashown in Fig. 4 where the relative lengths ofcalcium sorption fronts at different temperaturesare shown for Lewatit R 249-K. A similar effectwas observed for Lewatit R 250-K. In this fig-ure, lequil was obtained from the correspondingconcentration-volume histories by determiningthe volume of solution passed through the col-umn before achieving the equilibrium. The val-ues of Vequil Pp lied to calculate the data plottedin Fig. 4 were 1220 ml for Lewatit R 250-K(Q = 32 meq) 4 600 ml for Lewatit R 249-Km$ )n,=ii;l no;). The relative error for the deter-

equilwas around 5%. The variations4Note that the capa city of the resin bed used in this series ofexperiments was approximately twice less than that applied inthe frontal sepa ration of Ca 2+ and Mg2+ shown in Fig. 3.

;a 0.60z-Y 0.4

0.2

0.00.0 0.2 0.4 06 0.8 1.0

VN qu,,Fig. 4. Influence of temperature on relative length oif calciumsorption fronts for artificial sea water on Lew atit R 249-K:20C (circles), 40C (squares) and 80C (triangles).

of VesUi l with the temperature lied within thecorridor of experim ental errors.

The data shown in Fig. 4 demonstrate that thefront of Ca2 + shifts remarkably to the right withthe increase of the temp erature. This shift willresult on the increase of the separation ldegreeof Mg2+ and Ca2+ (which can be determined asthe ratio of their concentrations in the respectivesolutions samples) in the head part of thelbreak-through curve (see Fig. 3). This is confirmed bythe data presented in Fig. 5 where the separationdegrees of Mg *+ and Ca*+ for the head partsof concentration-volume histories obtained onLewatit R 250-K at different temperatunes areshown.3.1.2. N a t u r a l s ea w a t erIon-exchange equi l ibr ium. Experiments on Istudy-ing ion-exchange equilibrium in Mg 2+-C a2+-Na+ system on Lewatit R 250-K and Lewatit R249-K resins from natural sea water w ere carriedout at two temperatures: 10C and 80C ssum -?!ling the temperature dependences of a,:, c$ andQ$$ ~ o be practically linear as was demo nstratedin the experiments with artificial sea water. Re-


10/16

12 0 D. M uraviev et al. I Reactive & Functional P olymers 28 (1996) 111-126

60 1603

BE 40

0 I I I I I I I0 100 200 300 400 500 600 700Volume (ml)

Fig. 5. Ratios of Mg 2+ and Ca*+ concentrations (Mg 2+ purifi-cation degree) in solution samples collected in frontal sepa ra-tion from artificial sea water by Lew atit R 250-K in Na+-format different temperatures: 20C (circles), 60C (squares) and80C (triangles).

suits of this series of experiments are presentedin Table 5 where the values of cx for the three ioncouples studied for the ternary system and k aretabulated. For (;IIMga the values were determined byboth the direct and the indirect methods.

As seen in Table 5, cx values for C a+-N a+and Mg 2t--Na+ exchanges determined fromsamp les of natural sea water are much sm allerthan those determined from artificial sea water(cf. Fig. 1) while for Ca2+ -Mg 2+ exchange theyare similar to these shown in Fig. 2 for artifi-cial solutions . This difference can be attributed

Table 5Values of 01 for the binary exchanges Ca2+-N a+, Mg2+-Na+ and Ca2+-Mg2+ and k for the ternary Na+-C a2+-Mg2+exchange on carboxylic resins from natural sea water at 10and 80CLewatit resin t (C) a,: a,8 d$ k

direct indirectR 249-K 10 21.8 7.41 2.94 2.79 6.04

80 32.8 9.90 3.31 3.46 8.01R 250-K 10 31.0 10.2 3.05 2.60 8.0980 40.2 12.6 3.19 3.28 9.78

to the higher total ionic concen tration in nat-ural sea water samples received, in particularsodium concentration, in comparison with thatin artificial sea water prepare d by following thereported data in the literature (see, e.g.

iJ2, p.

261). Therefore the decrease of c$ and aNi canbe ascribed to the electroselectivity effect [39].As follows from Eq. 7 the increase of N,,, re-sults to the decrease of cz values. Nevertheless,the trend in changing a! values with tempera-ture for different ion couples is similar to thatobserved in artificial s ea water and also followsthe order of differential enthalpy values for therespective ion-exchange reactions (see above).Comp arison of ~~2s values determined by differ-ent methods (see Table 5) demonstrates that theindirect method allows for a better elucidationof temperature dependence of a! than the directmethod (as discussed above).

The influence of temperature on the separa-tion of Ca2 + and M g2+ from natural sea water byLewatit R 250-K resin under dynamic conditionsin column is illustrated by the data presented inFig. 6, where the positions of calcium sorptionfront at 10C and 80C (a) and magn esium purifi-cation degree (b) are shown. A s seen in Fig. 6athe increase of temperature in the system leadsto remarkable retardation of calcium front thatresults in a more effective separation of Ca2 +from Mg2+ (see Fig. 6b).

The results obtained for the ion-exchangeequilibrium of the ternary Ca*+ -Mg 2+-Na +system on carboxylic resins both acrylic andmethacry lic ion exchangers from artificial andnatural sea water can be used to carry outa dual-temperature ion-exchange concentration.The highe r capacity of the acrylic resin provide swith a better charac teristics for this application .K in e t i cs o f t h e m m fe s or p t io n . The results ofstudying kinetics of thermodesorption of Ca2 +and Mg*+ in eluting w ith natural sea water at10C from Lewatit R 250-K pre-equilibrated5Chronologically samples of natural sea water ha ve been re-ceived after carrying out experiments with artificial solutions.


11/16

D. Muraviev et al. IReact iv e & Functi onal Poly mers 28 (1996) 111-126 121

1 . 0

0 . 6

,o 063rH0 4

0 . 2

i-6 0

50 -

4 0 -

i s0s 3 0 -z

2 0 -

1 0 I

0 . 6 0 . 6 1 . 0 0 1 0 0 2 0 0 3 0 0 4 0 0 5 0 0 6 0 0 7 0 0

JN,,,, Volume (ml)Fig. 6. Influence of temperature on relative length of calcium sorption fronts, (a), and Mg2+ purification degree (see Fig. .5), (b),in frontal separation of CaZ+ and Mg2+ from natural sea water by Lewatit R 250-K resin in Na+-form at 10C (circles) and 80C(squares).

0.0 2.5 5.0 7.5 10.0time (min)

Fig. 7. Kinetics of thermodesorption of Ca2+ and Mgz+ fromLewatit R 2.50-K with natural sea water (see text).

with sea water at 80C are presented in Fig. 7where the kinetics curves for ions under studyare sho wn. As seen in Fig. 7 Mg 2+ demonstratesa higher rate of release than Ca2+ . The valuesof the time of half conversion (ta.5) estima ted

from the respective kinetic curves appelared tobe 114 s for M g2+ and 222 s for C a2+. The dif-fusion coefficients, 0, of Ca2+ and Mg 2+ in theresin phase have been estimated assum ing thekinetics of thermo desorp tion to be controlled bythe intraparticle diffusion and by applying thefollowing equation [40]:

r:D = 0.03-& (15)where r,, is the radius of the resin beads in aswollen state = 0.0265 cm for the resin fractionused.

The calculated values have been found tobe x 1.0 x 10e7 cm2 s-l for Mg 2+ and 1.9 x10e7 cm2 s-l for Ca2+ , which correlates +ith therespective data for alkali-earth metal ion$ [26, p.901. The results obtained indicate that noi kineticdifficulties in carrying out the therm ostrm ping ofCa2+ and M g 2+ from the resin can be exF/ected.3.2. Dual-temperature ion-exchange concelttration

Concentration-volume histories obtajned bythermostripping at 10C from Lewatit 4 250-Kequilibrated at 80C with natural and artificial


12/16

12 2 D. Muraviev et al. IReact iv e & Functi onal Poly mers 28 (1996) 111-126

1 . 6

1 . 4

c / c , 1 . 2

1 . 0

0 . 8

(4

I I / I I0 2 0 4 0 6 0 8 0 1 0 0

Vo l u me ( ml )

1 . 6

1 . 4

1 . 2

1 . 0

0 . 8 I I I I0 2 0 4 0 6 0 8 0 1 0 0

Vo l u me ( ml )Fig. 8. Thermostripping breakthrough curves for natural (a) and artificial (b) sea water from Le watit R 250-K resin. Na+(triangles), Cazf (circles) and Mg 2+ {squares).

sea water are sho wn in Fig. 8a and b, respec-tively. The total cap acity of the resin bed usedin this series of experiments was 57 mg-equiv.Cooled sea water was passed at 1.5 cm3/minof flow rate. As seen in Fig. 8 thermostrippingleads to a selective desorption of both C a2+and Mg2+ from the resin while Na+ ions aresorbed. This results in the increase of divalentmetal ions concentration in eluate and decreaseof sodium con centration. This situation may beconsidered as unique in polythermal concentra-tion in comparison with, e.g. conventional evap-oration technique which allows for increasing theconcentration of all com ponen ts of the solutionunder treatment.

In the case of Lewatit R 249-K resin, results ofthermostripping have shown that the maximu mconcentration degree (C/C,) of Ca2 + and Mg 2+for this ion exchanger do not exceed 38% whilefor Lewatit R 250-K this parameter is muchhigher (see Fig. 8). This confirms the conclusiongiven above ab out the better applicab ility of theacrylic resin for dual-temperature ion-exchangeconcentration.

After the thermostripping, the resin phaseappears unloaded with Ca2+ and Mg2 + and be-comes able to sorb them again without any addi-

w, 0 . 9

0 . 8

0 . 7

0 . 60 5 0 1 0 0 1 5 0 2 0 0 2 5 0 3 0 0

Vo l u me ( ml )

Fig. 9. Thermoso rption breakthrough curve for natural seawater on Lewatit R 250-K resin. Na+ (triangles), Ca2+ (cir-cles) and Mg2+ (squares).

tional treatment (regeneration). Passing hot seawater through the resin bed depleted after ther-mostripping leads to thermosorption of Ca2 +and Mg2+ as shown in Fig. 9.

Solution collected during thermostripping(with increased concentration of Ca2+ and


13/16

D. Murav iev et al. /React iv e & Functi onal Poly mers 28 (1996) III-126 123

0.1 Ca2

0.0 t-e0 1 2 3 4

Number of cyclesFig. 10. Concentrations of Cazi, Mg*+ and Na+ obtainedin consecutive thermo-sorption-stripping cycles vs number ofcycle.

M g+) can be underwent the repetitivethermosorption-thermostripping cycle wherethe next concentrate of Ca2+ and Mg2+ isyielded. Carrying out several concen tration cy-cles has to result to the further increase of Ca2+and Mg2+ and decrease of Na+ concentrations.The results of 4 consecutive thermo-sorption-stripping cycles carried out are shown in Fig. 10.As seen in Fig. 10 both Ca2+ and Mg2 + concen-tration grows practically linear wh ile also a lineardecrease of Na+ concentration takes place. Thisallow s for the following description of the dual-temperature ion-exchange concentration processobserved:

(16)where aj is the concentration factor; c, and ciare the concentrations of Ca2+ , M g2+ or Na+in the initial sea water and in the stripping so-lution a fter i cycle, respectively. The valuesof concentration factor ai determined for Ca2+ ,M g2+ and Naf in consecutive thermo-sorption-stripping cycles are tabulated in Table 6, wherethe products of ai values for each metal ion arealso presented.

Table 6Concentration factors (a,) for Ca , Mg*+ and Na+ obtainedby consecutive thermo-sorption-stripping cycles on Lewa tit R2.50-K resinCycle CaZ+ Mg+ -Na+1 1.64 1.69 0.952 1.44 1.43 0.953 1.30 1.27 0.944 1.27 1.22 0.97nP_, a, 3.90 3.74 0.82

As follows from the data given in Table 6 theconcentration factors of Ca2+ and M g2+ demon-strate a clear trend to decrease from cycle tocycle while this parameter remains practicallyconstant in the case of Na+. On the other hand,the values of ai can be predicted from the resultsof the first cycle in the following way:

Ci Co i- NACai=Ci= C, + (N - 1)ACAfter rearrangements one obtains:

ACai=l+ C, + (N - 1)AC (18)

(17)

where AC is the absolute increase of a givenmetal ion concentration in thermostripping solu-tion after the first cycle, C, is the sam e ass n Eq.16, and N is the numb er of cycles carried out.

As follows from F ig. 10, AC values remainconstant for both Ca 2+, M g2+ and for Na+ .This allows for computing ai vs N dependencesfor each component of the sea water. The re-sults of these calculations and the experimen-tal ai values are show n in Fig. 11. Comp iarisonof the computed (curves) and the experimen-tal (points) results show s that E q. 18 can besuccessfu lly applied for predicting concen trationdegrees for all metal ions under study after anythermosorption-thermostripping cycle. The de-crease of concen tration factors in the case ofCa2+ and Mg2+ can be attributed to the increaseof their concentration in solution after each cyclethat results in the decrease of CY ,~nd CX,,alues(see above). This can be observed from the datagiven in Table 7, where cr values for Ca *+-Na +,


14/16

12 4 D. Muravi ev et al. I React iv e & Functi onal Poly mers 28 (1996) 111-126

I I I 1 I1 2 3 4 5 6 7 6Number of cycles

Fig. 11. Comp uted (curves) and experimental (points) valuesof concentration factors vs number of thermosorption (SOC )/thermostripping (1O C) cycles. Naf (triangles), Ca 2+ (circles)and Mg2+ (squares), using Lewatit R 2.50-K.

Table 7Separation factors of Ca2+-N a+, Mg2+-N a+ and Ca2+-Mg2+ exchanges on Lewatit R 250-K from thermostrippingconcentrates obtained I = 10C.N o . cycle UNaCa Mg(YNk? Total concentration

(g-equiv/dm3)1 38.2 11.03 3.41 0.722 33.2 9.79 3.39 0.783 26.4 1.95 3.32 0.844 21.3 6.66 3.33 0.95

Mg2+-Na+ and Ca2+-Mg2+ ion-exchange reac-tions on Lewatit R 250-K resin at 10C from therespective thermostripping solutions (artificiallyprepared) are tabulated.3.3. Separation of Mg2+ and Ca2+

As follows from the data shown in Table 7a$$ values remain practically constant (and suffi-ciently high) in all thermostripping solutions ob-tained. This allows to carry out the final purifica-tion of Mg2+ from Ca2+ which can be carried outin applying frontal separation on acrylic cationexchanger in Na+-form. The concentration-

volume history of this process is similar to thatshown above in Fig. 3 and is characterized bythe formation of a solution zone containing pureMg2+ free from Ca2+ impurities. Several ex-periments carried out with stripping solutions onfrontal separation of Mg2+ and Ca2+ have shownsimilar results to those presented in Fig. 3. Thewidth of pure Mg2+ zone depends on the heightof the resin bed and could be easily increased,e.g. by carrying out the process in the countercurrent column [41].

One additional problem has to be solved forcarrying out the final separation of Mg2+ andCa2+. This problem deals with the preparation ofthe resin in Na+-form, i.e. the resin regeneration.This process can be fulfilled in two stages: (1)removal of all ionic species from the resin withacid solution, and (2) conversion of the resinfrom H+ into Na+-form by treatment with alkalisolution. The completeness of each stage can befollowed by controlling pH of solution leavingthe column.

The results of the experiment on the regen-eration of Lewatit R 250-K resin after carryingout the separation of Mg2+ and Ca2+ from ther-mostripping solution concentrate (3rd cycle) arepresented in Fig. 12.

As seen, in Fig. 12a the complete removalof all metal ions sorbed (Na+, Ca2f and Mg2+)needs practically equivalent expenditure of HClsolution. As follows from the data shown in Fig.12b the final conversion of the resin into Na+-form needs also equivalent amount of NaOHsolution.

The composition of the solution yielded afterregeneration with acid allows for mixing it withthermostripping solution used in separation ofMg2+ and Ca2+ so it can be returned back intothe process. This means that even during theregeneration of the resin practically no wastesare produced.Acknowledgements

This work has been carried out with the fi-nancial support of CICYT, the Spanish Com-


15/16

D. Murav iev et al. I React iv e & Functi onal Poly mers 28 (1996) 111-126 125

00 _;

6PH

6

i

Fig. 12. Regeneration of Lewatit R 250-K with HCI and NaOH solutions after thermostripping (see text): V and C are volumeand concentration of the regenerant solution; Q is the capacity of the resin bed.

mission for Research and Development, projectPTR930009. The authors are indebted to Mrs.Maria Oleinikova for her assistance in prepara-tion of the manuscript. Bayer Hispania Indus-trial, S.A. is gratefully acknowledged for supply-ing with samples of Lewatit resins.

References

121[31

M

[51

W I

[7 1

PI

[9 1

R. Khamizov, D. Muraviev and A. Warshawsky. In: J.Marinsky and I. Marcus (Eds.) Ion Exchange and Sol-vent Extraction, Vol. 12. Marcel Dekker, New York,1995, p. 93.J.L. Mero. The Mineral Resources of the Sea. ElsevierPublishing Co., New York, 1965, p. 34.M. Marhol and K.L. Cheng, Anal. Chem., 42(6) (1970)652.M.D. Arguello and J.S. Fritz, Anal. Chem., 49(1 I) (1977)1595.D. Barba, V. Brandani and P.U. Foscolo, Desalination,48 (1983) 133.H.T. Chen. In: PA. Schweitzer (Ed.) Handbook of Sep-aration Techniques for Chemical Engineers. McGraw-Hill, New York, 1979, p. 467.G. Grevillot. In: N.P Cheremisinoff (Ed.) Handbook forHeat and Mass Transfer. Gulf Pub]., West Orange, NJUSA, 1985, p. 427.D. Tondeur and G. Grevillot. In: A.E. Rodrigues (Ed.)Ion Exchange: Science and Technology, NATO AS1 Vol.107. Martinus Nijhoff, Dordrecht, 1986, p. 369.PC. Wankat. In : A.E. Rodrigues and D. Tondeur (Eds.),Percolation Processes, Theory and Applications. Sijthoff

and Noordhoff, Alphen aan den Rijn, 1978, p. 443.[lo] T. Szanya, L. Hanak and R. Mohila, Zhur. Prikl. Khim.,

56, (1986) 2194 (in Russian).[l I] T Szanya, L. Hanak and R. Mohila, Hung. J. Ind. Chem.

Veszp., 16 (1988) 21.[12] B.M. Andreev, G.K. Boreskov and S.G. Katalnikov,

Khim. Prom., 6 (1961) 389 (in Russian).[13] V.I. Gorshkov, A.M. Kurbanov and N.V Apolonnik,

Zhur. Fiz. Khim., 45 (1971) 2969 (in Russian).[14] VI. Gorshkov, M.V. Ivanova, A.M. Kurbanov and VA.

Ivanov, Vestnik Moskov. Univ., Ser. Khim., 5 (1977) 535(in Russian); English translation in Moscow Univ. Bull.,32 (1977) 23.

[15] M. Bailly and D. Tondeur, J. Chromat., 201 (19$0) 343.[16] M. Bailly and D. Tondeur, J. Chem. E. Symp. Ser., 54

(1978) Ill.[17] G. Klein, M. Villena-Bianco and 7: Vermeulen, Process

Design Develop., 3 (1964) 280.[18] V.D. Timofeevskaya, VA. Ivanov and V.I. Gorshkov,

Russ. J. Phys. Chem., 62(2) (1988) 1314.[19] V.A. Ivanov, V.D. Timofeevskaya, V.I. Gorshkov and TV.

Eliseeva, High-Purity Substances, 4(2) (1990) 369.[20] VA. Ivanov, V.D. Timofeevskaya, VI. Gorshkov and TV!Eliseeva, Russ. J. Phys. Chem., 65(9) (1991) 1296.

[21] VA. Ivanov, V.D. Timofeevskaya and V.I. Gorshkov,React. Polym., 17 (1992) 101.

[22] Lewatit Product Information, Edition Jan. 1993, Bayer,Leverkusen.

[23] F.H. Spedding, J.E. Powell and H.J. Svec, J. Am. Chem.Sot., 77 (1955) 6125.

[24] VI. Gorshkov, In: B. Nykolsky (Ed.) Thermodynamics ofIon Exchange, Nauka, Minsk, 1968, p. 122 (in Russian).

[25] D.N. Muraviev, A.V. Chanov, A.M. Denisov, I? Gmarovaand S.R. Tuikina, React. Polym., 17 (1992) 29.


16/16

12 6 D. Muraviev et al. /React iv e & Functi onal Poly mers 28 (1996) 111-126[26] K. Dorfner. In: K. Dorfner (Ed.) Ion Exchangers. Wa lter

de Gruyter, Berlin-New York, 1991, p. 94.[27] K. Dorfner. In: K. Dorfner (Ed.) Ion Exchangers. Wa lterde Gruyter, Berlin-New York, 1991, p. 339.[28] D. Muraviev, Chem . Scripta, 29 (1989) 9.[29] E. Ekedahl, E. Hogfeldt and L.G. Sill&t, Acta C hem.

Stand., 4 (1950) 556.[30] E. Hogfeldt, Arkiv Kemi, 5 (1952) 147 .[31] E. Hogfeldt. In: K. Dorfner (Ed.) Ion Exchangers. Wal-ter de Gruyter, Berlin-New York, 1991, p. 581.[32] Z.A. Samchenko, E.F. Nekrjach and O.D. K urilenko,

Zhur. Fiz. Khim., 41(10) (1975) 1042.[33] K.A. Kraus and R .J. Raridon, J. Phys. Chem., 63 (1959)

1901.[34] O.D. Bonner and J.R. Gverton, J. Phys. Chem., 65(1961) 1599.[35] G.L. Starobinets and VS. Soldatov, Z hurn. Fiz. Khimii.

37 (1963) 294.[36] VS. Soldatov and VA. Bychkova. Ion Exchange Equi-libria in Multicomponent Systems. N auka i Tekhnika,Minsk, 1988, p. 86.[37] P Fletcher and RX Townsend, J. Chem. Sot. Farad.Trans. 2,77(6) (1981) 955.

[38] P Fletcher and R.P T ownsend, J. Chem. Sot. Farad.Trans., 77(11) (1981) 2077.[39] F. Helfferich. Ion Exchange. McGraw-H ill. New York,1962.[40] F.G. Helfferich and Hw ang Y ng-Long. In: K. Dorfner(Ed.) Ion Exchangers. Walter de Gruyter, Berlin-NewYork, 199 1, p. 128 8.[41] VI. G orshkov. In: J. Marinsky and I. Markus (Eds.)Ion Exchange and Solvent Extraction, Vol. 12. Marc elDecker, New York, 1995 , p. 29.

Ca e Mg da agua do mar remoçao

Documents

Transcript of Ca e Mg da agua do mar remoçao