Preparation and characterization of transition metal ion-selective membrane electrodes

f s K Mcno ,~,~ S Amarc h a~t &\Y K Agrawc;' ;

Departlllent o r C hellli st ry. School o r Sciences, Guj arat Un ive rs it y, Ahllledabad 380 009 y.;a

;' I?eceil'ed .J Septclllber 2002; rel'ised received 12 Se/!telllher 2003; accepted 30 j mll/WT 200.J

\/ Ion se lec ti ve elec trodes o r trans iti on meta l ions Co(l l). C u( ll ) and Zn( lI ) have heen fabri cated us ing the

co rrespond in g hydrox i1 mate complt:xes. These e lec trodes have been systelllati ca lly charact e ri zed hy studYIng the

response tillle. e lec trode response. pH range and se lec ti vi ty coeffi c ient s over c losely associated me tal s . The e lec trode

respon se in non-aqueous and the stab ilit y ove r the mediulll have bee n studied .

IPC Code: GOIN271333 Keyword s: Ion se lec ti ve e lec trode!. Transi ti on Illeta l ion ' , Co(II ), Cu(ll ) ;lIld Zn( II ), Hydroxalllate cO lllplexes --- r __

The increasing demand for chemi ca l surveillance in environmental protecti on, medicine and many industri al processes has created the need fo r sensors wi th features such as hi gh se lecti vity. sensiti vity, re li ab ility and sturdiness. These demands can often be sati sfied by ion se lec ti ve elec trodes (TSEs), which are co mmonl y used ow ing to their sim pli cit y, lower cost and fas t provision fo r analyti ca l results U

Copper ion selective electrodes are used as sensor"). Convent ionalliqu id membrane e lec trodes~6 were replaced by carbon rod supported polymer membrane electrodes and coated wire ISES7.'} A few electrodes have been reported for Co(l l) and Zn(II )IO.13

In the present in vesti gation , so lid state electrodes were fab ricated fo r sensin g Co(ll) , Cu(II ), and Zn(II ). The chelating ability of hydroxaillic ac id to form very strong complexes with transiti on meta l l~ has prompted the use of metal hydroxamates as the electro-act ive material for the preparation of membranes and furth er for the fabri cati on of membrane ion se lective electrode (MISE).

Experimental Pr'ocedure

C hemicals

All the chemicals used were of A. R. or G.R. grades of E. Merc :lnd B.D. H., respec ti vely unl ess

" For cnrrcspond~E-llla il : drykagrawal@yahoo.com; Fax 079-( 308545)

otherwise specifi ed. Ion free double distilled water was used throughout the experiment.

Metal salt solu tions

The stock solutions of the metal ions were prepared by dissolving weighed quantiti es of the nitrates and final concentrat ions were determined volumetricall y I) and by atomic absorpti on spectrophotometry.

Instruments

A Philips DC-Micro Model PP 9004 with a saturated ca lomel electrode as re ference electrode were used for potenti al measurements controlled pH analyser. PHAN 6E488, was used for the pH measurements. The conductance was measured on a Systroni cs Model 303 direct measuring conduct ivity meter. The metal ion concentrations were determined by GBC 902 atomic absorption spectrophotometer. All measurements were made at 25 ± 1 "C.

Preparation of membranes

The electro active materi als required for the preparation of MISE were the corresponding metal­hydroxa mates of N-pheny lbenzo hydroxami c ac id (PBHA). The reagent PBHA was synthesised by the repo rted method 1(, . The meta l hydro xamate were prepared by followin g th e ge ne ral me th od of prec ipitation l7 by the proper control of pH (Co : 5.5-6.5 pH, Cu : 3.6 - 6.0 pH and Zn: 5.3-6. 8 pH). The

202 IN DIA N J. CHEM. TECHNOL. , MARCH 2004

]J urity of the complexes was checked by elemental analysis.

The elect ro acti ve material (200rng) was mi xed wi th PVC ( 100 mg) in 6 Il1L terrahydrofuran (THF) and fi nally mixed wi th two drops of dioctylphthalate. The resulting solution was carefully cast into a slide glass and left for slow evaporation to get a thin membrane1x.

Characterization of memhranes

The physicochemica l properti es of the membrane, \'i-; .. thickness, water content , exc hange capac ity etc. were determ ined l ~." 1 after conditi oning the mem brane as described ea iier"".

Conditioning of the membrane

The membranes were conditi oned fo r tes ting by equilibra ting with I M sodium chl oride and a few mL of sodium acetate to adj ust the pH 5-6.5 (to neutrali ze the acid present in the fi lm ) fo r 24 h, prolonged equilibration in 0. 1 M sodium chl oride preceded all mea. 'uremen ts .

Electrolyte a hsorption

For the dete rmination of electrolyte absorption , the membranes were immersed in I M sod ium chloride so lution unti l equ ilibrat ion was attained, afterwards taken out, wiped free of adhering electrolyte, and dipped in 25 mL of co ndu cti vity water for 2-4 h with in termittent shak ing. The solution was transferred into a 100 mL measuring fl ask. The process was repeated th rice and th e entire so lution was co ll ec ted in a measuring f'lask and made up to mark with conductivity water. The conducti vity of the solution was measured.

Speci fic conductance

For elec tri cal conductan ce measurements , the mem branes were ccmcnted between two half ce ll s with o I M sodiu m chl ori de so luti on. After equilibrati on, the solution of sod ium chloride was replaced by mercury previous ly equ ilibrated with sod ium chl oride solution.

The conductance was measured by connect ing plati nu m electrodes dipped in the mercury to the conducti vity bridge.

Fabl'ication of MISE

The MISE were fab ri cated by th e repo rted procedurel X, Membranes of appropriate sizes were cut fro m these mas ter membranes and were mounted at the lower end of the g lass tube. A 0.1 mol dnl" metal soluti on was used as internal reference so luti on. Functioning of electrodes

The electrochemica l behaviour of prec ipitate ion­selective electrodes is based on so lu bili ty equiiibria and precipitate exchange react ions, ex ist ing at the phase boundaries of the electrode membrane. The phase boundary potenti a l of th e e lectrode in so lu ti ons containing onl y the primary ion, i, as the prec ipitate of the electrode membrane, is g iven by,

RT (o i ) E = E + - In -' : where

II ZiF (o i ) III

E is the electrode [Jotential, E the standa rd electrode " potenti al, ((Ii), and (oi)1ll are th e ac ti vities of the primary

ion in the so lution and in the membrane phase. The selecti vi ty of MISE towards primary ion

lover th e inte rfe rin g ion k of activity O k can be expressed (IS E = constant + S lo a (0 +KI'I'f({ III'k)

'"- . b i I~ k

where KP'" is the se lectivity coefficient, Zi and Zk are ,k

the valences of i and K and S is the slope of the electrode per decade change in activity.

Results and Discussion The propert ies of the membranes, VI Z.. water

content, porosity, swelling, electro lyte absorption and specific conductance are given in Table I. The electrodes for the transiti on metals Co(lT).Cu(JI) <l nd Zn(II) , have been prepared and were characterized by stud ying the electrode respo nse, respo nse time, pH range, and selec tivity factor. The data are g iven in Table 2.

Tahle I-Characteri 7.a tion or membranes

Na m.:! or th e hydmx~lm~ltc

mcmb ranc

C'll( II )

Cudl)

Thickness (mm )

1.38

0.67

1.25

Water contcnt as o/r weight or wet membrane

25.70

25.60

14.70

Po rosity

0.207

0.207

0.065

Swelling % w I. or wet

mcmbrane

0.5 13

0.513

0.337

Elec tro lyte Speci ric absorption cunci Llc t i vi t y NaCl g" or wet ol1n1" Cill "

membrane pg

27 2...13 x i O'

56 2.44 x 10-1

45 2.43 x 10-'

MENON 1'1 al.· TRANSITION METAL ION SELECTIV E MEMBRANE ELECTRODES 203

It can be seen from Table 2, that , the ex ten t of swe lling depends upon the thick ness of the membrane; as the thickness decreases, the swelling increases. The order of poros ity and water content of the membrane are, Porosity: Zn<Co "" Cu Water content : Zn<Cu "" Co

Generall y, an ideal membrane should have less thick ness, moderate swelling, porosity and water content capacity. To prepare a functional membrane for use in ion-selec ti ve electrodes, it is usually desirable to select a plasticizer in which the ionophore is so luble. For thi s reason , several plasticizers were studied. All ionophores were more so luble in dioctylphthalate and showed best respo nse co mpared to di octyl adipate, dioctyl phenyl phosphonate, diphenyl phthalate and bi s (2-ethyl hexyl) sebacate and hence dioctylphthalate was used as the plasti cizer th roughout the study. The MISE fabri cated from these membranes were characterized by studying the response, electrode response, electrode response in non-aqveous so lvents, pH range and select ivity factor.

Response time

The electrode res ponse was studi ed for th e electrodes Co(l l), Cu(ll) and Zn(II ). The response time is usuall y measured by reco rding the emf of the electrodes as a funct ion of time, when it is immersed in the so lution to be studied. The electrode is first dipped in 0. 1 to 0.0 I dm-·l solution of the ions studied and immediately sh ifted to another solution of ten fold concentrati on of the ion. The potential of the so lution gives the read ing at zero second and after immed iate ly dipping the electrode in the second solution, the potential is noted at 5 s intervals. The potentials are then plotted again st time. The time taken by the electrode to obtain a steady potential is the response time of the elec trode (Table 2). All the electrodes studied showed very prompt response, with response time between 20 and 40 s.

Electrode response

To determine the electrode response of MISE, the

electrode potential was measured in a series of solut ions with the concentrati ons ranging from 0.1 to I x 10 7

mol dm-·l .

The response curve is linear up to a particular concentrat ion, after which the curve tends to become parallel to X-ax is. Suitable concen trations were chosen corresponding to the sloping portion of the linear curve fo r the measurement of potentials . The point when the curve becomes parallel to the activ ity or concentration ax is is known as lower detecti on limit of the electrode (Table 2).

A lin ea r respo nse was o bserv ed in th e concentrati on range I x 10.1 - I x 10-4 mol dm·.\ in the case of Cu( I1 ) electrode and up to I x 10-5 mol dm ; for Co(l l) and Zn(ll ) electrodes (Fig. I). The slopes derived from the linear cu rves have a near Nern stian slope for Co(ll) and sub-Nern sti an response for Cu(l l) and Zn(lI ) electrodes. The MISE electrode response depends upon the chemica l stJ1Jcture, so lubility products, stab ili ty and th e ph ys ica l properties of th e I iga nd a nd th e precipitate:! .l :!4. The deviation from th e Nernstian response for Cu(lI) and Zn( I1 ) e lectrodes may be due to

160 y=·10x+150

R2 = 1 140 Co (1·5)

120

100 Zn (1-5)

> 80 E

60 _ ><_. __ * _--x

Cu (1·5)

40

20

0

0 2 3 4 5 6

pM

Fig. I- Calibral ion cu rvc for Cn( II ). Zn(lI ) & Cu(lI )

Tab le 2- Characleri slics of l he eleclrodes

Eleclrode Response li me pH range Slope mV Deleclion Life lime Dcteclion (in seconds) decade" range (momhs ) limil

Co(lI ) 30 2-R 33 10'; - 10" 4-5 3.5 x 10';

Cu(lI ) 20 3- 10 17.5 10-1- 10' :1 7.9 x 1 0~'

Zn(lI ) ·W 4- 10 23 10 ;- 10-' 3-4 6.3 x 1 0~'

204 I DI AN J. CH EM . TECHNOL., MARCH 2004

principally ionic conduction through the membrane and al so to the point defects in the crystal lattice which leads to a dev iation":']!'. However. since a linearity is observed in the electrode response, these electrodes are used fo r the quantitati ve determination of these metal s.

The electrode response was studied in various non-aq ueous so lvents, to find out th e viability of applicati on in non-aqueous systems. The solvents chosen were ethanol, methanol, di oxane, DMF, DMSO and acetone in 30-70 ifr: in water.

The calibrati on curve is pl ott ed in each case and the slope of th e curve is compared with the aqueous system. The studi es were repeated after every 5 h to find out stabilit y of the elec trode in the vari ous non­aqueous solve nts.

The respo nse stud ies showed, t hat, the electrodes are stab le only in ethanol Hndmethan ol systems whereas in all other so lvent sys tems the dev iati on of slopes is observed compa red to aqueous sys tems.

The e lec trode res ponse of Co( ll ) in variou s so lvent s and in S0 9'c methanol and ethan ol systems wit h time is given here in Table 3 as a representati ve case, whic h shows that the response is steady up to 30 h after whi ch the potenti al ge ts dev iated.

The electrode response was checked peri odi call y to ge t the lifetime of the elec trodes . The electrodes ha ve a li re time be tween 3 and 5 months (Table 2).

Effect of pH

The pH studi es have also been carried out in the t \vo co nce ntrati on ranges IO-e and 10·-1 M in different pH betwee n 1.0 and ! 3.0 and results are given in Tab le 2. The data revea ls that the electrodes have a \V ide wo rking range of pH 2- 10 (Fig.2).

Selectivity coefficients

The se lec tiv ity coeffi cient s depend upon the ex perimental conditi ons, predominating on the activities of the primary ions (al) and of interfering ions (Ok) and the ioni c strength.

Th e se lec ti vit y coe ffi c ie nt s of th e variou s

e lectrodes we re studi ed with respec t to c lose ly associated metals by the mixed solution method, and the K,~" l is calculated as discussed elsewhere l x

. In most of the cases the fix ed concentration of 10.2 and 10·-1 M inte rferin g ions we re us ed aga inst th e varyin g concentrati on of primary ion. The selecti vity coefficient obtained by the calculati on method and graphi cal method were co mparable (Table 4).

Of th e e lec trodes, Zn(ll ) ex hib its maxi mum selectivity and can be co rrelated with the low poros ity of the membrane, water content. swell i ng and spec i fi c conducti vity which are theoretically responsible for high se lectivity. Here the diffusi on will be on ly through the electroacti ve material and not through the interstiti al liquid film.

100

90 CO(II)

80

50 ZI1(II)

40

30

Cu(ll)

:i--- , - - . 6 10 12

pH

Fig. 2- EITcC L or pH on thc poLenLi al or Co(ll ). Zn(lI ) & ClI ( lI )

T,IOIc :I-ElecLrode response or Co(ll ) in SO'7p meLhanol/cLh anol-water media Hnd 50% waLer organi c so lvenL media Co( lI ) response

Time (!lr~) -,

Slope (mV decadc ) Siopc (mV decade -,

)

ELhanol M eth anol 5 33 33 DimcLh yl fo rm<lmidc 16 10 33 33 15 33 33 20 :13 33 AceLone 15 ~"i 32.S 30 Dioxane 12 30 33 28 -w 2ll 28 Dimcthyl su lphoxidc 10

50 30 25 60 27 22 70 18 20

MENON el a/. ' TRANSITION META L ION SELECTIV E MEMBRANE ELECTRODES 205

These electrodes were selective over their closely assoc iated metals. Zn(lI) electrode exhibited se lectivity over Pb( lI ) and Cd(lI) which are the metal s qu ite often associated with eac h other. The interference of Pb(lI ) in Cu(lI) electrode was very small with Kr::'Ph =0.078. Further the Zn(lI) electrode, with very closely assoc iated metals like Cd( lI ) and Hg( lI ) show very negligible interference Kzl':,'.'H ." = 0.029 and KI''''

- L.: Zn . l-I ~

=0.m2. The extent of interference of Be(lI) in Cu(lI)

electrode may be due to the similarity of their ioni zation potentials [Be( II)=27.52eV; Cu( II )=28.03eVl and so also is the case with Hg( lI ) in Cu(lI) electrode [Hg( lI ) = 28.94eV]. Ionization potentia ls are only one of the factors for the prec ipitate exchange reac ti ons at the phase boundary of the electrode during the sensing of

co-ion by the precipitate based membrane. The precipitate based membranes are prepared from metal co mp le xes by the reaction of a complex ing agent with meta l cations and the acti on of the prec ipitate undergoes prec ipitate exc han ge reac ti on with the co-ions of the tes t soluti on. Since the formati on and stab ility of the hyd roxalllate complexes depend on the ioni zati on potential of the metal cation , the prec ipitate exc hange may depend on the ioni za tion potential of the cation of the prec ipitate and the co-ions, which are getting exc hanged.

Similarly, ioniza tion potenti al is onl y one of the factors for interference. Zn(lI) and Be( lI ) possess alillost sa me ioni za ti on pot e nti a ls [Zn ( II )=27.2 ge V: Be(II)=27.S2eV]. However, Be( lI ) does not interfere \V ith Zn(lI) electrode. These responses may be related with the stabi lity of their respecti ve metal hydroxamare and

Table 4-Selectivit y coeffi cient KP'\ for transition metal ion sensors

Sensor I x 10-' 11101 dnl" solution of interfering ion

Ba(II ) Be( ll ) Ca(II ) Cd(II ) Co(l l) Cu( II ) Hg( ll ) Mg(ll ) Pb(ll ) Zn(l l )

Co(l l) 0. 165 0.835 0. 165 0.580 0.46 1 0.792 0.052 0. 360 0. 51 0

Cu(ll ) 0.461 0.lD5 0. 164 0.461 0.256 0.576 0 . .1 55 oon O-l() I

Zn( II ) 0.Q21 0.000 0.20 0.029 0.025 0.029 0.032 0.000 0.035

Table 5-Applicabi lit y of ISE in environ l11ental samplc analysis

Samples' Metals found , ppm

Co( lI ) Cu(II ) Zn( lI )

ISE AAS ISE AAS ISE AAS

App le 0.02 ± 0.01 0.02 1.20 ± 0.02 1. 18 0.50 ± 0.02 0.52

Orangc 0.04 ± 0.02 0.03 1.2 1 ± 0.02 1.20 2.50 ± 0.02 2.-19

Potato 0.04 ± 0.01 0.05 2. 80± 0.03 2.80 6.50± 0.05 6.5 1

I1e,lIl 0.04 ± 0.02 0.04 .1 .50 ± 003 .1 .49 850 ± 0.06 ,). 50

Corn O.OJ ± 0.01 0.03 4.00 ± 0.05 4.00 7.50 ± 0.06 7. -1 :-:

Urinc 0.05 ± O.DI 0.05 5.20 ± 0.03 5. 19 l .flO ± 0.03 1.60

Blood 0.03 ± 0.01 0.02 0.07 ± 0.04 0.06 3. 80 ± 0.08 U;O

Ri ver wa t cr ~ 0.07 ± 0.02 0.06 0.20 ± 0.01 0.20 0.52 ± 0.02 0.50

EfIlucllt ' .1 .50 ± 0.05 348 7.50 ± 0.05 7.50 10 0 ± 0.07 IOO()

, Averagc of , 1

15 determinations, . Saharmati river, Ahmedabad. Efiluents from Dyes Industries of Vat V<l , Ahmedahad.

206 INDIAN J. CHEM. TECHNOL. , MARCH 2004

their relative solubilities . The less se lectivity of Cu(JI ) membrane elec trodes compared with the other elec trode may be due (0 the high porosity of the membrane.

The select ivit y of the elec trodes was also inves tigated in solutions of several in organic anions such as chl oride, sulphate, acetate , nitrate, bromide, iod ide and thi osulphate using the same mixed solution method and found that almost 100 times excess of these ion s do not change the potenti al and cause intcret'e rnce.

The fa bricated electrodes ha ve been used for the determ ination of Co(l l), Cu(l£) an d zinc(TI ) in blood and environmental samples . The known we ight of the sample (0.5-1.0 g) was trans t'erred in to 200 mL Borosil beaker, 25 mL of perchlori c acid (60%) and 5 mL of co nce ntrated nitri c acid was added , fo ll owed by diges ti on on a sand-bath t'or I h and evaporation to c1ryness. The res idue was heated with 10- I 5 mL concentrated hyd rochloric acid along with 0.5 g of ammonium persulphate, centrifuged and the solution was dilu ted with di still ed water to 100 mL. The concentrat ion of t he metal ions were determined with ion se lect ive electrodes. The concentrati on of zinc and cobalt in the sa mpl es was determined by standard additi on method27

. The reliability of the method was compared with AAS and results given in Table 5 are in good agreement.

Conclusion The analytical characteristi cs of the membrane

elec trodes in ves ti ga ted here, mostl y depend on the physical properti es of the membrane, the stabi lity of ,he respective metal hydroxamates and their relati ve sol ubiliti es . The selectivity and sensiti vity of Zn( ll ) electrode was better than Co(l l) and Cu(lI ) electrodes. The membrane electrodes fabricated are stable and can be successfully used for non-aqueous studi es and stabil ity constant de termi nat ionsc2

References I Convigton A K, lOll -se lect i ve U CClmde Methodolog.\', Vol

2, (CRe. Boca Raton , FL), 1079. 90.

2 Oe~ch U, Amman D & Simon W, Clill ClII'III. 32 ( 1978)i'+-I'-

3 Archer S, FASI:'N ) , 7 (1993) 749.

4 Ca rnrn an K. Fres(,lIill.l· Z Allol ClII'III, .l20 ( 1088) 691 .

5 Wang E & Kamata S, Allol Chelll AC/a, 261 ( 19lJ2) :199.

6 Ca rd we ll C J, Catt ratt R W & Hauser P C. Allol Chilli

ACla , 241 ( 1988) 359.

7 Analdi A & Epste in S I. Allol Chelll, 45( 1(73) 595 .

8 La wton R S & Yacynych A H, Allol Chelll Acta, 160 ( 108.+)

1'+9.

9 Rocheleac h M J & PLirdy W e. Tola ll !ll. 37 ( 1000) 307.

10 Asw<l th i S p, Tlwillni T C, Kulkarni V T & Sundaresan 1\1

J , L'leclmchelll Sec. 50( I ) ( 109 I ) 23.

II Viksna, La/ll, 2(6) ( 1092) 6X4.

12 J:lber AMY, Moody G J & Thol1las J D R, Allalrst . t 02

( 1977) 943 .

13 Hampton M D, Peters C A & Wellington LA. Allal Ch('lll

Acta, 17 1 ( 1987) 1994.

14 Ros hania R & Agrawal Y K, .I Chl'lII I:'II ,~ /)alll. 23 ( 197R)

269. 15 A Vogel's Text Book oj' QIlWili /(il i l'(' Ch(,lIl iml Allall·sis. 5th

eel n (Longman, U. KJ. I <)X9.

16 Agrawal Y K & Tandon S G, .I Ch<'111 Ellg {J IIW. 16 ( 1971 )

371. 17 Bhatt K & Agrawal Y K. Snilh III()/ :~ (vl('/(I1 (hg Ch<'/Ii, 2

( 1972) 175.

18 Craggs A. Moody G J & Thomas J D R, .I Chelll Edll. 51

( 1074) 541. 19 Sri vas tavil S K, Jain A K, Agrawa l S & Singh R P. TIIIIIII/a.

25 (1978) 157. 20 Jain A K & Singh R p, Ilidiall } T<'chllol. 10 ( 19X I ) 192.

2 1 Grego r H p, Jacobson H, Shair R C & \Ve~tons D M.} Phn

Chelll , 61 ( t957) 141.

22 Amarchanel S, Menon S K & Agrawa l Y K, UeCll'Oww/rsis.

12 (2000) 522.

23 Bailey P L, Allo lrsis allll l oll Sell'ctil'l' £I<'C/IVIil's (Heydan.

Londo n), 1976, 44.

24 Moody G J, Owasu R K & Thol1las J D R. ;\1111/.1'.1'1. 113

( 1988) 65.

25 H-Friser, IOIl -Sele('t il '(' /:'I('ctrodes ill A llolrlical Chellli.I·lrr:

Vol s I & 2 (Plenum, New York ), 1980.

26 Pun go r E, l Oll-Select i ve L/('(' tl'Odl's. Vol. 22 (E lsevier.

Amsterdam) 1985.

27 Durst R A, Microchilll Acta , 6 11 ( 1969).